Meet the World\'s Challenges

2014 ASNT Annual Conference Paper Summaries

Meet the World’s Challenges Charleston Convention Center, Charleston, SC, USA • 27–30 October 2014

Copyright © 2014 by The American Society for Nondestructive Testing. Exclusive of those papers that are a work of the federal government and not subject to copyright. The American Society for Nondestructive Testing, Inc. (ASNT) is not responsible for the authenticity or accuracy of information herein. Published opinions and statements do not necessarily reflect the opinion of ASNT. Products or services that are advertised or mentioned do not carry the endorsement or recommendation of ASNT. No part of this publication may be reproduced or transmitted in any form, by means electronic or mechanical including photocopying, recording or otherwise, without the expressed prior written permission of The American Society for Nondestructive Testing, Inc. IRRSP, NDT Handbook, The NDT Technician and www.asnt.org are trademarks of The American Society for Nondestructive Testing, Inc. ACCP, ASNT, Level III Study Guide, Materials Evaluation, Nondestructive Testing Handbook, Research in Nondestructive Evaluation and RNDE are registered trademarks of The American Society for Nondestructive Testing, Inc. First printing 12/14 Errata, if available for this printing, may be obtained from ASNT’s web site, www.asnt.org. ISBN-13: 978-1-57117-355-3 Printed in the United States of America Published by: The American Society for Nondestructive Testing, Inc. 1711 Arlingate Lane Columbus, OH 43228-0518 www.asnt.org ASNT Mission Statement: ASNT exists to create a safer world by promoting the profession and technologies of nondestructive testing.

ii

ASNT Annual Conference Program Chairs Welcome to the ASNT 2014 Annual Conference. Some members may remember when the Annual Conference was held in Charleston in 2008. Today, like then, the ASNT Annual Conference is where NDT professionals Meet the World’s Challenges by sharing information on cutting edge techniques and applications, networking with peers and industry innovators, and showcasing new products and services. Attendees, presenters and exhibitors come from more than 30 countries representing the full-spectrum of NDT user industries. We have an outstanding program this year that represents the NDT industry internationally across a broad spectrum of topics. And, of course the NDT Annual Conference is more than just a technical program. More than 165 companies are participating in the exhibit, which by the way is where you will find lunch every day. As your program chairs, we are proud and pleased to be members of ASNT and to have the opportunity and privilege to work on the Annual Conference Committee. We hope you each find this conference a valuable and rewarding experience.

Claudia V. Kropas-Hughes Program Chair

Trey Gordon Program Co-Chair

Keynote address Tuesday 28 October 8:00 – 8:45 am

Boeing South Carolina Capabilities and Technologies Dan Mooney was appointed vice president of engineering for the Boeing South Carolina Engineering Design Center in June 2013. He is responsible for guiding the growth and the capability of the newly established design center, as well as overseeing all matters pertaining to the engineering statement of work for BSC. Prior to coming to South Carolina, Mooney served as the vice president of Aviation Safety and Engineering Function, 787-8 Development, Regulatory Affairs for Commercial Airplanes, the 747/747-8 program, product development, and was chief project engineer for the 767 and 757300 programs. He has a Bachelor of Science degree in civil engineering from Pennsylvania State University. He is a Fellow of the Royal Aeronautical Society and an Associate Fellow of the AIAA. Daniel P. Mooney Vice President of Engineering, South Carolina Engineering Design Center, Boeing Commericial Airlines

Mooney will provide an overview of operations at Boeing South Carolina – from the decision to locate a facility in South Carolina, to the site’s strategy and plans for growth. He will discuss some of the advanced manufacturing and inspection processes used in the fabrication and assembly of Boeing’s flagship 787 Dreamliner, as well as the advanced manufacturing and inspection challenges for the future.

iii

PLENARY ADDRESSES Wednesday 29 October 8:00 – 8:45 am 2013 Lester Honor Award Lecture

The Journey of Nondestructive Testing (NDT) Looking back at early, primitive NDT applications, which date back to World War II, it is obvious that NDT methods and techniques have achieved tremendous expansion and advancements since their inception. These achievements, which have come about primarily due to industry needs, are the result of the tireless work of NDT pioneers, advancements in the industry, and the digital age. These advancements have been and are still being confronted with ever growing challenges as the role, reliability, and significance of NDT to the industry increases. This paper is an attempt to address the past use of NDT, current NDT applications, and prospects for NDT in the future.

Nat Y. Faransso KBR, Houston, TX

Nat Faransso is Corporate NDT Level III, Kellogg Brown & Root LLC, Houston, Texas. He is a past president of ASNT, and a member of several ASME and ASTM NDE committees. He is also an AWS Certified Welding Inspector and an ASNT Level III in MT, PT, RT, UT, and VT. Nat Faransso is Corporate NDT Level III, Kellogg Brown & Root LLC, Houston, Tex. He is a past president of ASNT, and a member of several ASME and ASTM NDE committees. He is also an AWS Certified Welding Inspector and an ASNT Level III in MT, PT, RT, UT, and VT.

Thursday 30 October 8:00 – 8:45 am 2014 NDT Industry Salary Survey – The Numbers

Don’t Lie: A Strong Recovery in NDT Continues As the global economy continues to heat up, the nondestructive testing industry has been a bellwether of renewed economic activity. Employment recovery within the NDT industry, as established by the benchmark annual PQNDT NDT Salary & Benefits Survey, showed a marked upturn well in advance of the more general rebound, serving as an indicator for the increased production that followed. Michael Serabian, the author of PQNDT’s Salary & Benefits Survey, will report the results of the 2014 edition of the survey, which confirm and reinforce this positive trend. Full-time NDT employment is up, compensation is increasing, and benefits continue to be strong for full-time workers. Mr. Serabian will also address the messages between the numbers, including an analysis of which industry segments are leading the way, which geographical regions are thriving, and which certifications are in the highest demand. Michael P. Serabian President, PQNDT Inc., Arlington, MA Michael Serabian, is the president of PQNDT, Inc., a specialized recruitment and job placement firm headquartered in Arlington, Massachusetts. Mr. Serabian is a nationally recognized speaker on career issues in the field of NDT, quality and inspection personnel. PQNDT has been serving the personnel recruitment and placement needs for the NDT industry since 1967.

iv

SPECIAL PLENARY ADDRESS Thursday 30 October 11:00 – 11:45 am

The Georgiana: Mystery Ship of the Confederacy Edward Lee Spence was born in Germany in 1947 and found his first 5 shipwrecks by the time he was 12 years old. He is a pioneer in underwater archaeology for the study of shipwrecks and sunken treasure.

Dr. Edward Lee Spence Underwater Archaeologist, Shipwreck Consultant, NOGI Award Winner

Dr. Spence's talk will cover his overlapping research and discoveries of the steamer Georgiana and the "real Rhett Butler." The Georgiana was a British blockade runner that was rumored to have been the most powerful Confederate cruiser ever built. Other records say the iron hulled steamer was so lightly built that if gun were fired from her decks that she would shake from stem to stern. The Georgiana was not fortunate. She was sunk on her maiden voyage while trying to run through the Union blockade into Charleston. Spence discovered the wreck his senior year of high school and eventually salvaged over a million individual artifacts from the wreck. Officially, the Georgiana and her cargo were owned by shipping magnate George Alfred Trenholm. Trenholm was a true merchant prince from Charleston and was variously described as brave, tall, and handsome. Trenholm is best known as the last Treasurer of the Confederacy and the head of the three most important blockade running firms in the Civil War. Trenholm's companies made today's equivalent of several billion dollars in less than five years of blockade running. But Trenholm's greatest claim to fame is that he was the historical basis for the dashing blockade runner captain Rhett Butler in "Gone With The Wind."

v

Conference Program: Tuesday 28 October 2014 Keynote Address: 8:00–8:45 am – Boeing South Carolina Capabilities and Technologies Daniel P. Mooney, Vice President of Engineering, South Carolina Engineering Design Center, Boeing Commercial Airplanes

BALLROOM B

Eddy Current I

NDE of Composites I

General NDT I

BALLROOM B

BALLROOM C2

BALLROOM C3

3D Eddy-Current-Based Inspection of Lightweight Components – from Idea into Realization S. Hillmann, M. Schulze, M. Pooch, H. Heuer, Fraunhofer IKTSMD

A Comparison of Acoustography with Other NDE Methods for F.O. Inclusion Detection in Graphite Epoxy Laminates A.Poudel, S. Shrestha, and T. P. Chu, Southern Illinois University; J. Sandhu, Santec Systems, Inc.; C. Pergantis, The U.S. Army Research Laboratory

Application of Flash Thermography in Automotive Carbon Fiber Composites H. Zhao, P. Blanchard, Ford Motor Company

Developing New Generation ECT Flaw Detectors to Meet Client Needs J. Hansen, ETher NDE

A Nondestructive Evaluation Technique for Detecting, Locating and Quantifying Damage in Large Polymer Composite Structures Made of Electrically Nonconductive Fibers and Carbon Nanotube Networks A. Naghashpour, S. Van Hoa, Concordia University

Evaluation of Friction Stir Welds with Advanced Ultrasonic and Eddy Current Techniques E. Todorov, R. Spencer, H. Castner Edison Welding Institute

Eddy Current Transfer M. Collingwood, A-Scan Laboratories Inc.

Acoustic Emission for Damage Evaluation in Realistic CFRP Components R. Austin, Texas Research Institute Austin Inc.; M. ElBatanou, M. Abdelrahman, P. Ziehl, University of South Carolina

Automatic Conductivity Scanning of Rolled Aluminum Plates for Aerospace Applications H. Ghaziary, Advanced NDE Associates; W. Johnson, AUT Consulting Inc.; A. Haszler, Altech Consulting GMBH

Tracing Defects in Glass Fiber Polypropylene Composites using Ultrasonic C-Scan and X-Ray Computed Tomography Methods A. Hassen, M. Yester, U. Vaidya, University of Alabama at Birmingham; A. Poudel, T. Chu, Southern Illinois University

Chair: A. Poudel, Southern Illinois University

Chair: M. Collingwood, A-Scan Laboratories Inc.

Chair: G. Garcia, EVRAZ North America

9:30 am 10:00 am 10:30 am

11:00 am–1:00 pm Lunch in the Exhibit Hall vi

9:00 am

Reducing Arc Flash Risks with Electrical Maintenance Safety Devices M. Robinson, IRISS Group

NDE of Composites II

Infrastructure

Chair: G. Garcia, EVRAZ North America

Chair: A. Poudel, Southern Illinois University

Chair: M. Miceli, Miceli Infrastructure Consulting LLC

BALLROOM B

BALLROOM C2

BALLROOM C3

Eddy Current Inspection of Twisted Tube Heat Exchangers / Field Trial Test Results T. Rush, Mistras Group, Inc.; S. Hoyt, Marathon GBR; O. Lavoie, Eddyfi, Koch Heat Transfer Company

Application of Advanced Non-Contact Ultrasound for Composite Material Qualification A. Bhardwaj, K. Patel, The Ultran Group

The Role of NDT in Risk-based Inspection for Highway Bridges G. Washer, M. Nasrollahi, University of Missouri

Examination of Tube-toHeader Welds with Flexible Eddy Current Probes J. Bartlett, G. Burkhardt, Southwest Research Institute; S. Walker, EPRI

Inspection of Composite Stingers Using the SAUL (Surface Adapted Ultrasound) Method with Circular Arrays G. Neau, M2M; I. Ivakhnenko, ATK Aerospace Systems

Does Structural Health Monitoring (SHM) Provide Safety and Maintenance or Confusing Data? T. Tamutus, R. Gostautas, Mistras Group; M. Johnson, California Dept. of Transportation;

2:00 pm

Eddy Current II

Health Assessment of Heat Exchanger Tubes through Eddy Current Testing (ECT) and Internal Rotary Inspection System (IRIS) and Their Comparative Study Performed on Various Tube Materials. A. Al-Shamari, M. Al-Shaiji, G. Kumar, A. Gupta, Kuwait Oil Company

Numerical Analysis of Acoustic Spark Source Focused by an Ellipsoidal Reflector for Air-Coupled Ultrasonic Excitation (2012 Fellowship Award Winner) Y. Tsai, J. Zhu, M. Haberman, University of Texas

2:30 pm

1:30 pm

1:00 pm

Conference Program: Tuesday 28 October 2014

The Pulsed Eddy Current Testing Technique Research for Austenitic Stainless Steels H. Wang, J. Wang, Nanjing University of Aeronautics

Using Guided Waves to Monitor for Fatigue Damage on Electrical Transmission Tower Line Hangers G. Light, S. Vinogradov, Southwest Research Institute

3:00-3:30 pm Refreshment Break in the Exhibit Hall – Sponsored by Mandina’s

vii

Conference Program: Tuesday 28 October 2014

Chair: T. Chu, Southern Illinois University

Chair: K. Schmidt, Evisive, Inc.

BALLROOM B

BALLROOM C2

BALLROOM C3

A Brief History of the Magnetic Particle Inspection Method G. Hopman, Quality Control Council of the United States

Digital Image Correlation Techniques for Aerospace Applications T. Chu, A. Poudel, Southern Illinois University

Laser Ultrasound: Inspecting Next Generation CFRP Structures M. Osterkamp, K. Yawn, D. Kaiser, PaR Systems

Verifying Magnetic Field Strength for Magnetic Particle Examination of Some Complex Parts J. Brunk, Consultant

UT Camera Imaging: An Alterna-tive to Phased Array for Aerospace & Petrochemical Ap-plications B. Lasser, R. Scheib, D. Rich, O. Mallaug, Imperium, Inc.

Outside Diameter Laser Measuring Apparatus H. Prejean, NOV Tuboscope

New Known Defect Standard for the Verification of Penetrant System Performance D. Geis, B. Collins, K. Boden, Magnaflux

Visualizing Internal Health of Frame Turbines Remote 3D Surface Scanning & Analysis Verify & Validate Before it Fails P. Thompson, GE Inspection Technologies

Microwave to Detect Cold Fusion Joints in High Density Polyethylene Pipe R. Stackenborghs, Evisive Inc.; K. Murphy, Exova; B. Gray, Spectrum NDT

Minimization of Impacts on the User’s Health and the Environment by PT and MT Consumables K. Alward, K. Lessmann, Pfinder KG

Aging Aircraft NDT R. Davis, L3 Mission Integration Division

Microwave Inspection of Fiber Reinforced Plastic Products for Absolute Thickness and Remaining Wall R. Woodward, URS Corporation; K. Schmidt, Evisive Inc.

viii

5:00 pm

Chair: D. Moore, Sandia National Laboratories

4:30 pm

Laser Methods/ Microwave

4:00 pm

Aerospace

3:30 pm

MT/PT

Conference Program: Wednesday 29 October 2014 Plenary Address: 8:00–8:45 am Lester Honor Lecture: The Journey of Nondestructive Testing (NDT) Nat Faransso, Office of the Chief Engineer-KBR BALLROOM B

Chemical and Petroleum I

10:30 am

10:00 am

9:30 am

9:00 am

Chairs: R. Nisbet, QualSpec; S. Hoyt, Marathon Petroleum

NDT Engineering Panel Discussion Chairs: R. Shannon, Siemens Energy Inc.; A. Poudel, Southern Illinois University

Reliability Chairs: D. Forsyth, Texas Research International

BALLROOM B

BALLROOM C2

BALLROOM C3

Inspecting Multiwalled Ammonia Converters with AET C. Allevato, Stress Engineering Services

Panel Discussion Panelists: J. Duke, Virginia Polytechnic and State University; G. Georgeson, Boeing Research & Technology; R. Waldrop, U. S. Coast Guard; M. Gehlen, Uniwest; E. Lindgren, U.S. Air Force, Wright Patterson Air Force Base

What’s Missing in NDE Capability Evaluation? C. Annis, Statistical Engineering

Inspection of Subsea Pipelines Using Guided Wave D. Alleyne, Guided Ultrasonics Ltd.

NDE Characterization Metrics for Optimization, Validation and Quality Control J. Aldrin, Computational Tools

3D Reality Capture for Asset Management T. Taylor, Acuren Group Inc.

Using Modeling to Improve Inspection Reliability in Practice M. Warchol, MFAW NDT LLC

What to Do When ASME Code Ultrasonic Exam Requirements Are Not Enough to Find Critical Flaws D. Bajula, Acuren Group Inc.

Give Inspectors a Chance D. Forsyth, Texas Research International

11:00 am-1:00 pm Lunch in the Exhibit Hall

ix

Conference Program: Wednesday 29 October 2014

Digital Radiography I Chair: K. Bruer, Amee Bay LLC

BALLROOM B

BALLROOM C2

BALLROOM C3

Codes and Specifications for Minimizing Corporate Risk Using Proper PMI Practices D. Mears, Analytical Training Consultants

From the Aerospace to the Petroleum Industry: Fourier Transform Infrared Spectroscopy as a New NDE Tool F. Prulliere, J. Seelenbinder, F. Higgins, G. Miller, C. Sasso, J. Fitzpatrick, L. Tang, Agilent Technologies

Film vs. Digital Radiography on Thin Walled Orbital Arc Tube Welds J. Elston, B. Bartha, P. Vona, W. Cheng, PaR Systems

Advancements in Digital Radiography for the Inspection of Welds CR, DR and DICONDE J. Gibson, R. Mills, GE Inspection Technologies

Optical Coherence Tomography Using On-Chip Spectrometers A. Nitkowski, K. Preston, B. Schmidt, A. Hajian, Tornado Spectral Systems

Comparison of Industrial Process Control CT Systems VTomex C and Speed/Scan J.Gomez, GE Inspection Technologies

Review of Magnetostrictive Sensors for Guided Wave Screening of Heat Exchanger Tubing S. Vinogradov, C. Duffer, G. Light, Southwest Research Institute

Pulsed Terahertz Methods for Non-contact Inspection T. Tongue, B. Schulkin, Zomega Terahertz Corp.

TDI (Time Delayed Integration) Technique and Device in NDT of Digital Radiography (DR) L. Yang, C. Want, N. Luu, D. Meng, XScan Imaging Corporation

Developing Effective NDT Training D. Mandina, Mandina’s Inspection Services Inc.

Improvements in Barkhausen Sensor Design with Application to Magnetic Nondestructive Evaluation (2013 Fellowship Award Winners) N. Gaunkar, I. Nlebedim, D. Jiles, Iowa State University

3:00-3:30 pm Refreshment Break

x

2:30 pm

Chairs: D. Ryan, Siemens

2:00 pm

Chairs: R. Nisbet, QualSpec; S. Hoyt, Marathon Petroleum

1:30 pm

Emerging Methods I

1:00 pm

Chemical and Petroleum II

5:00 pm

4:30 pm

4:00 pm

3:30 pm

Conference Program: Wednesday 29 October 2014

Chemical and Petroleum III

Emerging Methods II

Digital Radiography II

Chairs: R. Nisbet, QualSpec; S. Hoyt, Marathon Petroleum BALLROOM B

Chairs: G. Garcia, EVRAZ North America BALLROOM C2

Chair: K. Bruer, Amee Bay LLC

Importance of a Systematic Review Process for Data Analysis of Advanced NDT Data L. Mullins, P. Tremblay, Zetec Inc.

Effect of Varying Inspection Parameters in Crack Depth Measurements Using Potential Drop Method D. Utrata, D. Enyart, Iowa State University

X-Ray Backscatter Tomography for Large Scale Constructions A. Floet, P. Krueger, Fraunhofer IKTSMD

Rapid Screening of Insulated Pipes for Corrosion Under Insulation using Advanced Electromagnetic Technique A. Vajpayee, D. Russell, Russell NDE System Inc.

Ultrasound NDT Camera and Team-Based NDT J. Endrerud, E. Skoglund, Dolphitech

Optimization of Image Processing with Cone Bean CT Software Tools for 2D and 3D Digital Radiography D. Shedlock, M. Hu, D. Nisius, J. Star-Lack, Varian Medical Systems

Inspection of Welds From One Side Only as Found in Taper Neck Flange T. Armitt, Lavendar International NDT Consultants

Evaluating Dual-Mode Pulse Reflectometry Y. Harel, AcousticEye

NDT Personnel Certification Requirements in the 2015 ASME Code L. Mullins, Zetec Inc.

Ultracloud – A New Ultrasonic Nondestructive Testing Technology Z. Chen, Y. Zheng, J. Wu, L. Hong, SIUI

xi

BALLROOM C3

Conference Program: Thursday 30 October 2014 Plenary Address: 8:00–8:45 am - 2014 NDT Industry Salary Survey Michael Serabian, President, PQNDT, Inc. LLROOM B BALLROOM B Chairs: D. Moore and C. Nelson, Sandia National Laboratories

Chair: F. Spears, Laser Technology

Chair: E. Charpia, Bluegrove NDT Consulting

BALLROOM B

BALLROOM C2

BALLROOM C3

Inservice Inspection for Enhanced RBI of Above Ground Storage Tanks S. Ternowchek, Mistras Group Inc.

Shearography NDT of Aerospace Composites J. Newman, Laser Technology Inc.

Load-Enhanced Methods for Lamb Wave in situ NDE of Complex Components (2012 Fellowship Award Winners) X. Chen, J. Michaels, T. Michaels, Georgia Institute of Technology

The Limitations of Magnetic Flux Leakage Scanning of Aboveground Storage Tank Bottoms D. Carden, Zuuk International

High-Frequency Eddy Current System for Analyzing Wet Conductive Coatings during Processing I. Patsora, H. Heuer, Technical University of Dresden; S. Hillmann, Fraunhofer IKTSMD Dresden; B. Foos, J. Calzada, A. Cooney, Air Force Research Laboratory

Designing an Integrated, Affordable and Versatile Ultrasonic Solution using Advanced Technologies D. Hopkins, BERCLI Phased Array Solutions; M. Brassard, Techno Diffusion NDE

Applications of Industrial Internet to Visual Inspection T. Merry, GE Measurement and Control

Classification of Alkali-Silica Reaction Distress using Acoustic Emission M. Abdelrahman, M. ElBatanouny, P. Ziehl, University of South Carolina

Environment–Assisted Corrosion Cracking in Carbon Steels Utilizing Advanced Ultrasonic Techniques M. Abu Four, M. Rosa, Y. Al Munif, Saudi Aramco

NDT Joins the IoT Revolution! M. Wijtkamp, Librestream

10:00 am

Ultrasonic Inspections

10:30 am

The Research of Corrosion Monitoring on Pipe Based on the Flexible Ultrasound Sensor J. Wang, H. Wang, Nanjing University of Aeronautics

11:00 am - 11:45 am Special Plenary Lecture xii

9:30 am

General NDT II

9:00 am

Power Plant NDT

Conference Program: Thursday 30 October 2014 Special Plenary Address: 11:00-11:45 am The Georgiana: Mystery Ship of the Confederacy Dr. Edward L. Spence, Underwater Archaeologist, Shipwreck Consultant, NOGI Award Winner BALLROOM B

Pipeline Inspections I

Education and Training Issues I

Chairs: R. Stanley, ItRobotics BA

Chair: J. Chen, Schlumberger

Ultrasonic Phased Array Inspections

BALLROOM C2

Chair: D. Alleyne, Guided Ultrasonics BALLROOM C3

2:30 pm

2:00 pm

1:30 pm

1:00 pm

BALLROOM B Inspection Range of Guided Wave Testing of Pipeline S. Kim, H. Kim, Guided Wave Analysis LLC

Introducing Young People to NDT P. Trach, Laboratory Testing Inc.

Design and Fabrication of Curved Ultrasonic Phased Array Probes for Customized Applications S. Hillmann, J. Fischer, J. Michauk, T. Herzog, H. Heuer, Fraunhofer IKTSMD

A Reconfigurable System Design for Pipeline Inspection using Guided Waves Ultrasound L. Zhang, University of Regina

When We Get it Wrong: Why NDT Matters D. Davies, Alcoa Fastening Systems

Further Development of an Optimized Array Wheel Probe for Inspection of Fiber Glass Composites J. Buckley, Sonatest Ltd.

Monitoring of Time-Dependent Degradation in Pipelines with Ultrasonic Guided Waves using Permanently Installed Sensors P. Mudge, TWI Ltd.; P. Jackson, K. Thornicroft, Plant Integrity Ltd.

Lessons Learned from Failed Radiographic Qualifications S. McClain, ARDEC Picatinny

DAC Recording using Phased Array Probes Made Easy W. Kleinert, Y. Oberdoerfer, GE Sensing & Inspection Technologies

Some New Results in Coiled Oilfield Tubing Inspections R. Stanley, NDE Information Consultants

Practical Guidelines for Conversion of Existing Ultrasonic Techniques for Use of Phased Array Technology J. Turcotte, Sonatest Ltd.

AWS D1.1D1.5 Phased Array Examinations P. Furr, University of Ultrasonics

3:00-3:30 pm Refreshment Break

xiii

Conference Program: Thursday 30 October 2014

BALLROOM B

BALLROOM C2

BALLROOM C3

Chemical and Petroleum EMI Inspection Standards W. Averitt, New Tech Systems USA

Condition-Based Maintenance of Reinforced Rubber Cooling System Expansion Joints Utilizing a Microwave Nondestructive Inspection Method R. Woodward, URS Corporation

Small, Flexible and Advanced Phased Array Module for Customizing NDT Applications G. Dao, Advanced OEM Solutions

EMA Technology as a Way to Ensure Quality Ultrasonic Inspection of Coated and Rough Surfaced Objects O. Iurchenko, O. Skorba, Ultraconservice LLC

Embedded Modelling Tools Significantly Improve Educational And Training Classes for Weld Inspection Using Phased Array and Conventional UT Technologies J. Turcotte, Sonatest

A Simple, Efficient and Evolving Solution to Improve the Reliability of Weld Inspection using Phased Array UT & TOFD P. Tremblay, L. Enenkel, L. Mullins, J. Berlanger, Zetec Inc.

Differentiation of 3D Scanners and Their Positioning Method When Applied to Pipeline Integrity J. Lavoie, J. Beaumont, M. Maizonnasse, Creaform 3D

The Art of Technical Writing O. Lewis, West Penn Testing Group

Phased Array Ultrasonic Testing of Closed Rib (U-rib) Welds R. Boundouki, M. Foerder, Alta Vista Solutions

In-situ Nondestructive Positive Material Identification Testing for Determining Carbon Steel Pipeline Material Properties K. Greene, G. Donikowski, TD Williamson Inc.

The C.A.P. System: Clarity, Accuracy, and Punctuality in NDT Reports J. Taylor, Nova Data Testing Inc.

Ultrasonic Measurement of Residual Stresses in Welded Elements and Structures Y. Kudryavtsev, J. Kleiman, Structural Integrity Technologies Inc.

5:00 pm

Chair: A. Al Hassen, University of Alabama Chair: D. Braconnier, Advanced at Birmingham OEM Solutions

4:30 pm

Chairs: K. Greene, TD Williamson Inc.

xiv

Ultrasonic Weld Inspections

4:00 pm

Education and Training Issues II

3:30 pm

Pipeline Inspections II

PAPERS

1

Minimization of Impacts on the User’s Health and the Environment by PTonand Consumables Minimization of Impacts theMT User’s Health and the Environment by PT and MT Consumables Karsten Lessmann and Kersten Alward Pfinder KG, Rudolf-Diesel-Strasse 14, 71032 Böblingen, Germany Karsten Lessmann and Kersten Alward +49-7031-2701-73; fax +49-7031-2701-51; e-mail [email protected] Pfinder KG, Rudolf-Diesel-Strasse 14, 71032 Böblingen, Germany +49-7031-2701-73; fax +49-7031-2701-51; e-mail [email protected]

ABSTRACT

Users of MT and PT-consumables are forced to think about the use of label free and environmentally accepted products. The REACH-legislation is fully in force and the new labelling regulation CLP has to be applied as well and brings an additional pressure to this situation. Leading manufacturers of NDT consumables have already improved their products in order to match these requests without influencing the general performance. While the main focus is still on the potential health hazards of the MT- and PT- materials, another important aspect is the environmental impact of the products. As one of the first companies Pfinder has done a carbon footprint evaluation of widely used NDT consumables in order to optimize their environmental influence. This enables the users of NDT process materials to make decisions not only based on technological and monetary criteria. The minimization of impacts on the user’s health and the environment can now be taken into consideration better than ever before. Keywords: penetrant testing, carbon footprint, sustainability

INTRODUCTION

Beside the general technical applicability the influence on humans and the environment has to be considered as well, when using consumables for penetrant and magnetic particle testing. This holistic view is getting more and more important to ensure that the material can also be used in the future without causing restrictions concerning labor and environmental safety. Adding the process costs, one has considered the essential aspects that lead to a sustainable product use.

RESULTS Definition Sustainability

A common definition of sustainability is the simultaneous review of ecological, economical and social aspects [1]. If these fundamental issues have to be applied to the evaluation of concrete consumables, it is advisable to investigate the influence of the consumable to the costs, the environment and the human. Only when all of these three factors have been evaluated, one can speak of a sustainable approach. So for the methods the following conclusions can be drawn: Ecology – Environmental impact Economy – Costs Social aspects – Impact on the staff

Individual Aspects of Sustainable Product Approach for Penetrant Consumables Within the higher-ranking effects there are now some individual aspects which are of interest for the user that can be considered.

2

Environmental impacts:     

Solvent-free, VOC-free Readily biodegradable Low GWP (global warming potential) / Carbon Footprint Raw materials based on renewable/recyclable sources Process free of hazardous waste

Costs:     

Low investment costs Low material consumption Low process monitoring efforts Low energy costs Low service-/ maintenance costs

Impacts on the staff:     

Easy interpretability of indications Labeling free Not flammable No allergic skin reaction No health hazard (Inhalation/Ingestion)

These individual aspects can now be rated from 1 = unincisive to 10 = definitive. The result of this is a first matrix to rate processes and to achieve improvements or to manage future developments. In particular aspects, there is also the opportunity to accept disadvantages consciously, if this is beneficial for other individual factors.

Results of the Single Evaluation

As an exemplary process evaluation the effect of the penetrant PFINDER 800 was evaluated. PFINDER 800 is a red and fluorescent penetrant (sensitivity class 2) which is free of mineral oil, directly water washable and readily biodegradable. It shows low background fluorescence because of the good removability within the penetrant removal step. This product is available in barrels/cans and also as aerosol with propane /butane as propellant. The result of comparing these two systems considering the above mentioned aspects is as follows (Table 1).

3

Table 1: Rating of PFINDER 800 according to its sustainability: 10 = definitive, 1 = unincisive GREEN: YELLOW: BLUE:

Environmental impacts Costs Impacts on the staff

Solvent-free, VOC-free

Pfinder 800 Aerosol

5 7

Readily biodegradable

4

Low Carbon Footprint Raw materials based on renewable/recyclable sources Low investment costs

5

10

Low energy cost

10

Low service-/maintenance cost

2

Labeling free

1

Not flammable

6

No allergic skin reaction

3

No health hazard (Inhalation/Ingestion)

9

Easy interpretability of indications

8

10 9

Low process monitoring efforts

10

2

2

Low material consumption

10

2 3

Process free of hazardous waste

Pfinder 800 Liquid

8 9 7 8 8 6 9 6 8 9

Illustrating this table as a graphic, it shows clearly the strengths and weaknesses regarding the specific criteria (Image 1) The full line shows the evaluation of the PFINDER 800 as aerosol whereas the dotted line the PFINDER 800 as a liquid in a can. This shows clearly that the aerosol spray can has its advantages predominantly on the side of investment, energy and service costs or inspection effort. This is because of the easy handling of the aerosol spray can for the application of the product. On the other hand the aerosol spray can has significant disadvantages concerning the environmental impacts, such as waste, and it is not free of VOCs (by the propellant gases). Additionally the fire hazard as well as the danger of inhalation increases due to the propellant gases and the spraying. The purchasing costs of the penetrant in aerosol cans are significantly higher relatively to the surface to be tested with aerosol cans. In all of these aspects the bulk material of PFINDER 800 has significant advantages. By this one can clearly realize, that fundamental differences are not caused by the product itself but by the way of application.

4

Full Line: Pfinder 800 bulk Dotted Line: Pfinder 800 Aerosol Image 1: Evaluation of the particular sustainability aspects.

Evaluation of the Carbon Footprints

The meaning of the application form appears very well when evaluating the carbon footprints. This figure considers the production of carbon dioxide during the life cycle of a product from the raw materials to the production, the logistic expenses and the use until disposal. With this the carbon footprint indirectly provides information about the consumption of fossil fuels. The carbon dioxide (CO2) has a well-known high importance to the global warming and is therefore classified as one of the most critical gases in our atmosphere. Thus, the carbon footprint and therefore the GWP (global warming potential) needs a separate evaluation. If you want to determine the carbon footprint, you should first define the system for which you want to create a balance. In this case the product was balanced from the mining of the raw materials to the factory’s gate for balancing. The use and application of the products may be strongly varying from customer to customer, so that an evaluation for the determination of the carbon footprint is not possible. A separate evaluation would be necessary for each customer. A scheme of these two possible balancings is shown in image 2.

5

Chosen balance Image 2: Overview of the possible balancings [2]. The first approach is called: "From the cradle to the grave", the second of "From the cradle to the gate." For the system PFINDER 800 we have chosen the balancing to the gate (second balance). As part of a master's thesis, the individual steps of the origin of PFINDER 800 (as well as the pre-cleaner, the penetrant remover and the developer as parts of the product family), including raw materials, packaging, active agents were balanced and their carbon footprint was defined. For the essential process and production steps, there are already a lot of data in databases and related software for evaluation. For evaluation the four aspects of production were evaluated separately: 1. Supply of raw materials of the active agent 2. Supply of raw materials for packaging 3. Manufacturing of the products 4. Transportation and logistics Two types of packaging have been analyzed: 5 Liter PE canister PFINDER 800 500 ml aerosol spray can PFINDER 800 The results of the analysis are shown in image 3 and 4 [3]:

6

Image 3: % Carbon footprint PFINDER 800 5 Liter PE canister.

Image 4: % Carbon footprint PFINDER 800 aerosol spray can. It shows that for aerosol spray can the highest percentage (52.3%) of the carbon footprint is related to the supply of the raw material of the packaging. The tin cans play the most important part concerning raw material extraction and manufacturing. With 30.2% of the raw material for the active agent/propellant and 16 % for the manufacturing the percentages are comparatively low. Together with its low percentage for transport the single aerosol can has a Carbon Footprint of 3.08 kg. The impact of waste and application at the customer’s side has not been evaluated due to the evaluation approach up to the factory’s gate.. Comparing these results with the 5 Liter PE canister, as expected the main percentage of the carbon footprint lies in the supply of raw material for the active agent (78%). The percentage of the packaging is now only 3%. The total carbon footprint for 5 Liter PFINDER 800 in a PE canister is about 20.45 kg. When relating these figures to the quantity of the active agent the result is the active-agent-related carbon footprint (Table 2)

7

Table 2: Summary of the Carbon Footprints of PFINDER 800 [3]. Packaging PFINDER 800, 500ml Aerosol PFINDER 800, 5 Liter canister

Carbon Footprint/ Packaging 3,08 kg 20,45 kg

Carbon Footprint/ kg (active agent) 12,0 kg 4,2 kg

Comparing the active-agent-related carbon footprint, there is a significant advantage for the packaging in canisters compared to the aerosol spray cans. This carbon footprint consideration shows clearly, which consumption of resources the aerosol spray cans mean. This shows the obvious disadvantages which come with the comfortable use of aerosol spray cans. For improvement of the carbon dioxide emissions users can switch to canisters or invest into compensation payments in case a different packaging is not be possible due to technical reasons. Based on these results PFINDER has decided to minimize the carbon footprint of the PFINDER NDT products filled in aerosol spray cans by supporting compensation efforts. In this way all users of penetrant and magnetic particle products who are not able to use other packaging than aerosol spray cans have the opportunity to make a conscious decision for a significant lower impact to the environment.

SUMMARY

For a safe use of consumables for penetrant or magnetic particle testing also in the future, the sustainability has to be considered as well beside the pure technical performance. Particular attention should be laid to the impact on the employees and the environment. To show rooms for improvement of sustainability a rating system has been developed by PFINDER. PFINDER will continue to work consistently at its improvements of product development. The aspect of the carbon footprint shows easily the impact of penetrants to carbon dioxide emissions. It is obvious that especially using aerosol spray cans the influence of the active agent (the actual penetrant) is low. Improvement is possible by changing to bulk ware or by payment for compensation efforts. Pfinder supports all users with a strong focus on worker’s safety health and safety and on environment protection with numerous activities.

REFERENCES 1. 2. 3.

Final report of the Enquete-Commission „Protection of the human and the environment -- Aims and frame conditions of a sustainable future compatible development“, German Bundestag: printing 13/11200 of 26th June 1998 British Standards Institution (Hg.) (2008a): Guide to PAS 2050. How to assess the carbon footprint of goods and services. London T. Karsunke, Masterthesis, master’s programme ecology, University of Economics and Environment, Nürtingen-Geislingen, 2014

8

Acoustic Emission for Damage Evaluation in Realistic CFRP Components Acoustic Emission for Damage Evaluation in Realistic CFRP Components 1 2 Russell Austin , Mohamed K. ElBatanouny , Marwa Abdelrahman, and Paul H. Ziehl 1 Texas Research 2 Institute, Russell Austin1, Mohamed K. ElBatanouny , Marwa Abdelrahman and Paul H. Ziehl 9225 Bee 1 Cave Road, Austin, TX 78733 Texas Research Institute, (512) 263-2101; fax (512) 263-4085; e-mail [email protected] 9225 Bee Cave Road, Austin, TX 78733 (512) 263-2101; fax (512) 263-4085; e-mail [email protected] 2 Dept. of Civil and Environ. Engr., University of South Carolina St., Columbia, SC 29208 2 Dept. of Civil300 andMain Environ. Engr., University of South Carolina (803) 348-5182; 300 fax (803) 777-0670; e-mail Main St., Columbia, [email protected] 29208 (803) 348-5182; fax (803) 777-0670; e-mail [email protected]

ABSTRACT

Fiber reinforced polymers (FRP) materials are widely used in many industries including aerospace, automotive, marine, and construction. Numerous nondestructive evaluation (NDE) techniques have been used to assess the condition of FRP coupons. However, the complex geometries and irregular shapes of FRP components in actual applications limit the applicability of NDE for in-service monitoring. In this paper, acoustic emission (AE) monitoring was implemented for damage evaluation in a realistic carbon fiber reinforced polymers (CFRP) specimen. The specimen had irregular shape and thickness. Stepwise loading was applied to the specimen and an AE based numerical condition assessment criteria for real-time monitoring was used to evaluate the damage. Ballistic impact test was also conducted to assess the ability of AE in locating the damage. AE data filters were developed to reject signals from noise and wave reflections. The methodology used to detect, evaluate and locate damage using AE is reported. Preliminary results for CFRP damage classification (matrix cracking, fiber breakage, and delamination) based on AE data are also presented. Keywords: graphite epoxy composites, nondestructive evaluation, acoustic emission, structural health monitoring, ballistic damage, flaw locating, flaw typing

INTRODUCTION

Fiber reinforced polymers (FRP) materials are widely used in many industries including aerospace, automotive, marine, and construction. The material is manufactured through merging fibers, such as glass or carbon, with a polymer matrix, such as epoxy and vinylester. The low weight-to-strength ratio of FRPs along with its long life expectance and corrosion resistance makes it a suitable substitute for metals, such as steel and aluminum. FRP materials are used in critical parts and applications in many industries; therefore, it is important to develop a reliable non-destructive testing (NDT) method that can assess the structural integrity of such members in real-time. During the last two decades, numerous studies have been conducted to develop NDT methods for FRP materials [13]. Among these methods is acoustic emission which is defined as transient stress waves (in the kHz range) emitted as a result of sudden energy release, such as crack growth or friction [4]. AE has been previously used to detect damage in FRP tanks and vessels [5]. This damage quantification method is included in an ASTM standard [6]. Recently, Austin and Coughlin 2006 [7] proposed a numerical rating to evaluate damage in structural members. The method is based on the results of four criteria including number of high amplitude hits/events, historic index, cumulative energy, and severity. The criteria and scoring system are used and discussed in this study. This study examines the ability of AE to detect damage in a realistic FRP specimen. The specimen was first tested under stepwise increasing displacements. After completion of the load test, the specimen was shot three times at three different locations to assess the ability of AE to detect hit locations. Finally, the load test was repeated using similar displacement cycles as those used in the first load test. The results show that AE is able to assess and quantify damage in the specimen. Additionally, AE enabled the detection of the locations were the bullets penetrated the specimen during the shooting test.

9

EXPERIMENTAL PROGRAM Load Test

An FRP specimen with thickness ranging from 0.25 to 0.50 inch was tested. The specimen had an irregular shape with geometry as shown in Figure 1. The specimen was tested under 3-point bending using a hydraulic ram. The load was applied at the centerline of the specimen with the magnitude measured using a load cell. The displacement at the centerline was measured using a string potentiometer. Cyclic displacements were applied with magnitudes as follows: 0.50, 0.75, 1.00, 1.25, and 1.45 in. The load and deflection results for the load test are shown in Figure 2.

Void Void

Void

Void

Figure 1: Specimen geometry and dimensions.

Figure 2: Load and deflection versus time for the first load test.

Figure 3: Test setup and AE sensor layout.

10

A total of 16 AE sensors manufactured by Mistras Group, Inc. were used to collect AE signals. The sensor array included twelve AE sensors with 150 kHz resonant frequency and 40 dB integral pre-amplification (R15i) and four 300 kHz resonant sensors (Micro 30) connected to four 40 dB pre-amplifiers with frequency band pass of 100 kHz – 1200 kHz. Sensors were attached to the specimen using a two part epoxy. AE data was collected and displayed using a 16-channel data acquisition system (Sensor Highway II, Mistras Group, Inc.). Based on experience with similar materials and specimens, the collection threshold in all channels was set to 50 dB. The test setup and sensor layout are shown in Figure 3. All channels were tested using pencil lead breaks (PLB) prior to the application of load as recommended by ASTM E1316 [4]. The load test was repeated using the same sensor layout after the shooting test. The five load cycles completed in the first test were repeated by applying the same load and four additional cycles were performed. The measured deflections in cycles 6 through 9 were 1.85, 2.10, 2.35, and 3.10 in., respectively. The load and deflection results for the repeat load test are shown in Figure 4.

Figure 4: Load and deflection versus time for the repeat load test.

Figure 5: AE sensors layout for shooting test with hit locations in white.

Shooting Test

A shooting test was conducted to examine the ability of AE sensors to detect the impact and locate it using source triangulation. This technique has been used to locate cracks successfully in previous research [8,9]. A wood frame was built to hold the specimen in place at a distance of 30 ft. The specimen was shot at three predetermined locations using a rifle with a Winchester bullet of diameter 0.270 in., 130 grain, and muzzle velocity of 3,060 ft/s. A total of nine AE sensors were used; 7 R15i sensors and 2 Micro 30 sensors. To perform source location, AE wave

11

speed in the material was determined prior to the test to be 200,140 in/s. Figure 5 shows the sensor layout for the shooting test with the hit locations in white.

RESULTS AE Data Filtering

Three data filters were applied to eliminate spurious data. The first filter accepted signals with peak frequency and frequency centroid between 20 and 400 kHz. The second filter removed signals with rise time exceeding 300 µseconds. The third filter accepted data with duration less than 10,000 µ-seconds. These filters were based on analysis of waveforms collected during testing.

Load Test

The load-deflection response of the specimen during the first and second load test is shown in Figure 6. No visual damage was seen in the specimen after the first load test while two cracks developed during the second load test as shown later in the paper. As seen in Figure 6, the stiffness of the specimen decreased during the second load test. The maximum load achieved during the first load test was 805 lb. while a maximum load of 945 lb. was achieved during the second load test. In the second load test, the deflection increased during the additional cycles with drops in the load as seen in Figure 6. This is attributed to breaking of fibers and cracking which was visually observed during the additional load cycles (cycles 6 through 9).

Figure 6: Load-deflection response during load tests. AE data was collected continuously during the load tests. The deflection and AE plots of amplitude versus time are shown in Figures 7(a) and 7(b) for the first and second load tests, respectively. From Figure 7(a) it can be seen that the number of AE hits increases with the increase of the applied load/deflection. Also, higher amplitude hits (> 75 dB) are detected with the increase of deflection/load. A considerable increase in the number of AE hits was detected in the third cycle with deflections exceeding 0.75 in. During this cycle a popping sound was heard and continued to occur during the following cycles. This may be attributed to matrix cracking and fiber debonding which is considered the first step in the failure of FRP materials. During the second load test, a smaller number of high amplitude hits (>75 dB) was detected in the first four cycles as compared to the first load test. At cycle 5, the maximum load measured during the first load test was achieved at a higher deflection value. Therefore, high amplitude hits (>75 dB) were detected during this cycle as new damage occurred. A crack initiated in cycle 6 which led to high AE activity. The crack continued to propagate during the remaining three cycles where high AE activity was recorded. In general, it was observed that the number of hits increased with the increase of the load/deflection.

12

The presence of a significant number of high amplitude hits during the fifth cycle of the second load test shows that the Felicity Effect exists, the presence of detectable acoustic emission at a fixed predetermined sensitivity level at stress levels below those previously applied, which indicates that significant damage had occurred during the first load test [4,10].

(a) (b) Figure 7: Deflection and AE signals amplitude versus time: (a) first load test, and (b) second load test.

Damage Evaluation Using AE

A numerical rating method was used to quantify damage in the specimen. This method was proposed by Austin and Coughlin 2006 [7] and uses four criteria including number of hits, historic index, cumulative energy, and severity [11]. Detailed explanation of the method can be found in Austin and Coughlin 2006. To assess the condition of the monitored structure, the scores of the four criteria are summed and then divided by 5. The final score, with a maximum of 4 for the most severely damaged, is used to determine the condition of the structure as shown in Table 1. This numerical rating enables less acquainted users to characterize the damage in the monitored structure and decreases the subjectivity of the results [7,11]. Final Score 3.7

Table 1: Condition assessment limits [7]. Assessed Condition Comments Insignificant Emission source is structurally insignificant Minor Minor Emission. Note for future reference – visually inspect accessible areas Significant Significant structural defect requiring follow-up evaluation. As a minimum, evaluation should include further data analysis and visual inspection of accessible areas Major Major structural defect. Immediate shutdown and additional nondestructive examination

AE data collected in each cycle in both load tests was evaluated based on the described criteria. The results are shown in Table 2. For the first test, the results showed that insignificant damage was detected in cycle 1 while major damage was detected in the last cycle (cycle 5). This shows that significant damage had taken place in the specimen during the load test. In the second load test, major damage was detected in cycle 5 where the maximum load reached in the first load test was achieved. This is attributed to occurrence of new damage as the deflection during this cycle was higher than that achieved at the same loading level in the first load test. The specimen continued to score major damage through the remaining cycles (cycles 6 through 9) as cracking occurred and propagated. It is noted that major damage was first detected by the damage evaluation method at 85% of the maximum applied load, where visual cracking occurred. This shows the ability of the criteria to give early warning prior to visual observed damage.

13

Table 2: Damage evaluation results. Cycle number First test Second test 1 2 3 4 5 6 7 8 9

Insignificant Significant Significant Significant Major N/A N/A N/A N/A

Significant Minor Insignificant Insignificant Major Major Major Major Major

Damage Classification

The damage failure mechanisms in FRP mainly include fiber breakage, matrix cracking, and delamination [1,2]. In general, matrix cracking (MC) and delamination (DL) occurs prior to fiber breakage (FB). Different studies investigated the ability of AE to detect different failure mechanisms in FRP specimens. This investigation uses the AE data collected during the reload test to classify damage in the specimen using signal strength and duration. Previous studies reported that the source-to-sensor distance and attenuation affects the damage classification parameters. Therefore, only data collected from the closest 300 kHz resonant sensor is used. Figure 8 shows the signal strength versus duration collected during the reload test. During the reload test, new stress levels were applied starting from cycle 5. Therefore, AE data collected in cycles 1 through 4 are mainly related to matrix cracking since fiber breaking will only occur with new stress levels. In Figure 8, the data is divided into two sets related to the anticipated damage mechanism where only MC should occur in cycles 1 through 4 while all three mechanisms will exist in cycles 5 through 9.

1.00E+08

1.00E+08

8.00E+07

8.00E+07

Signal Strength, pVs

Signal Strength, pVs

As seen in Figure 8, AE signals related to MC have a low signal strength and low duration while signals related to DL and FB have high signal strength. The literature indicated that FB signals should have smaller duration than DL signals. Based on this knowledge, the signal strength-duration chart may be used to classify cracks as shown in Figure 8.

6.00E+07

4.00E+07 2.00E+07 0.00E+00

0

6.00E+07

4.00E+07

FB

2.00E+07 0.00E+00

5000 10000 Duration, micro-s

DL

0 MC 5000 10000 Duration, micro-s

(a) (b) Figure 8: Signal strength and duration versus time; (a) cycles 1-4, and (b) cycles 5-9.

Load test-Source Location

To detect the location of cracks formed during the test, the AE wave speed for the FRP material used was determined experimentally as 200,140 in/s. Since the wave speed is very high proper filters must be used to eliminate wave reflections and emissions not related to the test; thus, minimize error in source location. DurationAmplitude filters (Swansong II) [8,9] were employed for this purpose as shown in Table 3. The filters were

14

determined using manual inspection of recorded AE waveforms. The results of source location from AE sensors and visual observed cracks are shown in Figure 9. AE was able to locate the general vicinity at which the cracks developed. The AE events detected around the crack may be attributed to fiber breakage and delamination occurring within the specimen and is not visual yet. Another possibility for scattering in the AE events is the irregular shape of the specimen as well as the presence of voids which may have an effect on the detected AE events. Table 3: Rejection limits for data filters. Amplitude (dB) Duration (µs) 1,000 60-67 >2,000 Duration-Amplitude 68-75 >4,000 76-83 >6,000 84-91 > 8,000 92-100 >10,000 Filter Type Amplitude

Figure 9: Photograph showing the formed crack (in black) and Ae source location (in red).

(a)

(b) (c) (d) Figure 10: Photographs showing the shooting test: (a) AE source location results, and (b-d) bullet penetration locations.

15

Shooting Test-Source Location

The specimen was shot three times at three different locations. At each hit, a 100 dB AE signal was detected followed by wave reflections. AE was able to locate the points at which the bullets were hit with reasonable accuracy as shown in Figure 10. The close-ups for the bullet hit locations [Figures 10(b), 10(c), and 10(d)] shows that the three bullets were able to penetrate the specimen. The bullet at the hat stiffener was able to penetrate the specimen at the stiffener as well as the specimen surface.

CONCLUSIONS 1. 2. 3. 4.

AE has the ability to detect damage in FRP materials. The condition of the member can be evaluated using available damage evaluation techniques. The used AE damage evaluation method has the ability to give warning of impending failure prior to the occurrence of visual damage. This was achieved at 85% of the maximum applied load. Classification of damage in FRP specimen can be achieved using AE monitoring. AE source triangulation can effectively determine the location of damage during load tests and due to ballistic impact.

REFERENCES 1.

Carey, S., “Damage Detection and Characterization in CFRP Composites Using Acoustic Emission and Acousto-Ultrasonics”, PhD Dissertation, University of South Carolina, 2008. 2. Harvey, D. W., “Acoustic Emission in a Aerospace Composite”, Thesis, University of Texas at Austin, 2001. 3. Ziehl, P., Fowler, T., “Fiber Reinforced Polymer Vessel Design with a Damage Approach”, Journal of Composite Structures, 61(4), 395-411, 2003. 4. ASTM E1316, “Standard Terminology for Nondestructive Examinations”, American Standard for Testing and Materials, 1-33, 2006. 5. Fowler, T.J., Blessing, J., Conlisk, P., “New Directions in Testing”, Proc. 3rd International Symposium on AE from Composite Materials, Paris, France, 1989. 6. ASTM E1067/E1067M-11, “Standard Practice for Acoustic Emission Examination of Fiberglass Reinforced Plastic Resin (FRP) Tanks/Vessels”, American Standard for Testing and Materials, 1-15, 2011. 7. Austin, R.K., and Coughlin, C., “Disturbed Mode System for Real Time Acoustic Emission Monitoring”, U.S. Patent No. US 7,080,555 B2, July 25, 2006. 8. ElBatanouny, M.K, Larosche, A., Mazzoleni, P., Ziehl, P.H, Matta, F., and Zappa, E., “Identification of Cracking Mechanisms in Scaled FRP Reinforced Concrete Beams using Acoustic Emission”, Experimental Mechanics, V. 54, Issue 1, 69-82, 2014. 9. Abdelrahman, M., ElBatanouny, M.K., and Ziehl, P., “Acoustic Emission Based Damage Assessment Method for Prestressed Concrete Structures: Modified Index of Damage”, Engineering Structures, V. 60, 258–264, 2014. 10. Pollock, A. A., “Inspecting Bridges with Acoustic Emission—Inspection Details about In-Service Steel Bridges and Monitoring Weld Operations: Application Guidelines”, Prepared by: Physical Acoustics Corporation, 1995. 11. Austin, R., Forsyth, D., Yu, J., ElBatanouny, M., and Ziehl, P., “Damage Evaluation for High Temperature CFRP Components Using Acoustic Emission Monitoring”, 40th Annual Review of Progress in Quantitative Nondestructive Evaluation (QNDE Conference), Baltimore, Maryland, AIP Publishing, V. 1581, No. 1, 501505, 2014.

16

Application of Advanced Non-Contact Ultrasound for Composite Material Qualification Application of Advanced Non-Contact Ultrasound for Composite Material

Qualification

Anuj Bhardwaj, Kashyap Patel, Mahesh C. Bhardwaj1, and Konstantine A. Fetfatsidis2 1 The Ultran Group 1 , and A. Fetfatsidis2 Anuj Bhardwaj, Kashyap Mahesh C. Bhardwaj 2380 Patel, Commercial Blvd, State College, PAKonstantine 16801 1 The Ultran Group (814) 861-2001 2380 Commercial Blvd, State College, PA 16801 2 (814)Flight 861-2001 Aurora Sciences 4 Cambridge 2Center, Suite 11, Cambridge MA 02142 Aurora Flight Sciences (617) 229-6818 4 Cambridge Center, Suite 11, Cambridge MA 02142 (617) 229-6818

INTRODUCTION Ultrasound Testing (UT) is a popular and effective method of non-destructive materials analysis. However, conventional UT has certain limitations, primarily the need for coupling liquid or direct contact with the test material. Such limitations have prompted the desire for air-coupled or Non-Contact Ultrasound testing (NCU), which can be directly applied to high volume inspection, supporting manufacturing quality control (QC) processes. For a long period though, NCU has proved challenging due to inefficient coupling between ultrasonic transducers and air. Inefficient air coupling prohibits ultrasonic signals from propagating through air and into materials, thus making meaningful analysis a near impossibility. However, over recent years, improvements to air coupling in ultrasound have opened up the ability to propagate and detect ultrasonic signals comparable to those achieved in immersion ultrasound. Further improvements have even increased the frequencies of efficient air transmission to include a wide range from 30 kHz to 5 MHz. Efficient NCU in this frequency range provides applicability for a wide variety of composite materials. We will review actual composite quality measurements resulting from improvements to NCU and discuss the applicability to production QC within composite manufacturing.

Non-Contact Ultrasonic Advancements As mentioned above, core product improvements to ultrasonic transducers have opened the door for efficient air coupling. One of the key improvements is proper matching between piezoelectric material (the heart of an ultrasonic transducer) and air.

Figure 1: Ultrasonic transducer composition. A schematic of an ultrasonic transducer is depicted in Figure 1, above. From this image, we can see that a piezoelectric transducer can be composed of the following elements: Piezoelectric material, transition layers, matching layers, and sometimes damping and electrical matching. The piezoelectric material is the primary element, which makes the key conversion from electrical to mechanical energy, and vice versa. However, it is the transition and matching to the test medium (i.e. air) which is of key importance. In this section of the transducer, we are able to affect the efficiency of acoustic propagation into the desired medium. These advancements have opened the door to possibilities which we will now discuss.

17

Composite Material Analysis Composite manufacturing volumes have reached new heights and quality control is ever more important. As composites are expensive and difficult to reproduce, non-destructive testing (NDT) is preferred over other destructive methods. Ultrasound has often been a favored method of NDT as it is safe and reliable. However, conventional UT is often not applicable to high volume manufacturing QC as immersion or contact is required. The advancement of NCU can now allow us to inspect high volumes, even 100%, of manufactured parts and material. Composite materials which can be analyzed using NCU include the following: Prepreg, honeycomb composites (Nomex and aluminum core), GFRP, CFRP, and C-C composites. Defects and non-uniformities include porosity and density variation, delamination, foreign object detection, cracks, holes, and more.

EXPERIMENTATION Material Transmittance The transmittance of a material can be measured by relating the strength of its main transmission signal relative to that through air alone as depicted in Figure 2, below.

Figure 2: Representation of transmittance Where Ta and Tc are measured in dB, the transmittance of material, m, can be represented as follows: Tm= Tc-Ta. The transmittance of the material is a representation of how easily it allows ultrasound to pass through it. As our experimental results will show, this can be used to determine whether or not a defect, such as delamination, is present, as well as porosity in certain instances.

Test Conditions In addition to domains of measurement, we also have the option to analyze materials under different base conditions, including transducer frequency, size, and focus. With advancements, as previously described, we can analyze materials at frequencies as low as 30 kHz and as high as 5 MHz in non-contact mode. The sizes of such transducers can vary from as small as 1 or 2 mm active diameter to as large as 100 mm active diameter or more (frequency dependent). Certain materials and applications are well suited for higher frequencies, these would include high resolution (detectability of small material property variances) requirements. However, such materials may need to be relatively thin, especially at very high NCU frequencies, such as 3-5 MHz and receptive to high frequency ultrasound. Highly attenuative and thick materials, such as concrete and certain composites, are often best analyzed at lower frequencies, such as 50 kHz to 500 kHz. In the following section we will observe results upon composite materials measured under various test condutions and across different domains of analysis.

RESULTS While non-contact ultrasound has only recently begun practice across various industries in manufacturing conditions, a large amount of data have been collected through analysis of composites and other materials. Such tests have been conducted across the frequency range mentioned above in both the velocity and attenuation domains. In the following sections we will explore variations in material properties and correlations developed against ultrasonic units of measurement. Such results will allow us to draw conclusions upon which conditions are ideal for certain applications and materials, enabling for detection during manufacturing processes for quality control.

18

Direct Transmission in NCU In order to demonstrate the variation in material detectability under different frequencies, we can turn our attention to a test material of PMMA (acrylic) which was analyzed at the frequencies of 200 kHz, 500 kHz, 1 MHz and 2 MHz in through transmission NCU.

Figure 3: Detectability as a function of frequency. As observed in Figure 3, above, the detectability of drilled holes in acrylic material vary significantly with the frequency measured. We can visually see that the results at 500 kHz exhibit the ideal condition for detection of the induced defects within this material. When encountering new applications for NCU, such trials must be conducted in order to experimentally confirm the best fequency for analysis.

NCU Transducers Acoustic Profile The images in Figure 4 below exhibit characteristics of high efficiency NCU transducers in ambient air at select frequencies. These data were collected in direct transmission mode and the transducer specifciations and distance are listed under each plot. The blue trace represents the A-scan in time while the red trace is the frequency profile. Transducer 50 kHz at 50 mm diameter 100 mm air; S = -40 dB

Transducer 500 kHz at 19 mm diameter 20 mm air; S = -43 dB

Transducer 4 MHz at 6.3 mm diameter 3 mm air; S = -76 dB

Figure 4: Characteristics of transducers in air at select frequencies (50 kHz 500 kHz, and 4 MHz). The data in Figure 4 demonstrate that ultrasonic signals in air can be produced efficiently at a large range of frequencies. Perhaps especially impressive is the propogation of NCU in air at 4 MHz as shown in the final trace of the above figure. Similar results are achieved at 5 MHz using both planar and focused transducers.

Materials Analysis After determining ideal frequency and transducer conditions for certain composites, we can perform X-Y scans to create 2-dimensional images (C-scans) on various materials.

19

Figure 5: Composite materials analyzed using NCU. Figure 5 represents materials successfully analyzed using NCU. These materials include CFRP, GFRP, prepregs, CC composites, foam core sandwich composites, honeycomb core composites, wood composites, and more. We will now explore results from analysis upon select materials.

Honeycomb and Foam Core Panels

The below honeycomb panel image shown in Figure 6 below depict core damage within the composite structure. The damaged area is represented by the blue circular regions. The resolution of this image is high, as demonstrated by visibility of individual honeycomb cells.

Figure 6: High resolution imaging of CFRP skin Al Honeycomb Core (25mm thick) with fine embedded defects and broken honeycombs imaged with focused 200 kHz transducers (left). Honeycomb panel analyzed at 200 kHz with planar transducers (center) and point focused transducers (right). Using high resolution ultrasound with small spot size transducers (Figure 6 - left), we can resolve detail including individual honeycomb cells and small defects. The center and right images in Figure 6 demonstrate the effect of using focused and unfocused transducers upon the same material sample. The sample analyzed is a 25 mm thick honeycomb panel with delamination defects of varying size and a recessed core on the right side. The image to the left represents analysis conducted with planar transducers at 200 kHz, while the image to the right was created using focused transducers, also at 200 kHz. While both analyses can detect all embedded defects, the visual difference between the two samples is stark. The focused analysis resolves variation in the material to a much finer spatial resolution. Depending upon the desired information, either analysis could be useful. For example, if the variation within individual cells is an undesirable artifact of the analysis, a planar set of transducers may be preferred. However, if visual representation of individual cells is preferred then focused analysis can be advantageous.

C-C Composites

Carbon-Carbon composites are traditionally difficult to to analyze under various conditions, including ultrasound, due to high attenuation. However, with efficient through transmission ultrasound at relatiely low frequency (between 100 kHz and 500 kHz), substantial progress has been achieved. In Figure 7, below, the representative images show extreme delamination within the central portion of a C-C aircraft disk brake (left). The blue and dark blue areas represent low signal amplitude in the attenuation domain, which is attributed to poor bonding between layers.

20

Carbon-Carbon Plates for Oven Fixtures

Major Delamination

Non-Uniform Bond Quality

Well Bonded Layers

Figure 7: C-scan images of C-C disk brakes (left) and oven fixtures (right).

Figure 7 also depicts C-scan images of C-C composites used as high temperature oven fixtures (right). Using NCU, we can identify relevant defects in the C-C plates. In the above image, major delamination was recorded in the leftmost part, noticeable in the dark blue regions. In addition, uneven bonding was observed in the center unit as compared to a defect free sample on the right.

CFRP (Carbon Fiber Reinforced Plastic) Composites

Bonding between layers within CFRP panels is a key quality metric. NCU is largely receptive to bond quality, as poor bonding highly attenuates signal amplitude.

Figure 8: CFRP composite with embedded defects (4mm thick) with line scan to the right. As seen in Figure 8, various embedded defects within CFRP are easily detected using NCU. While the upper image represents the C-scan, the lower plots are line scans through selected sections of the material. The upper and lower line scans pass through the defective portions while the middle represents the more uniform central section of the material.

Figure 9: CFRP composites with varying porosity (4mm thick) with line scan to the right.

NCU is also highly sensitive to porosity variation. In Figure 9 we observe a notable difference between samples of higher porosity (left) and lower porosity (right). Increased porosity tends to attenuate ultrasound more heavily.

21

Figure 10: High resolution C-scan images of CFRP lamiantes with embedded defects (6mm thick). High resolution NCU allows for detection of small embedded defects within materials, such as CFRP lamiantes. Figure 10 demonstrates the detection of defects smaller than 1mm in thickness. These results were obtained using focused ultrasound at 500 kHz.

Carbon Fiber Prepreg

A key ingredient material for aerostructures, the quality of carbon fiber prepreg is of great importance. Traditional ultrasonic detection methods, such as contact and immersion, are not applicable due to the adverse effect upon uncured prepreg from direct contact or water exposure. For this reason, NCU is a desirable method for quality inspection.

Figure 11: C-scan images of aerospace prepreg material with detects and varying resin content. Figure 11 represents images of aerospace prepreg material. The left image contains overt defects (red spots) and lower porosity (blue regions). The right image does not contain defects and is of relatively normal porosity. To validate the correlation between ultrasonic amplitude in carbon fiber prepreg and material impregnation level, an experiment involving a secondary test method was used. Two samples of expected variation in porosity were measured using NCU. Results from these measurements are seen below in Figure 12 below. Sample A exhibited higher ultrasonic apmplitude than Sample B with readings ranging from 0 to -10 dB and -20 to -25 dB, respectively.

Figure 12: NCU scans of sections of unidirectional prepreg with expected differences in impregnation level. Following the measurements in NCU, the samples were evaluated under a secondary, destructive, method of water pickup. During the water pickup test, the samples are immersed in water for a period of time under controlled conditions. The samples are weighed prior to water exposure and immediately after. It is expected that any increase in weight is due to porosity or inverse impregnation within the prepreg material. This is caused by “soaking” of water within material that is not fully impregnated.

22

Table 1: Water pick-up test results of unidirectional prepreg specimens. Sample A B

NCU Amplitude High (0 dB to -10 dB) Low (-20 dB to -25 dB)

W1 (g) 2.977 3.022

W2 (g) 3.086 3.558

WPU (%) 3.7% 17.7%

Table 1 represents the results from water pickup upon the prepreg samples from Figure 12, where W1 represents the weight prior to water pickup, W2, after water pickup, and WPU being the change in weight on a percentage basis. From these results, we see that Sample A, which exhibited higher ultrasonic amplitude has a smaller change in weight as compared to Sample B, which exhibited lower ultrasonic amplitude. These results seem to validate the claim that lower amplitude of prepreg in NCU represents lower impregnation, and vice-versa.

Figure 13: C-scan images of wind turbine blade prepreg material with multiple defects: Wave defect (left), Fuzzball (center two), dry region (right). Figure 13 above depicts carbon fiber prepreg for wind turbine blades with various defects. The upon the left side represents a wave defect in blue. As can be verified in the photograph of the actual material, this wave defect (or fiber misplacement) is clearly detected using NCU. In the center, two images depict a “fuzzball” defect, which is a misplacement of fiber in a circular ball. The right image contains a dry region in blue towards the lower section. While unanticpated during manufacturing, it was confirmed that this section is notably more porous than the surrounding area.

Manufacturing Quality Control While NCU has demonstrated applicability for measurement of key material characteristics, such as porosity, delamination, and foreign objects, it presents a unique opportunity to serve as a safe, reliable, and cost effective quality control method. 2-dimensional imaging, recording key NCU values such as transmittance or velocity, is a preferred output for quality control measurements. Such images can be produced using X-Y scanning as well as multi-channel static arrays.

X-Y Scanning

For high resolution imaging, X-Y scanning can provide a large amount of detail in a cost-effective manner. In its simplest form, as seen in Figure 21 below, X-Y scanning consists of a motion control system capable of moving a transmitting and receiving transducer across a test material in two axes to create a 2D image at the desired resolution.

Figure 14: Example of X-Y Scanning system for 2D image creation (left) and software/data analysis (right).

23

Depending upon throughput requirements, NCU X-Y imaging can be a powerful method of quality control, creating detailed images with statistical processing capability for quick accept/reject decisions. It also allows for human operators to visually process C-scans to determine if defects or non-uniformities may be present within products. To the right in Figure 14 we see an example of the kind of statistical and graphical analysis which can be created from X-Y imaging. These include general measures such as average, minimum, maximum, standard deviation for a C-scan as well as histograms and line scans across desired areas. While X-Y scanning can provide detailed information, QC for high throughput applications are sometimes well served using static array imaging systems.

Multi-Channel Arrays

Continuous inspection during manufacturing can be best suited using multi-channel arrays in NCU. Each channel in the array can produce a line scan or rolling C-scan to provide constant and instanteous feedback. The spatial resolution is determined by the number of channels and their spacing relative to each other.

Figure 15: 32-channel non-contact ultrasonic array (left) and representation of modular linear array (right). Each production application may require a different configuration of ultrasonic transducers in a multi-channel array. Figure 15 provides an example of a 32-channel array, arranged in a brick pattern for continuous coverage during webline manufacturing. This particular example consists of 500 kHz ultrasonic transducer elements, each with dimensions of 19 mm x 19 mm active area as receivers (visible on the lower portion of the photograph). The transmitting unit is a large oversize transducer with dimensions of 4 cm x 25 cm. The transmitter transmits ultrasonic energy across all receiving elements but the image resolution is determined by the receiver array pattern. In addition to a brick pattern, transducer arrays can also be constructed in linear form. While continuous coverage is not obtained, the effectiveness of a linear set of trasnducers can often provide enough information. To the right of Figure 19 we see a linear pattern comprised of multiple channels per transducer pair. In this case several pairs of transducers can be set up alongside each other and depending upon the coverage area and size of element, more or less pairs can be used. Like the previous brick pattern example, the transmitting units are oversized to cover the area of the receiving transducer arrays.

CONCLUSIONS Through significant advancements in non-contact ultrasonic technology, inspection opportunities to support quality control in manufacturing are now possible across a number of products in the composites industry. As demonstrated above, measurements such as porosity, density, delamination, and more can be achieved through NCU. Based upon the nature of the desired measurement, this can be accomplished at frequencies between 30 kHz and 5 MHz through X-Y scanning or by use of multi-channel arrays for continuous inspection. Continuous inspection provides beneficial feedback during manufacturing and full information regarding product quality following fabrication. While a number of materials have been analyzed, the technology is relatively new and there are a multitude of materials an applications which have yet to be explored. Further work is required and a number of industrial applications should be pursued using non-contact ultrasound for the purpose of production quality control.

24

REFERENCES AND RELATED WORK 1.

Bhardwaj, M.C., “Non-Destructive Evaluation: Introduction of Non-Contact Ultrasound,” Encyclopedia of Smart Materials, ed. M. Schwartz, John Wiley & Sons, New York, 690-714 (2002).

2.

Bhardwaj, A.M., “Application of Non-Contact Ultrasound for In-Line Inspection and Material Qualification,” Manufacturing 4 the Future conference, 2014, Hartford, CT.

3.

K. Fetfatsidis, Bhardwaj, A.M. “Correlation of Prepreg Resin Impregnation levels to Resulting Composite Part Porosity Using Non-Contact Ultrasound (NCU).” CAMX/SAMPE Conference, 2014, Orlando, FL.

4.

Cantavella, V., Llorens, D., Mezquita, A., Molti, C., Bhardwaj, M.C., Vilanova, P., Ferrando, J., and Maldonado-Zagal, S., “Development of Non-Contact Ultrasound Techniques for Measuring the Bulk Density for Optimization of the Pressing Process,” QUALICER 2006 IX World Congress on Ceramic Tile Quality, Vol. 2, 2006, Valencia, Spain.

5.

Bhardwaj, M.C., “Non-Contact Ultrasonic Characterization of Ceramics and Composites,” Proceedings Am.Cer.Soc., Vol. 89 (1998).

6.

Kulkarni, N., Moudgil, B., and Bhardwaj, M.C., “Ultrasonic Characterization of Green and Sintered Ceramics: I, Time Domain,” Am. Cer. Soc., Cer. Bull., Vol. 73, No. 6, (1994); II, Frequency Domain,” Am. Cer. Soc., Cer. Bull., Vol. 73, No. 7, (1994).

7.

Bhardwaj, M.C., “High-Resolution Ultrasonic Nondestructive Characterization,” Cer. Bull., v. 69, n. 9, (1990).

25

The Limitations Limitations of of Magnetic Magnetic Flux Flux Leakage Leakage The Scanning of of Aboveground Aboveground Storage Storage Tank Tank Bottoms Scanning Bottoms Daniel Carden Daniel Carden Zuuk International Zuuk International 3567 Meeting Street Rd 3567Charleston, Meeting Street Rd North SC 29405 North Charleston, SC 29405 (843) 414-9570; fax (843) 566-7258; e-mail [email protected] (843) 414-9570; fax (843) 566-7258; e-mail [email protected]

INTRODUCTION

Magnetic Flux Leakage (MFL) scanning can be an extremely effective method of detecting areas of isolated corrosion pitting in aboveground storage tank (AST) bottoms. This method has been proven to be effective by numerous different means (i.e. Ultrasonic (UT) prove-up, coupons) and has become the industry standard for assessing the condition of a tank bottom when an aboveground storage tank has been removed from service. However, the effectiveness of this inspection method is almost entirely dependent on the knowledge and experience of the MFL operator. The MFL operator’s knowledge of the theory of MFL as well as the MFL operator’s knowledge of his/her specific equipment is imperative, but something that is often overlooked is the limitations to MFL that are often present when scanning a tank bottom. Numerous limitations are often present on a tank bottom that can prohibit a thorough tank bottom scan. The objective of this paper is to educate storage tank owner / operators as well as MFL operators on these limitations in order to improve the overall quality of MFL tank bottom scanning.

Insufficient Cleanliness

Insufficient tank bottom cleanliness is one of the most obvious limitations to MFL scanning but is often something that MFL operators are confronted with due to tank owner / operator imposed time and money constraints. The most ideal surface preparation for MFL scanning of tank bottoms would be a white blast that leaves the tank bottom free from product residue, standing water, loose scale, dirt, debris and failed coating.

Inherent Physical Limitations

Limitations to the area of a tank bottom that cannot be effectively scanned due to the physical geometry of the scanning equipment and / or the tank are often referred to as inherent physical limitations. These areas are commonly present in the corners of each bottom plate, adjacent to lap welded bottom seams, adjacent to the shell-tobottom weld and adjacent to any internal obstructions (i.e. sumps, floating roof leg striker plates, fixed roof column bases, and piping).

Excessive /Variations in Coating Thickness

When tank bottoms have internal linings, it is crucial that the MFL operator is aware of the thickness of these coatings. A coating thickness that is beyond the range of a particular MFL system creates an excessive amount of lift-off between the sensor bar and the bottom plate that drastically decreases the sensitivity of the equipment. Additionally, variations in coating thickness can drastically affect the sensitivity of the equipment.

26

Undulations in Bottom Plates or Edge Settlement

Undulations, or more simply known as waviness or unevenness, of bottom plates can affect the sensitivity of MFL equipment in the same manner as variations in coating thickness. The additional lift-off created when the magnetic bridge of an MFL instrument is across a low-lying undulation drastically decreases the sensitivity of the system and the possibility of significant corrosion going undetected exists.

Variations in Nominal Bottom Plate Thickness

MFL technology is designed to detect localized abrupt changes in plate thickness. Areas of tank bottoms that have experienced gradual metal loss, such as knife-edge corrosion in the critical zone, cannot be effectively detected with MFL. Additionally, if general corrosion is present on a tank bottom that the nominal thickness varies throughout the tank, the sensitivity of the MFL system will be affected.

Numerous Existing Lap Welded Patch Plates

When a tank has previously been removed from service and repairs have been made to the bottom, if lap welded patch plates are in close proximity to each other or in close proximity to lap welded bottom plate seams, limitations may exist to the physical space necessary for normal scanner travel.

27

General Product Side Pitting

Tank bottoms that contain widespread product side pitting can have an adverse affect on the calibration of MFL equipment. When widespread product side pitting is present, it can be extremely inefficient for the MFL operator to attempt to identify each individual product side pit.

Severe Soil Side Pitting

In situations where severe soil side pitting is present on a tank bottom in the form of numerous through-thickness holes and numerous pitted areas in close proximity to each other, limitations may be present that could inhibit the MFL system from thoroughly detecting each individual area of corrosion.

CONCLUSION

Although many limitations to MFL scanning exist, it has proven to be the most effective means of assessing the suitability of an aboveground storage tank bottom. When aboveground storage tank owner / operators and MFL operators are aware of these limitations and properly address them, a thorough and effective tank bottom inspection can be performed.

28

Ultracloud – A New Ultrasonic Nondestructive Testing Technology Ultracloud – A New Ultrasonic Nondestructive Testing Technology ChenZhifa, Zhifa,Zheng ZhengYanchun, Yanchun,Wu WuJinhu, Jinhu,and andHong HongLixiang Lixiang Chen Shantou Institute of Ultrasonic Instruments Co., Ltd. (SIUI) Shantou Institute of Ultrasonic Instruments Co., Ltd. (SIUI) No.77 Jinsha Road, Shantou 515041, Guangdong, China Jinsha Road, Guangdong, China +86No.77 754 88250150; fax Shantou +86 754 515041, 88251499; e-mail [email protected] +86 754 88250150 ; fax +86 754 88251499; e-mail [email protected]

ABSTRACT

By applying cloud computing to ultrasonic nondestructive testing, it provides users a more secure, convenient

and efficient ultrasonic nondestructive testing technology and working method. This paper briefly describes the implementation and working method of this technology.

The “Ultra” device, integrated with multiple technologies, collects and sends data to the cloud server for storage and analysis; meanwhile it can acquire testing solutions and technical support from cloud server. The cloud

server features solution design (simulation analysis), storage, data processing, evaluation, communication and

management. An “Ultracloud” testing product is also described in this paper.

Keyword: nondestructive testing, ultracloud testing, Wi-Fi network, cloud computing

INTRODUCTION

In the field of non-destructive testing, inspection methods and systems such as ultrasound, X-ray and eddy current were developed to address a wide variety of engineering issues. With the diversification of testing

methods and systems, there has been increasing needs for simulation and calculation, safety storage, fast search and processing as well as effective management of mass data.

On the other hand, with the continuous development of Internet and computer technology, “Cloud computing”, a new distributed computing model, is entering all aspects of life, bringing convenience to people's life. As

defined by the National Institute of Standards and Technology (NIST), cloud computing is a model that can be easily accessed on demand to a public collection of configurable computing resources (for example, network, server, storage device, application or service). Such resources can be provided and released quickly, while minimizing the management cost or service provider intervention.

This paper describes the combination of cloud computing and the current non-destructive testing methods and

systems, which constitutes a non-destructive testing method based on cloud computing, serving as an important future trend. The method can make full use of computers, communications and other electronic technology

development outcomes, by effective organization and management of the current testing methods, systems and test results, so as to render the users with more secure, reliable, convenient and efficient service.

ULTRACLOUD ULTRASONIC NONDESTRUCTIVE TESTING It is generally agreed that cloud computing has both narrow and broad definitions.

The narrow definition of cloud computing refers to vendors building data centers or supercomputers through distributed computing and virtualized technologies, and provide service such as data storage, analysis and

29

scientific computing to technical developers or enterprise customers on demand and in a scalable method through the network free of charge or for lease on demand.

The broad definition of cloud computing refers to vendors building network server cluster to provide customers

with a variety of service such as online software service, hardware leasing, data storage, calculation and analysis. The broad definition of cloud computing involves more vendors and service types. The network providing

resources is called “cloud”, while the resources in the “cloud” seem to be infinitely scalable for users and can be used on demand and is pay-per-use[1,2].

In combination with cloud computing, the paper proposes the concept “Ultracloud” ultrasonic nondestructive testing.

“Ultra” refers to the innovation derived from the interdisciplinary efforts of modern physics, biology, acoustics, electronics, computer technology and network technology. It breaks the limitation of traditional technology

platform, achieves unlimited expansion of technical processing resources, improves processing speed of the system and optimizes resources allocation, so as to enhance the ability to adapt to future needs. It enables

customers to meet different complicated needs by simple operations. “Cloud”, as the carrier, with the aim to

achieve interoperability between various terminal systems, enables customers to acquire secure, convenient and efficient service.

The technical architecture of “Ultracloud” platform is shown in Fig. 2-1, which consists of three layers: IaaS (Infrastructure as a Service), PaaS (Platform as a Service) and UaaS (UltraCloud as a Service).

IaaS is the infrastructure capacity (servers, storage, computing power, etc.) available to users via the Internet, and a user-paid service for resources based on actual usage or occupancy. IaaS is to provide all kinds of

hardware service required by the ultracloud platform, including cloud storage, cloud computing servers and network connectivity environment.

PaaS is the service of application software deployment and operating environment based on cloud computing

infrastructure and provided to users[3]. It is a development environment, based on which the Ultracloud platform

provides users with all the hardware and software service.

UaaS includes all interfaces of hardware and software platform service that all types of users have direct access and application, through which the users may log in the Ultracloud platform to use service functions such as identification, cloud computing and cloud storage.

30

Figure 2-1. Technical Architecture of Ultracloud platform.

COMPOSITION AND FUNCTIONAL DESCRIPTION OF ULTRACLOUD ULTRASONIC TESTING Composition

Ultracloud ultrasonic testing described in the paper mainly has three components: ultra-testing system and user

terminal, cloud server and the Internet. See Fig. 3-1.

Figure 3-1: Composition of Ultracloud Ultrasonic Testing. “Ultra” device is the testing system at user terminal. In addition to general features on ultrasonic testing system, the most important feature of an “Ultra” device is supporting network communication, including wired and

wireless network. During testing, the users may use the test system to log in and connect to the cloud server to obtain the testing solution specific to the test object, or upload the test data realtime to the cloud server for storage, or have online discussions with other users and experts through the cloud server.

31

User terminal refers to all types of PC system, including personal computers and smart handheld devices. The users may use such devices anytime anywhere to access to the cloud server for searching test data; analysis,

processing and management; have online discussions with site testing personnel through the cloud server; or acquire status of the test device.

Cloud Server is the functional center of the whole “Ultracloud”, with the main functions of cloud storage and

cloud computing capabilities. All types of user data can be saved and cloud computing may be performed as per user needs.

Internet, either wired or wireless, is the interconnect channel of the three “Ultracloud” components above.

Features

This section describes key features of Ultracloud ultrasonic testing based on cloud computing, which mainly includes:     

Cloud simulation for testing program or process design implementation; Cloud storage for data storage and sharing; Cloud processing for data post-processing;

Cloud management for monitoring and management of testing process, systems and personnel; and Cloud maintenance for providing customers with efficient service.

Cloud simulation

With the development of testing process, there is an increasing demand for testing simulation and calculation,

and the authenticity and reliability of simulation software are improved significantly. However, although the simulation software brings convenience to users, its high cost and demanding requirements for operating

environment become an obstacle to the growth of simulation software, especially for small-size testing users, the

expensive operating cost constrains the demand for such user needs.

By combining simulated computing with cloud computing, the large-scale simulation software runs at the cloud

server, which not only decreases the operating cost of single users, but enabling users free from temporal and spatial limitations, as a result promoting the development of simulation software.

The user may use the client terminal to login the cloud server, run the simulation software as required, such as

sound field simulation software and specific testing process simulation software. The simulation results can be stored in the cloud server, or downloaded to the connected testing system. See Fig. 3-2 for cloud simulation approach.

32

Figure 3-2: Cloud Simulation.

Cloud Storage

See Fig. 3-3 for Cloud storage. As the user performs testing with the device at site, he may also login the cloud

server to download the special testing process, with the test results sent to the cloud server simultaneously. The

storage method is a change to the current method, with the latter saves the test results temporarily in the system

or other storage medium first, and loads to the computer after the testing. During testing, other authorized users may use the server to monitor the test process and test result, which is very important for online testing. The

real-time high capacity storage and reliability of cloud storage server improves user testing program and test data storage speed, shares the range and withstand unexpected risks.

Figure 3-3: Cloud Storage.

Cloud Processing

Upon finishing the testing, the user may login the cloud server, and perform data processing of the test results, such as test result measurement, evaluation, statistics and analysis, test report generation. Since the test results are stored in the cloud server, the cloud processing function can be accomplished simultaneously by another person located in a different region or even another country. See Fig. 3-4 for Cloud processing.

33

Figure 3-4: Cloud Processing.

Cloud Management

Each system or person connected to the cloud server needs authorization and verification to log in. Through user interaction, the authorized user may perform management functions such as monitoring test process, tracking

the status of test system, checking the working condition of the test operator. Meanwhile, the management

function allows the user to create customized test approval process, with an aim to tackle management difficulty due to geographically dispersed testing personnel and equipment.

Cloud Maintenance

The Cloud Maintenance function offers a new channel for the user to communicate with the system or software

provider. By connecting to the cloud server, the system or software provider may communicate directly with the user, or even operate directly on the device or software with user authorization. This approach significantly

shortens the distance between the user and the vendor, providing a virtual site solution. The user can also obtain system or software update or change information via the Cloud Maintenance function.

EXAMPLE

Here is an application example of the Ultracloud ultrasonic testing equipments developed by the Shantou

Institute of Ultrasonic Instruments Co., Ltd (SIUI). The ultrasonic testing equipment was a general purpose ultrasonic flaw detector, and the test was for the oil and chemical industry. In the application, due to the

confidentiality requirements of the testing company, the cloud server was set up in the testing company, with the

testing equipment, the user terminal and the cloud server connected via Ethernet connection. See Fig. 4-1 for the

network connecting structure.

Figure 4-1: Network Connection.

34

In addition to ultrasonic testing, physical and chemical methods such as low magnification inspection,

metallurgical testing, scanning electron microscopy (SEM), fracture test and spectral analysis were adopted to

examine flaw location, size and morphology, which provided anatomical analysis of different specimen defects and identified the disqualified reasons of the components on the oil and gas drilling system.

In this application, up to 10 units of ultrasonic flaw detectors were operated manually for testing work pieces,

and the data collected include system status and flaw detection waveform information (either identified with or without flaws). The test results were transmitted to the “cloud server” via the wireless (Wi-Fi) network.

After the testing, the operator logged in the cloud server from the client terminal. Under the customized

workflow, the test data was collated, such as review system status, test process effectiveness, flow detection waveforms and work pieces. See Fig. 4-2.

The test data approved by the review was classified and saved in the cloud server in accordance with the

intrinsic properties. And then the testing engineer would analyze the data information to identify the defect

property and underlying causes, and give feedback on raw materials, forging and heat treatment production based on the specific process. The statistical information such as defect count and cause classification was generated.

Figure 4-2: Cloud Computing Data Processing. Upon finishing the analysis, the cloud server automatically generated an ultrasonic flaw detection data report based on the operation, as shown in Fig. 4-3, and printed via the networked printer.

35

Figure 4-3: Cloud Computing Report.

CONCLUSIONS AND OUTLOOK

Ultrasonic nondestructive testing based on cloud computing is a new concept. In the future, this technology can also include other physical and chemical testing methods, achieving effective combination and sharing of

information. While providing users with a safe and reliable storage environment, login to the cloud server is basically free from any temporal or geographical restrictions, bringing users convenient experience of data

processing and technical exchange. Consequently the users may manage testing equipment and learn the test personnel status timely and efficiently, which is an important trend of future nondestructive testing.

REFERENCES 1. 2. 3.

Zhang Weimin, Tang Jianfeng, Luo Zhiguo, et al. Cloud Computing Profoundly Changes the Future,

Science Press, Dec. 1, 2009

Written by George Reese, translated by Cheng Hua, Cloud Computing Application Architecture, Electronic Industry Press. July 2010

Lei Baohua, Rao Shaoyang, Jiang Feng, et al. Cloud Computing Decoding: Technical Architecture and Industrial Operations, Electronic Industry Press, April 2011

36

Digital Image Correlation Techniques for Aerospace Applications Digital Image Correlation Techniques for Aerospace Applications Tsuchin Philip Chu and Anish Poudel Philip Chu and Anish Department Tsuchin of Mechanical Engineering andPoudel Energy Processes, Southern Illinois University, 1230 Lincoln Drive, Carbondale, IL 62901 Department of Mechanical Engineering and Energy Processes, (618) 453-7003, fax (618) 453-7658, e-mail [email protected] Southern Illinois University, 1230 Lincoln Drive, Carbondale, IL 62901 Tel. (618) 453-7003, Fax: (618) 453-7658, Email: [email protected]

ABSTRACT This paper discusses the history, current state-of-the-art, theory, and application of digital image correlation (DIC) techniques in various aerospace applications. First, the history of DIC followed by the current state-of-the-art of DIC techniques are presented. Second, brief working principle of DIC is presented. After that, several examples of DIC technique for different aerospace applications are presented. Some of the examples include strain measurement in friction stir weld (FSW) samples, measurement of high strain regions in particulate composites, Poisson’s ratio measurement in spray on foam insulation (SOFI) samples, measurement of strain and deformations in NASA’s massive large rocket test section, full-field strain evaluation in helicopter drone composite rotor blade, and measurement of strains in shear panels of an aircraft. Finally, this paper also discusses on-going effort for measuring the in-plane bulk displacement and deformation in composites and adhesive joints by using hybrid NDE DIC techniques. Keywords: DIC, NDE, Ultrasonic C-scans, Composites, Deformation, Strain.

INTRODUCTION Digital Image Correlation (DIC) is a non-contact whole-field imaging technique that employs tracking and image registration for accurate twodimensional (2D) and threedimensional (3D) deformation and strain measurements. It is a Original Location technique based on surface analysis Location/Shape Starting sub-image Starting sub-image (after loading) methods which involve monitoring and identifying changes in a pattern (a) (b) applied to the surface of objects Figure 1: Speckle pattern for DIC. (a) Before deformation; and (b) under observation, that have been after deformation. subjected to some form of mechanical pressure or strain as shown in Figure 1 . DIC method was developed in the early 1980’s for measuring surface displacements and deformation [1, 2]. This method requires a digital imaging system to optically record images of the surfaces before and after deformation. The pair of gray level functions of the images were then compared using advanced image correlation and processing algorithms to determine the displacement and deformation gradients. Unlike laser speckle techniques, which required an optically rough, reflective surface and minimal vibration, the only requirement for surface condition was a visually “speckled” surface. If not inherent to the material, this could be attained by application of a suitable random pattern such as with spray paints. A novel approach that determined the local displacements and deformation gradients was later developed by Chu et al. [3]. A faster approach for image correlation was later developed by Sutton, et al. [4] Instead of using the iterative approach, such as a coarse-fine method, a second order Newton-Raphson method was employed. This method improved the computation speed of the correlation. However, it presented some convergence difficulties. A special technique was then added to achieve a higher probability of convergence. This correlation method was later evolved into measuring in-plane and out-of-plane displacements using two video cameras separated by a predetermined distance. The lines of sight from two cameras may either be parallel to each other [5] or form a pan angle between them [6]. As soon as two image points are matched by correlation methods, the three-dimensional

37

coordinates of the physical point may be determined. A surface contour may be generated from one pair of stereo images, while a deformation measurement requires the correlation of four images: a stereo pair before deformation and another after deformation. DIC of speckle patterns for engineering applications has been used extensively in many applications to measure displacement components and deformation gradients of an object’s surface due to deformation for over four decades [1-28]. The development and broad applications of this method can be attributed to the rapid evolution of computing power, CCD camera or recorders, and frame grabbers. This paper presents several examples that DIC technique have been successfully applied for different aerospace applications. Some of the examples include strain measurement in friction stir weld (FSW) samples, measurement of high strain regions in particulate composites, Poisson’s ratio measurement in spray on foam insulation (SOFI) samples, measurement of strain and deformations in NASA’s massive large rocket test section, full-field strain evaluation in helicopter drone composite rotor blade, and measurement of strains in shear panels of an aircraft. Finally, this paper also discusses on on-going effort of authors for measuring the in-plane bulk displacement and deformation in composites and adhesive joints by using hybrid NDE DIC techniques.

Working Principle of DIC The basic correlation algorithm uses two image data sets, typically in the form of a speckle pattern, to extract deformation profiles from tiny changes in the images. Figure 2 shows three examples of typical speckle patterns.

(a) (b) (c) Figure 2: Typical speckle patterns for DIC. (a) Painted speckle pattern, (b) natural surface pattern, and (c) Acoustography fiber patterns. Figure 2(a) is a spray painted pattern over an aluminum specimen, Figure 2(b) is an image of a particulate composite, so the pattern occurs naturally, and Figure 2 (c) is an Acoustography image of a composite panel, so the pattern, once again, occurs naturally. Now consider any of the three specimens shown in Figure 2 being subjected to some loading condition. Two images are taken, one being taken before loading (undeformed state) and another during loading (deformed state). The two images are then analyzed to estimate displacements and strains. The aim of this method is to find the displacements and deformations of small subsets from the second image relative to the first one. This is accomplished by comparing the intensity levels (0-255) of the subsets in the images.

Figure 3: Local deformation.

Figure 4: Interpolated subset.

38

If the second image represents only a pure displacement then the search is fairly simple, but if the subset is deformed due to the loading conditions, then the search is a little more complex as explained next. Local deformation of a subimage, the corresponding displacement components, and deformation gradients are shown in Figure 3. To start the analysis, a subset from the digitized intensity pattern of the undeformed object surface centered at P o is chosen. An interpolation scheme, typically bilinear, is applied to the selected subset to obtain sub-pixel accuracy as shown in Figure 4. The method for comparing two subsets is commonly given by the use of the cross-correlation coefficient, C:

 u u v v   * f x, y  f * x   , y   dA C   (u, , ), (v, , )   x y x y     f x, y 2 dA  f * x   , y   2 dA      *

(1)

where,  = subset in the undeformed image, * = subset in the deformed image

 u  v

The values of u, v,

u u x  y x y

v v x  y x y

(2) (3)

u u v v which maximize C are the local deformation gradients for the selected , , , and x y x y

subset. The main objective of the image correlation process is to find these six values for the subset under investigation and then repeat it for all subsets in a given region so as to find the whole-field deformation profile.

DIC APPLICATION EXAMPLES DIC is a proven technique both in terms of the technical advancement and in terms of the applications for which it is applied. Several examples where DIC technique have been successfully applied for different aerospace applications are presented in this section.

Strain measurement in friction stir weld (FSW) samples Friction Stir welding is solid state welding process for joining metals by plasticizing and consolidating materials around the bond line using thermal energy produced from localized friction forces [29]. This welding method is applied in many aerospace applications and is often required to inspect the welded region for the integrity of the welds. Strains induced in the FSW samples made with dissimilar materials, i.e. Al 2195 and Al 2219 alloys were determined by using Sub-Pixel Digital Image Correlation (SPDIC) technique. For this, uni-axial tensile loads were applied in longitudinal direction of the samples until they failed. The material properties of the weld samples such as the Young’s modulus, yield stresses were also determined from the stress-strain graphs obtained by using the SPDIC tool. In addition, the pseudo colored strain contour maps as shown in Figure 5 indicated the area of maximum strain and thereby predicting crack initiation in the sample.

Figure 5: Strain contours of FSW sample at different loading condition.

39

Figure 6: Plot of the stress-strain curves for FSW sample at different locations. The Young’s modulus of elasticity determined for the FSW sample from the stress-strain graphs as shown in Figure 6 was around 62 GPa. This shows that the strength of the friction stir weld samples was quite good and was almost equal to the strength of the parent materials. From the stress-strain graphs obtained at various locations on the samples it was also evident that the friction weld material behaves similar to its parent materials.

Measurement of high strain regions in particulate composite sample Particulate composites used for rocket fuels were analyzed by using DIC techniques. The main objective of this work [12] was to determine the high strain regions under low loading conditions that would possibly indicate regions of eventual failure under high loading conditions. This is important since the underlying causes of failure such as microcracks or other surface/bulk defects are not always detected by traditional NDE techniques. Figure 7 shows the strain fields generated by using DIC for the particulate composite test sample. From the results obtained, it was demonstrated that there is a statistical distribution of high strain regions within the particulate composites. Based on the result obtained, it was demonstrated that DIC method may be used to identify very early on regions of eventual failure, and may also be used to change the distribution of particles in the particulate composite so as to obtain improved material properties.

(a) (b) Figure 7: Strain map (εyy) for particulate composite sample showing: (a) 5% average strain for subset size = 15, and resolution of 5 pixels; and (b) 5% average strain for subset size = 21, and resolution of 2 pixels.

40

Poisson’s ratio measurement in spray on foam insulation (SOFI) samples The Space Shuttle's external tank is covered with special spray-on foam insulation (SOFI) which serves to insulate the tank before and during launch. It is very difficult to measure the Poisson’s ratio of SOFI by using mechanical methods of by using electric gauge, speckle interferometry, and holography techniques. Because, SOFI has a loose structure and is made of tiny bubbles or foam cells which makes it tough to measure the deformations using the above mentioned procedures. Poisson’s ratio of SOFI samples were calculated by averaging strains determined at different locations on the specimen by using DIC technique [28]. Figure 8 shows the speckle pattern of the SOFI samples before and after deformation. The measurement locations were chosen in such a way that they covered whole area of the test sample. Nine different locations of size 20x20 pixels (a) (b) (0.0324x0.0324 in2) were selected to calculate the local Figure 8: Digital images of the SOFI test samples. (a) displacements in the sample. These locations were Undeformed state; (b) deformed state. correlated using the SPDIC software to obtain the local displacements. The average displacements in both transversal and longitudinal directions were then calculated for these selected locations. Hence, the average strains in two directions, εx and εy were obtained. The average Poisson’s ratio determined for the SOFI test sample from SPDIC was 0.37 with standard deviation of 0.015. To validate the results, the exact same test sample was turned around after the first test and the same test procedure was followed on the other side of the sample. For the second case, the Poisson’s ratio determined was 0.29 with standard deviation of 0.012. The difference in the Poisson ratios between the two tests can be accounted to the fact that in the second test loads are applied on the elongated sample so there is a possibility that the sample have reached the state of plasticity.

Measurement of strains and deformations in NASA’s massive large rocket test section NASA's Engineering and Safety Center (NESC) Shell Buckling Knockdown Factor (SBKF) Project was established in March 2007 to develop and validate new analysis-based shell buckling design factors (i.e. knockdown factors) for Ares I and V metallic and composite launch vehicle structures. During the most recent test conducted in NASA's Marshall Space Flight Center in Huntsville, Al., a massive 27.5-foot-diameter and 20-foot-tall aluminum-lithium test cylinder as shown in Error! Reference source not found.(a) was loaded with one million pounds of force until it failed [30]. Advanced image correlation techniques were used to monitor tiny deformations over the entire outer surface of the test article.

Full-field strain evaluation in helicopter drone composite rotor blade Rotor blades are usually subjected to high stresses due to their weight and the force created by irregular winds. DIC technique has been successfully employed to analyze these rotor blades under normal to shear stress conditions as well as in the bending fatigue test. GOM recently demonstrated the application of DIC technique to monitor the strain fields on composite rotor blade for helicopter drones in the laboratory environment subjected to bending test [31].

Measurement of strains in shear panels of an aircraft DIC technique has also been successfully implemented to inspect the shear panel subjected to shear load for mapping whole-field out-of-plane displacement (buckling). GOM recently demonstrated that the out-of-plane displacement map generated from ARAMIS DIC software was in good agreement with the FEA results [31].

NEED FOR ADVANCED NDE DIC TECHNIQUE

41

With the use of advance materials such as fiber reinforced composites for various military/commercial aerospace/aircraft applications, there is increasing demand for the novel NDE techniques that can characterize these materials at different stages of their lives. Existing NDE/DIC techniques are already proven both in terms of the technical advancement and in terms of the applications for which they are applied. However, DIC techniques now are basically implemented to monitor the surface deformation only. On the other hand, NDE techniques are often applied to detect and characterize flaws in test objects. NDE technique lacks the capability to provide deformation/strain map in test objects. Similarly, existing DIC technique still lacks the capability to provide the realistic information for the bulk of the composites or the adhesively bonded joints. By combining DIC with the existing NDE methods, it is possible to obtain a bulk deformation profile of the composites and adhesive joints. This will be a key for damage monitoring and failure prevention in future. Very few literatures have made an attempt to combine NDE and DIC methods to measure deformations in solids. Sachse [32] demonstrated how ultrasonic spectroscopy could be used to detect both dimensions and mechanical properties of flaws or impurities inside elastic solids. Schaeffel, et al. [33] developed a one dimensional ultrasonic speckle interferometry by employing the cross correlation technique to determine the displacement of a region on the surface or in the body of an object. Ranson [34] extended the ultrasonic speckle interferometry to determine the displacement components with a two-dimensional cross correlation scheme. Hong, et al. [35] have demonstrated that the displacements could be measured by ultrasonic speckle interferometry, but determination of deformation gradients and strains has not been demonstrated. Zhang, et al. [14] implemented ultrasonic C-scan acoustical speckle images to make translation, rotation, and deformation predictions either on front or back surface, and to discuss the usefulness and limitation with the digital correlation method in determination of the surface displacements and the in-plane strain fields. The experimental results indicated that the correlation method with ultrasonic speckle pattern could make the correlated two-dimensional displacements within 1.2 mm (19 pixels) and small angle rotation within 1.2° accurate. Current efforts are also being made to implement the state-of-the art NDE techniques such as Acoustography, linear array ultrasonic phased-array, and x-ray backscatter techniques for the NDE DIC improvements.

CONCLUSION

This paper presented the history, current state-of-the-art, and working principle of digital image correlation (DIC) techniques in many aerospace applications. Several examples are presented to show the viability of this method in aerospace applications. However, existing DIC techniques are basically implemented to monitor the surface deformation only. It still lacks the ability to provide the realistic information for the bulk of the composites or the adhesively bonded joints. Because, today, polymer matrix composites (PMCs) such as graphite/epoxy composite laminates or Carbon Fiber Reinforced Plastics (CFRPs) are increasingly used in many structural applications which ranges from aerospace to automotive, industrial, sports industry and many other consumer products. In addition, the increasing use of composites in aerospace/aircraft industries have also driven the use of adhesives in the assembly of these structures. These include bonding of similar and dissimilar materials such as metal-metal, composite-composite, and composite-metal joints. The use of these advanced materials in many structural applications have raise challenges to the existing NDE/DIC techniques. Also, there is a big demand for a novel technique that can not only characterize these advanced materials but also predict the life of the part/structure to ensure the safety and reliability of composite structures. By combining DIC with the existing NDE methods, it is possible to obtain a bulk deformation profile of the composites and adhesive joints. This will be a key for damage monitoring and failure prevention in future. REFERENCES [1]

Peters, W. H., and Ranson, W. F., 1982, "Digital Imaging Techniques in Experimental Stress Analysis," Optical Engineering, 21(3), pp. 427-431.

[2]

Peters, W. H., Ranson, W. F., Sutton, M. A., Chu, T. C., and Anderson, J., 1983, "Application of Digital Correlation Methods to Rigid Body Mechanics," Optical Engineering, 22(6), pp. 738-742.

[3]

Chu, T. C., Ranson, W. F., Sutton, M. A., and Peters, W. H., 1985, "Applications of Digital Image Correlation Techniques to Experimental Mechanics," Experimental Mechanics, 25(3), pp. 232-244.

[4]

Bruck, H. A., McNeill, S. R., Sutton, M. A., and Peters, W. H., 1989, "Digital Image Correlation Using NewtonRaphson Method of Partial Differential Correction," Experimental Mechanics, 29(3), pp. 261-267.

42

[5]

Kahnjetter, Z. L., and Chu, T. C., 1990, "3-Dimensional Displacement Measurements Using Digital Image Correlation and Photogrammic Analysis," Experimental Mechanics, 30(1), pp. 10-16.

[6]

Luo, P. F., Chao, Y. J., Sutton, M. A., and Peters, W. H., 1993, "Accurate Measurement of 3-Dimensional Deformations in Deformable and Rigid Bodies Using Computer Vision," Experimental Mechanics, 33(2), pp. 123132.

[7]

Sutton, M. A., Cheng, M. Q., Peters, W. H., Chao, Y. J., and McNeill, S. R., 1986, "Application of an Optimized Digital Correlation Method to Planar Deformation Analysis," Image and Vision Computing, 4(3), pp. 143-150.

[8]

Perng-Fei, L., Chao, Y. J., and Sutton, M. A., 1994, "Application of Stereo Vision to Three-Dimensional Deformation Analyses in Fracture Experiments," Optical Engineering, 33(3), pp. 981-990.

[9]

Helm, J. D., McNeill, S. R., and Sutton, M. A., 1996, "Improved three-dimensional image correlation for surface displacement measurement," Optical Engineering, 35(7), pp. 1911-1920.

[10]

Mahajan, A., Pilch, A., Chu, T. C., 2000, "Intelligent Image Correlation Using Genetic Algorithms for Measuring Surface Deformations in The Autonomous Inspection of Structures," Proceedings of the 2000 American Control Conference, Vols 1-6, pp. 460-461.

[11]

Chu, T., Mahajan, A., and Liu, C. T., 2002, "An Economical Vision-Based Method to Obtain Whole-Field Deformation Profiles," Experimental Techniques, 26(6), pp. 25-28.

[12]

Chu, T. C., and Mahajan, A., 2002 "Measurement of High Strain Regions in Particulate Composites Using Digital Image Correlation," Proc. 10th U.S.-Japan Conference on Composite Materials, Fu-Kuo, ed., DEStech Publications, Inc.

[13]

Zhang, Y., Chu, T. C., and Mahajan, A., 2002, "Analysis of Short Crack Growth in Particulate Composites," Proc. SEM Annual Conference & Exposition on Experimental and Applied Mechanics.

[14]

Zhang, Y., Byrne, C., Chu, T. P., and Mahajan, A., 2002, "Intelligent Correlation of Ultrasonic C-scan Images to Extract Bulk Deformations," Proc. SEM Annual Conference & Exposition on Experimental and Applied Mechanics.

[15]

Pilch, A., Mahajan, A., and Chu, T., 2004, "Measurement of Whole-Field Surface Displacements and Strain Using a Genetic Algorithm Based Intelligent Image Correlation Method," Journal of Dynamic Systems Measurement and Control-Transactions of the Asme, 126(3), pp. 479-488.

[16]

Adapa, S., Chu, T. P., and Schneider, J. A., 2005, "Evaluation of Friction Stir Welds," Proc. SEM Annual Conference & Exposition on Experimental and Applied Mechanics - Experimental Mechanics Applied to Advanced Materials Systems.

[17]

Pradhan, S., Chu, T. C., 2005, "Tensile Test Apparatus for Micro-Scale Specimens on Scanning Electron Microscope Using Sub-Pixel Digital Image Correlation," Proc. SEM Annual Conference & Exposition on Experimental and Applied Mechanics - Experimental Mechanics Applied to Advanced Materials Systems.

[18]

Bastawros, A. F., Bart-Smith, H., and Evans, A. G., 2000, "Experimental analysis of deformation mechanisms in a closed-cell aluminum alloy foam," Journal of the Mechanics and Physics of Solids, 48(2), pp. 301-322.

[19]

Lu, H., Vendroux, G., and Knauss, W. G., 1997, "Surface deformation measurements of a cylindrical specimen by digital image correlation," Experimental Mechanics, 37(4), pp. 433-439.

[20]

Vendroux, G., and Knauss, W. G., 1998, "Submicron deformation field measurements: Part 1. Developing a digital scanning tunneling microscope," Experimental Mechanics, 38(1), pp. 18-23.

[21]

Tong, W., 1997, "Detection of plastic deformation patterns in a binary aluminum alloy," Experimental Mechanics, 37(4), pp. 452-459.

[22]

Bay, B. K., Smith, T. S., Fyhrie, D. P., and Saad, M., 1999, "Digital volume correlation: Three-dimensional strain mapping using X-ray tomography," Experimental Mechanics, 39(3), pp. 217-226.

[23]

Peddle, J., Goudreau, A., Carlson, E., and Santini-Bell, E., 2011, "Bridge displacement measurement through digital image correlation," Bridge Structures, 7(4), pp. 165-173.

[24]

Caminero, M. A., Lopez-Pedrosa, M., Pinna, C., and Soutis, C., 2013, "Damage Monitoring and Analysis of Composite Laminates with an Open Hole and Adhesively Bonded Repairs Using Digital Image Correlation," Composites Part B-Engineering, 53, pp. 76-91.

43

[25]

Sirohi, J., and Lawson, M. S., 2012, "Measurement of Helicopter Rotor Blade Deformation Using Digital Image Correlation," Optical Engineering, 51(4).

[26]

Revilock, D. M., Thesken, J. C., Schmidt, T. E., and Forsythe, B. S., 2007, "Three-Dimensional Digital Image Correlation of a Composite Overwrapped Pressure Vessel During Hydrostatic Pressure Tests ", G. R. Center, ed., National Aeronautics and Space Administration Cleveland, Ohio.

[27]

Leser, W. P., Newman, J. A., and Johnston, W. M., 2010, "Fatigue Crack Closure Analysis Using Digital Image Correlation," L. R. Center, ed., National Aeronautics and Space Administration Hampton, Virginia.

[28]

Adapa, S., Zhang, Y., and Chu, T. P., 2005 "Poisson’s Ratio Measurement of SOFI using Image Correlation," Proc. SEM Annual Conference & Exposition on Experimental and Applied Mechanics.

[29]

Adapa, S., Chu, T. P., Schneider, J. A., 2005, "Evaluation of Friction Stir Welds," Proc. Proceedings of ASME International Mechanical Engineering Congress and Exposition, pp. 727-733.

[30]

Braukus, M., Harrington, J. D., and Stanfield, J., 2011, "NASA's Successful 'Can Crush' Will Aid Heavy-Lift Rocket Design."

[31]

09/03/2014, "The Use of Optical Metrology in the Aerospace Industry," GOM: Optical Measurement Technique, Braunschweig, Germany.

[32]

Sachse, W., 1974, "Ultrasonic Spectroscopy of a Fluid Filled Cavity in Elastic Solid," Journal of the Acoustical Society of America, 56(3), pp. 891-896.

[33]

Schaeffel, J. A., Ranson, W. F., and Swinson, W. F., 1980, "Acoustical-Speckle Interferometry," Experimental Mechanics, 20(4), pp. 109-117.

[34]

Ranson, W. F., 1984, "Theoretical Properties of Acoustical Speckle Interferometry," Journal of the Acoustical Society of America, 75(2), pp. 640-640.

[35]

Hong, S. K., and Ohr, Y. G., 1998, "Ultrasonic Speckle Pattern Correlation Interferometry Using a Pulse-Echo Method," Journal of Physics-Applied Physics, 31(11), pp. 1392-1396.

44

Aging Aircraft NDT Aging Aircraft NDT Russell A. Russell A. Davis Davis L-3 L-3 Mission Mission Integration Integration 10001 10001 Jack Jack Finney Finney Blvd, Blvd, CBN CBN 201 201 Greenville, TX 75402 (903) 457-4816; 457-4816: email [email protected]

INTRODUCTION

The sustainment of aging aircraft often requires novel application of standard NDT techniques, combinations of techniques and sometimes just plain judgment. This ever-changing, sometimes “rough and tumble” world makes close partnership of NDT technicians and sustainment engineers a must. Close communication between the two merges in-depth knowledge of the inspection technology with an understanding of structural load paths, aircraft usage and material properties to make an informed “decision without disassembly.” The element of the unknown often looms large in the aging aircraft field. The deterioration of the aircraft system may come from outside sources: fatigue or environmental attack (i.e. corrosion). Sometimes it is self-induced: inadequate maintenance practices, poor material selection in the design phase or changes in usage of the platform. Despite the cause of damage, the sustaining organization must remain ever vigilant; not just relying on the inspections in the book (although those are necessary), but going out and “looking for trouble.” This approach requires the right technique(s) in the right place at the right time – with just a little luck thrown in.

AGING AIRCRAFT

Aircraft age from the very first day of operation and sometimes even prior to that. Recent operators, both commercial and military, seek to extend the lives of their fleets due to myriad economic concerns and uncertainties. The costs of recapitalization can be substantial and development cycles for new aircraft can span a decade or more. The average age of the US and Canadian narrow body commercial fleets is 15 years, with a worldwide average age of 12 years [1]. In the first half of 2013, the average age of commercial aircraft upon retirement was 23 years [2]. The US military fleets tend to notably greater calendar age than commercial aircraft; however, they will generally have many fewer pressure cycles and flight hours than a commercial aircraft of similar age. Specifically, the US Air Force fleet has an average age of 25.2 years, with some aircraft types over 50 years old (see Figure 1) [3]. As

Average Age of USAF Fleet Number of Aircraft

2500 2000 1500 1000 500 0 0-10

11-20

21-30

31-40

41-50

51-55

Aircraft Age Figure 1: Average Age of USAF Fleet (as of 30 September 2013) [3]

45

regards the data presented in Figure 1, it is noteworthy that 285 of the aircraft under ten years of age are unmanned vehicles or drones. Reduced recapitalization and program development delays will “require current fleets to have lives extended ten to thirty years into the future if the current force structure is to be maintained.” [4]. In fact, Major General Robert M. Worley, USAF, commented to the Air Force Scientific Advisory Board that “the fleet of tomorrow may very well be today’s.” Aging aircraft, especially in the military fleets, are a long term prospect. The assessment and sustainment of them will see even greater emphasis in the future with a key role for Nondestructive Testing and specialized engineering.

Collaboration Between NDT and Engineering

The uncertainty inherent in maintaining aging aircraft requires close cooperation between NDT professionals and engineers supporting maintenance and depot activities. The National Academy of Science’s Examination of the U.S. Air Force’s Aircraft Sustainment Strategy to Meet Future Needs challenges the sustainment community to “deploy a collaborative engineering methodology to determine the required processes, tasks and frequency of maintenance actions.” [5] This collaboration brings a wider scope of knowledge to the problem, often allowing effective “brainstorming” of anomalous situations (non-standard indications, uncommon material behavior and so forth), sometimes avoiding costly and time-consuming disassembly for full evaluation. An illustrative example of such collaboration occurred in an anomalous inspection finding in a fastener hole common to the Wing Terminal Fitting of a C-135 aircraft undergoing depot maintenance. The Wing Terminal Fitting is a major structural component of the aircraft with limited rework allowances and a substantial cost for replacement. The initial finding was a minor eddy current indication and the standard rework action, a small upsize of the hole to remove the flaw, was undertaken. The subsequent inspection gave an indication of increased magnitude – not the typical result, generally upsizing of the hole eliminates or at least reduces the size of the defect. Moreover, the NDT technician reported uncommon readings that were inconsistent with the earlier indication. All of this led the sustainment engineer and NDT technician to examine the hole by borescope (see Figure 2). They found discontinuities that did not match

AFTER REAMING HOLE, EDGES OF CRACK ARE ROUGH AND OFFSET FROM THE REST OF THE HOLE BORE

OUTBOARD AFT

Figure 2: C-135 Wing Terminal Fitting With Anomalous Indication expected behavior of gouges or cracking. Comparing notes, they investigated further, eventually physically probing the hole and separating a thin partial layer of material from the circumference of the hole. The discontinuity and thin

46

layer of material came from a “freeze plug” – a repair method which shrink fits a repair bushing into an oversized hole to bring it back to a smaller diameter. This previous rework action had not been recorded in the aircraft records and the plug had been damaged during removal of the fastener, with further damage from upsizing the hole. During this entire evaluation period (approximately two days), other organizations, from Production to the customer, held their collective breath due to the potential costs of a replacement. However, the teamwork of an NDT technician and sustainment engineer allowed “decision without disassembly.”

Deterioration of the Aircraft System

Age and usage deteriorates not only the human body, but aircraft as well. Design of the aircraft presupposes a certain environment and usage. At times, these presuppositions (generally based on the best information at the time) are incomplete: the aircraft is used more harshly, perhaps in a more aggressive environment than anticipated or with inadequate maintenance. A telling example combining all of these is the Aloha Airlines Flight 243 incident on April 28, 1988 (see Figure 3). [6] On the day in question, the upper lobe of fuselage section 43 tore away from the aircraft,

Figure 3: Aloha Airlines Flight 243, Boeing 737-200 (Courtesy of Honolulu Advertiser) rapidly decompressing the cabin and ejecting a flight attendant (the only fatality). Aloha Airlines operated a fleet of eleven Boeing 737 aircraft, typically on short-hop flights between the various Hawaiian Islands, turning numerous flights a day. The mishap aircraft N73711 had 35,496 flight hours and 89,680 flight cycles at the time of the incident – the second highest total cycles in the worldwide 737 fleet (the worldwide leader was also an Aloha aircraft). The aircraft was built in 1969 and had been operated by Aloha its entire life. While the short flights did not allow full pressurization on every flight, the numerous flight cycles contributed to significant fatigue cracking at multiple locations of the riveted fuselage lap joints. The tropical operating environment, coupled with some of the bonding methods used in the aircraft construction, led to widespread corrosion. These issues were exacerbated by inadequate inspections (including NDT) and maintenance by the operator that were noted in the final NTSB report. Previous Service Bulletins had noted similar concerns and had been made a part of the maintenance program. Of note, a passenger later reported seeing a longitudinal crack on boarding the aircraft for the mishap flight (the ninth flight for the aircraft that day), but did not notify the crew or ground personnel.

The “Unknown” Unknown

NDT discovery of deterioration can sometimes be haphazard. Unless an area was identified as a concern in the design process or has been previously found on other aircraft (the known unknown), there will generally be no

47

planned, scheduled inspections. In many cases, deterioration, whether fatigue or corrosion, will be evidenced in other ways and may then be confirmed with NDT technologies. Fatigue cracks are often found visually, either directly or through their influence on other components. Often cracks and corrosion in integral fuel tanks are found during maintenance for leaks. If a part is cracked and can no longer carry load, it may lead to distortion of adjacent components which can be more visually noticeable. The corrosion process tends to destroy a portion of the material, creating a byproduct which is visible in its contrast to the base material, especially when painted. In joints where corrosion is present between the layers, buildup of the byproduct will often swell the material between the clamping action of the fasteners or spot welds, creating a “pillowing” effect (see Figure 4) seen visually. Sometimes, seeing

(b)

Figure 4: Pillowing of Spot Welds on C-135 Lower Fuselage Skin the cause can lead to the damage. For example, evidence of leaking lavatories (blue water stains) or galleys can often point to trapped areas where corrosion is likely (see Figure 5). Depots still rely heavily on the “Mark 1 Eyeball” for finding the “unknown” unknown.

(a)

(b)

Figure 5: Trapped Blue Water From a Leaking Lavatory (a) and the Result- a Corroded Stringer (b)

48

Material Considerations

The aerospace industry has a long history of emphasizing performance over other considerations, pioneering new manufacturing processes and new materials that would be prohibitively expensive in other commercial ventures. Many of our current aging aircraft were designed in the 1950’s and 1960’s, in a time when high strength materials were eagerly sought, sometimes to the exclusion of all else, in a quest for expanded structural performance at lower weights. Abbreviated development cycles (especially on military programs) often did not allow for full characterization of the latest “wonder” material prior to its incorporation on the next aircraft program. Furthermore, generous budgets and rapidly advancing technology tended to make service lives short, pushing a model aside for something newer and more capable. As such, methods and material evaluations currently deemed necessary for long term sustainment of a fleet would have been likely dismissed as unneeded and even wasteful in this earlier time. Unfortunately, a number of these transitional or interim design aircraft are still operating today, some fifty or more years later.

Aluminum

The 7000 series aluminum alloys, particularly in the T6 heat treat form, were seen as the answer to the need for high strength at lower weight. The material was implemented in several highly loaded areas of the aircraft such as bulkheads, floor beams and wing skins. However, subsequent usage and operation found that the alloys tended to exhibit poor fatigue resistance and damage tolerance with a key susceptibility to corrosion, particularly stress corrosion and other intergranular forms. Stress corrosion is a synergistic phenomenon that combines sustained loading and corrosion to rapidly grow cracks. Some 7000 series alloys showed such low resistance to stress corrosion that parts could crack sitting on the shelf just from the minimal residual stresses imparted from forming or machining. Intergranular corrosion typically manifests itself between the grains of the metallic structure, often serving to almost “delaminate” layers of the solid metal. This can be a key failure mode of the 7000 series alloy. The C-135’s BL 70 Rib is a key member that joins the center and inboard wings with the fuselage. It is a 7075-T6 extrusion. During depot maintenance on an aircraft, a hole common to the BL 70 Rib yielded a “crack like” indication. Two subsequent oversizes of the hole failed to eliminate the defect. At this point, knowing the material characteristics of the 7075-T6, the sustainment engineer requested a follow up with ultrasonic inspection. The UT found a sizable discontinuity, emanating from the hole, at approximately half the thickness of the member – intergranular corrosion (See Figure 6). Often, the uncertainty inherent in never before seen defects of aging aircraft requires a willingness to follow up with multiple techniques.

Figure 6: C-135 BL 70 Rib, Eddy Current Indication Followed Up With Ulrasonic (Shaded)

49

Steel

Steel is also sometimes a material of choice for performance in the aerospace industry. The F-111 aircraft presented a particular structural challenge to the General Dynamics designers because of the loading from the variable geometry or “swing” wings. They addressed this concern through the use of D6AC steel, a very high strength alloy, in portions of the wing carry through structure. However, after investigation of the catastrophic failure of an aircraft in 1969, engineers determined that the high strength D6AC had such low fracture toughness (i.e. was brittle) that critical sized cracks for the wing loading conditions were smaller than could be detected with current NDT methods. This led to a grounding action of the F-111 fleet until a reliable and safe method of inspecting the aircraft could be determined. General Dynamics developed a cold proof test to evaluate each aircraft for the presence of a defect significant enough to cause a failure in service (see Figure 7). The test was based on the fact that extreme cold

Figure 7: F-111 Cold Proof Test at RAAF Amberley (Courtesy of L. Molent, DSTO) significantly reduces the normal fracture toughness of steel. By reducing the fracture toughness of the steel and applying sufficient load to the aircraft in a fixture, it was possible to assure that no flaws of critical size were present in the structure. The presence of such a flaw would lead to a catastrophic failure (essentially the wing breaking off) during the testing and not during operation. Over the course of the life of the F-111 fleet with the US Air Force and Royal Australian Air Force, eleven aircraft failed the cold proof test but none were lost in service to this failure after its implementation.[7] The RAAF retired the aircraft in 2010.

Looking For Trouble

Every day in the life of an aircraft it gets one day older. Even when not being actively operated, corrosion can still be a consideration. With aging aircraft, sometimes beyond original design service lives, every day can be an adventure – blazing a new trail. As was mentioned before, developing a set-piece, established inspection for all eventualities is impractical for many reasons, often because we are simply unaware. This requires a unique attitude and outlook to the business of NDT. Success requires an open-minded and opportunistic approach that seeks maximum situational awareness of the aircraft’s condition, regardless of the method. New technologies are always of interest, but they must be harshly judged with an eye for practicality. Speed and actionable information can often

50

be key in decision making. Techniques and methods that can survey large areas quickly are highly sought after, allowing follow up of specific areas with other more exact techniques later. At times, we stumble on a new way or capability when we least expect it. In testing Thermal Wave Imaging’s (TWI) thermography system, the aim was to find the differing heat signature of corrosion byproduct trapped in the layers of lap joints on the C-135. However, in the course of evaluation, the system also found the discontinuity in heat conduction from broken spot welds (see Figure 8), an opportunity for even more situational awareness.

Broken Spot Welds

Figure 8: Thermographic Images Showing Broken Spot Welds (Courtesy of Thermal Wave Imaging)

CONCLUSION

The uncertain world of aging aircraft is a natural environment for a free-thinking partnership of NDT technician and sustainment engineer. The NDT and repair manuals are only the starting point for effective support of depot and fleet operations. Publications and procedures cannot always keep up with the entropy that is deterioration of aging aircraft. Awareness of the “sins of the past” gives a better feel for potential problems – the “unknown” unknown, all the while continuing to support the standard maintenance that is necessary for continued operation.

REFERENCES 1. 2. 3. 4. 5. 6. 7.

Compart, Andrew. Africa and Caribbean Home to Oldest Narrowbodies Aviation Daily. August 13, 2012. Compart, Andrew. Data Indicate That Age of Retired Commercial Jets Is Stabilizing. Aviation Daily. May 21, 2013. 2014 USAF Almanac. Aircraft Age. Air Force Magazine. May 2014 United States Air Force Scientific Advisory Board. SAB-TR-11-01. Sustaining Air Force Aging Aircraft into the 21st Century. Dated 1 August 2011. National Research Council. Examination of the U.S. Air Force’s Aircraft Sustainment Needs in the Future and Its Strategy to Meet Those Needs. Washington, DC: The National Academies Press, 2011. National Transportation Safety Board. AAR 89-03. Aircraft Accident Report – Aloha Airlines, Flight 243, Boeing 737-200, N73711, near Maui, Hawaii, April 28, 1988. Dated June 14, 1989. Redmond, Gerard. From “Safe Life” to Fracture Mechanics – F111 Aircraft Cold Temperature Proof Testing at RAAF Amberley.10th Asia-Pacific Conference on Non-Destructive Testing. September 17-21, 2001.

Broken Spot Welds

51

AWS D1.1 and D1.5 Phased Array Examination AWS D1.1 and D1.5 Phased Array Examination Parrish A. Furr ParrishofA.Ultrasonics Furr University 2159 Rocky Ridge Road – Suite 103 University of Ultrasonics Hoover, Alabama 2159 Rocky Ridge Road – Suite 103 (205) 822-5203 Hoover, Alabama (205) 822-5203

INTRODUCTION

Phased Array Ultrasonics (PAUT) is an expanding and growing technology which provides an alternative means of volumetric inspection. From its industrial origins in Nuclear Power in the 1980’s to today, the technology has become recognized as a viable alternative for determining asset integrity in several industrial sectors. Neither the current American Welding Society (AWS) D1.1 nor D1.5 Codes specifically address PAUT. The AWS D1.1 code addresses “Advanced Ultrasonic Systems” in Section 6 Part F, but provides no direct reference to PAUT which leaves questions regarding how to implement these types of systems. AWS has recognized the benefits this technology can offer and the need for a specific direction. The committee is currently working to incorporate requirements for PAUT into an annex of the D1.1 and D1.5 Codes to provide users with specific guidelines on implementation and evaluation techniques. This paper will address current ways to implement PAUT to AWS D1.1 or D1.5 and provide some detail to the direction of the proposed AWS Phased Array annex.

PHASED ARRAY OVERVIEW

Phased Array Ultrasonic Testing (PAUT) is an advanced form of ultrasonic testing evolved from the medical field with initial experimentations in nondestructive testing beginning in the early 1980’s1. PAUT is the same technology used today in several medical assessments such as sonograms and echocardiograms. In the industrial section, the majority of the applications for PAUT were at first limited to nuclear pressure vessels (nozzles), large forgings, shafts, and low pressure turbine components.1 The versatility of the technology provided alternatives to these more complex inspections. The limitations which held the technology back from growing in the industrial sector related mostly to equipment costs, computer speed, computer size, and data storage limitations. With the evolution of computer technology, these systems are now available in battery-powered portable units which can easily be taken into even the toughest field conditions and at more practical costs. PAUT implements multiple element transducers, rather than single element transducers typically used in conventional ultrasonics, providing the ability to sweep the sound field through multiple angles (Sectorial Scans – see Figure 1) or raster through a series of the same angle from one probe (Electronic Scan – see Figure 2). The systems can be configured to collect data from multiple probes in sequence to optimize data collection from opposite sides of the weld or multiple positions from the same side of the weld in one scan (see Figure 3). These modern advanced computerized systems have the ability to store all the collected information from the weld into a data file which can then be reanalyzed after the acquisition of data is completed. Figure 1: Phased Array Sectorial Scan.

52

Figure 2: Phased Array Electronic Scan (E-Scan).

Figure 3: Multi-Probe Data Collection.

The benefits of this technology also extend to the system’s ability to form the collected data into 2D and in some cases 3D imagery to assist the inspector in analysis. There are several versions of software which vary in complexity, display and processing options. The most commonly used displays consist of A-Scan (source of all other displays), S-Scan (cross-sectional/side view), B-Scan (back view), and C-Scan (top view). These views allow the operator to sift through the collected information efficiently and help to build confidence regarding locations and types of flaws. The perspective of the common display options are shown in Figure 4. Figure 4: Typical PAUT Imaging Views.

S

C

A

B

APPLICATION OF PAUT TO CURRENT AWS DB RATING SYSTEM

Both AWS D1.1 and D1.5 were developed with Conventional Ultrasonics in mind and the standard method of implementation and evaluation is based on basic angles of 45, 60, and 70 degrees. Both Codes also utilize a dB rating system which was derived in 1969 off of a single point calibration in the International Institute of Welding (IIW) Standard. The rate of sound attenuation in materials associated with these Codes, which is only Carbon Steel, was deemed constant so a standard formula was used to calculate the equivalency of an indication regardless the sounds path distance away from the probe.

53

Current PAUT systems can be made to adapt to the rules of D1.1 and D1.5 Section 6 methodology by using E-Scans at an angle(s) in accordance with the procedure chart detailed in Table 6.1 and with the aperture size controlled to maintain the physical size requirements, typically 5/8” x 3/4”. The necessity of E-Scans to fit the current methodology however provides complications in this application. The complications are primarily due to larger probe apertures required for E-Scans (60 or more elements and larger pitch sizes which reduce quality) and reduced resolution of E-Scans at higher angles (70 degrees) which are commonly used in the Codes. In additional to that above, other complications exist when trying to encode data using the dB rating system which has forced current users to find an alternative means to perform PAUT under the current Code.

APPLICATION OF PAUT TO CURRENT AWS D1.1 ANNEX S

AWS D1.1 Annex S provides an alternative and more conducive avenue for applying PAUT in accordance with the current code and may be used with Engineers Approval. Acceptance of its use is based upon the Engineers approval of the written procedure which must be qualified on a mock up block to prove detection and sizing capabilities to the satisfaction of the Engineer. Annex S covers “UT Examination of Welds by Alternative Techniques” and was created with the intent to provide users a method for inspecting materials outside the limitations of the main body of the Code. The most common industry use of this annex is for inspection of materials below 5/16” or above 8”, which is the limitation of Section 6 Part F of D1.1. PAUT has been found to adapt well to this methodology and has been approved and used successfully on several projects to date. Annex S utilizes a Distance Amplitude Correction (DAC) based inspection methodology which is much more conducive to typical PAUT setups and configurations. Because DAC based system measure sound losses through the material, additional aperture and frequency options are available for use. This gives the PAUT user the ability to use S-Scans, which can be configured with smaller footprint probes and enhanced resolution over the E-Scan counterparts. The other advantage of this DAC based system is that it allows for the conversion to and use of Time Corrected Gain (TCG) which is more adequate for PAUT inspection systems, especially when encoding the data. TCG adds decibels to the returned signal at specified time intervals which equalizes the amplitude response off of a given flaw size at various depths. This results in the calibration reflector yielding a response of approximately 80% Full Screen Height (FSH) regardless of depth in the material or the configured angle. This balance in color is important to detection capabilities of PAUT since flaw detection is based on the color and color is based off of amplitude. This also simplifies the evaluation as will be discussed later in this section. A correlation from the Annex S DAC illustration and the corresponding TCG calibration detailed below is shown is Figure 5 below. Figure 5: DAC to TCG Correlation.

ARL +5 dB SSL DRL (-6 dB) DAC

TCG

54

Calibration to Annex S is based off the 0.06” (1.5mm) diameter side drilled hole as in the International Institute of Welding (IIW) Standard. The reflector may be placed in a specially designed calibration block, weld mockup or in a production piece. Carbon Steel PACS® type blocks, as shown in Figure 6 with 0.06” (1.5mm) diameter side drilled holes at multiple depths, can be used to establish the TCG through entire thicknesses range to be tested and provide relatively easy calibration reduced potential disturbances from other nearby holes or reflectors. Blocks of the same material with other hole diameters can still be used for basic linearity and TCG calibration, but final reference level sensitivity must be established off the 0.06” (1.5mm) diameter hole. Figure 6: PACS® Block.

Once the TCG is established, the 0.06”(1.5mm) diameter hole is maximized and placed at 50% FSH which establishes the Standard Sensitivity Level (SSL) of the test system as illustrated in Figure 5 above. This is a slight modification from the standard reference level threshold establishment of 80% FSH, but modified to allow all relevant thresholds to be visible on the screen at reference level dB. With a proper TCG calibration, all focal laws will equal approximately 50% FSH off of the reference reflector regardless of depth in the material. An additional Disregard Level (DRL) is then specified to be set at 6 decibels below the SSL which equates to 25% FSH and is illustrated in Figure 5 above. The acceptance criteria of Annex S call for rejection of all indications exceeding 5 dB over SSL regardless of length. Using the basic dB ratio formula of [dB = 20 log (h1/h2)], we can calculate this corresponding screen height division of 89% FSH and establish an additional threshold mark we’ve called the Automatic Reject Level or ARL and illustrated in Figure 5 above. CURRENT INDUSTRY ACCEPTANCE PAUT fits AWS D1.1 Annex S very well as described above which makes Engineer approval obtainable, provided a proper procedure and qualification is in place. The Bridge Welding Code, AWS D1.5, however does not have a similar Annex and currently only addresses conventional ultrasonic techniques. With that stated, Engineer approval for using PAUT has been successfully obtained in some D1.5 projects and using the Annex S methodology detailed in AWS D1.1. In any case, we must be mindful when trying to implement PAUT to either AWS D1.1 or D1.5 that there are still specific variables which are not detailed in the current Code. Fundamentals related to scan plans, specific focal law requirements, data collection requirements, and data integrity are not specified in the Annex but should be specified in the procedure by a UT Level III with PAUT experience and training. Due to this lack of specific information, the qualification process required in Annex S is of paramount importance to assure the relevant reflector size can be detected in the given joint being inspected. Qualification is then used to verify that the inspection system works as designed and configured. There had been concerns in the AWS community with regard to the comparison of PAUT with the two common volumetric inspection methods commonly used, radiography and conventional ultrasonics. Research performed by the University of South Florida and sponsored by the Florida Department of Transportation showed that the overall reject rate for PAUT using Annex S criteria was nearly identical to that of radiography and ultrasonics over a reasonable sample size. In addition, cost savings between 2 to 4 million dollars per year were estimated by substituting PAUT for Radiography. Table 1 below shows some of the data taken from the research as related to reject rate of discontinuities on two separate projects.

55

Table 1: Reject Comparisons of RT versus PAUT2.

Technique RT PAUT UT

2 Bridge Welding Projects2 # of Welds Tested # of Rejects 108 10 92 8 54 4

% of Rejects 9.3% 8.7% 7.4%

PROPOSED AWS PAUT ANNEX

AWS is currently seeking adoption of a PAUT specific annex into the D1.1 and D1.5 Codes. A great deal of research has gone into finding a methodology and acceptance standards which best suites the technology and the AWS D1.1 and D1.5 expected quality levels. The end goal is to define the essential parameters and acceptance criteria to eliminate ambiguity in regard to PAUT application and allow users to access the technological and cost savings benefits under these defined set of rules. The initial research for the proposed Annex began with trying to make PAUT work with the basic conventional acceptance criteria. E-Scans at the basic 45, 60, and 70 degree angles were experimented with initially with the attempt to duplicate the current main body conventional techniques as closely as possible. The disadvantages noted in the previous section of this paper in regard to probe aperture size requirements, resolution issues, and complications of encoding forced us to take another avenue. Additional research was performed with S-Scans and the conventional dB rating acceptance to help eliminate the large aperture and resolution issues of the E-Scans. Patterns were identified in the AWS D1.1/D1.5 acceptance tables between 70, 60, and 45 degree angles. Extrapolation of indication ratings between the common angles (i.e., 61 to 69 degrees) were experimented with and showed to be somewhat comparable but new issues arose with complications of adapting common PAUT calibrations and assuming attenuation rates at the intermediate angles. Since Annex S in AWS D1.1 is already a code approved, yet alternate inspection technique, and the technology better fits that typically used in PAUT, it was chosen as the direction to take with the current proposed PAUT annex. Numerous comparisons were made by the University of Ultrasonics comparing each proposed methodology as it relates to AWS conventional ultrasonic acceptance. This research helped to verify that the PAUT version of the Annex S technique nearly duplicated the inspection results obtained from the conventional UT inspection on the same weld discontinuities. A subset of these comparisons can be seen in Table 2. Table 2: AWS Acceptance Criteria Comparison.

Table Notes: 1. Red = Differs from the conventional assessment – rejected an acceptable discontinuity 2. Green = Differs from conventional assessment but would be identical if rejected on crack characterization.

56

A draft version of a PAUT annex is currently working its way up through the AWS committee approval process with hopes of adoption into the 2015 AWS D1.5 Code and later in the AWS D1.1 Code. The proposed annex is following the lines of D1.1 Annex S methodology as described in the Application of Annex S section above as is designed for the encoded line scanning. Additionally, Sectorial Scans are being proposed as the primary inspection technique with E-Scans allowed for supplemental coverage. The proposal also includes a requirement of Time Corrected Gain (TCG) calibration off a 0.06” diameter side drilled hole and details that all focal laws must be calibrated. The proposal is to use the same basic Annex S acceptance tables with terminology modifications to make it more familiar to conventional AWS UT inspectors as shown in Figure 7 and Table 3. The proposed annex also looks to remove procedure qualification requirements in which costs would often deter some users from implementing the technology. To allow the elimination of qualification, users would have to follow a very specific method of configuration and scanning requirements as detailed in the annex. This includes specific focal law configuration requirements, aperture size restrictions, beam sweep restrictions, and joint coverage requirements among others. Figure 7: Proposed AWS Calibration/Setup.

A

ARL (5 dB over SSL)

B SSL

C D

Table 3: Proposed AWS Acceptance Criteria.

57

DRL (6dB under SSL)

Maximum Discontinuity Amplitude Level Obtained Class A (Greater than ARL) Class B2,3 (Between SSL and ARL)

Maximum Discontinuity Lengths by Weld Stress Category 1 Statically Loaded

Cyclically Loaded

None allowed

None allowed

3/4 in [20 mm]

1/2 in [12 mm]

Class C2,3,4 (Between SSL and DRL)

2 in [50 mm]

Class D (Equal to or less than DRL)

Middle half of weld: 2 in [50 mm] Top or bottom quarter of weld: 3/4 in [20 mm]

Disregard

Disregard

CONCLUSION

PAUT is a viable alternative to radiographic or conventional ultrasonic inspections on structural components. Although the technology can be adapted to the current version of the AWS D1.1 and D1.5 Codes, the lack of detailed information could cause substantial variance from one project to another. The adoption of specific requirements into the AWS D1.1 and D1.5 Codes will be paramount to assure that users understand the proper way to achieve quality results and provide Engineers a level of comfort in allowing users to opt for the technological benefits. Like with any new technique or application, education regarding the technology from all sides of the table will be a vital part of assuring quality is maintained or elevated. Phased Array is a powerful tool, but one that must be carefully carried out to assure that insufficient or excessive quality levels are not a result.

REFERENCES

1. 2. 3. 4.

Olympus NDT, “Introduction to Phased Array Technology Applications” Steve Duke, CPM & Stuart Wilkinson, PhD – University of South Florida “Advanced Ultrasonic Testing NonDestructive Testing Techniques in Accordance with the AWS D1.5 Bridge Welding Code”, 2014 ; AWS D1.1:2010 “Structural Welding Code – Steel” AWS D1.5:2010 “Bridge Welding Code”

58

Automatic Conductivity Conductivity Scanning Scanning of of Rolled Rolled Aluminum Aluminum Plates Plates for for Automatic Aerospace Applications Aerospace HormozGhaziary Ghaziary11,,Wayne Wayne Johnson Johnson22,, and and Alfred Alfred Haszler Haszler33 Hormoz 1 Advanced NDE Associates 1 NDE Associates 3474Advanced Voyager Circle, San Diego, CA 3474 Circle, Diego, CA (858) Voyager 350 8630; e-mailSan [email protected] (858) 350 8630; e-mail [email protected] 2 Castro Valley, CA 94552 2 Castro Valley, 94552 (510) 282 4855; e-mailCA [email protected] (510) 282 4855; e-mail [email protected] 3 Altech

Consulting, Koblenz, Germany Altech Consulting, Koblenz, Germany e-mail [email protected] e-mail [email protected]

3

INTRODUCTION

The use of eddy current techniques for the measurement of electrical conductivity has long been a major test requirement for rolled aluminum plates destined for aerospace applications. The major reason for conductivity testing is to detect “soft spots” that may occur as a result of improper heat treatment or macro-segregations initiated during casting process. Aerospace aluminum plates are typically made of “heat treatable” aluminum alloys, e.g. Al-Zn (7xxx), Al-Cu (2xxx) and Al-Si-Mg (6xxx). In order to impart optimum mechanical properties, plates are solution heat-treated. This process consists of heating the plate to an appropriate temperature/time cycle which puts the soluble elements in solid solution. This is followed by rapid quenching, usually in water, which oversaturates the structure followed by precipitation kinetics, termed “aging”. The process of quenching the plate is done by a matrix of nozzles which spray water over the entire area of the plate from top and bottom, as it moves over horizontal rolls. The next step is stretching or cold compression after which the plate is kept at a low temperature/time cycle to accelerate precipitation hardening - termed artificial aging- so that the desired and stable mechanical and corrosion properties can be achieved.

Figure 1: Production route for heat-treatable rolled aluminum plate. As shown in figure 1, electrical conductivity of a heat-treatable plate is measured twice during the production process for different purposes. After stretching, conductivity measurement is done (on both sides of the plate) to verify the uniformity of heat treatment- quenching over the entire plate and detect any possible soft spots. Also, areas of different chemistry and structure i.e. macro-segregation are detected at this stage. After the completion of artificial aging, when the mechanical properties of the plate are fully stabilized, it is once more tested for conductivity (on one side) to verify that a proper temper has been achieved.

TEST PRINCIPLES

Electrical conductivity of aluminum –as well as other non-ferromagnetic metals- is sensitive to variations in its physical and mechanical properties, e.g. there is a good inverse correlation between mechanical properties and

59

electrical conductivity of a specific aluminum alloy. Of particular interest in production of rolled aluminum plates are: 1) variations in chemistry and structure to detect macro-segregation, 2) effects of thermal processing, and 3) effects of mechanical processing. The influences of the above factors on material conductivity are significant enough to make for useful and reliable test applications. A well established technique for measuring conductivity of aluminum plates is the use of eddy current test instrument. Conductivity measurement using eddy currents is based on the fact that the impedance of a test coil varies with the conductivity of a nearby material. As shown in fig. 2, the magnitude of impedance decreases with increasing conductivity. In fig. 2, the inductive reactance of the coil is plotted on the Y axis and the coil’s resistance on the X axis. The 0% conductivity point represents the probe in the air; it also represents maximum reactance. 100% represents the conductivity value for International Annealed Cupper Standard (%IACS). This relationship is used for conductivity measurement by calibrating the output of an eddy current instrument. Eddy current signal can be displayed in X-Y format with X being the resistive component and Y being the reactive components of the signal. Using a set of calibration standards, both X and/or Y axes may be scaled for a range of conductivity values.

Figure 2: Measured conductivity relationship in MS/m (%IACS).

Test Procedure

All test practices for conductivity measurement of aluminum plates, set forth by specification societies and aerospace customer requirements, are based on spot measurements in a specified grid over the plate. For example, a 7050-T7451 plate is tested every 500 mm in length direction and every 100 mm in the width direction, on both sides, immediately after solution heat treatment, quenching, and stretching. After the completion of artificial aging, the same plate is tested again, for temper verification, in a grid of 100mm by 100 mm but only from one side. Typically, a hand held EC instrument is used and the test results are recorded in a data sheet along with their respective positions. While this is quite practical for a small production environment, it becomes quite labor intensive in a busy production environment, especially where larger plates are produced. Today’s aircraft designs

60

necessitate the use of long and wide aluminum plates. For example plates intended for aircraft wing structure could be well over 25 meters long and over 2 meters wide. For a measurement grid of 100mm by 500 mm, more than one thousand measurement points must be made, on each side. Thus, it is not hard to imagine the amount of time needed to mark the plate for each measurement point, perform one thousand conductivity measurements, and record the results. Furthermore, the plate must be turned over for testing the back side and this requires special equipment. Another important issue with manual conductivity measurement is related to the number of measurement points (distance between measurement points in length and width direction). What is prescribed by specifications is not intended for 100% inspection, i.e. full coverage of the plate; it’s simply a sampling procedure based on which the uniformity of thermal processes can be estimated. Yet at the same time, the plate manufacturer must guarantee the uniformity of the distribution of conductivity values within 1.5% IACS over an entire plate. As such, there is always a possibility that a defective area falls between two measurement points and remain undetected while all required conductivity tests have been done in accordance with the most stringent specifications; the plate manufacturer will be responsible for all consequences.

Automatic Conductivity Scanning

The idea of automating conductivity measurement on aluminum plates and the ability to produce a conductivity map as a useful analysis and process control tool has been entertained by a few rolling mills in the past two decades. However, this has resulted in only limited action due to issues arising from the extreme sensitivity of eddy currents to liftoff (distance between the sensor and the plate surface) which becomes a major factor when scanning a plate. The authors describe a prototype conductivity measurement system in which conductivity values are dynamically corrected for liftoff while the plate is being scanned from top and bottom simultaneously. Besides performing standard testing procedures, the capability of the system to produce a conductivity map makes it a powerful quality and process control tool to analyze trends of conductivity variation related to heat treatment e.g. quenching nozzles, as well as casting issues such as micro segregation.

Figure 3: Automatic Conductivity Scanner

CHALLENGES AND SOLUTIONS

A picture of the conductivity scanner is shown in fig. 3. Plates to be tested move on a roll table, through the scanning frame. Upper and lower probe assemblies are attached to the scanner frame. The probe assemblies come into contact with the plate and perform a raster scan. Conductivity measurement is done on both upper and lower sides of the plate simultaneously. The position of each measurement reading is registered by an encoder wheel and motor encoders in longitudinal and transverse directions respectively. The system is controlled and operated by a PC through an interface software specifically created for this purpose. The major challenges to deal with in building an automatic scanner are mostly related to the extreme sensitivity of eddy current signal to “liftoff’ i.e. the distance between the measurement coil and the test object i.e. aluminum plates. In manual testing, the probe is placed directly over the plate with zero nominal liftoff. As such, sensitivity to liftoff distance is negligible. In an automatic scanner, on the other hand, the probe cannot come into contact with the plate and must maintain a nominal distance from it during scanning. At test frequencies suitable for conductivity measurement of aluminum (50-60 KHz), the sensitivity of eddy currents to conductivity variations diminishes

61

exponentially as liftoff increases and is practically limited to about 1 mm or 0.04” liftoff distance. A liftoff distance of 0.5 mm (0.02”) has been found practicable. However, one must take into account the fact that rolled aluminum plates are not always completely flat. For example, according to aerospace specifications transverse flatness can be usually up to 0.2 % of the width, i.e. up to 4 mm for a 2 meter wide plate; with additional allowances for short cycle flatness. Obviously, maintaining a constant liftoff distance of 0.5 mm would not be possible without mechanical adjustments.

Figure 4: Scanning Head and probe assembly upper head (a) and lower head (b).

Scanning Head

In order to mechanically maintain a constant liftoff distance, the eddy current probe is inserted in the center of a polymer disc as shown in fig. 4. In order to mechanically maintain a constant liftoff distance, the eddy current probe is inserted in the center of a polymer disc as shown in fig. 4. The scan head rolls on high tolerance bearings with a preloaded force on the plate. The heads are free to articulate in two axes so that they can follow and conform to the small variations in surface flatness of the plate. The head preload, and rotational freedom, help to keep the measurement probe with in a viable liftoff range for reading conductivity.

Correction for Liftoff Variations

Despite the improvements achieved by mechanical design, small liftoff variations can still occur. In order to optimize measurement accuracy it is necessary to correct measured conductivity values for liftoff variations. High accuracy conductivity measurement requires that the actual probe liftoff be measured, and corrected for, before the actual conductivity is measured. This can be done by establishing conductivity-liftoff relationship for the appropriate conductivity range. Figure 5 shows a family of liftoff curves for conductivity range of 18-60% IACS. These curves are produced using certified conductivity standards. Furthermore, for the area between curves a mathematical procedure is used to simulate unlimited number of liftoff curves. As such, a Conductivity-liftoff data pair is registered for every measurement point during conductivity scanning.

62

Figure 5: Variation of Conductivity due to liftoff.

Correction for Temperature Variations

The response of electrical conductivity measurement instruments is highly sensitive to ambient temperature variations. For test purposes, the eddy current instrument must be turned on and placed in the actual test environment to warm up for at least an hour. Furthermore, the temperature of the test piece and calibration standards must stabilized in the test environment before actual tests can take place. Most conductivity measurement instruments provide tools for correcting for ambient temperature. When using an automatic scanner and thus using a liftoff distance, the effect of temperature is further complicated as sensitivity to temperature varies with liftoff and conductivity. Therefore the liftoff curves shown in fig. 5, must be established for various temperatures, typically from 5oC to 30oC at 5oC intervals.

CONCLUSIONS

The advantages that the automated conductivity scanning system can offer (based on 6 months of application in production environment) are listed below.

Routine Operation 



 

Measure accuracy of dynamic measurement was found similar to that of spot measurements made by a properly calibrated portable EC instrument. Scanning speed in the width direction is about 150 mm (6”) per second at which a smooth rolling of the scan head over the plate can be maintained. The time needed to perform a complete conductivity scan is significantly shorter than that of manual testing. E.g. testing a 2x10 meter plate (on both sides) takes about 10 minutes including test report generation. The system can be operated by one LII operator. The system generates a C-scan type display showing conductivity values at any given measurement position as shown in figures 6 and 7. All of the appropriate data for top and bottom surfaces of the plate are constantly updated during the scan process.

63

Figure 6: Conductivity scan of a 1000mm by 3200 mm 7050 plate. Measurement increment is 100 mm by 100 mm.

Figure 7: Conductivity scan of a 1000 mm by 3200 mm 7050 plate. Measurement increment is 100 mm by 25 mm.

64

The Use of the Automatic Conductivity Scanner as a Tool for Process Control

Besides measuring conductivity in compliance with customer specifications, the system can also be used as a tool for process control as it can display conductivity variations in greater detail. Acquiring measurement data in smaller increments, especially in the width direction may reveal variations that could have been missed if larger increments were used. For example, the plate scan shown in fig. 6 is taken in 100mm by 100 mm increments. The same plate is re-scanned with 25 mm increments in the width direction as shown in fig. 7. Bands of lower conductivity in the length direction which couldn’t be detected at 100 mm increment, may well be an indication of clogged or malfunctioning quench nozzles. Even if the general conductivity variance is within acceptable level, i.e. 1.5% IACS, such information may be used to correct the problem and avoid more serious issues. This enables the plate manufacturer to establish robust processes by early identification of process deviations and their correction. Being able to acquire conductivity date in smaller increments raises an important issue. Typically, the measurement grid is specified by aerospace customers and the related specifications, based on which, maximum, minimum, and variance for conductivity values over a plates is determined and an accept-reject decision is made. An example is 100 mm LT and 500 mm L. However, it is quite possible that a plate that has been found acceptable with this measurement grid (< 1.5% IACS), fail if smaller measurement grid is used. It is therefore important to know that the plate manufacturer is ultimately responsible for the uniformity of conductivity distribution, regardless of the required number of measurements.

REFERENCES 1.

2. 3.

Blitz J. “Electrical, Magnetic, and Visual Methods of Testing Materials” Vol. 2 “Eddy Current Methods” Butterworth, 1969. ASTM E1004-09 “Standard Test Method for Determining Electrical Conductivity Using the Electromagnetic (Eddy Current) Method” MIL-STD-1537C “Test method Standard for Electrical Conductivity test for Verification of Heat Treatment of Aluminum Alloys Eddy Current Method”, June 25 2002

65

ASNT Fall Conference 2014 October 27 - 30, 2014, Charleston, In-Situ Nondestructive Positive Material Identification Testing SC, USA For Determining Carbon Steel Pipeline Material Properties

IN-SITU NON-DESTRUCTIVE POSITIVE MATERIAL IDENTIFICATION Kenneth J. Greene1 and Greg Donikowski2 1 TESTING FOR DETERMINING T.D.CARBON Williamson, Inc.STEEL PIPELINE MATERIAL 2256 N. Pagosa St.#200 PROPERTIES Presenting Author:

Aurora, Colorado, U.S.A. (303) 884-5614; e-mail [email protected]

Co-author:

2 T.D. Williamson, Inc. Kenneth J. Greene Greg Donikowski 6747 S. 65th West Ave. T.D. Williamson, Inc. T.D. Williamson, Inc. Tulsa, Oklahoma, U.S.A. 2256 N. Pagosa St.#200 6747 S. 65th West Ave. (918) 200-4422; e-mail [email protected] Aurora, Colorado, U.S.A. Tulsa, Oklahoma, U.S.A. (303) 884-5614 (918) 200-4422 [email protected] [email protected]

ABSTRACT: TDW has developed and validated a non-destructive Positive Material Identification test method used to identify unknown carbon steel line pipe in the oil and gas industries to determine material grade based on the American Petroleum Institute 5L Specification. This process identifies both Yield and Tensile strength values along with chemical analysis and carbon equivalence. When these PMI methods are used for material evaluation, within the scope, limitations and guidelines of the validated PMI procedures, accuracy tolerances have been validated as follows:  Ultimate Yield Strength (UYS) +/-10% with a 95% Confidence Level  Ultimate Tensile Strength (UTS) +/-10% with a 95% Confidence Level  Carbon percentage (C) +/-25% with an 85% Confidence Level  Manganese percentage (Mn) +/-20% with a 90% Confidence Level Our process requires exposing a minimum three feet section of pipe and coating removal for 360 degrees around the pipe. Prior to this development, customers generally had to rely on destructive test methods to determine these values which may have included a line shut-down, destructive removal of a test coupon and off-site laboratory testing. This process was time consuming, expensive and labor intensive (often requiring a line by-pass). This PMI procedure is truly non-destructive and is performed in-situ, providing material property assessment without pressure reductions, while the line remains in service.

NOMENCLATURE Automated Ultrasonic Testing – AUT Department Of Transportation - DOT Engineered Tensile Strength - ETS Engineered Yield Strength – EYS Instrumented Indentation Testing - IIT Integrity Verification Process - IVP Non-Destructive Evaluation – NDE Magnetic Particle Testing – MT Material Test Report - MTR

66

Mechanical Properties Assessment – MPA Optical Emissions Spectrometry - OES Phased Array Ultrasonic Testing – PAUT Pipeline Hazardous Material Safety Administration - PHMSA Positive Material Identification – PMI Remaining Wall Thickness - RWT Ultrasonic Testing – UT Ultrasonic Thickness Testing – UTT Visual Testing - VT

INTRODUCTION Herein discussed is a Proprietary Analytical Method as described in ASTM A751-Sect.3.1.1.4. Conventional laboratory test methods provide a means of determining Yield Strength, Tensile Strength, Chemical Analysis & Carbon Equivalence material property values. In the pipeline industry, great time and expense is incurred for the cut-out of materials that can provide coupons for laboratory testing. Additionally, great efforts have gone into developing alternative testing techniques that allow sub-sized samples that have been removed from hot-tap coupons to be used as laboratory test specimens. Yet both of these options still remain ‘destructive’ test methods. They both require a line shut-down, interruption of service or a doubleSTOPPLE® bypass (Figure 1), a cut-out and welding of replacement pipe or an in-service hot-tap fitting and associated welding which sacrifices the line integrity at the coupon removal location. These coupons are then sent off-site to a testing laboratory where results are often delayed for several days or weeks.

Figure 1: Double-STOPPLE® bypass illustration The need for truly ‘non-destructive’ testing, which provides material properties assessments engineered to be equivalent to conventional destructive laboratory testing or certified mill test reports, prompted an internal research project that began in Q1-2011.

Challenges “Out of the box” utilization of surface Instrumented Indentation Testing (IIT) Mechanical Properties Assessment (MPA) equipment provided inconsistent and scattered test data results when applied to API-5L carbon steel line pipe. A process was needed that provided consistent results, mathematically culling poor outlier data and accounting for skewed data due to variables in non-homogenous material from conditions such as carbon segregation, pearlite interlamellar spacing and grain size variations.

67

Searching the industry for equipment with the potential to be modified or adapted for use within a particular industry and for specific applications can be challenging. We found our most effective results when working with manufacturers who are most willing to understand your vision for their equipment and work at modifying their equipment until the particular application is realized. Figure 2 illustrates generational improvements over multiple process revisions which included modifications to equipment, software, parameter controls, techniques and procedural specifications.

Figure 2: Generational progression of MPA test data accuracies Pipe wall cross-sectional variability must be understood, located, identified and avoided as potential test sites. Material preparation must account for surface variables and prevent the contamination of the selected test locations. Figure 3 is an AUT screenshot image of a selected test location, which upon evaluation requires the re-location due to an internal wall loss anomaly.

Figure 3: AUT imagery of proposed test site location requiring relocation Additional challenges include the recognition and acceptance, by governing or code development agencies, of new technologies, new applications of existing technologies or newly developed processes.

Validation The establishment of tolerance ranges and confidence levels are the result of ongoing validation processes that will include confirming the validity of manufacturer’s claims. As the manufacturer works with the end user to develop applications outside the original intent of the product design, the end user will likely perform the R&D roll for the manufacturer. Providing materials, test data and feedback to the manufacturer is an important part of a thorough validation process. This would include in-house, as well

68

as 3rd Party results being compared with certified mill test reports. Data is then analyzed and scrutinized against conventional laboratory results during the Referee Analysis as per ASTM A751-Sect.3.1.1.5 Historical validation data must be tracked to establish stated tolerance ranges and confidence levels. Equipment manufacturers must be willing to continue evolving their equipment until these tolerance ranges are acceptable and useful for the particular application within the desired industry. The validation process must be documented for each component and step within the process. The ultimate validation in the process discussed, is that it remains ‘non-destructive’. A non-destructive test, in the terms used herein, is a test that provides the stated engineered material properties through a ‘Certified Positive Material Identification Report’, which is equivalent to (within the stated tolerances) those that would otherwise be provided through a certified mill test report or through conventional destructive laboratory testing results. Any potentially detrimental surface indentions and OES burns are removed and the Remaining Wall Thickness (RWT) remains within stated allowable mill tolerances as specified within API-5L-Table 11. Figure 4 demonstrates a portion of the OES burn depth validation process. OES burns were sent to an independent testing laboratory where OES burn cross-sectionals were Macro-etch with the maximum depth measured. This allowed confirmation of OES equipment manufacturer claims, established a depth target for OES burn removal and confirmed the Chemical Analysis and Carbon Equivalence portion of the PMI process to be non-destructive.

Figure 4: OES Burn Depth Validation: 3rd Party Laboratory Macro-Etch

Solutions Overcoming the problems with inconsistent data began with the selection of manufacturers who were willing to engage in the endeavor of new possibilities for the equipment and willing to working through the R & D process which was being generally directed by an end user service provider company. Compiling a host of material samples with a wide range of material grades and various vintages, both traceable identified materials with MTR’s, as well as unidentified materials for use as blind tests, is necessary. Tracking and eliminating or controlling technician variables is an essential factor that will allow consistency in datasets to become a reality. Equipment modifications are generally in vain until all adjustable or variable parameters have been identified, monitored, measured and either controlled through procedural application or eliminated. Once parameters are identified and controlled, test datasets can be compared with MTR’s, 3 rd Party laboratory test results, vendor technician test results and in this case vendor laboratory test results. Ultimately an industry recognized Referee Analysis 3rd Party testing laboratory will generate the final analysis on the accuracy and validate the stated tolerance claims to be prescribed within the written procedure. The table in Figure 5 list “blind test” samples in which PMI was performed under the scrutiny of Kiefner & Associates acting as a Referee Analysis 3rd party laboratory.

69

Lab vs Lab vs TDW TDW Yield Tensile Diff Diff 3 35000 47900 69500 50540 65560 5.40% 5.80% 6 45000 52000 75000 52770 77850 1.50% 3.70% 13 45000 53500 74000 51480 72320 3.80% 2.30% 16 45000 57000 81500 54540 76760 4.40% 6.00% 18 45000 52000 73000 53380 67570 2.60% 7.70% 20 45000 51000 73500 54680 80350 7.00% 8.90% 30 35000 48500 69000 50530 77210 4.10% 11.20% *31 52000 63500 77000 54860 86490 14.60% 11.60% 32 45000 55000 75500 51850 78050 5.90% 3.30% *33 45000 63500 90000 58630 81870 8.00% 9.50% 35 45000 54000 74000 51740 76120 4.30% 2.80% *items in red were outside of the range of application for the PMI process

Pipe Sample

Specified Yield Strength

Lab Yield Strength

Lab Tensile Strength

TDW Yield Average

TDW Tensile Average

Figure 5: 3rd Party Testing Laboratory Referee Analysis Report The recognition and acceptance of new technologies, new applications of existing technologies or newly developed processes, by governing or code development agencies, can be a long arduous process. The development of the PMI NDE process provides obvious solutions to a specific industry. Providing regulator review teams with consistent progress updates of process improvements and 3 rd Party Referee Analysis results, has allowed this PMI NDE process inclusion into the D.O.T./PHMSA IVP DRAFT (Figure 6). Future pending approval of the PHMSA DRAFT IVP by the D.O.T. would put the reference to in situ NDE into 49 CFR 192 Regulation.

Figure 6: PHMSA DRAFT IVP – Now includes in situ NDE as alternative to destructive laboratory testing

70

SUMMARY Positive Material Identification or PMI as herein described, involves identifying the pipe’s coating, diameter and nominal wall thickness. MT is performed on exposed pipe for the detection of SCC or long-seam defects. Randomly selected variable test locations are chosen to account for non-homogenous materials. Once the areas are selected for testing, actual wall thickness (AWT) is determined by UTT. Ensuring that the selected test locations are free from laminations or internal mill anomalies is crucial, thus an AUT C-scan is performed. Careful and consistent preparation of the selected test location is critical to avoid bringing contaminants from the coating or even from the technician’s finger prints into the test area which would cause the collection of skewed data. MPA is performed with a minimum of 5 readings per dataset to collect EYS/ETS data. OES is performed with a minimum of 5 readings per dataset to collect Chemical Analysis and Carbon Equivalence. The MPA-IIT indentions, as well as the OES burns are removed. Etching is performed to ensure that the removal of the OES burns is complete. MT is once again performed at the selected test locations to ensure that the PMI process was non-destructive and non-detrimental to the pipe tested. A final UTT is performed to measure the final RWT, to ensure that the PMI process did not remove more material than API-5L-Table 11 ‘Tolerances For Wall Thickness’ allows for mill tolerances. VT & PAUT are performed to identify the long-seam weld type. A Certified Positive Material Identification Report is generated. PMI of Carbon Steel requires the use of equipment from multiple vendors and the utilization of Level II technicians qualified in VT, UT (UTT, UT, AUT & PAUT), MT, & In-house PMI (OES & MPA) certifications. This PMI procedure is truly non-destructive and is performed in-situ, providing material property assessment of API-5L line pipe without pressure reductions, while the line remains in service.

REFERENCES

1. 2. 3. 4.

API-5L Specification for Line Pipe ASTM-A751-14 Standard Test Methods. Practices, and Terminology for Chemical Analysis of Steel Products ASTM E415-08 Standard Test Method for Atomic Emissions Vacuum Spectrometric Analysis of Carbon and Low-Alloy Steel ASTM E2546-07 Standard Practice for Instrumented Indentation Testing

71

Tracing Defects in Glass Fiber/Polypropylene Composites Using Ultrasonic C-Scan and X-Ray Computed Tomography Methods Tracing Defects in Glass Fiber/Polypropylene Composites Using Ultrasonic C-Scan and X-Ray2 Computed Tomography Methods 1 2 3 1 Ahmed Arabi Hassen , Anish Poudel , Tsuchin Philip Chu , Michael Yester , and Uday K. Vaidya 1 Department of Materials Science Engineering Poudel2, Tsuchin Philip Chu2,and Michael Yester 3, and Uday K. Vaidya1 Ahmed Arabi Hassen1, Anish 1 Materials Processing & Applications Development (MPAD) Center, University Department of Materials Science and Engineering of Alabama at Birmingham 1150 10th Ave. South, Birmingham, 35249 of Alabama at Birmingham Materials Processing & Applications Development (MPAD) Center, AL University th (205) 934-8450; fax (205) 934-8485; e-mail [email protected], 1150 10 Ave. South, Birmingham, AL [email protected] (205) 934-8450; fax (205) 934-8485; e-mail [email protected], [email protected] 2 Intelligent Measurement and Evaluation Laboratory 2 Intelligent Measurement Evaluation Laboratory Department of Mechanical Engineering and and Energy Processes, Southern Illinois University Department of Mechanical 1230 Engineering Processes, Southern Illinois University Lincolnand Dr.,Energy Carbondale, IL 62901 Lincoln Carbondale, IL 62901 (618) 453-7049;1230 fax (618) 453Dr., -7658; e-mail [email protected], [email protected] (618) 453-7049; fax (618) 453 -7658; e-mail [email protected], [email protected] 3 Division of Physics and Engineering, Department of Radiology 3 Division of Physics and of Engineering, of Radiology University Alabama atDepartment Birmingham University of Alabama at Birmingham 619 19th St. South, Birmingham, AL 35249 619 19th fax St. South, Birmingham, [email protected] 35249 (205) 934-3213; (205) 975-4679; e-mail (205) 934-3213; fax (205) 975-4679; e-mail [email protected]

ABSTRACT

The presented work concerns identifying various defects in glass fiber/polypropylene (glass/PP) composites parts. Comprehensive comparison between two Non Destructive Evaluation (NDE) techniques, namely X-ray Computed Tomography (X-CT) and ultrasonic C-scan techniques are presented. Artificial defects of different shapes, sizes, and materials were embedded in a glass/PP panel comprising of eighty (80) layers to assess the capabilities of NDE systems and to compare the detection sensitivities. The X-CT technique provided a volumetric map through the thickness of the specimen, which resulted in detecting the size and the shape of the defects. Fiber orientation and misalignment were successfully identified using this technique. However, the X-CT technique was not able to identify objects with relatively close material densities. Similarly, ultrasonic pulse-echo C-scan could not map all the embedded artificial defects. However, the through transmission Ultra Sonic (UT) technique showed the capability to identify the location and shape of the defects in glass fiber reinforced thermoplastic composite. The defects size was overestimated by the through transmission technique and fiber related defects could not be detected. The ultrasonic techniques effectively detected delaminations and porosity, while X-CT was not effective in detecting these.

INTRODUCTION

Thermoplastic composites are being increasingly used in aerospace, mass transit, military and automotive applications [1-3]. Thermoplastic composites have advantages over conventional materials in terms of their high strength-to-weight and stiffness-to-weight ratio, low fatigue susceptibility, ability to recycle, and superior damping capacities [4, 5]. With increase of use of thermoplastics in applications that require high standards of safety, there is an important need for implementing Non Destructive Testing/Evaluation (NDT/E) techniques for quality control and in service monitoring to avoid any future structural or catastrophic failure. Thermoplastic composites are prone to defect mechanisms that can occur during manufacturing processes. Some of the common defects include foreign object inclusions, fiber wrinkling, porosity, and ply misalignment that can progress during the service time and lead to other damages such as delaminations, cracks, and disbonds [6, 7]. Many NDT/E techniques, such as X-ray radiography, infrared thermography, ultrasonics, and Acoustography are widely employed today to detect internal anomalies in composites [8-15]. Conventional X-ray radiography suffers from lack of providing three-dimensional information, because the final image is a shadowgraph of the entire part (i.e. structural superposition). Infrared Thermal NDT techniques provide rapid assessment of discontinuities in composites, but are limited to identifying only subsurface discontinuities [11]. Using A-scan ultrasonic inspection technique for detecting defects in composite is quite challenging, as the interpretation of the scanned signals is difficult. This is due to the fact that composites have many signal dispersing/refraction interfaces [16]. Acoustography is another emerging novel near real-time area-scanning technique that can be employed for rapid scanning of larger and complex parts [12].

72

X- CT technique is a state of art inspection technique that provides two-dimensional (2D) density map of the part cross section that is directly related to the amount of the attenuated X-ray beam. These 2D density images can be stacked together to provide accurate 3D information hence the absence of structural superposition. Conventional XCT scanner has a spatial resolution of 0.2-4.5 line pairs per millimeter (lp/mm) [16], excellent dimensioning capability and the contrast resolution (density discrimination) of X-CT less than 1% [16]. This gives the system the ability to differentiate between similar attenuating materials such as polymeric matrix materials. Several researchers have used X-CT technique to detect several types of defects and characterize different composite materials and structures, including honeycomb composites, fiberglass power poles, and carbon foams [17-20]. Ultrasonic Testing (UT) is widely used as an active NDT/E technique for materials quality inspection and quality control in major industries. The technique was used in monitoring different structures such as aircraft structures, jet engine, ship hulls, automotive components, bridges, piping and welds [6, 11, 21-23]. Stefanos et al. found that using A-scan data alone would not result in sufficient diagnosis and characterization for the entire component [24]. C-scan imaging is a standardized and effective defect detection method. The present work focuses on different defects in glass/PP thermoplastic composite using two different NDT/E techniques (i.e. X-CT scan and UT C-scan). The ability and limitation of each technique for identifying the location, shape, and size of the internal defects are discussed.

MATERIALS AND METHODS

Glass /PP (Polystrand Inc.) of density of 1.88 g/cm3 and 80 wt.% as glass fiber weight fraction was the material used in this study. X-ply [0/90]40 configuration were used with a total of 80 plies were stacked and molded into plaque of dimension 152.4 mm (6”) x 152.4 mm (6”) x 14.1 mm (0.55”). The mold was heated up to 170.5°C (339°F) at a constant rate of 5°C/min to a temperature of 65.5°C (150°F). When the temperature reached 65.50C, 10 tons of pressure was applied using a 150 metric ton press (Pasadena Hydraulics Inc.) The dwell time was 40 minutes and then was allowed to cool in air. Six artificial defects with different shapes and materials were embedded in the panel, as shown in Table 1. The distance between the defects was carefully designed to avoid interference between each defect and its neighbor during the inspection process, as shown in Figure 1. The Kapton and Teflon were chosen as materials to impart embedded defects. Defects A3 and A5 were introduced to represent delamination defects at two locations. The A6 defect was created by oversize cut of the last 12 plies that created a fiber wash defect during the compression molding process. Table 1: Artificial defects location, size and material for glass/PP panel. Defect #

Shape

Size (mm) +

Square 10 X 10 X0.13 A1 Square 10 X 10 X 0.26 A2 Circular 15.8* X 1.09 A3 Square 10 X 10 X 0.26 A4 Square 20 X 20 X 1.1 A5 Fiber-wash N/A A6 +This value denotes length X width X thickness *This value denotes diameter of the circle

Location in Z-direction

Defect Material

Between ply No. 14 and 15 Between ply No. 34 and 35 Between ply No. 40 and 41 Between ply No. 48 and 49 Between ply No. 60 and 61 Oversize cut of ply No. 68 to 80

Kapton Teflon Steel plate Teflon Aluminum plate -

EXPERIMENTAL X-ray CT-scan

A medical CT scanner, GE Discover 750 HD, was used to scan the panel in a helical mode with energy of 80 kV, current of 200 mA, Display-Field-of-View (DFOV) of 200 mm and scan resolution of 350 μm. In this system the source and detector were rotated 360 degrees while the part only moves in a horizontal direction, as shown in Figure 2. Two-dimensional density images of cross sections through the plaque are acquired through tomographic reconstruction. The density can be correlated to the point-by-point linear attenuation coefficient in each slice of the final image and is given by [25]:

73

(1) where, I0 is the intensity of the un-attenuated radiation, I is the intensity of the attenuated radiation over the integrated path length and is the linear attenuation coefficient. However, the medical CT-scanner used in this work does not use the attenuation coefficients directly. Instead it uses the CT-number (i.e. Hounsfield unit) [26]. The Hounsfield unit is used to standardize the original linear reconstructed attenuation coefficient ( to linear attenuation coefficient of water at a photon energy of 37 Kev [26]: (2) According to Equation 2, a change of one Hounsfield unit (HU) represents a change of 0.1% of the attenuation coefficient of water [26]. The higher the density of the material, the higher will be the gray-scale of CT images of material. Multi Planar Reconstruction (MPR) method was used to visualize the volumetric data attained by the CT. The conversion of the two dimensional images to MPR volumes, segmentation and visualization of the volumetric data set were performed using Osirix imaging software [27]. Post processing was carried out using Image J and Matlab software to enhance the clarity of the images and removal of noise and other artifacts [28, 29].

A2 �� �

��� �

��

10�

R�3.45�mm�

A3�

20�

A1 �� �

20�

20�

Φ15.8�mm�

76.2�mm�

152.4�mm�

�� � A4

10�

152.4�mm�

��

A5�

� X � Y

20�

Figure 1: Schematic for defect locations in glass/PP panel.

Ultrasonic C-scan Pulse-echo Ultrasonic C-scan

Pulse-echo ultrasonic C-scan technique was used to study the effect of changing the wavelength on the detectability of the shape and size of the embedded defects. The tests were conducted at two different frequencies, 2.25 MHz and 5 MHz respectively. Low frequency flat transducer, 2.25 MHz, with an element diameter of 12.7 mm (0.5 inch) and high frequency focused transducer, 5 MHz, with an element diameter of 12.7 mm (0.5 inch) were used. The data acquisition system, transmitting-receiving unit was Olympus Omni scan system and the immersion tank and the automated step motors drivers were designed and fabricated by Marietta NDT, GA as shown in Figure 3. The scanning speed in the X-Y axis was 200 mm/s (7.8 inch/s) with a resolution of 0.5 mm (0.019 inch). In order to prevent the variation of the acoustic pressure when the ultrasonic beam travel through the water bath, a distance between the transducer and the spacemen was set to 78 mm (3 inch). The 2.25 MHz scan was operated in a fullwave-rectified signal mode with 0 dB overall gain and 44 dB Time Corrected Gain (TCG) at 9.058 mm. The TCG is usually used while scanning composite materials to add an additional amount of gain relative to the depth in the part to account for attenuation. This gain is zero at the front wall position and increases linearly further into the part. However, the 5 MHz scan was carried out using a 6 dB overall gain and 52 dB time corrected gain at 8.717 mm. The TCG point was placed at or behind the back wall to obtain a consistent increase in gain throughout the thickness of the part. The gate was placed at the back wall and the maximum peak strategy was used to identify and allocate the defects at the specimen.

74

Through-transmission Ultrasonic System

A 10 Axis gantry scanner consisting of a steel frame scanner, two X axis scanning bridges, two Y axis carriages, two Z axis columns, and two nozzle gimbal / swivel axis assemblies designed and manufactured by Marietta NDT was used. The plates were scanned using two different frequencies (i.e. 1 MHz and 5 MHz) at a single scan. The low frequency sensor, 1 MHz, was a focused transducer with a 38.1 mm (1.5 inch) outer diameter and 23.9 mm (0.94 inch) inner diameter. The high frequency sensor, 5 MHz, was a flat transducer with a diameter of 12.7 mm (0.5 inch). The acquisition system used was Olympus MultiScan MS5800 scan 8 channel system. The scanning speed in the X-Y axis was 200 mm/s (7.8 inch/s) with a resolution of 0.5 mm (0.019 inch). In order to focus on small features in the plaques a 3.1 mm (0.125 inch) diameter water nozzle was used to enforce the ultrasonic beam to travel through narrow water bath. The plaques were scanned using 40 dB as an overall gain.

Transmitter/ receiver system

UT sensor Specimen

Water tank

Figure 2: Schematic for helical CT- scan setup; a) X-ray source, b) Detector, c) Stage, d) Composite plate, e) Direction of the rotational motion of the source and the detector.

Figure 3: Typical UT C-scan setup.

RESULTS AND DISCUSSIONS

Figure 4 shows the CT scan for the glass/PP panel. The embedded defects were clearly resolved as can be observed in Figure 5(b-e). The high attenuation of the material results in streaks at the edges of the defects and the polymer matrix. However, it can be noticed that defects A2 and A4 were not detected. This could be attributed to the low differences in the attenuation of the X-ray beam between the Teflon inserts and the GFRP (i.e. Teflon CT-number is around 1000 H and GFRP CT-number is around 1200 H) [26, 30]. The Teflon insert thickness is of the order of the CT scan slice thickness that is also make it hard to detect theses defects while the tomographic reconstruction of the sample. The UT C-scan pulse- echo system showed that most of the defects could not be detected, with both frequencies, as shown in Figure 5. Figure 5a presents amplitude C-scan of the glass fiber /PP specimen with defects at scanning frequency of 2.25MHz. It was observed that only defects near the scan surface (i.e. A5 defect) were detected. However, the exact shape and size of the defects could not be determined as shown in Figure 5a. The Eglass fibers have high reflectivity of the UT signal and the signal is scattered and attenuated. The glass content in these samples is 80 wt. %, which produce more intense echoes in comparison to the ones produced by the defects. Figure 5b shows a 5MHz amplitude C-scan for the glass/PP plate. The signal has a smaller wavelength that is attenuated more sharply, specially in relatively thick parts used in this research, to where the defects could not be detected using the pulse-echo technique. Through transmission UT technique was used to overcome the high attenuation problem in the pulse-echo technique. The through transmission technique is based on the amplitude of the received signal, as opposed to the attenuation of the back wall signal. Lower frequency (i.e. 1MHz) permits deeper signal penetration through the sample. Figure 6 shows amplitude C-scan, in gray scale, at 1MHz scan for the glass/PP plate. It was noticed that all of the defects were captured, which was not the case with the pulse-echo and X-CT techniques. The UT C-scan does not depend on the material density differences that limit the detectability of the X-CT system. However, the exact shape and size of some of the defects (i.e. A2, A3 and A4) could not be obtained and analyzed precisely in

75

comparison to the X-CT data. The defects cause separation at the adjacent areas of the defects and form a local bending in the composite laminate above the defect area, which result in echoes that are recognized as a defect area. This will lead to an overestimation of the area of the embedded defects. The amplitude of the defects at these areas are mixed with echoes coming from the fibers [6]. The results for 5MHz scan for the glass fiber was the same as the results obtained by the pulse echo technique, the signal was attenuated resulting in the absence of the defect detectability.

Figure 4: CT-scan slices for glass/PP plate; a) between ply No. 75 and 80, b) between ply No. 60 and 61, c) between ply No. 40 and 41, d) between ply No. 14 and 15. It was noticed that the X-CT technique has an advantage over the pulse-echo ultrasonic C-scan in detecting the composite texture and fiber related defects. Figure 4 shows the orientation pattern off the glass fibers (i.e. 0 0/900) in the composite specimen. This pattern is clearly defined due to the density difference, X-ray attenuation differences, between the glass reinforcement and the PP matrix. Figure 4a (at the upper and lower right area) shows the fiberwash, defect A6, that resulted from the over size cut of the plies 68 to 80. The out-of-plane waviness is usually difficult to be perceived from the part surface and regularly not detected until a part is destructively examined [16]. Due to the high attenuation and signal desperation of the material of the material the pulse-echo C-scan technique could not identify the fiber orientation configuration that was clearly detected by the X-CT-scans. Nevertheless, this was overcome with the use of the low frequency through-transmission UT technique as shown in Figure 6. However, the composite texture was captured using the UT through-transmission splash system; the fiber wash defect (i.e. A6 defect) was not identified. The 2.25 MHz pulse-echo scan provided information about the delaminated areas that could not be detected by the X-CT scan technique, Figure 5a. Areas of relatively higher amplitude were observed surrounding the embedded defect areas that can be attributed to the change on the thickness or delamination in the composite layer at these places due to the presence of the defects. Figure 5 shows white areas, which represent loss of the back wall signal attributed to porosity in the samples. The pulse-echo system successfully detected porosity and showed that the 5MHz frequency (i.e. smaller wavelength) is more sensitive to smaller pores causing the signal to attenuate more sharply. The limited resolution of the medical CT used in this work affected the detectability of the porosity in the composite plates.

76

Figure 5: Amplitude pulse-echo C-scan of glass /PP specimen; a) 2.25 MHz and b) 5MHz, (1) Areas represent porosity in the specimens and (2) Areas represent delamination and separation in the specimens.

Figure 6: Amplitude through-transmission C-scan at 1 MHz of glass /PP specimen, (1) Areas represent clamping device.

CONCLUSIONS

Different NDE/T techniques were explored to characterize thermoplastic composites (i.e. embedded defects, delamination, porosity and fiber orientation). In both of the ultrasonic methods it was challenging to scan composites with glass fiber reinforcement due to the high signal attenuation characteristic of the E-glass fibers. The results showed that (a) the pulse-echo ultrasonic C-scan detected A5 defect (i.e. aluminum plate insert) however it could not identify the exact size and shape of the defects, fiber misalignment (i.e. fiber-wash) and fiber orientation, and (b) The through-transmission ultrasonic C-scan resolved all of the embedded defects location and shape in addition to the texture of the composite. The through-transmission ultrasonic system overestimated the defect size and fiber related defects could not be detected. (c) The X-CT technique provided a volumetric map through the thickness of the specimen, and resolved the size and the shape of the A1, A3 and A5 defects. In addition fiber orientation and misalignment were also identified using X-CT technique. Nevertheless the CT-scan could not identify objects that have relatively close material densities. The study suggested that a combination of both of XCT and UT C-scan techniques would provide optimum results in identifying a range of defects and fully characterize glass reinforced thermoplastic composite parts.

77

ACKNOWLEDGMENT

All of the ultrasonic testing was performed using Marietta NDT Inc. facility at GA, USA. The authors would like to acknowledge the support and help of Ryan McCarthy and Daryle Higginbotham at Marietta NDT that made this research paper possible. Support from the Department of Energy Graduate Automotive Technology Education (DOE GATE) is gratefully acknowledged.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

Haibin Ning, Selvum Pillay, and Uday K. Vaidya, Design and development of thermoplastic composite roof door for mass transit bus. Materials & Design, April, 2009. 30: p. 983-991. Carrillo J. G, et al., Ballistic performance of thermoplastic composite laminates made from aramid woven fabric and polypropylene matrix. Polymer Testing, 2012. 31(4): p. 512-519. Anahi Pereira da Costa, E.C.B., Michelle Leali Costa, Nilson Eiji Narita, José Ricardo Tarpani, A Review of Welding Technologies for Thermoplastic Composites in Aerospace Applications. Journal of Aerospace Technology & Management, 2012. 4(3): p. 255-265. Chung, D.D.L., Reviwe: Materials for viberation damping. Journal of materials science, 2001. 36: p. 57335737. Akonda, M.H., C.A. Lawrence, and B.M. Weager, Recycled carbon fibre-reinforced polypropylene thermoplastic composites. Composites Part A: Applied Science and Manufacturing, 2012. 43(1): p. 79-86. Theodoros H., Efstratios B, and Nicolaos G. T., Application of Ultrasonic C-Scan Techniques for Tracing Defects in Laminated Composite Materials. Journal of Mechanical Engineering, 2011. 57: p. 192-203. T. Khan, et al., Novel Damage Diagnosis Algorithms for Aerospace Nondestructive Testing Data Using Ultrasonic Testing Technique, in ASNT 23rd Research Symposium, ASNT, Editor. March, 2014: Minneapolis,MN ,USA. p. 67-71. Chu, T.P., A. Poudel, and P. Filip. C/C Composite Brake Disk Non-Destructive Evaluation by IR Thermography. in Conference on Thermosense - Thermal Infrared Applications XXXIV. 2012. Baltimore, MD. Chu, T.P., et al., Defect characterization in commercial Carbon-Carbon composites World Journal of Engineering, 2012. 9(6): p. 481-486. Poudel, A. and T.P. Chu, Intelligent Nondestructive Testing Expert System for Aircraft Carbon/Carbon Composite Brakes Using Infrared Thermography and Air-coupled Ultrasound. Materials Evaluation, 2012. 70(10): p. 1219-1229. Poudel, A., J. Strycek, and T.P. Chu, Air-Coupled Ultrasonic Testing of Carbon/Carbon Composite Aircraft Brake Disks. Materials Evaluation, 2013. 71(8): p. 987-994. Poudel, A., et al. Porosity Measurement in Carbon Fiber Epoxy Laminates by Using Acoustography. in 23rd Annual Research Symposium and Spring Conference. 2014. Minneapolis, MN: ASNT. Brady, R.P., et al., Thermal image analysis for the in-situ NDE of composites. Journal of Composites Technology & Research, 1999. 21(3): p. 141-146. Liu, C.-T.B., et al., Detection of Discontinuities in Carbon-carbon Composites using X-ray Compton Backscatter Radiography: Radiography by Selective Detection. Materials Evaluation, 2012. 70(3): p. 367377. Pan, Y.P., et al., Thermal Diffusivity Mapping of Carbon/Carbon Composites. Materials Evaluation, 2009. 67(5): p. 540-546. Green, W.H. and P. Sincebaugh, Nondestructive Evaluation of Complex Composites Using Advanced Computed Tomography (CT) Imaging, A.r. laboratory, Editor. Feberuary, 2001: Maryland, USA. Moore, D.G. and C.L. Nelson, Assessment of Advanced Ultrasonic and Infrared Inspection Methods to Detect Delaminations and Water Ingress in Composite Honeycomb Materials, in 13th International Symposium on Nondestructive Characterization of Materials (NDCM-XIII). May, 2013: Le Mans, France. Persson, S., and E. Ostman, The Use of Computed Tomography in Non-Destructive Testing of Polymeric Materials, Aluminum and Concrete. II. Applications. Polymer Testing, 1986. 6(6): p. 415-446. Lowrey A. R, K. D Friddell, and D.W. Cruikshank, Nondestructive Evaluation of Aerospace Composites Using Medical Computed Tomography (CT) Scanners, in American Societyfor Nondestructive Testing (ASNT) Spring Conference. 11-14 March, 1985: Washington, DC, USA.

78

20. 21. 22. 23. 24.

25. 26. 27. 28. 29. 30.

H. Hocheng and C.C. Tsao, Computerized Tomography and C-scan for Measuring Drilling-Induced Delamination in Composite Material Using Twist Drill and Core Drill. Key Engineering Materials, May, 2007. 339: p. 16-20. Seshu D. Rama and M.N.R. Dakshina, Non Destructive Testing of Bridge Pier - A Case Study. Procedia Engineering, 2013. 54(0): p. 564-572. Wagner, D., et al., Ultrasonic fatigue tests at high temperature on an austenitic steel. Propulsion and Power Research, 2012. 1(1): p. 29-35. Zhen-Guo, Y., L. B., and S. J., Ultrasonic testing technique for small-calibre piping containing defects. International Journal of Pressure Vessels and Piping, 1996. 68(3): p. 325-330. Stefanos Giannis, Vladimír Matěják, and R. Freemantle, Maximizing the benefit of non-destructive inspection data to evaluate the damage tolerance performance of composites, in SAMPE Tech conference, S.f.t.A.o.M.a.P.E. SAMPE, Editor. October, 2013, Society for the Advancement of Material and Process Engineering (SAMPE): Wichita, KS, USA. (ASTM), A.S.f.T.a.M., Standard Guide for Computed Tomography (CT) Imaging, in 3. 1986, ASTM international: Columbus, OH, USA. Oster, R., Computed Tomography as a Non-destructive Test Method for Fiber Main RotorBlades in Development, Series and Maintenance, in International Symposium on Computerized Tomography for Industrial Applications and Image Processing in Radiology. March, 1999: Berlin, Germany Rosset A, S.L., Ratib O., OsiriX: an open-source software for navigating in multidimensional DICOM images. Journal of Digit Imaging., Sep 2004. 17(3): p. 205-216. Tiago Ferreira and Wayne Rasband, ImageJ User Guide, N.I.o. Health, Editor. October, 2012, National Institutes of Health: USA. MathWorks, Matlab Primer R2014a manual 2014, Mathwork, Inc.: MA, USA. Peter, E., et al., Interphantom and interscanner variations for Hounsfield units—establishment of reference values for HU in a commercial QA phantom. Physics in medicine and biology, 2010. 55: p. 5123-5135.

79

A Nondestructive Evaluation Technique for Detecting, Locating and A Nondestructive Evaluation Detecting, Locating Quantifying Damage in Large PolymerTechnique CompositeFor Structures Made of Electrically and Quantifying Damage in Large Polymer Composite Structures Made Non-Conductive Fibers and Carbon Nanotube Networks of Electrically Non-Conductive Fibers and Carbon Nanotube Networks Ali Naghashpou Naghashpou and and Suong Suong Van Van Hoa Hoa Ali Concordia Centre for Composites (CONCOM) and Center for Applied Research on Polymers and Composites (CREPEC), Concordia CentreDepartment for Composites (CONCOM) Center Engineering, for Applied Research onUniversity Polymers and Composites of Mechanical andand Industrial Concordia (CREPEC), Department of Mechanical andMontreal, IndustrialQuebec, Engineering, 1455 De Maisonneuve Blvd.W., CanadaConcordia H3G1M8.University 1455 De Maisonneuve Blvd.W., Montreal, Quebec, Canada +1 514 848 2424 ext 7228; fax +1 514 848 3175 H3G1M8. +1 514 848 2424 ext 7228; fax +1 514 848 3175. e-mail [email protected], [email protected] e-mail [email protected], [email protected]

ABSTRACT A novel and practical nondestructive evaluation (NDE) technique is presented to detect, locate and quantify damages that occur at one or several locations in large polymer composite structures made of electrically nonconductive fibers and carbon nanotube networks. In this technique, multiwalled carbon nanotubes were embedded into epoxy resin to make electrically conductive matrix. This modified matrix was used to prepare glass fabric/epoxy/MWCNT composite plates and kevlar fabric/epoxy/MWCNT composite plates. The large plate was mounted with grid points made from silver-epoxy paste. The electrical resistance values between the grid points were measured and used as a reference set. Drilled holes and impact damages were created in the large plates. These damages were detected, located and quantified based on the significant local variations in the distribution of electrical resistance change.

Keywords: carbon nanotubes, damage detection, damage location, large polymer composite structures INTRODUCTION

Damage detection in large polymer composite structures (LPCSs) is essential in industrial applications that provide safety and proper performance for the LPCSs. This is due to susceptibility of composite materials to different types of damages such as matrix microcracking, fiber/matrix interface debonding and inter-ply delamination [1]. Many non-destructive evaluation (NDE) techniques such as X-ray, ultrasound scanning, thermography, acoustic emission, piezoelectric active sensors, fiber optics and resistance based sensors have been used to detect damage in the LPCSs. However, they are not much used due to poor in-situ capabilities and low spatial resolution [2-8]. Adding carbon nanotubes (CNTs) into polymer composites creates potential to not only for the modification of polymer but also solving sensing challenges. This is because of their small size, high aspect ratios, exceptional electrical conductivity [9], thermal conductivity [10] and outstanding mechanical properties [11]. Incorporating CNTs as nanoscale conductors at low concentration into a resin will create electrically conductive networks distributed around the structural fibers which also act as in-situ piezoresistive sensors. Upon application of a mechanical load, the CNTs network is deformed and if the load is high enough to create cracks in the matrix, the configuration of the CNTs network is affected that induces a change in the electrical resistance. This displays their piezoresistive behaviour. Recently CNTs were added to glass fiber/epoxy composites and their piezoresistive properties were exploited to assess both deformation and microstructural damage by electrical resistance measurement (ERM) [12-19]. Chou et al. [18-21] found that cracking in glass /epoxy/CNTs composite coupons of about 4×6 inch2 or smaller correlates well with the change in electrical resistance. Nofar et al. [22] indicated that failure region in glass/epoxy/CNTs composite coupons can be detected by significant change in electrical resistance during tensile and fatigue testing. Naghashpour et al. [23] found that the measurement of through-thickness electrical resistance can be used to detect through-thickness strain of glass/epoxy/CNTs composite laminate where it is subjected to a thicknesswise load. Baltopoulos et al. [24] described the forward and inverse methods for detecting the location of cracks in 4×4 inch 2 glass /epoxy/CNTs composite sample using electric resistance tomography (ERT) technique. They mounted electrodes around the boundary of the sample. They were able to detect damages around the boundary of the sample, but the technique fails to detect damage at the center of the sample. Proper et al. [25] indicated that when 4×6 inch 2

80

Kevlar /epoxy/CNTs composite sample is damaged by mechanical impact, there is a correspondence between the change in voltage across the grid points spaced at 0.25 inches apart and the impact damage. Wardle et al. [26-27] developed alumina fiber/CNTs/epoxy composite and found that the changes in resistance across the grid lines spaced at 0.118 inch apart correspond to impact damage in the 4.5×1 inch2 sample. All the works in this area have been limited to detect damage in composite coupons of about 4 inch by 6 inch or smaller. Real engineering structures are large and much bigger than the lab size samples presented in most works. Here, we present a new NDE technique for large polymer composite structures using CNT networks.

EXPERIMENTAL METHODS Materials

Multiwalled carbon nanotubes (MWCNTs) with 95% of purity, diameters of 2-20 nm and lengths of 1 µm to more than 10 µm were purchased from Bayer Material Science. Plain weave Kevlar and glass fabrics purchased from HL. plasto company, Epon 862 and EPIKURE W purchased from Miller-Stephenson chemical company were utilized as reinforcement, epoxy resin and curing agent respectively.

Methods

Fabrication of Composite Plates

To manufacture fabric/epoxy/MWCNT composite plates, the epoxy resin and curing agent (26.4 wt %) were first mixed. Then 0.30 wt% MWCNT [28] was added into epoxy matrix. The mixture was processed on three roll mill (EXAKT 80E, EXAKT Technologies Inc) to disperse the MWCNTs within the epoxy matrix. The modified epoxy matrix was heated up to 600C for 20 min in a vacuum oven to remove air bubbles. The modified epoxy matrix was dispersed between three layers of woven fabrics by hand lay-up method to make glass/epoxy/MWCNT composite plates and kevlar/epoxy/MWCNT composite plates. The plates were cured using an autoclave.

Plate Specification and Arrangement of Electrical Connections

A grid of forty conductive electrical contact points made from silver-epoxy paste spaced at 3 inch apart was mounted on the surface of 22×13 inch2 plates for ERM. Electrical wires were attached to the contact points to make electrodes. Then the electrodes were connected to the data acquisition system (Vishay Micro-Measurements System 7000) and a computer with the program (see figure 1). The spacing between the grid points was 3 inch [28]. 13"

3"

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

22"

3"

1

DA PC

Figure 1: Schematic illustration of plate mounted with grid points and wires.

81

Electrical Resistance Measurement (ERM) ERM was performed by the two-probe method using multimeter (Agilent 34401A) and data acquisition system. A constant source voltage was directly applied and the electrical current was measured to calculate resistance. Electrical resistance change (ERC) is defined as

ሺΨሻ ൌ ሺ ˆǡ‹ǡŒ−  ǡ‹ǡŒ Ȁ ǡ‹ǡŒ ሻ ൈ ͳͲͲ

(1)

Where  ǡ‹ǡŒ and  ˆǡ‹ǡŒ are initial and final electrical resistance values between grid points i and j respectively.

DRILLED HOLES AND IMPACT TESTS

Two types of damage were made. The first type was performed by drilling holes of different sizes at different locations in the plates. The second type was carried out by impact caused by collision with high velocity projectiles and drop weights on the plates. The high velocity projectiles impact tests were done using a gas gun with the impact energy of 78J created by 318 mg aluminum particle travelling at 700 m/sec. Experimental set-up to detect and locate damage due to the high velocity impact test is shown in Figure 2. The low velocity impact tests were carried out by dropping weight on the clamped plates placed on electrically non-conductive rigid supports. Impact damage detection and location

Gas gun

Figure 2: Experimental set-up to detect and locate damage due to high velocity impact test.

RESULTS AND DISCUSSION Glass fabric/epoxy/0.3wt%MWCNT composite palates Three glass fabric/epoxy plates containing 0.30wt% MWCNT with the dimension of 22×13 inch2 were considered. Holes of sizes 1/16, 2/16, 3/16, 4/16, 5/16 and 6/16 inch respectively were drilled in glass fabric/epoxy/0.3wt%MWCNT plate 1 as shown in Figure 3(a). Figure 3(b) shows the locations and values of the changes in electrical resistance after six holes were drilled. By comparing Figures 3(a) and 3(b), it is clear that the significant local variations in distribution of the electrical resistance change correspond exactly to the locations of holes of different sizes drilled at different locations in plate 1. Figure 3(c) presents the influence of hole volume on the change in electrical resistance. In figures 3(c) and 5(c), the numbers below the curve represent the pairs of electrodes for the ERM. This pair of electrodes is closest to the hole or damage region. As it can be observed from figure 3(c), there is a clear correlation between the hole volume and the change in electrical resistance proving that

82

Electrical Resistance Change (%)

this technique can also be used to quantify the extent of damages, in addition to the detection and location capabilities.

Hole 5 (5/16 inch) Hole 2 (2/16 inch) Hole 1 (1/16 inch)

Hole 5 (5/16 inch)

Hole 2 (2/16 inch)

10 9 8 7 6 5 4 3 2 1 0

20

19

18

17

16

15

14

13

12

11

10

9

8

7

6

Hole 1 (1/16 inch)

5

4

3

2

1

25

24

23

Hole 3 (3/16 inch)

30

21

29 28

Hole 4 (4/16 inch)

27

Hole 4 (4/16inch)

Electrical Resistance Change (%)

(a)

40

31

39

38

Hole 6 (6/16 inch)

37

36

Hole Quantification

11

Hole 6 (6/16 inch)

34 33

32

(b) Hole 3 (3/16 inch)

35

26

10 6

9 8 7

5

6 5 4

4 3

3

2 1

2 1

0

(c)

0

0.0005

0.001

0.0015

0.002

0.0025

0.003

0.0035

0.004

1 : Electrodes (14-15): Electrical Resistance Change between Electrodes (14-15) at hole location 2 : Electrodes (6-7) 3 : Electrodes (34-39) 4 : Electrodes (17-18) 5 : Electrodes (4-5) 6 : Electrodes (31-36)

0.0045

Hole Volume (inch3)

Figure 3: (a) Glass fabric/epoxy/0.3wt%MWCNT plate 1 with six drilled holes, (b) electrical resistance change (ERC) distribution for plate 1 after drilling of holes1, 2, 3, 4, 5 and 6 (1/16, 2/16, 3/16, 4/16, 5/16 and 6/16 inch) and (c) Effect of hole volume on ERC. (Data are presented as mean ± standard deviation (SD) from three experiments.) Glass fabric/epoxy/0.3wt%MWCNT composite plate 2 was impacted using gas gun with 78J at two different regions as shown in Figure 4(a). Figure 4(b) shows the locations and values of the changes in electrical resistance due to the collision with high velocity projectiles (78J each) in plate 2. Comparing Figures 4(a) and 4(b), impact damages 1 and 2 are detected and located according to the sharp local variations in distribution of the electrical resistance change. Figure 4(c) reveals the locations and values of the changes in electrical resistance due to the collision with low velocity projectiles in glass fabric/epoxy/0.3wt%MWCNT plate 3. The energy levels vary from 1J to 10J as produced by drop weight impact tests. Some of the damages at the lower energy levels (1J to 4J) are barely visible by naked eye. These locations correspond exactly to the location of the damages created and barely visible damage zones. It is clear from Figure 4(c) that six impact damages created with different energy levels at different locations in plate 3 are detected and located distinctly based on the significant local variations in distribution of the electrical resistance change. Figure 4(d) shows the relation between the change in resistance and the energy level which indicates impact severity.

83

Damage 1(78J)

Damage 2 (78J)

(b)

(a) Damage 5(5J) Damage 3(3J)

Damage 1(1J)

Damage 6(10J)

(c)

Electrical Resistance Change (%)

Damage 2(2J)

Impact Severity

0.9 0.8

6

0.7 0.6 0.5

5

0.4 0.3

4 3

0.2

2

0.1

1

0

Damage 4(4J)

(d)

0

1

2

3

4

5

6

7

8

9

10

1 : Electrodes (35-40): Electrical Resistance Change between Electrodes(35-40) at Impact point 1 2 : Electrodes (1-6): Impact point 2 3 : Electrodes (24-25): Impact point 3 4 : Electrodes (31-36): Impact point 4 5 : Electrodes (5-10): Impact point 5 6 : Electrodes (17-22): Impact point 6 11

Impact energy (J)

Figure 4. (a) Glass fabric/epoxy/0.3wt%MWCNT plate 2 after impact damages 1 and 2 (78J each) were made, (b) Electrical resistance change (ERC) distribution of plate 2 after impact damages 1 and 2, (e) ERC distribution of glass fabric/epoxy/0.3wt%MWCNT plate 3 after impact damages 1, 2,3,4,5 and 6 (1J, 2J, 3J, 4J, 5J and 10J) were created and (d) Effect of energy level on the change in electrical resistance. (Data are presented as mean ± SD from three experiments.)

Kevlar fabric/epoxy/0.3wt%MWCNT composite palates The new technique was demonstrated for Kevlar fiber reinforced epoxy composite plates containing 0.3wt%MWCNT. Three kevlar fabric/epoxy plates containing 0.30 wt% MWCNT with dimension of 22×13 inch2 were prepared. Kevlar fabric/epoxy/0.3wt%MWCNT plate 4 was drilled to make holes of sizes 1/16, 2/16, 3/16, 4/16, 5/16 and 6/16 inch respectively as shown in Figure 5(a). Observing Figures 5(a) and 5(b), it is found that holes of sizes 1/16, 2/16, 3/16, 4/16, 5/16 and 6/16 inch respectively drilled at different locations of plate 4 are detected and located distinctly according to the sharp local variations in electrical resistance change distribution. The effect of hole volume on the change in electrical resistance is indicated in Figure 5(c). A clear correlation is observed between the hole volume and the change in resistance. This demonstrates that the new technique is capable of quantifying the extent of damages, in addition to the detection and location capabilities.

84

Electrical Resistance Change (%)

Hole 3 (3/16 inch) 8 7 6 5 4 3 2 1 0

Hole 1 (1/16 inch) Hole 5 (5/16 inch)

30

21

29

28

Hole 6 (6/16 inch)

(b)

25

24

23

22

20

19

18

17

16

Hole 2 (2/16 inch)

14

13

12

11

10

9

8

7

6

5

4

3

2

1

27

26

35 34

33 32

40

31

39 38

37 36

Hole 4 (4/16inch)

(c)

(a)

Figure 5: (a) Kevlar fibric/epoxy/0.3wt%MWCNT plate 4 after six drilled holes, (b) ERC distribution for plate 4 after drilling of holes1, 2, 3, 4, 5 and 6 (1/16, 2/16, 3/16, 4/16, 5/16 and 6/16 inch) and (c) Effect of hole volume on ERC. (Data are presented as mean ± standard deviation (SD) from three experiments.) Kevlar fabric/epoxy/0.3wt%MWCNT plate 5 was subjected to high velocity projectiles with 78J to create impact damage 1 and 2 as shown in Figure 6(a). Figure 6(b) shows the locations and values of the changes in electrical resistance due to the collision with high velocity projectiles (78J each) in plate 5. Observing Figures 6(a) and 6(b), good correspondence is found between the locations of impact damages 1 and 2 and significant local variations in ERC distribution. The locations and values of the changes in resistance due to low velocity projectiles with different energy levels ranging from 1J to 10J in kevlar fabric/epoxy/0.3wt%MWCNT plate 6 are shown in Figure 6(c). Some of the damages at the lower energy levels are barely visible impact damages (BVID) and cannot be detected by visual observation. As it can be seen from Figure 6(d), the good relation is found between the change in electrical resistance and the energy level which indicates impact severity.

85

Electrical Resistance Change (%)

Damage 2 (78J)

2 1.5 1 0.5 0

21

25

24

23

22

20

19

18

17

16

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

30

29

28 27

Damage 1(78J)

(b)

Electrical Resistance Change (%)

(a) Damage 1(1J) 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

35

26

34

33 32

40

31

39 38 37 36

Damage 3(3J) Damage 6(10J)

2

1

14

13

12

11

25

24

23

22

21

Damage 4(4J)

19

18

17

16

10

9

8

7

6

5

4

3

30

29

28

Damage 5(5J)

(c)

27 26

35

34 33

32

40

31

39 38

37 36

Damage 2(2J)

(d)

Figure 6: (a) Kevlar fabric/epoxy/0.3wt%MWCNT plate 5 after impact damages 1 and 2 (78J each) were made, (b) ERC distribution of plate 5 after impact damages 1 and 2, (c) ERC distribution of kevlar fabric/epoxy/0.3wt%MWCNT plate 6 after impact damages 1, 2,3,4,5 and 6 (1J, 2J, 3J, 4J, 5J and 10J) were created and (d) Effect of energy level on the change in electrical resistance. (Data are presented as mean ± SD from three experiments.)

CONCLUSIONS

A new, practical NDE technique has been proposed to detect, locate and quantify damages in the large polymer composite plates made of electrically non-conductive fibers and CNT networks. In this technique, electrical resistance distribution for undamaged plate was used as refrence map. An actual map of electrical resistance distribution was compared with refrence map to calculate ERC distribution. Drilled holes, impact damages and barely visible impact damages were detected, located and quantified in the large polymer composite plates according to significant local variations in electrical resistance change distribution.

REFERENCES

[1] Chou TW “Microstructural Design of Fiber Composites” (Cambridge: Cambridge University Press),1992. [2] McCombe GP, Rouse J, Trask RS, Withers PJ and Bond IP “X-ray damage characterization in self-healing fibre reinforced polymers” Composites: Part A 43 613–20, 2012. [3] Takeda S, Aoki Y, Ishikawa T, Takeda N and Kikukawa H “Structural health monitoring of composite wing structure during durability test” J.Compos Struct 79, 133–9, 2007.

86

[4] Masters JE “Damage detection in composite materials” Philadelphia PA American Society for Testing and Materials p 35, 1992. [5] Rizzo P and di Scalea FL “Acoustic emission monitoring of carbon-fiber-reinforced polymerbridge stay cables in large-scale testing” J.Exp Mech 41, 282–90, 2001. [6] Park G, Faffar CF, di Scalea FL and Coccia S “Performance assessment and validation of piezoelectric activesensors in structural health monitoring” J.Smart Mater Struct 15, 1673–83, 2006. [7] Kirkby E, de Oliveira R and Mänson JA “ Impact localization with FBG for a self healing carbon fibre composite structure” J.Compos Struct 94, 8–14 , 2011. [8] Balageas D, Fritzen CP and Guemes A “Structural health monitoring” ISTE L td, 2006. [9] Ebbesen TW, Lezec HJ, Hiura H, Bennett JW, Ghaemi HF, Thio T “Electrical conductivity of individual carbon nanotubes” Nature 382 54–6, 1996. [10] Berber S, Kwon YK and Tomanek D “Unusually high thermal conductivity of carbon nanotubes” Phys Rev Lett 84 4613–6, 2000. [11] Ruoff RS and Lorents DC “Mechanical and thermal-properties of carbon nanotubes” Carbon 33, 925–930, 1995. [12] Fiedler B, Gojny FH, Wichmann MHG., Bauhofer W and Schulte K “Can carbon nanotubes be used to sense damage in composites?” Ann Chim Sci Matér 22, 81–94, 2004. [13] Böger L, Wichmann MHG, Meyer LO and Schulte K “Load and health monitoring in glass fibre reinforced composites with an electrically conductive nanocomposite epoxy matrix” J.Compos Sci Technol 68,1886–1894, 2008. [14] Li C, Thostenson ET and Chou TW “Sensors and actuators based on carbon nanotubes and their composites: A review” J. Compos Sci Technol 68, 1227-1249, 2008. [15] Gao SL, Zhuang RC, Zhang J, Liu JW and Mäder E “Glass Fibers with Carbon Nanotube Networks as Multifunctional Sensors” Adv Funct Mater 20, 1885–1893, 2010. [16] Gao SL, Zhuang RC, Zhang J, Liu JW and Mäder E “Glass Fibers with Carbon Nanotube Networks as Multifunctional Sensors” Adv. Functional Materials 20, 1885-1893, 2010. [17] Alexopoulos ND, Bartholome C, Poulin P and Marioli-Riga Z “Structural health monitoring of glass fiber reinforced composites using embedded carbon nanotube (CNT) fibers” J. Compos Sci Technol 70, 260–271, 2010. [18] Gao L, Chou TW, Thostenson ET and Zhang Z “ A comparative study of damage sensing in fiber composites using uniformly and non-uniformly dispersed carbon nanotubes” Carbon 48, 3788–3794, 2010. [19] Thostenson ET and Chou TW “Carbon nanotube networks: sensing of distributed strain and damage for life prediction and self-healing” Adv Mater 18, 2837-2841, 2006. [20] Thostenson ET and Chou TW “Real-time in situ sensing of damage evaluation in advanced fiber composites using carbon nanotubes networks” Nanotechnology 19, 215713, 2008. [21] Gao L, Chou TW, Thostenson ET, Zhang Z and Coulaud M “In situ sensing of impact damage in epoxy/glass fiber composites using percolating carbon nanotube networks” Carbon 49, 3382–3385, 2011. [22] Nofar M, Hoa SV and Pugh MD “Failure detection and monitoring in polymer matrix composites subjected to static and dynamic loads using carbon nanotube networks” Compos Sci Technol 69, 1599-1606, 2009. [23] Naghashpour, A and SV. Hoa. 2013” In-situ monitoring of through-thickness strain in composite laminates using carbon nanotube sensors”. Composites Science and Technology, 78: 41-47. [24] Baltopoulos A , Polydorides N, Vavouliotis A, Kostopoulos V and Pambaguian L “Sensing capabilities of multifunctional composite materials using carbon nanotubes” IAC-10.C2.9.2 Prague CZ 61st International Astronautical Congress, 2010. [25] Proper A, Zhang W, Bartolucci S, Oberai A and Koratkar N “In-Situ Detection of Impact Damage in Composites Using Carbon Nanotube Sensor Networks” Nanoscience and Nanotechnology Letters 1, 3–7, 2009. [26] Wicks S, Barber D, Raghavan A, Dunn CT, Daniel L, Kessler SS and Wardle B L “Health monitoring of carbon nanotube (CNT) hybrid advanced composite for space applications” MIT, 2009. [27] Raghavan A, Kessler SS, Dunn CT, Barber D, Wicks S and Wardle B L “ Structural health monitoring using carbon nanotube (CNT) enhanced composites” Stanford University (Stanford, CA) 7th International Workshop on Structural Health Monitoring, Sep 9-11, 2009. [28] Naghashpour A. and Hoa S.V” A technique for real-time detection, location and quantification of damages in large polymer composite structures made of electrically non-conductive fibers and carbon nanotube networks” Nanotechnology, 24 (455502): 1-9, 2013.

87

High-Frequency Eddy Current System for Analyzing High-Frequency EddyWet Current SystemCoatings for Analyzing Conductive Coatings during Conductive duringWet Processing Processing2 1 2 3 3 Iryna Patsora , Susanne Hillmann , Henning Heuer , Bryan C. Foos , and Juan G. Calzada 1 Technische Universität Dresden, Fakultät Elektrotechnik und Informationstechnik, Iryna Patsora 1, Susanne Hillmann 2, Henning Heuer 2, Bryan C. Foos 3, Juan G. Calzada 3 Institut für Aufbau- und Verbindungstechnik, Dresden, Germany 1 Technische Universität Dresden, Fakultät Elektrotechnik und Informationstechnik, e-mail [email protected] Institut für Aufbau- und Verbindungstechnik, Dresden, Germany [email protected] 2 Fraunhofer Institute for Ceramic Technologies and Systems, 2 Branch for Material (IKTS-MD); Dresden, Germany Fraunhofer Institute for Ceramic Technologies andDiagnosis Systems, Branch for Material Diagnosis (IKTS-MD); Dresden, Germany 33Air

Air Force Force Research Research Laboratory; Laboratory; Dayton, Dayton, Ohio Ohio

ABSTRACT

Due to the growing usage of carbon fiber composites in the aircraft industry, the necessity of their protection against damage caused by a lightning strike increases. Instead of networks of copper wires located on the surface of these materials, wet conductive coatings are now being used as an alternative. The conductivity as well as the thickness of wet conductive layers has to be controlled during the application, which becomes possible by knowing their drying behaviors during processing. A nondestructive Eddy Current method is indicated for use due to the following advantages: works quickly, does not require direct contact with a deposited layer and responds to slightest changes in the electro-physical properties of the coatings. The system allows automatic data acquisition over the drying time of the wet coatings, so their drying behaviors can be visualized and analyzed. This paper explores a principle of the analysis of wet conductive coatings used for an algorithm on an example of high and low conductive coatings deposited on isolating and low conductive substrates with different parameters.

INTRODUCTION

Due to special properties, carbon fiber composites (CFRP) are being widely used in many industry branches, and are especially valuable for aircraft, wind power, automotive industry, etc. In order to protect these structures against damages caused by a lightning strike, wet conductive coatings are potentially effective to be used. However, the properties of wet conductive coatings, and especially their drying behaviors, are not sufficiently investigated at this time, which is important from the viewpoint of controlling the application process and namely conductivity and thickness changes during drying. Therefore, the aim of this work is to explore and analyze the drying behaviors of wet conductive coatings during processing. Such coatings are liquid after depositing and solid once dry. In order to investigate drying behaviors of the wet conductive coatings over the entire drying time, a nondestructive method should be selected, that can be applied in a wet state as well. A nondestructive Eddy Current (EC) method is indicated for use due to the following advantages: works quickly, does not require direct contact with a deposited layer and responds to slightest changes in electro-physical properties of coatings [1-4]. A High-Frequency EC System based on the EddyCus® system, designed and manufactured by Fraunhofer IKTS, was used in this work for coating analysis. The system allows automatic data acquisition over more than 24 hours, and it operates in a wide frequency range and allows recording of smallest changes in the specimen conductivity due to the specific configuration of the EC coil. Results of High-Frequency EC measurements on two types of wet conductive coatings (low and high conductive) deposited at different thicknesses and with dissimilar coated areas on ceramic Al2O3 and CFRP substrates are shown and discussed in this paper.

EXPERIMENTS High-Frequency Eddy Current System

The High-Frequency EC System used in this work is based on the EddyCus® system, designed and manufactured by Fraunhofer IKTS, is a prototype and consists of the EC sensor, electronics, laptop and software for data analysis and visualization. The sensor is integrated into a precision positioning table (Figure 1), that allows an exact adjustment of the lift-off with an accuracy of 10µm. This system permits automatic data acquisition over more than 24 hours, and operates at a frequency range of 100kHz up to 20MHz, and records smallest changes in the specimen conductivity due to the new specific configuration of the EC coil (Figure 2).

88

Figure 1: Picture of the mechanical housing of the Eddy Current sensor integrated into a precision positioning table.

Figure 2: Picture of the Eddy Current sensor (left), composition of the coil and the pre-amplifier inside of the sensor-box (middle) and picture of the Eddy Current coil (right). The EC coil used for these experiments has a special configuration. In most cases, EC coils consist of two wounded coils that are combined with each other in different ways. By reducing the influence of the exciting field in the measurement coil, the measurement signal can be increased in a large way, because the actual information about the electrical conductivity of the sample is included in the secondary field, which is detected by the measurement coil. The influence of the exciting field on the measurement coil can be reduced by a composition of both coils, where preferably, all electric flux lines of the exciting coil are closed inside of the measurement coil. In this case, only very small tensions are induced inside the measurement coil. To reach this goal, the exciting coil is placed inside of a ferrite cup core, which bundles and focusses the exciting field. The measurement coil is wound outside of the core. The spatial resolution of this configuration is affected by the size of the exciting field, related to the size and form of the exciting coil. The measurement coil can be much larger without increasing the spatial resolution. Additionally, a larger measurement coil increases the induced field to be measured. The measurement signal can be tripled using this type of coil. The preamplifier is located close to the inside of the coil inside the sensor-box to stabilize the signals und to decrease parasitic effects of stray fields and similar effects. Other advantages are higher sensitivity of the coil and higher spatial resolution in comparison to other high-frequency EC coils.

Properties of Wet Conductive Coatings

Wet conductive coatings are liquid after deposition on the substrate. While they are in a liquid state, they do not provide any electrical conductivity and the thickness of the coatings is at its maximum. During drying, chemicals evaporate, the layer shrinks and conductive particles move closer together. After curing, coatings are conductive, having reached a minimum thickness. This means, that the conductivity as well as the thickness of the wet conductive coatings are changing over the entire drying process, which strongly influences the EC measurements [2-4]. Due to the EC sensor used in this work operates at high frequencies, capacitive effects and displacement currents of the wet conductive coatings in a liquid and wet states influence the EC measurements additional to the inductive and conductive effects.

Preparation of Wet Conductive Coatings

Two types of wet conductive coatings are used in this work: low and high conductive pastes. Low conductive pastes are polymer-based lacquers, reinforced by silver-coated glass conductive particles; high conductive pastes are polymer-based lacquers as well, but reinforced by silver-coated copper conductive particles (Table 1). The curing time for both types of coatings is 24 hours at room temperature under normal conditions. Wet conductive coatings are deposited on the ceramic substrate by screen printing technique, but using the frame instead of the mesh, in three thicknesses (80µm, 160µm, 240µm) in accordance with the thickness of the copper frame used by the application. 89

Ceramic substrate is used because it does not have its own conductivity and the EC measurements are influenced only by the changes in the wet conductive coatings. Knowing the drying behaviors of the wet conductive coatings on ceramic substrate, the influence of the CFRP substrate on EC measurements can be investigated. Table 1:

Properties of additional particles for completing the sample plan (allocated by Potters Industries). Silver-coated Copper

Silver-coated Glass

Form of Particles

Flakes

Flakes

Averaged Size

3µm – 15µm

23µm

Photograph from datasheet

In accordance with requirements, next wet conductive coatings were deposited and measured: (1) High conductive coatings with coated area of 4×4cm on ceramic substrate. (2) High conductive coatings with coated area of 4×4cm on CFRP substrate. (3) High conductive coatings with coated area of 2×2cm on ceramic substrate. (4) Low conductive coatings with coated area of 4×4cm on ceramic substrate. After depositing on a substrate, each wet conductive coating is placed under the High-Frequency EC sensor and measurements start immediately over 24 hours at 30 second intervals and a lift-off of 80µm. Data obtained by EC measurements are represented in three dependences: real part of the complex voltage over the drying time, imaginary part of the complex voltage over the drying time and in the complex voltage plane. Due to the manual application process, the thickness of the coatings deviates from a desired value. Therefore, reference measurements on these conductive coatings were used for analysis: the final sheet resistivity and the final thickness of the conductive coatings were measured after curing. The thickness was measured using the Laser Profilometer and the sheet resistivity using the Keithley multimeter with a four-point probe [5].

RESULTS AND DISCUSSION

Based on EC measurements, the drying behaviors of the wet conductive coatings can be visualized and are described as follows: there are three dominating states during drying and these three states have both high and low conductive coatings (Figure 3):  State 1 occurs when the coatings are liquid and the chemicals used as a thinner evaporate and particle percolation starts.  State 2 occurs while the coatings are still wet, the added chemicals are evaporated and polymerization is in progress.  State 3 characterizes the end of the polymerization processes of the coatings. State 1 can also be called a “percolation threshold” [6]. This is the time of the drying, when the properties of the coating are being changed strongly according to the rapid evaporation of the chemicals. High conductive coatings have a distinctive feature in State 1 – a “characteristic point”. Low conductive coatings do not have it. It was experimentally established, that the “characteristic point” provides an opportunity to characterize coatings while they are liquid in accordance to the correlations between amplitude and final sheet resistivity and time of the “characteristic point” and final thickness. Correlations for the silver-coated glass-based layers are not possible in this state. State 2 occurs after the percolation threshold, when the added chemicals are evaporated and properties of the conductive coatings are changing slowly. The polymerization process is in progress. The coatings are wet in this State 2 so that repairs become possible. There are correlations for all measured conductive coatings and the characterization of each of them is possible. Sometime after polymerization process has started, the coatings are hardened enough and their properties are almost unchanged or are not changed at all. This characterizes a State 3, the end of the polymerization. In this state, no repairs are possible but correlations for all measured conductive coatings and the characterization of each of them is possible. 90

The time when the States 1, 2 or 3 occur depends on the chemical composition, environmental conditions, temperature of the curing, thickness of the coating, coated area etc.

Figure 3: Schematic representation of the drying behaviors for low (top) and high (bottom) conductive coatings. Figure 3 shows drying behaviors for silver-coated glass-based (top) and silver-coated copper-based (bottom) layers. Schematic representations show States 1, 2 and 3 for these coatings. For high conductive coatings it is seen, that the thicker the coating, the later a “characteristic point” occurs and the lower the amplitude is of the EC signal for them. In Figure 3 it is seen that the type of the particles used as a conductive filling strongly influences percolation processes in the wet coatings during drying and their final conductivity. It is also realized, that the time when States 1, 2 and 3 occur stays the same for layers based on copper and on glass particles. This means that the time of percolation, polymerization start and end does not depend on the type of the particles, but on the type of polymer matrix and the chemicals, used as a thinner. Conclusion: final sheet resistivity and drying behaviors in State 1 depends on the type of particles; the duration of each state and the time when the “characteristic point” occurs, depends on the type of a polymer and a thinner. Knowing the drying behaviors, the application process of the wet conductive coatings can be controlled.

Correlation Scope Characterization in State 1

Characterization in State 1 is only possible for layers having a characteristic point. Using the reference measurements the next correlations are possible: (1) High conductive coatings with coated area of 4×4cm on ceramic substrate:  Correlations between a real part of the complex voltage at the characteristic point at a frequency of 10MHz and the final sheet resistivityprediction of the final sheet resistivity of the layers, EC measurements should be started immediately after deposition and continued until the characteristic point has occurred.  Correlations between a time of the characteristic point of the real part of the complex voltage at a frequency of 10MHz and the final thicknessprediction of the final thickness of the layers, EC measurements should be started immediately after deposition and continued until the characteristic point has occurred. 91

(2) High conductive coatings with coated area of 4×4cm on CFRP substrate:  Even though the coatings have a characteristic point, there are no correlations possible in State 1 because CFRP substrates have their own conductivity and the conductivity of wet coatings in the beginning of drying is very low, which influences the EC measurements strongly. (3) High conductive coatings with coated area of 2×2cm on ceramic substrate:  Results show that it is only possible to characterize these conductive coatings using the correlations between a time of the real part of the complex voltage at the characteristic point at a frequency of 10MHz and final thickness prediction of the final thickness of the layers. EC measurements should be started immediately after deposition and continued until the characteristic point has occurred. (4) Low conductive coatings with coated area of 4×4cm on ceramic substrate:  Low conductive coatings do not have a “characteristic point” during the drying and thereby no correlations are possible for them in State 1. Figure 4 below illustrates correlations between the final sheet resistivity and the amplitude of the real part of the complex voltage at the characteristic point at a frequency of 10MHz for high conductive coatings deposited on ceramic substrate with coated area of 4×4cm.

Figure 4: Correlations between the final sheet resistivity and the amplitude of the real part of the complex voltage at the characteristic point at a frequency of 10MHz for high conductive coatings deposited on ceramic substrate with coated area of 4×4cm. Figure 5 shows correlations between the final thickness and the time of the characteristic point at a frequency of 10MHz for high conductive coatings deposited on ceramic substrate with coated area of 2×2cm.

Figure 5: Correlations between the final thickness and the time of the characteristic point at a frequency of 10MHz for high conductive coatings deposited on ceramic substrate with coated area of 2×2cm. Reference measurements used for these correlations are given in Table 2. 92

Table 2:Reference measurements for high conductive coatings deposited on ceramic substrate with different parameters.

Screen printing frame thickness

Final sheet resistivity of the high conductive coatings deposited on ceramic substrate with coated area of 4×4 cm

Final thickness of the high conductive coatings deposited on ceramic substrate with coated area of 2×2 cm

Thin d = 80µm

RF1-1 = 71.67–138 mΩ/□

d3-1 = 82.1-88.1µm

Middle d = 160µm

RF1-2 = 28.05-43.18 mΩ/□

d3-2 = 88.3-101.6µm

Thick d = 240µm

RF1-3 = 22.7-27.16 mΩ/□

d3-3 = 144.5-167.5µm

Characterization in States 2 and 3

Characterization of each of coating is possible in States 2 and 3. The time when State 2 occurs is approximately 70 minutes after deposition for all coatings used in this work. EC measurements can be performed 70 minutes after depositing, without performing them during the first hour. For different coatings, different frequencies should be used in accordance with a task and for getting the higher accuracy by predicting the final parameters. (1) High conductive coatings with coated area of 4×4cm on ceramic substrate:  Correlations between the amplitude of the real part or the amplitude of the absolute value of the complex voltage at a frequency of 13MHz of the complex voltage at a frequency of 13MHz and the final sheet resistivity.  Correlations between the amplitude of the imaginary part or the real part of the complex voltage at a frequency of 10MHz and the final sheet resistivity. (2) High conductive coatings with coated area of 4×4cm on CFRP substrate:  Correlations between the amplitude of the real part or the absolute value of the complex voltage at a frequency of 13MHz and the final sheet resistivity.  Correlations between the amplitude of the real part of the complex voltage at a frequency of 13MHz and the final thickness. (3) High conductive coatings with coated area of 2×2cm on ceramic substrate:  Correlations between the amplitude of the real part or the absolute value of the complex voltage at a frequency of 13MHz and the final Sheet Resistivity.  Correlations between the amplitude of the imaginary part of the complex voltage at a frequency of 10MHz and the final Sheet Resistivity. (4) Low conductive coatings with coated area of 4×4cm on ceramic substrate:  Correlation between the amplitude of the real part of the complex voltage at a frequency of 10MHz and the final sheet resistivity. Conclusion Coatings based on the same particles can be characterized at the same frequencies. By characterizing the high conductive coatings on both CFRP and ceramic substrates at the same time, a frequency of 13MHz (absolute value of the complex voltage | |) should be used as shown in Figure 6, for example. If silver-coated copper-based layers only on the ceramic substrate are characterized, the frequency of 10MHz (imaginary part of the complex voltage) is optimal, so that the highest accuracy can be reached. If only coatings on CFRP have to be characterized, a frequency of 13MHz (real part of the complex voltage) is the optimum. In regard to the low conductive coatings, only a frequency of 10MHz (real part of the complex voltage) is good.

93

Figure 6: Correlations between the final sheet resistivity and the amplitude of the absolute value of the complex voltage at a frequency of 13MHz for high conductive coatings with coated area of 4×4cm deposited on ceramic and CFRP substrates. Reference measurements used for these correlations are shown in Table 3. Table 3: Reference measurements for high conductive coatings deposited on ceramic and CFRP substrates with different thicknesses.

Screen printing frame thickness

Final sheet resistivity of the high conductive coatings deposited on ceramic substrate with coated area of 4×4 cm

Final sheet resistivity of the high conductive coatings deposited on CFRP substrate with coated area of 4×4 cm

Thin d = 80µm

RF1-1 = 71.67–138 mΩ/□

RF2-1 = 67.7-89 mΩ/□

Middle d = 160µm

RF1-2 = 28.05-43.18 mΩ/□

RF2-2 = 40.1-54.8 mΩ/□

Thick d = 240µm

RF1-3 = 22.7-27.16 mΩ/□

RF2-3 = 26.7-44.5 mΩ/□

By representing all measured layers in the complex voltage plane at a frequency of 10MHz, it can be observed that the changes of each parameter (type of particles, type of substrate and coated area) causes a shift of the drying curves, which provides an opportunity for separating coatings having different parameters, without knowing the time of the deposition as shown in Figure 7.

Figure 7: Drying curves of silver-coated copper-based and silver-coated glass-based complex voltage plane at a frequency of 10MHz. 94

All correlations will be recorded in an algorithm and in accordance to the task (parameters are known and should be evaluated), the optimum frequency will be automatically selected. This part of the work, namely the design/development of the algorithm is currently in progress.

CONCLUSIONS

High-Frequency EC measurements were performed on wet conductive coatings having different parameters. It was experimentally proven that there are three dominating states in wet conductive coatings during drying, and these states have both low and high conductive coatings deposited on ceramic and CFRP substrates. Drying behavior in State 1, called a percolation threshold, depends on the type of the particles used as a filling. Thus, silver-coated copper-based and silver-coated glass-based layers have different drying characteristics in State 1. Only silver-coated copper-based layers can be characterized in State 1, due to their characteristic point. Final sheet resistivity and drying behaviors in State 1 depend on the type of particles; the duration of each state depends on the type of polymer and thinner used. States 2 and 3 provide an opportunity for the characterization of all the coatings used in the work. Coatings based on same particles can be characterized at the same frequencies. There is an opportunity for separating coatings having different parameters, without knowing the time of deposition, by characterizing them in the complex voltage plane at a frequency of 10MHz. Based on results presented in this work, an algorithm can be developed allowing the prediction of the final parameters of wet conductive coatings in a wet state.

REFERENCE 1. 2. 3. 4. 5. 6.

I. Patsora; H. Heuer, S. Hillmann, D. Tatarchuk, Bryan C. FOOS, “Experimental setup for the characterization of the percolation behavior of wet conductive coatings by high frequency Eddy Current spectroscopy”, 978-1-4799-0036-7, 2013 IEEE, 36th Int. Spring Seminar on Electronics Technology. S. Hillmann, M.Klein, and H. Heuer, „In-Line thin film characterization using eddy current techniques“, Studies in Applied Electromagnetics and Mechanics Volume 35, ISBN 978-1-60750-749-9, 2011. S. Hillmann “Evaluation eines zerstörungsfreien Prüfverfahrens zur Ermittlung des Flächenwiderstandes flüssiger, leitfähiger Schichten; Dresden International University, Studienrichtung „Zerstörungsfreie Prüfung“; Masterarbeit, 2013. D. Lu, Q. K. Tong, C. P. Wong, “Conductivity Mechanisms of Isotropic Conductive Adhesives (ICA`s)”, IEEE Transactions on electronics packaging manufacturing, Vol. 22, No. 3, July 1999, pp. 223-227. F. M. Smits, “Measurement of sheet resistivities with the four-point probe”, The bell system technical journal, May 1958, pp.711 718. Stauffer D. and Aharony A. 1991 Introduction to Percolation Theory (London: Taylor & Francis).

95

A Comparison of Acoustography with other NDE Methods for F.O. for Inclusions A Comparison of Acoustography with other NDE Methods F.O. Detectiondetection in Carbon Epoxy Laminates Inclusions in Carbon Epoxy Laminates 1 1 1 1 1 1, and Charles Pergantis Anish Poudel , Shashi Shrestha , Jaswinder TsuchinPhilip PhilipChu Chu Anish Poudel , Shashi Shrestha , Jaswinder Singh Singh Sandhu Sandhu22,, Tsuchin , and Charles Pergantis3 3 1 Department of Mechanical Engineering and Energy Processes, Southern Illinois University 1 Department of Mechanical Engineering and Energy Processes, Southern Illinois University 1230 Lincoln Drive, Carbondale, IL 62901 1230 Lincoln Drive, Carbondale, IL [email protected] 62901 (618) 453-7049; fax (618) 453-7658; e-mail Tel. (618) 453-7049, Fax: (618) 453-7658, Email: [email protected] 2 Santec Systems, Inc. 2 Santec Systems, Inc. IL 60005-4726 2924 Malmo Drive, Arlington Heights, 2924 Drive, Arlington Heights, IL 60005-4726 (847)Malmo 215-8884; e-mail [email protected] Tel. (847) 215-8884, Email: [email protected]

U.S. Army Research Laboratory 3 Charles Pergantis Aberdeen Proving Ground, MD U.S.(410) Army306-0688; Research e-mail Laboratory, Aberdeen Proving Ground, MD [email protected] Tel. (410) 306-0688, Email: [email protected] 3

ABSTRACT In this research, the through-transmission ultrasonic (TTU) Acoustography non-destructive evaluation (NDE) method was employed to detect foreign object inclusion (FOI) defects embedded in carbon fiber epoxy composite laminates. The study employed three different test specimen with varied size FOI defects embedded at varying depth within the composite laminates. For validation, Acoustography results were compared with conventional immersion TTU and infrared thermography (IRT) methods. Also, quantitative measurements based on Signal-to-Noise Ratio (SNR), defect sizing, and inspection times were carried out to compare and correlate between different NDE methods that were applied. The Acoustography method, operating at 5MHz, was easily able to detect all FOIs defect and was also able to resolve two Teflon defects placed side by side (separation = 1.5 mm) at different layers within the composite test panels. SNR measurements for Acoustography were more than 6:1 and were in good correlation with TTU UT and IRT results. The defect sizing ability of TTU Acoustography for FOIs defect in graphite epoxy composite laminates were also in strong correlation with the TTU UT and IRT techniques. Finally, TTU Acoustography inspection time was about 1 minute to scan 300 x 300 mm2 (11.81 x 11.81 inch2) area which was faster compared to conventional TTU C-Scan, and IRT techniques. This is a very significant finding because Acoustography is being developed as a faster, more efficient, and affordable alternative to traditional ultrasonic inspection systems for composites. Keywords: NDE, Acoustography, TTU C-Scans, IRT, Composites, and F.O. Inclusions.

INTRODUCTION Advanced composites such as carbon fiber reinforced plastics (CFRP) are increasingly being used in many structural applications ranging from aerospace to automotive, industrial, sports industry, and many other consumer products. The main reason for this is, unlike traditional metals and their alloys, composites offer outstanding thermal and physical properties which include high strength and stiffness to weight ratios, low coefficient of thermal expansion, high fatigue resistance, inherent corrosion resistance, and low electromagnetic reflectance [1-5]. However, composites are prone to defect mechanisms which can occur either during the processing stages or during in-service operations, and repair environment. In addition, due to the heterogeneous nature of composites, the form of defects is often very different from those typically found in traditional metals and their alloys, and the fracture mechanisms are much more complex [6]. Some of the common manufacturing defects that can occur during the production process of composites include micro-cracks, fiber breakage, voids, delamination, porosity, and foreign object inclusions (FOIs). FOIs usually occur during the manufacturing process mainly due to foreign debris accidentally included in parent material during manufacture. Some of the common examples of FOIs include pre-preg backing paper or separation film, which is inadvertently left between plies during layup, and peel ply. Moreover, FOIs can come in many different forms and existing NDE techniques are often encountered with difficulties in detecting and characterizing FOIs when they have similar acoustic impedance compared to the composite parts. FOIs can have

96

degrading effect on mechanical properties and may act as sites of stress concentration and potential initiation sites for more serious defects such as delaminations and disbonds in composites. This may produce disastrous effects if not identified and corrected in timely manner. In severe cases, FOIs can also directly threaten safety of flight crews and integrity of the aircraft. TTU C-scan, followed by ultrasonic pulse echo A-scan inspections are the primary NDE techniques utilized by industry as quality assurance (QA) checks for the production of composite parts and structures. These inspection techniques require tedious point-by-point inspection of parts that can be time-consuming. Acoustography NDE provides an alternative to point-by-point scanning and offers new capabilities to the NDE engineer. It could be a simple, fast, and economical alternative to conventional UT for the inspection and evaluation of certain composite components [7-10]. This paper examines and compares the application of TTU Acoustography with conventional immersion TTU UT and Infrared Thermography (IRT) for FOIs detection in CFRP laminates. In addition, quantitative comparisons were also carried out to correlate Acoustography with other NDE techniques.

How Acoustography Functions Acoustography is a broad-area, nearly real-time ultrasonic imaging technique that provides an alternative to point-bypoint UT [11-14]. In this approach, a novel, wide-area acousto-optic (AO) sensor is employed to provide whole-field ultrasonic images similar to real time x-ray imaging, as shown in Figure 1.

(a) (b) Figure 1: Comparison of the whole field imaging technology (a) Acoustography; (b) X-ray. Through-transmission ultrasonic (TTU) Acoustography employs an AO-sensor (detector) made from proprietary mesophase liquid crystal (LC) material. In the TTU Acoustography setup, the front-side of the AO-Sensor is exposed to an acoustic field and the backside of the AO-Sensor is viewed by a camera while exposed to polarized light. As ultrasonic beams propagates from the sound source towards the AO-Sensor, any flaws or voids that it encounters will produce a differential attenuation. As the beam propagates past the flawed/anomalies region, it caste an ultrasonic “shadow” along its path which is instantly converted into a visual image by the acoustic-optic sensor located underneath in near real time as shown in Figure 2.

Figure 2: Working Principle of TTU Acoustography.

97

The physical principle by which the AO-sensor converts ultrasonic waves into visual images is based on the birefringent properties of LC materials contained in the sensor [9]. The LC layer exhibits no birefringence in the absence of ultrasound and exhibits a uniform dark field under cross-polarizer viewing. But, when an ultrasonic beam is exposed on the AO-sensor, the LC layer becomes birefringent, showing the brightness change (optical density change) under cross-polarizer viewing as shown in Figure 4. The brightness change can be related to the ultrasonic intensity by AO-sensor acousto-optic transfer curve [14].

Figure 3: Effect of ultrasound exposure on AO-sensor. (a) Without ultrasonic exposure; and, (b) with ultrasonic exposure.

EXPERIMENTAL DESCRIPTION This section illustrate on the description of composite test panels that were used for this research work. It also describe the laboratory test setup of an Acoustography system.

Sample Description

Three CFRP test panels with engineered FOI defects were considered for this study. Panel A and panel B were fiberreinforced 16-ply composite laminates, thickness/ply = 0.21 mm (0.008inch), with a symmetric orientation, [[0/45/0/45]2]S as shown in Figure 4. Similarly, panel C was a step-wedge laminate which had similar ply orientation as panel A and B but its plies varied from 2 to 28 plies as shown in Figure 4. Panel A and panel B sizes were approximately 300 mm x 300 mm (11.81 inch x 11.81 inch) with a cross-sectional thickness of approximately 3.36 mm (.13 inch). Similarly, panel C was approximately 150 mm x 350 mm (5.91 inch x 13.78 inch) with a crosssectional thickness that ranged from approximately 0.42 mm - 5.88 mm (0.02 inch - 0.23 inch).

(a)

(b)

98

(c) Figure 4: Engineered defect map in CFRP panels. (a) Panel A; (b) Panel B; and, (c) Panel C.

Laboratory TTU Acoustography Inspection System Figure 5 shows the actual laboratory test set-up for TTU Acoustography NDE. The sound source used was fabricated using a 76.2 mm x 76.2 mm (3 inch x 3 inch) Piezo-electric plate with a center frequency of 5 MHz. The AO sensor used was selected to have an operating frequency of 5 MHz to match that of the sound source.

Immersion Tank

Workstation and PC

Optics & Electronic Housing

Sound Source

AO-Sensor

Test Panel

Figure 5: Laboratory test set-up of TTU Acoustography technique; the experimental Acoustography system and close-up view of the sound source, test panel, and scanning system. For the comparison purpose, immersion UT and IRT described in an outside work [1, 15] were also used during the experimentation.

99

RESULTS AND DISCUSSION This section presents the experimental results obtained for the composite test panels by using different NDE techniques. It also presents the results on the quantitative comparison between NDE methods that were applied.

TTU Acoustography Results Figure 6 shows the TTU Acoustography result for the CFRP test panels containing various simulated FOI defects. Before testing, the AO sensor was calibrated to relate AO sensor brightness to ultrasound power. This was done by measuring the brightness as the function of ultrasound power without part and with test part respectively. The 255 gray shades in the images shown in Figure 6 represents the ultrasound attenuation variation across the sample; darkest shade (i.e. 0) representing the highest attenuation and brightest shade (i.e. 255) representing the lowest attenuation. FOIs in the test panels exhibiting higher acoustical impedance than the CFRP attenuated the transmitted ultrasonic signals. All images were acquired in near real-time. In order to generate images of an entire part, a series of images recorded in sequence were stitched using the Acoustography software.

1

2

3

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(i)

100

Figure 6: 5 MHz TTU Acoustography results for CFRP panels with embedded FOIs defect. (a)-(c) Panel A at 0-10dB, 4-10dB, and 0-30 dB; (d)-(f) Panel B at 0-16dB, 4-16dB, and 0-30 dB; and, (g)-(i) Panel C at 0-12dB, and 0-16 dB respectively.

TTU UT Results For the direct comparison and validation of TTU Acoustography results, 5 MHz TTU UT and active infrared thermography methods were also co-inducted in all test panels. Figure 7 shows a typical TTU UT C-scan results at 5 MHz. The scans were conducted by using a pair of flat transducers with element size = 9.53 mm (.375 inch) and placing the test panels in the far-field of the transducers (approx. 76.2 mm = 3 inch). The sampling rate = 125 MHz, scan/index increments = 1 mm (40/1000 inch), and scan speed = 5.1 mm/sec (.2 inch/sec) were used during the TTU scans.

1

2

3

(a)

(b)

(c)

(d)

(e)

(f)

(g) (i) Figure 7: 5 MHz immersion TTU results for CFRP panels with embedded FOIs defect. (a)-(c) Panel A at 010dB, 4-10dB, and 0-30 dB; (d)-(f) Panel B at 0-16dB, 4-16dB, and 0-30 dB; and, (g)-(i) Panel C at 0-12dB, and 0-16 dB.

101

IRT Results Figure 8 shows a typical infrared thermography (IRT) results for the CFRP panels with embedded FOIs defect. During experimentation, a continuous heat flux was applied for approximately 5 seconds. The samples were then allowed to cool, and the temperature responses were recorded, which allowed for a temperature variation within the discontinuous areas. The infrared camera recorded the data between the time interval from approximately 3 seconds prior to heat flux application and up to 12 seconds after the heat flux was removed. Best image in the sequence with the highest thermal contrast captured around 8.5 seconds i.e., during the initial cooling stage, were considered. The highest thermal contrast images were used for further analysis.

1

2

3

(a)

(b)

(c) Figure 8: IRT results for CFRP panels with embedded FOIs defect. (a) Panel A; (b) Panel B; and, (c)-(i) Panel C. From the results presented in Figure 6-8, it is demonstrated that the ability of TTU Acoustography to detect all FOIs defect in graphite epoxy composite laminates is in strong correlation with the TTU UT and IRT techniques. It is also demonstrated that Acoustography technique was able to detect the smallest inclusion (diameter 3 mm) embedded at different layers as shown in Figure 6 (a-c). In addition, this novel technique was also able to resolve two FOI defects embedded at different layers which were separated at 1.5 mm (.06 inch) distance as shown in Figure 6 (d-f) as the corresponding TTU C-scan and IRT results as shown in Figure 7-8. However, the definition of the defects seemed better in the Acoustography method. Also, the fibrous nature of composites were also more evident in the Acoustography imaging results. The superior definition of defects and the fibrous nature of the specimen may be attributed to the superior pixel resolution of the AO-sensor used in this novel technique: sensing LC molecules in the AO-sensor are in the order of 20 angstroms. Although, the AO-sensor pixel resolution is very fine, the pixel resolution of the camera and monitor is much lower compared to that offered by the AO-sensor. Accounting for the lower resolution of the camera and monitor and the magnification factor, the effective resolution of the acoustographic images were around 0.23 mm.

102

Quantitative Comparisons For quantitative comparisons of results obtained by three different NDE techniques, the Acoustography technique was directly compared with conventional immersion TTU C-scan and Infrared Thermography (IRT) on the basis of Signalto-Noise Ratio (SNR), defect sizing, and inspection times. The SNR was calculated by using: 𝑆𝑆𝑆𝑆𝑆𝑆 = (

𝜇𝜇𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 − 𝜇𝜇𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 𝜎𝜎𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠

(1)

)

where, µsound is the mean gray scale value for the sound region, µdefect is the mean gray scale value for the defect region, and σsound is the standard deviation value of gray scale for the sound region. The histograms showing the mean and standard deviation value for the sound and defect regions in panel A for all NDE techniques utilized are shown in Figure 9. These were used to calculate SNR for different NDE techniques that were applied for this research work.

(a)

(b)

(c)

(d)

(e)

(f)

Figure 9: Histograms showing mean and standard deviation values for the sound and defect regions in panel A. (a) Sound region-Acoustography 0-10 dB result; (b) Defect region-Acoustography 0-10 dB result; (c) Sound region-TTU UT 0-10 dB result; (d) Defect region-TTU UT 0-10 dB result; (e) Sound region-IRT result; and (f) Defect region-IRT result. Table 1 compares the SNR of three NDE methods, specifically, Acoustography, conventional immersion TTU C-scan, and Infrared Thermography methods for three different CFRP panels with embedded FOIs defect. From the results presented SNR measurement for Acoustography were more than 6 and are also in good correlation with TTU UT and IR results. Table 1: Signal-to-Noise Ratio (SNR) measurements. Panel A Panel B Panel C

TTU Acoustography 6.68 11.35 6.39

TTU C-Scan 4.36 4.54 0.81

IRT 2.74 10.18 7.97

Similarly, the defect sizing measurements were also conducted in panel A row 1 defects for each of the NDE techniques. The defect sizing were calculated by first calibrating the image scale to its original dimension and then by calculating the pixels in the defect regions. Figure 10 shows the line profile plots for CFRP panel A row 1 defects as shown in Figure 6(a), 7(a), and 8(a) for each of the NDE methods.

103

(a) (b) (c) Figure 10: Line profile plots for CFRP panel A row 1 defects. (a) TTU Acoustography; (b) TTU UT; and, (c) IRT. Table 2 lists the defect sizing measurements for panel A row 1 defects for each of the NDE techniques applied. From the results presented in Table 2, it is demonstrated that the defect sizing ability of TTU Acoustography for FOIs defect in graphite epoxy composite laminates is in strong correlation with the TTU UT and IRT techniques. Table 2: Defect sizing measurements for panel A row 1 defects. Defect #

NDE Techniques

1

TTU Acoustography TTU C-scan IRT

2

3

TTU Acoustography TTU C-scan IRT TTU Acoustography TTU C-scan IRT

Measured diameter (mm)

Actual diameter(mm)

Percentage Error

19.18 19.33 19.93

20 20 20

4.06% 3.33% 0.37%

19.56 19.33 19.19

19.56 18.67 19.93

20 20 20

20 20 20

2.21% 3.33% 4.06%

2.21% 6.67% 0.37%

Finally, the inspection time for each of the NDE methods applied for this research were compared. Table 3 shows the inspection time results for 300 mm x 300 mm (11.81 inch x 11.81 inch) panels. The inspection time required by using TTU Acoustography was about 1 minute and was much faster compared to TTU C-Scan, and IRT techniques. Table 3 compares a typical inspection time for three NDE methods to scan 300 x 300 mm2 (11.81 x 11.81 inch2) composite panel. Table 3: Inspection time for three NDE methods. Laboratory TTU Acoustography TTU C-Scan IRT Scan Speed 76.2 mm x 76.2 mm /shot 5.1 mm/s 100 mm x 150 mm /shot Indexing Steps 76.2 mm 1 mm 100 mm Image Generation Time 10 sec* N/A 1 sec Image Capture Time 30 ms N/A 18 sec Image Storage Time 1 sec N/A 20 sec Number of Shots 56 N/A 6 Time between Shots 1 sec N/A 60sec Typical Inspection Time 5 mins 3 hrs. 10 mins Area = 300 x 300 mm2 (11.81 x 11.81 inch2); Panel Thickness = 3.36 mm (.132 inch) (Typical) It is to be noted that the TTU C-scan inspection time was relatively higher because tests were conducted by using old immersion UT system. With the modern state-of-the art gantry/squirter systems, the scanning speed in the X-Y axis can be set to 300 mm/s (11.81 inch/s) can be achieved. With 0.5 mm indexing steps, it will require 600 passes to scan 300 x 300 mm2 (11.81 x 11.81 inch2) composite panel. This will reduce the inspection time for TTU UT to approx. 10 minutes. It should be mention that Acoustography scan time can be dramatically reduced by using larger field of

104

view (FOV) Acoustography NDE system. For example, currently, a 12 inch x 12 inch field of view (FOV) Acoustography NDE system is undergoing development. This will allow 12 inch x 12 inch area to be inspected in only 10 seconds: 60 times speed advantage over traditional point-by-point TTU scans.

CONCLUSION The through-transmission ultrasonic (TTU) Acoustography NDE method was employed to detect foreign object inclusion (FOI) defects embedded in carbon fiber epoxy composite laminates. Three different composite test panels with varied size FOI defects embedded at varying depth were considered for this work. The Acoustography method, operating at 5MHz, and was easily able to detect FOIs defect in the test composite laminate. A side-by-side comparison of Acoustography technique with convention immersion TTU and IRT techniques showed a good correlation between three NDE methods in detecting foreign object inclusions (FOIs) defects in composite test panels. The lateral resolution of Acoustography was found to be superior to the conventional TTU C-Scan method. Also, the flaw detection sensitivities of Acoustography were also found in good agreement to the conventional TTU C-Scan and IRT methods. Finally, from the operator’s point of view, the Acoustography technique was shown to be significantly simpler to operate and determine results because of the minimal skill level required for the inspection process. Currently, a 12 inch x 12 inch field of view (FOV) Acoustography NDE system is undergoing development. This will further reduce the number of images required to inspect larger parts, thereby proving a, an even more dramatic increase in inspection speed compared to other existing techniques.

REFERENCES [1]

Chu, T. P., Poudel, A., and Filip, P., 2012, "C/C Composite Brake Disk Non-Destructive Evaluation by IR Thermography," Proc. SPIE: Thermosense - Thermal Infrared Applications XXXIV, Vol. 8354, Baltimore, MD.

[2]

Chu, T. P., Don, J., Pan , Y., and Poudel, A., 2012, "Defect characterization in commercial Carbon-Carbon composites " World Journal of Engineering, Vol. 9(6), pp. 481-486.

[3]

Poudel, A., and Chu, T. P., 2012, "Intelligent Nondestructive Testing Expert System for Aircraft Carbon/Carbon Composite Brakes Using Infrared Thermography and Air-coupled Ultrasound," Materials Evaluation, Vol. 70(10), pp. 1219-1229.

[4]

Poudel, A., Strycek, J., and Chu, T. P., 2013, "Air-Coupled Ultrasonic Testing of Carbon/Carbon Composite Aircraft Brake Disks," Materials Evaluation, Vol. 71(8), pp. 987-994.

[5]

Poudel, A., Sandhu, J. S., Chu, T. P., and Pergantis, C., 2014 "Porosity Measurement in Carbon Fiber Epoxy Laminates by Using Acoustography," Proc. 23rd Annual Research Symposium and Spring Conference, Minneapolis, MN.

[6]

Matzkanin, G. A., and Yolken, H. T., 2008, "Techniques for the Nondestructive Evaluation of Polymer Matrix Composites," The AMMTIAC Quarterly, 2(4), pp. 3-7.

[7]

Sandhu, J. S., Wang, H. H., and Popek, W. J., 1996, "Acoustography for Rapid Ultrasonic Inspection of Composites," Proc. SPIE: Nondestructive Evaluation of Materials and Composites, Vol. 2944, pp. 117-124.

[8]

Sandhu, J. S., Wang, H. H., Popek, W. J., and Sincebaugh, P., 1999, "Acoustography: A Side-by-Side Comparison With Conventional Ultrasonic Scanning," Proc. SPIE: Nondestructive Evaluation of Aging Materials and Composites III, Vol. 3585, pp. 163-172.

[9]

Sandhu, J. S., Wang, H. H., Popek, W., and Sincebaugh, P. J., 2001, "Acoustography: It Could Be A Practical Ultrasonic NDE Tool For Composites," Proc. SPIE: Nondestructive Evaluation of Materials and Composites V, Vol. 4336, pp. 129-134.

[10]

Sandhu, J. S., Wang, H. H., Sonpatki, M. M., and Popek, W. J., 2003, "Real-time full-field ultrasonic inspection of composites using acoustography," Proc. SPIE: Nondestructive Evaluation and Health Monitoring of Aerospace Materials and Composites II, Vol. 5046, pp. 99-104.

[11]

Sandhu, J. S., 1987, "Non-Coherent frequency source and sector scanning apparatus for ultrasonic imaging system using a liquid crystal detector cell," U. S. P. Office, ed.U.S.A.

105

[12]

Sandhu, J. S. and Thomas, R. E., 1988, "Acoustographic Nondestructive Evaluation," Proc. IEEE Ultrasonics Symposium, Vol.1052, pp. 1053-1056.

[13]

Sandhu, J. S., 1988, "Acoustography: A New Imaging Technique and its Applications to Nondestructive Evaluation," Materials Evaluation, Vol. 46(5), pp. 608-613.

[14]

Sandhu, J. S., 1995, Acoustograhy: Special nondestructive testing methods, Nondestructive Testing Handbook, 2nd Edition, Vol. 9, ASNT, Columbus, OH.

[15]

Poudel, A., Kanneganti, R., Gupta, L., and Chu, T. P., 2013, "Nearest Mean Classifier for Defect Classification in CFRP Panels," Proc. ASNT 22nd Research Symposium and Spring Conference, Memphis, TN.

106

Some New Results in Coiled Oilfield Tubing Inspections

SOME NEW RESULTS IN COILED OILFIELD TUBING INSPECTIONS RodericKKStanley, Stanley, Ph. Roderic Ph. D., D.,I.I.Eng Eng Coiled Tube Resources Management Coiled Tube Resources Management Houston, Texas 77035 Houston, Texas 77035 Abstract Coiled tubing (CT) strings are now regularly 35,000-ft. long. The seam weld in HSLA coiled oilfield tubing has been difficult to inspect for longitudinal imperfections and defects even with the use of UT. In-line ET units are ineffective on the tight defects found is the seam weld of this EW tubing. In this paper we present results from the use of phased array UT (PAUT) on the seam weld, and also on the 45° skelp-end welds that join the strips together before the tubes are milled. Magnetic flux leakage systems have evolved for both the final inspection after hydrostatic test, and for field inspections. Improvements to the suppression of excessive MFL from the seam welds is discussed, and some typical results of recent inspections are given.

Introduction Prior articles (Stanley 1998, 1999, 2001) have described coiled oilfield tubing as welded HSLA steel pipe, meeting the API 5ST specification (API 2010). Its use has grown tremendously in recent years, and its quality has improved. The physical properties of current grades are given in table 1. Note that there are now restricted maximum yield strengths for materials commonly used in H2S work. Chemistries are ASTM A607, A607, or 1101 Grade

Table 1: Physical Properties of Coiled Tubing Yield Strength (min.) Yield Strength (max.) Tensile Strength (min.) psi MPa psi MPa psi MPa

Hardness (max.) Body & Weld HRC

CT70 70,000 483 80,000 552 80,000 552 22 CT80 80,000 552 90,000 620 88,000 607 22 CT90 90,000 620 100,000 689 97,000 669 22 CT100 100,000 689 108,000 745 28 CT110 110,000 758 115,000 793 30 Note: Cycling on and off the storage reel causes the yield strength to lower, and eventually, transverse cracks to

occur prior to failure

NDT of Strip: Hot rolled steel strip (36-48-in.wide) is produced from slabs extremely fast, and other than being wall thickness gauged by eddy current sensors, is not inspected for mid-wall and small surface defects. Slag in the steel can lead to mid-wall laminations (fused or unfused) which would pass on into the final product (Fig. 1) if not removed, this certainly being the case. NDT of Strip Welds: In API 5ST, all strip welds are inspected by radiography, and the customer can request a re-inspection by UTSW after the weld has passed RT. Figure 2 shows the results of one such inspection with PAUT at 10 MHz (Stanley 2005), Figure1: Inclusion in strip rolled into in which a crack is detected. lamination. This is a very quick and simple inspection, performed with a hand-held device, since the weld crown has been removed prior to inspection. In-line NDT of Seam Welds: The seam weld requires careful attention. While most mills have in-line saturated-field eddy current units, experience shows that they just do not find the smaller tight imperfections common to seam welds (e.g. hook cracks, penetrators, figure 3 (L)) that can come apart either in the mill’s final hydrostatic test, or during

107

service. What they do find are seams open to the OD surface, and larger balls of spume either hanging from the seam, or lying inside the tubing. This is because since most coiled tubing has variable wall thickness (the thicker end being

Figure 2: Skelp end weld under inspection (L) and signal from crack in weld on 3rd leg

at the top of the string and the well, typically 0.204-in. to 0.125-in.), no solution to removing the internal flash from multiple thickness tubing has yet been devised. Instead, a restriction is placed on the size of the internal flash column (Figure 3) which is not always met during production, and because of this flash, cannot be tested for except at the ends of the string. Thus designing a UT system for this weld presents challenges.

Figure 3: (L) A row of penetrators in a cold weld section of a 0.750-in tube. The internal flash line runs along the centre of the picture, (R) the head of a PAUT system for CT inspection, with a reference standard.

Phased array UT at 5 MHz (McCoy et. al., 2002) was selected after testing several CT sizes, this proving to be a challenge. First the sound beam had to flood the seam weld area from both sides, but there had to be no signal from the internal flash column. This flooding had to cover all the common wall thicknesses, from 0.125-in. to 0.204-in. Then, the system, located after the weld seam had been water-cooled, had to detect 1/32nd holes in the flash line, and ID and OD machined notches 0.25-in. long, with the mill line speed of 100 ft/min. Further, the system had to be accurate even if the weld wandered between the 10.30 and 1.30 positions, as welds do. Figure 3(R) shows the system, as developed. It uses mill coolant, maintained at a constant temperature, fed into a box with 2 curved transducer shoes, each with 128 small PA sensors, and located in the OD reduction sizing section of the EW mill. These shoes stand roughly 1.00-in. from the pipe surface. Once the mill was running, the lead end of the tubing had to thread though the coolant box via rubber gaskets. Two “quickscans” were used for phasing the transducers, and collecting reflected seam weld ultrasound. A high PRD was used. Reflections from the root of the flash weld column are suppressed by comparing the signal from successive reflections (roughly 1 mm apart as the tubing proceeds through the inspection head) and making informed decisions as to their removal.

108

Standardizing the unit also presented a problem. To accomplish this, a notched sample is rotated under the transducer beam (Fig 3(R)), and the signal amplitude collected for reflected sound from the notch in all positions. The system then self-adjusts so that the amplitude response across the array is relatively flat, irrespective of the notch location. Images of “Quickscan” screens for the 2 sets of probes (on either side of the seam weld line) are shown in figure.

Figure 4: (L) The screens seen by the UT inspector. There is one set of 3 screens for each set of 128 transducers. (R) Reflections from each size on the seam weld line, showing an imperfection and the seam weld crystalline anisotropy.

Typical Results: The operator’s full screen is shown in figure 4 (L). One form of the seam weld screen is shown in figure 4(R). Each rectangle represents 1 mm along the length of the string. This scan shows a small reflection from and imperfection in or close to the seam, which is made visible by the scattering from the anisotropy created in the seam weld grains (darker blue). The light blue areas indicate little scattering from the steel tube wall. Seam weld defects readily detected. This imperfection is readily apparent and its length can be measured since the pixels are 1mm long. The unit marks the tubing for further prove-up with PT, RT or handheld PAUT. Off-line Final NDT: Several companies have offered offline NDT, as a final inspection following the mill’s hydrostatic test. This hydrostatic test can do one of 3 things: (a) nothing, if there are no seam weld flaws (Pass), (b) blow a hole in the seam weld (Fail), or (c) open up a seam weld imperfection without causing a leak (Pass). These last often come apart prematurely when the tubing is cycled, or exposed to acids (15% HCl is commonly used in CT operations). Axial MFL (Stanley 2012) is used for imperfection detection, and either UT or magnetic reluctance (Stanley 1996, 2004) is used for wall thickness. Electromagnetic (ref) or contact (ref) methods are used for diameter and ovality. In one tool, the radial MFL field is measured with an encircling array of hall elements in closely-fitting shoes, and the wall thickness is measured by an encircling array of hall elements measuring the tangential field at a more substantial lift-off. Diameter is measured by 6 eddy current sensors placed around the tube, and ovality is calculated at 0.5-in intervals from Θ = 200(Dmax – Dmin)/ (Dmax + Dmin) from a least-squares fit to the diameter data.

Figure 5: MFL (Threshold channel), wall thickness (max, min, avge). MFL is due to OD pitting in this new string. The wall contains 3 sections of continuously tapered strip.

109

No ultrasonically based NDT tool has evolved for offline new tbg inspection other than the measurement of wall thickness with UTCW, and for that the OD surface has to be very clean. Typical results: Unremoved internal spume, damage to the tube OD created after the in-line inspection, and anything found during the in-line inspection are also found. Since the skelp-end welds exhibit different magnetic permeability from the tube, they can generally be seen on the MFL channel, so that this examination of the tube provides baseline data for the purchaser. OD surface flaws are removed during this inspection, to within the wall thickness tolerances allowed in API 5ST. Figure 5 shows a long section of a tube with three continuously tapered strips, and a lot of OD pitting.

Figure 6: (L) Pitting in new tube (due to Houston’s atmosphere). ( R) MFL from a small area of the tube.

Figure 6 shows some of the round-bottommed pitting (L), along with a 3D zoom of the area shown. Figure 7 shows a skelp end weld (45 deg.) in the same tubing string and emitting a considerable amount of MFL. Seam weld magnetic noise (from the internal flash column) is also visible. Stretch is measured by detecting these welds. Field NDT: This has emerged from failures in the N. Sea (and elsewhere), and is mandated for all offshore CT operations in the Norwegian sector. It does not appear to have caught on in the western hemisphere, except in critical circumstances. Rust and scale exclude the use of UT, so the MFL/ET/MR methods are used for wall loss, pitting, gouges, diameter changes, ovality and stretch. The MFL sensitivity is less than that of the mill tool, as the sensors have a larger lift off because of the possibility of raised metal at gouges. Figure 8 shows part of a string in which there is little MFL but a whole lot of variation in its OD, and also therefore ovality. This is relatively common for strings that are nearing the

Figure 8: Record for part of a string with variable OD and ovality, reaching 5%.

end of their useful life. One company (Torregrossa M., et al 2014) has mounted such a tool on top of its CT rigs and inspects the tubing both entering and coming out of the well. The changes in the tube are therefore almost continually monitored during

110

its lifetime (Figure 9). Typically, increasing MFL determines flaw growth, the distance between skelp-end weld MFL signals determine the tubing stretch, and increasing ovallity is monitored. Figure 9 shows the operator’s screen, with (a) the fold out map of the tubing (open tubing), (b) MFL amplitudes, (c) a vibration channel, (d) the max, min and avge wall thickness, ( e) diameters, and (f) an ovality computation vs distance.

Figure 9: Operator’s screen with data over 11,000-ft of tubing. This is the first run off the storage reel.

The map shows the seam weld line (light blue). The amplitude channel shows large MFL from 125-134 mil and 134150 mil skelp-end welds, and less from gauge-gauge welds. The vibration channel is very helpful in determining whether a MFL signal originates from a flaw or serious tubing vibrations. Prove-Up: Figure 10 shows (L) signals from lack of fusion in a butt weld and (R ) H2S to which a tubing sample.

Figure 10: (L) LOF in a weld and (R) reflections from H2S damage in a used tubing sample

111

had been exposed. Plough marks (Figure 11) appear to be the most serious OD flaw, as they can be sharp bottomed and therefore quickly cause fatigue cracks, (and therefore premature tubing failure) especially in the presence of H2S. Closing Statements: Systems have been developed which inspect coiled tubing both in the mill and the field, and, along with repair of sharp bottomed imperfections, are beginning to prolong the life of this thin-walled tubing. Nomenclature API ASTM CT ET EW HSLA ID MFL PT NDT OD PAUT PRD RT UT UTSW

American Petroleum Institute American Society for Testing and Materials Coiled Tubing (as used in the oilfield) Eddy Current Testing Electric weld High Strength Low Alloy Inside diameter Magnetic Flux leakage Penetrant testing Nondestructive Testing Outside diameter Phased Array ultrasonic testing Pulse repetition Density Radiographic Testing Ultrasonic Testing Ultrasonic Shear Wave

Figure 11: Plough Marks (Caused by the CT rig

injector)

References Specification for Coiled Tubing, Amer. Petroleum Institute (API), Washington 2010. Stanley, R. K, “Results from Inspections of Coiled Tubing,” Proceedings of the SPE/Icota Conference, Houston, Texas, Society of Petroleum Engineers, April 1998 Stanley, R. K, “Results of Recent Inspections Performed on Coiled Tubing,” Proceedings of the SPE/Icota Conference, Houston, Texas, Society of Petroleum Engineers, May 1999. Stanley, R. K, “Problems Associated With Coiled Tubing,” Proceedings from the International Chemical and Petroleum Conference, Houston, Texas 2001. McCoy, T., R Rosine, C Aulert, J Martin and R. K. Stanley, “HPHT Wells Create New Challenges of Coiled Demanding Higher Levels of Inspection and Management,” SPE paper74830, Proc. SPE/Icota Mtg, April 2002, The Woodlands, Tx. Stanley, R. K., “Magnetic Methods for Wall Thickness Measurement and Flaw Detection in Ferromagnetic Tubing and Plate,” Insight 38(1), pp 51-55. Stanley, R. K, “Observations on Magnetic Wall Measurements on Coiled Tubing,” Materials Evaluation, Feb. 2004, pp 125-128. Stanley, R. K., “Some New Developments Applicable to Coiled Tubulars,” SPE Paper 94058, SPE/Icota meeting, The Woodlands TX April 2005. Stanley, R. K., “An Assessment Tool for Coiled Tubing,” Materals Evaluation 70(6), June 2012, pp 765-774. Torresgrossa, M., L Zsolt,Zwanenberg, M., “Optimizing Pipe Management with a New Approach of Coiled Tubing Integrity Monitoring,” SPE paper 168303, SPE/Icota meeting, March 2014, The Woodlands, TX.

112

Does Structural Health Monitoring (SHM) Provide Safety andand Maintenance or or Does Structural Health Monitoring (SHM) Provide Safety Maintenance Confusing Data? Confusing Data? 1 2 1 1 2 1 Jon Watson3 Terry Tamutus , Michael Johnson , Richard Gostautas , and Terry Tamutus , Michael Johnson , Richard Gostautas , and Jon Watson3 1 Mistras Group Princeton, NJ P 1 Mistras Group Princeton, NJ P (609) 468-5737; e-mail [email protected]

(609) 468-5737; e-mail [email protected] 2

Caltrans

2

Caltrans

4Mistras UK

3

Mistras UK

34

For over 20 years structural health monitoring has had a few sad stories. Several years ago in N.Y., a SHM system was installed on a new bridge. No one was looking at the reams of data, nor did anyone know what it meant. The bridge authority’s lawyer suggested if something happened, and no one looked at the data, it may be a liability. After the first year the smart bridge was retired. A smart bridge should give answers, not data. In some cases Structural Health Monitoring (SHM) can provide immediate answers. This is one remarkable case that illustrates a smart bridge example. The project was also peer reviewed by the Transportation Research Bureau (TRB) and proof tested, which was planned, approved, and supported by three universities.

Copy right protected

Background

In 2009, a 12-inch long crack was discovered on a fracture critical eyebar on the San Francisco-Oakland Bay Bridge. The crack extended from the outer edge of the eyebar head to the pin. The bridge had been inspected in 2007 with no indication of cracking. Following this discovery, dye penetrant testing was performed on the eyebars and Caltrans ultrasonically tested all critical pins. They also increased the inspection frequency from two years to every month. Bridge inspection always presents a significant challenge. First, periodic visual inspection can be time consuming and disruptive to bridge operations. According to Caltran’s website an average of 277,000 commuters use the bridge daily, so their inspections were conducted at night in poor lighting and weather conditions. Additionally, figure 2 shows a typical situation where the eye bar heads configuration renders sections inaccessible for inspection. As a result of these difficulties, a remote monitoring method was needed, which would provide continuous, real-time alarms for crack initiation. The SHM system would complement Caltrans inspections and monitor large areas; some of which were impossible to inspect.

Figure 1: Crack from 2009 along the eyebar.

Vendor Selection and Procurement

Caltrans initiated a comprehensive study for SHM companies. They used a list taken from a 2009 MinnDOT/University of Minnesota report titled, “Bridge Health Monitoring and Inspections Systems A Survey of Methods” which contain 77 SHM firms. Caltrans

113

Figure 2: Typical pin connection showing limited access to eyebar heads.

selected the top seven firms and used a performance based RFP so vendors could select the best method at their disposal. In July 2010, Caltrans selected Acoustic Emission and the largest manufacturer of these instruments. In this project, several of the performance specifications are listed below:  Real time monitoring and alarming  Location accuracy within 18 in  Crack size detection of >0.1 in.  Filter the crack signals from extraneous noise In 2010 MinnDOT/University of Minnesota published a second report titled Development of an Advanced Structural Monitoring System (MN/RC 2010-39). This further backed Caltrans decision to use Acoustic Emission and its selected vendor, as the vendor rate was 96% in the 2010 Survey. Part of MinnDOT’s grading matrix is listed below and how they scored various SHM companies.

What is Acoustic Emission (AE)?

As stress causes deformation to the material, Acoustic Emission (AE) is released. As such, AE is passive and relies on the detection of the rapid release of energy generated by sources within the material, such as crack initiation or propagation. It is then detected and recorded by AE instrumentation (Figure 3). However, if AE is improperly used, it may require significant data analysis and by experienced personnel. However, through specialized systems, software, and experience, these challenges can be overcome by proper selection frequency, number, placement/spacing of sensors as well as detection threshold settings, extracted feature-sets of the waveform and proper selection of high- and low-pass filters.

114

Figure 3: AE monitoring schematic.

Lab Based Proof Test

Prior to installation, MISTRAS was required to prove or demonstrate that its equipment and technology worked per the contract performance specifications. Two full size eyebars were made and two 12” diameter pins are seen in figure 4. The bottom end is fixed and the top end has a 60 KIP actuator for fatigue testing. In 2012 the Journal of Transportation Research Board, published a paper titled “Real Time Eyebar Crack Detection using Acoustic Emission Method” on this proof test. This laboratory demonstration program also developed a pattern recognition algorithm, for use in a software classifier, for differentiating between different types of AE sources (e.g. fretting, crack initiation or propagation, personnel activity, etc.). The classifier evaluates selected parameters of the waveform (e.g. amplitude, energy, frequency centroid) from each located AE event compares the parameters against a series of tests. The AE event is then assigned a source type classification based on which tests were passed. For the source type classification related to potential cracking, an additional algorithm monitors for a Figure 4: Full-Scale AE System Lab Proof Test. certain set of conditions, which if satisfied, triggers an monitoring alarm and immediately notifies selected personnel. SHM automation was crucial for safety, dataschematic. management and alarm notification to bridge engineers.

Installation

In mid-2011 installation began on the Bay Bridge. Caltrans personnel simulated crack signals verifying that the system worked in real time. Four teams of two people carried out sensor and system installation. On a large-scale installation like this, logistics of a project have to be closely considered. Small organizations typically cannot provide safe, large turnkey system installation due to bonds and insurance. Prior to bid power estimates, electrical power was thought to be available. But during the installation electrical engineers decided that a new line needed to be run, which incurred additional costs.

Answers, Not Data

On Monday, July 16, 2012, at 9:03 AM, an email alarm notification was triggered by system on the bridge. Within an hour Caltrans personnel arrived to inspect the location. The team discovered an electrical conduit rubbing against the eyebar as the bridge vibrated up and down (Figure 6). This sawing motion wore away approximately 0.125 in. of eyebar material. The alarm location was within several inches of the problem. Within four hours of the alarm notice, Caltrans remediated this issue with $40 worth of foam pipe insulation.

Figure 5: Sensor installation on selected eyebar.

115

SHM Complements Inspection and Monitors Large Areas Impossible to Inspect Even with the best inspectors, visual alone may not identify this type of mechanism during regular inspection cycles. This location was easily accessible from the deck and might be considered in plain sight, but the conduit itself had masked the deterioration. Regular painting and maintenance also concealed the wear. Caltrans determined that additional metal loss would have resulted in a large stress concentration at the edge of the eyebar; similar to the condition that led to the crack from 2009. The cost for the monitoring system (design to installation), which was critical for safety, was a small fraction compared to the approximately $14 million spent for the eyebar replacement.

The Old Bridge Retired Safely

On this massive structure, SHM pinpointed the problem in real time, without vendor intervention or costly data Figure 6: Eyebar material loss due to frictional wear analysis. To date this project is the largest known SHM caused by contact with adjacent conduit. application where only 640 AE sensors monitor the entire length of 384 eyebars. If these eyebars were stretched end-to-end they would total 3.94 miles. The AE systems proved cost effective on a large-scale SHM project and accurately identify structural issues with a high degree of accuracy —unlike vibration, strain and fatigue sensors mentioned earlier. Additionally, this success provided Caltrans a high level of confidence that allowed AE to be turned on for other bridges and not shut down because of overwhelming data. As a result, the 16 AE systems will be redeployed to other bridges. Meanwhile, the 78-year old Bay Bridge recently made it to full retirement after much wear and tear and many earthquakes.

Summary of Acoustic Emission Structural Health Monitoring

To summarize, Acoustic Emission Structural Health Monitoring has been proven to provide accurate real time location of micro-cracking on extremely large structures with no operator intervention. These system are selfcalibrating or self-testing to reduce time on-site. There have been 1,000s of AE SHM projects throughout the world but never on a scale this large. Other technologies were compared and evaluated against AE such as strain gages, vibration, fiber optical, etc. and Acoustic Emissions performance specifications and equipment costs, installation ease, results exceeded Caltrans expectations by taking a massive structure and pinpointing a small problem. Since the SFO Bay Bridge is now being demolished, the AE systems have been removed from the bridge and will be redeployed to other bridges and problems. Discussions are underway to monitor weld cracks, post tensioned wire breaks, suspension cable wire breaks, and eye bars on other Caltrans bridges.

ACKNOWLEDGMENTS Dr. Didem Ozevin of University of Illinois, Dr. Glenn Washer of University of Missouri, Dr. Steve Song and Dr. Kanji Ono of University of California Los Angeles for their support during this project. REFERENCES Average Daily Traffic: 270,000 vehicles http://www.dot.ca.gov/hq/esc/tollbridge/SFOBB/Sfobbfacts.html MinnDOT/University of Minnesota report titled, “Bridge Health Monitoring and Inspections Systems - A Survey of Methods” 2009 MinnDOT and University of Minnesota. Development of an Advanced Structural Monitoring System (MN/RC 2010-39). available at http://www.lrrb.org/pdf/201039.pdf

116

Arturo Schultz, Principal Investigator, Department of Civil Engineering, University of Minnesota, 2010 The Journal of Transportation Research Board, published a paper titled “Real Time Eyebar Crack Detection using Acoustic Emission Method” on this proof test. Authors: Johnson, Michael B, Ozevin, Didem, Washer, Glenn A, Ono, Kanji. Gostautas, Richard S Tamutus, Terry A Publication Serial: Transportation Research Record: Journal of the Transportation Research Board ISSN: 0361-1981 Date: 2012 Transportation Research Board: Selected Field Results: San Francisco Oakland Bay Bridge Acoustic Emission Monitoring Program Authors: Michael B. Johnson P.E., California Department of Transportation, Richard S. Gostautas, Mistras Group, Terry A. Tamutus, Mistras Group Report 14-2859, 2014

117

The C.A.P. System: Clarity, Accuracy, and Punctuality in NDT Reports

The C.A.P. System: Clarity, Accuracy, and Punctuality in NDT Reports John G. Taylor Nova Data Testing Inc. John G. Taylor 85107 Commercial Park Drive Nova Data Testing Inc. Yulee, Florida 32091 85107 Commercial Park (866) 674-6806 Drive Yulee, Florida 32091 (866) 674-6806

The CAP system stands for Clarity, Accuracy and Punctuality. These are the 3 fundamentals in producing a quality NDT report for the customer. It has to be clear so that even those who are unfamiliar with NDT can comprehend its intent. The results must be accurate and leave the reader without doubt that all procedures and instructions were followed in a proper manner. The report needs to be punctual; the results must reach the end user in a timely manner so decisions can be made. When used together each one of these fundamental principles will make for an effective report. As a professional your signature at the bottom of the report will reflect your commitment to producing quality workmanship. The report is your canvas. Sign it proudly. As the business consultant John C. Maxwell observes, “You will never change your life until you change something you do daily. The secret of your success is found in your daily routine.”

CLARITY

Your writing is a reflection of your thinking. Clear writing will make you a valuable asset to any team. One of the best ways to improve your writing skill is to read aloud from writers who write clear and distinct for 10 minutes a day. Don’t focus on content as much as style. Listen for the drastic contrast between the vague and careless styles often used in daily conversations and the carefully crafted words of great writers. Always use courteous, dignified and appropriate language without being stuffy. Use specific words and avoid clichés like “all set” or “operationally ready”. Avoid redundant phrases like “assembled together”, “final completion” or “total number”. Some words can actually insult the reader. Cut out the gibberish. Avoid words that appear to say something but don’t. These are words that require neither effort nor thought. Journalists are taught not to “Bury the lead”. They put the headline first and don’t hide it in the story. The same is true for the reports NDT professionals write. Let the customer know the results in the first few sentences. Let the details of the inspection do the talking. If the report delivers results the customer does not want to hear, wording should be calm and clear. Always remember the reader can’t read your mind, what is obvious to you may require an explanation to the engineer. Always use a logical and detailed approach when developing content. Once the reader becomes confused or misunderstands the information, the reader can become distracted. A clear and concise report prevents that. Present important information using short words and sentences. The report will be clear and memorable. The series of events taking place during the inspection should be in the proper order. This allows the reader’s mind to easily follow the sequence. Use an active voice where the subject does the action. Keep sentences orderly with subject, verb and then object. For example “The weld was cracked” instead of “There was a crack in the weld.” Keep the rule consistent within the report. The report must also reflect the terminology that the customer uses at their facility. A tank at plant A might be called simply the S.D.A. or spray dry absorber while at plant B it’s the sulphur dioxide atomizer. Avoid using acronyms or slang terms for the parts that are being inspected. Identifying the equipment with the proper nomenclature reinforces your commitment to quality reporting. It also enhances your understand of the process. Make reports flow smoothly and thoroughly review them when completed. Try running the report through readability statistic like the Flesch Reading Ease or FRE test. A readability score between 60 and 80 allows the reader to stay engaged with the topic.

118

One important question to ask when reviewing the report is what information does the reader want from the report. Certainly the reader wants results of the test, but be aware of the reader. The format may be dictated by the audience and their specific expertise in the parts that were inspected. A design engineer may want more photos and drawings than text. A maintenance manager may derive better information from a graph, while someone in operations may prefer a statistical forecast.

ACCURACY

A report can only be as good as the work that is being presented. Any flaws in the design or in the execution of the inspection that are detailed in the report cannot be corrected or disguised. Avoid the cumulative effect of errors. The more errors there are in the report the worse it will look and sound to the reader. Each individual error, in your opinion, may not reduce the accuracy by very much. However these errors may be very evident to the reader. A report peppered with mistakes can cause the reader to stop reading and they may never again read anything with your name or your company’s name on it. Poor grammar and sloppy writing contribute to lowering the reports credibility. Conclusions and results will be viewed as weak and inept. While conducting the inspection use a reliable system for keeping track of items to include in the report and make reliable and readable notes that help assure accuracy of the report. Many codes and procedures include check lists, but often the best lists are blended using experience and a consensus from other inspectors. Part of being accurate is being honest. Avoid using code words or statements such as “It is clear that more additional inspections will be required before a complete understanding of this phenomenon occurs”. Tell the customer you don’t understand. Instead of saying “Typical results from this inspection and the analysis of the effects show”. Rather than commenting on results, it may be productive to present the reader a colorful graph or images of the condition. Avoid using subjective adjectives and stick with highly specific words and phrases. Remember that statements of fact will command the customer’s attention. One problem all writers encounter is the reliance on the computer’s spell check tool. Is it a fillet weld or a filet weld? Both words are acceptable to your lap top. Is that weld twelve inches or pinches? Try to get a second reader to review the text. A good hint is to start at the end and read backwards. This technique makes any small errors scream for repair. To improve accuracy revisit the report after a time away from it. When required be sure to reference the most current procedures and codes that were used in developing the inspection and the report. If the included checklist is from the fourth edition but the fifth edition was published six month ago your reader may question its accuracy. Besides reviewing current procedures read old reports. The old report may reinforce the finding of your inspection. But it can also contradict the results. If this occurs a detailed explanation should be included to defend your analysis and your accuracy. To improve future reports follow up with the customer to see how effective it was in achieving the objectives and adjust future reports for that company accordingly.

PUNCTUALITY

Punctuality simply measures how well you are able to manage your time. If your time is not valuable to you then other people’s time will also not be valuable to you. Late reports disturb the experience of other people and it puts a strain on your relationship with a customer. It can hurt your career and it will add more stress to your life. It is usually caused by misperceiving the passage of time or underestimating how long a project will take. The first step in having on time reports is to begin planning your time. Start at the end and work backwards. If you have three days to complete a written report make your time line with specific goals that need to be achieved daily. The main focus to on time reports is to work on your own powers of concentration. Be proactive and manage your time to fulfill your promise to deliver an outstanding report. Give your work the highest priority and learn to say “No”, to other projects explaining your report is critical and constrained by a deadline. An honest discussion with your manager or the customer reinforces your professionalism and commitment to the task.

119

There are a number of ways to improve your time management. First see what time of day works best for you. Then schedule as large a block of time as necessary to complete the detailed task. Keep a picture of the project at hand and develop the report around that. Try to finish the report before you start to edit it. Take breaks and quit when you get tired. But quit in the middle of a sentence so that by the next day you can get started without hesitation. Write on a sticky note you will be completed by a given time and then post that note conspicuously where you and everyone else can see it. One effective technique for organizing a report is called “Mind Mapping”. This is simply a group of words with circles drawn around them. The groups are connected by lines with other groups. Write the main point of the report in the middle of the page and then let the ides for building the document grow from there. You can use pictures, symbols, words, color and images to take a list of monotonous information into an organized and memorable report. Strive to make the report flow with the brain’s natural way of processing information. Have a fresh document preformatted with the font, margins, spacing, footers and headers necessary. Develop a typical report layout that can add or delete sections depending on how comprehensive a report is required. The lap top and tablet are great tools for creating faster reports. This technology allows for rapid information transfers and on line results. Ten simple rules to create great reports. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Keep the report short while including all essential information. Remember to value your reader’s time. Keep it organized for the convenience of the reader. Avoid clichés, jargon and redundant phrases. Use a great summary that gives a picture in miniature. Check and check again for spelling, grammar, layout and accuracy. Write to improve the reader’s understanding of the content. Keep all references current. Provide results that can stimulate action. Improve your reports with customer feedback.

Why would anyone want to struggle to improve their report writing skills? Because your managers and customers will recognize you possess strong analytical skills. It shows a passion for your work, how you present facts, and the results you produce. This skill is what employers are looking for in their company. Ask yourself what types of inspections and what types of equipment will you be inspecting in the next 20 or 30 years? We can only guess what will be cutting edge and what will be obsolete? The only thing that is certain is results will need to be presented in a clear, accurate and punctual report.

REFERENCES

1. 2. 3. 4.

John Maxwell, The 15 Invaluable Laws of Growth: Live Them and Reach Your Potential, Center Street, 2012 Rudolph Flesch, A New Readability Yardstick, Journal of Applied Psychology, 1948 Tony Buzan, The Mind Map Book, Penguin Group, 1996 Clear Writing with Mr. Clarity, Blogspot.com, December 2013

120

Introducing Young People to NDT Introducing Young People to NDT Phillip W. Trach Phillip W. TrachInc. Laboratory Testing, Laboratory Testing, 2331 Topaz Drive Hatfield, Inc. PA 19440 2331 Topaz Hatfield, 19440 (800) 219-9095; fax (215)Drive 997-8294; e-mailPA [email protected] (800) 219-9095; Fax (215) 997-8294; e-mail [email protected]

INTRODUCTION

As the current population of NDT practitioners ages, it becomes necessary for the industry to recruit and train new practitioners. The author’s experience is unique in that there was an existing family connection to the NDT industry; most potential NDT professionals have little or no direct connection to NDT. Therefore, active recruitment of young people into NDT is necessary. One challenge presented by this requirement is that of giving NDT presentations to young people. “Young people” as used in this paper means children of elementary school age through undergraduate university students. This paper will give an overview of how to prepare, practice, and present an introductory lesson in NDT suitable for a school classroom with students who are not currently actively preparing for an NDT career. Three cases from the author’s experience will be examined.

BACKGROUND

In a recent study, it was found that the average age of an NDT technician is 48 years old with 21.6 years of experience (PQNDT, 2013). This puts the average starting point for an NDT career at age 26. At 48 years old, this average technician is 17 years from a typical retirement age of 65, well past the halfway point in his or her career. This also indicates that, on average, technicians are starting their careers well after their school careers end. This indicates a lack of information within schools; a student who learned about the NDT industry and had an interest in it would not wait until his or her mid-20’s to begin a career in NDT. Numerous articles and editorial columns in NDT publications have highlighted the need to recruit new personnel (Allgaier, 2010; Peloquin, 2013). However, there has been little activity in educating or training existing NDT practitioners to present NDT information to young people. Furthermore, the skills and experience of a skilled NDT practitioner do not usually include classroom teaching beyond conducting specialized training classes for existing employees, usually in a workplace setting. These classes also do virtually nothing to generate public interest in the field. In short, training a new hire, no matter how well it is done, does not generate interest in NDT by students outside the workplace. Furthermore, by the time a student has entered high school, he or she has already made a number of choices regarding his or her education which will have a strong influence on available options later. Ninth or tenth grade, or sooner, is not too early to introduce a young person to NDT (Sander, 2014).

COMPOSING THE PROPOSAL

Nearly every professional teacher uses a lesson plan. A lesson plan is an outline of the content of the day’s lesson, along with specific information regarding the objective of the lesson, the core information of the subject for that day, the materials and activities which will be used, and usually some kind of assessment of how well the students understood and retained the lesson. The author does not intend to give comprehensive instruction in lesson planning. However, knowing the essential elements of a lesson plan will allow an NDT professional to begin crafting an NDT lesson according to a format which a professional educator can understand, and from which a workable lesson plan can be made (with some assistance from the classroom teacher). Furthermore, following an existing template can help ensure that nothing gets missed.

121

The following is a typical lesson plan structure (Sander, 2014): Lesson: (what you intend to teach) Grade: (obtain student grade level from the teacher) Overview: (explain why it’s important and how it relates to the subject) Objective: (why your lesson is important) Essential Question: (what your lesson answers) Materials: (equipment and supplies you need) Activities: (things you and the students will do as part of the lesson) Assessment: (how you will know the students understood your lesson) The initial goal for constructing this lesson plan is to be prepared to discuss your proposal with a teacher. It is far more productive to discuss a concrete plan with a teacher and make changes to it, than to begin the conversation with only a vague idea of giving a lesson on NDT, particularly when the teacher knows little more about NDT than the students. The goal of this initial discussion is to answer the question: “Why is it worthwhile to me to have my students learn about it?” (Sander, 2014). The better prepared you are to answer this question, the better your discussion will go. It has been the author’s experience that teachers are generally receptive to having guest lecturers, especially when the guest brings real-world applications of the subject into the classroom. This puts an NDT practitioner in the position of giving a concrete answer to the constant student question, “Why do we have to learn this?”. The author has found that taking a “Six W’s” approach to crafting an NDT lesson works well for completing a lesson plan form such as the one above. The “Six W’s” are: Who What When Where Why How

(the author counts the “w” at the end)

Who will you be teaching? You will take a different approach to a third grade science class than you would with a high school social studies class. In general, older students, and mathematics and physical science classes allow a more technical approach. What will you be teaching? This is more than the NDT method you choose; this will include how you intend to connect your NDT experience with the classroom subject. When will you teach it? You may need to flexible. For example, an ultrasonics lesson with lots of trigonometry will need to wait until the students have developed the mathematics skills to perform the calculations. You will also need to work within the available class time. Where will you teach it? Some portions of your lesson may not lend themselves to a typical classroom. You will also need to know what equipment is available in the classroom, what equipment you will need to bring, and how to set up your equipment in the room. Why will you teach it? This question should be answered in terms the teacher understands and can apply to the students. How will you teach it? You, as a potential guest lecturer, need to display the connection between NDT and the subject; the teacher will want details on how you intend to do that. This will generally be the longest portion.

122

Once these questions are answered, any lesson plan form can be at least partially completed. Once that is done, the details of the lesson can be discussed with the teacher and modifications can be made. The more prepared you are for this initial discussion, the more likely your proposal will be received well.

PROPOSAL TO THE TEACHER

Ideally, you should get a copy of a typical lesson plan from the teacher of the class you intend to visit. This allows you to adapt your initial answers to the Six W’s to best meet the needs and objectives of that specific teacher. Be prepared to give an introductory explanation of both the importance of NDT in today’s world, as well as a brief overview of any NDT method(s) you intend to use (Trach, 2013). In the words of one teacher, “We’re not technical people. We like pictures.” (Sander, 2014.) Also be prepared to adapt your lesson to any specific needs the teacher has. Keep in mind that you will be an interruption to the teacher’s curriculum. While the author’s experience has been that teachers are generally quite willing to have guest lecturers, you will be making some extra work for them. Be ready to accommodate their needs. Be ready to both ask and answer the question “Are there other ways NDT can be tied to the curriculum?”

PREPARING THE LESSON

Once you and the teacher have worked out the basics of your approach to your lesson, you can begin crafting the details of your lesson. You will refer often to your lesson plan, always keeping in mind the central question: “Why is this important?” The central theme of your lesson will be the answer to this question. The author has found that a single sentence is most effective for focusing the other parts of the lesson. Each statement or activity will generally be linked to this sentence (Trach, 2012). A comprehensive description of all the possible lesson approaches is beyond the scope of this paper; indeed, it would be tantamount to several semesters’ work in educational theory and curriculum development. However, a few simple approaches are worth mentioning.

Telling Stories

Particularly with younger students, or with older students who would benefit from “real world applications”, telling stories of your experiences can generate interest in your lesson. The challenge here is to be exciting. Your story is familiar to you, but brand-new to your class. Be lively, emphasize the skill and technology in your story, especially to older students. Even the most technical paper is, at its base, a story.

Demonstrations

Demonstrations can be very effective, especially when the students can try your equipment for themselves. However, this requires additional time; make certain you will have enough time both for demonstrations and for the explanation of what you show the class. The danger here is to give merely a demonstration of “cool science stuff” without connecting it to the world in which the students live. It is important when considering a demonstration in a classroom to remember the limitations of the classroom. For example, it may not be possible to get a classroom dark enough for a fluorescent penetrant test to be easily visible. Radiography using a portable source would be extremely difficult (if not impossible) to perform inside a school building. The small screen on most portable ultrasound units may not be visible to the entire class. Consider also the time it takes. Waiting 15 minutes for liquid penetrant to dwell on a part may not be the best use of time. In general, you will be illustrating principles more than conducting NDT training, so it may be better to have specimens prepared for viewing without spending classroom time performing processing.

123

First Principles

Starting with the basics and showing how a few simple physical phenomena can be connected to make a system that reveals amazing material conditions can be effective and thorough, provided you can keep the students’ attention. Particularly in a mathematics class, showing the expressions which govern certain aspects of a nondestructive test, then explaining the effects when various variables are manipulated, can give a compelling aspect to otherwise dry equations, especially when combined with a demonstration.

Coordination with the existing curriculum

No matter how the lesson is crafted, it is important that the lesson generate specific outcomes. “What will the students learn?” is the key question. “How will you know?” is an important secondary question. Your task is to make sure that what the students learn from your presentation is a natural extension of what they’re already learning in their regularly scheduled class. This is an area where flexibility on your part combined with communication with the teacher will help. Be open to suggestions; the teacher is the expert at teaching. Rely on that expertise, but it is very helpful if you do as much preparation as you can. If necessary, review the subject so that you can better craft your own lesson to match what the students have already learned and will learn in the lessons following yours.

Organizing the lesson

Once you have answered the Six W’s and have some idea how you want the lesson organized, you must expand your lesson plan to include the specific details of your lesson. This includes elements such as: Introductory statements Stories Questions for the class to attempt answering Demonstrations Participatory exercises You must plan each step of your lesson from the teacher’s introduction of you all the way to answering the last question from the last student. You will lose the attention of your audience if you are unprepared or disorganized. Each step should have some connection to the previous step and to the next step. The nature of these connections will depend on the steps themselves, your teaching style, and other factors which will be specific to a particular lesson and presenter. It is far more important to have a systematic and logical progression in place than to adhere to any particular style of progression. Have a plan, even if it’s not an optimum plan.

PRESENTING THE LESSON

The lesson plan at this point is essentially complete. Discussion with the class teacher has happened, ideas have been investigated, and the Six W’s have been answered in an organized format. The task then becomes answering, in person, the core question which the students (ideally) will have: “Why is this important?” There are a few significant differences between instructing trainees in an NDT department and instructing school students. Most significantly is the level of educational interest which either a paycheck or a tuition bill generates in a professional training setting. The presenter’s task in a traditional classroom setting is to capture and hold the students’ attention. The novelty of a guest lecturer helps with gaining attention at the beginning of the lesson. A lively and energetic delivery can help to hold the students’ attention, though it is important not to sacrifice content for the sake of being merely entertaining. Practicing the lesson, even to an empty room, will help with timing and delivery.

124

Practicing also includes practicing any demonstrations you may have. Do not skip over them or merely rehearse them in your mind. Pick up and use each tool, prop, and system you plan to use. This will show you if anything is missing, or if anything should be prepared beforehand, and allow you to get used to how your lesson flows when you start demonstrating instead of merely talking. Above all, know your material, know your plan, then follow the plan.

EXAMPLES

The author presents three examples of classes he has taught, at the elementary, junior high school, and undergraduate levels. Each presented its own challenges. In all three lessons, ultrasonic testing was the NDT method used.

Elementary Lesson

The lesson was presented at an evening boys’ group at a local church. The one-sentence summary of the lesson was “What’s inside counts.” Several aluminum blocks, some with holes drilled in varying orientations, were covered with tape to obscure the hole locations. A portable UT system was used to demonstrate the differences between a “good” block (with no holes drilled in it) and a “bad” block (that had holes drilled in it). The kids saw that even though several items may look identical on the outside, it is the things inside which really make the difference.

Junior High Lesson

The lesson was presented in a sixth-grade mathematics class at a nearby private school. The one-sentence summary of the lesson was “Word problems happen in real life.” The basic equation “distance equals rate times time” was presented as a common word problem about a car driving from one place to another. Each aspect of the word problem was replaced by an element of ultrasonic thickness testing until the students saw that this common word problem has a real-world application which is used every day. Students then performed a thickness test, including manually calculating the part thickness from the acoustic velocity of the material and the measured time-of-flight of the ultrasonic signal.

Undergraduate Lesson

The lesson was presented several times at a nearby university. The one-sentence summary of the lesson was “How to make a living doing NDT.” A number of short stories of the author’s experience in NDT, from initial training to developing new internal practices, were told. Additionally, classroom descriptions and demonstrations of typical materials testing situations were displayed. These acted as an adjunct to and extension of the students’ work on system calibration earlier in the semester.

CONCLUSION

In order to “raise up the next generation” of NDT practitioners, educational presentations in the classroom by experienced NDT practitioners are needed at all grade levels. Thorough and thoughtful preparation is extremely important, both for a practitioner to be welcomed into a teacher’s classroom, and for the practitioner to effectively introduce young people to NDT.

REFERENCES 1. 2. 3. 4. 5. 6.

PQNDT, “Salary Survey 2013”, PQNDT, Arlington, MA. 2013. http://www.pqndt.com/NDT-Salary-Survey/PQNDT-2013-Salary-Survey.pdf Allgaier, M. (2010). “NDT Education and Training: Today and in the Future”, Materials Evaluation, Vol. 68, No. 9, pp. 972-977. Peloquin, E. (2013). “Education Fosters Growth in Nondestructive Testing’s Future”, Materials Evaluation, Vol. 71, No. 10, pp. 1034-1039 Sander, M. Personal interview, 24 June 2014. Trach, P. “Introducing NDT to Visitors”, ASNT Fall Conference. Las Vegas, NV. 2013. Trach, F. and Trach, P. “Developing Presentation Training Skills”, ASNT Fall Conference. Orlando, FL. 2012.

125

ACKNOWLEDGEMENTS

The author wishes to acknowledge the support of the following: Laboratory Testing, Inc. Michael Sander, Social Studies Department, William Tennent High School, Warminster, PA. Vladimir Genis, Ph.D., Engineering Technology Department Head, Drexel University, Philadelphia, PA New Life Presbyterian Church, Glenside, PA Calvary Christian School, Fairless Hills, PA Keywords: Education, High School, STEM

126

Effect of Varying Inspection Parameters in Crack Effect of Varying Inspection Parameters in Crack Depth Measurements Depth Measurements Using Potential Drop Method Using Potential Drop Method David Utrata and Darrel A. Enyart DavidState Utrata and Darrel A. for Enyart Iowa University, Center NDE Iowa State University, Center for NDE 1915 Scholl Rd., Ames, IA 50011 Scholl294-7771; Rd., Ames, IA 50011 (515) 294-6095;1915 fax (515) e-mail [email protected] (515) 294-6095; FAX (515) 294-7771; [email protected]

SUMMARY When working with industrial clients, the NDE Group of Company Assistance within the Center for Nondestructive Evaluation at Iowa State University often straddles the area between research and application. For example, it is well known that the potential drop method may be used to measure crack depth in metallic specimens. But of interest to our manufacturing clientele is to define how much sample preparation and data interpretation would be needed if they wanted to apply a commercially available off-the-shelf (COTS) device to their particular inspection need. This paper documents use of a COTS device that uses alternating current potential drop (ACPD) to measure crack depth. It was applied to a variety of calibration samples having known cracks and notches, as well as to crack depth measurement on a real industrial component. The calibration samples had varying contact area geometries and were intended to simulate a range of industrially relevant materials. Results with the real component indicated the great utility of such inspection methodology, although there are limitations to its application. The results discuss the degree to which the industrial user can expect to get meaningful performance out of such test devices. Caution for various test geometric configurations are indicated, and some basic guidelines for adequate surface preparation and probe placement are offered.

THEORETICAL BACKGROUND The alternating current potential drop (ACPD) has been known for years and its implementation well documented [1-4]. Figure 1 illustrates the parameters that are controlled and values that are measured in this test; this schematic is used with permission from Saguy & Rittel [2].

Figure 1. Schematic of values of interest in ACPD test from [2]. Very briefly, a current is injected into a test piece and the electrical circuit closed at some point along the surface. Between the points where the current flows in and out, a voltage difference (potential drop) is measured. If the injected current traverses a distance within which a crack is located, measurement of the voltage differential will

127

vary, dependent upon the crack depth. Using alternating electrical current confines the current flow to a “skin” on the sample. A break in the component surface creates a longer path for the electrical current to travel, thus increasing its resistivity and voltage. This difference in voltage is known as the potential drop, and COTS devices, when properly calibrated, will use this value to indicate the depth of such surface breaking cracks. Variables other than crack depth that will affect readings have been identified in the literature. The electrical conductivity of samples will determine the value of an effective skin depth in the material, dictating how deep the current will travel in different material. As the sample thickness decreases, the current density will likewise be affected. Manufacturers of ACPD devices recommend proper use of these tools with supplied calibration blocks. Additionally, users are encouraged to then calibrate the devices on both crack free regions on test material, as well as adjust device response over notches of known depth. Basic recommendations such as not measuring closer than a centimeter away from any edge, are included in general usage guidelines. Nothing was found in the current scan of literature that discussed the quantitative effect on crack depth measurement of width of the contact surface, L. To mimic application to specific industrial inspections, this variable was chosen for inclusion in this study. In addition, crack depth as a function of percentage thickness was monitored. Finally, samples of steel, zinc and aluminum were chosen, as these cover an anticipated range of electrical conductivity that could be of interest to manufacturers. It was desirable to look at the effects of surface condition or preparation on the accuracy of ACPD crack depth measurements. But the guiding principle was to see how inaccurate measurements might be if a conscientious user of such devices was forced into generating data in an unexpected manner than anticipated.

TEST METHOD A commercially available crack depth gage was used for this study, the Karl Deutsch RMG 4015. Two sample configurations were developed for this work. One was a bar of 4340 steel that measured approximately 150 mm x 40 mm x 40 mm, and had tapered notches cut on each face along the major axis. This created a block with four notches of varying depth that could then lend themselves to different surface treatments. In this study, one such surface was machine ground smooth, while a second was cleaned to “shiny metal” using a motorized wire brush. These faces are shown in Figure 2.

Numbers indicate notch depths at various points

Figure 2. Calibration block for ACPD studies showing surfaces that were ground smooth (left) and wire brushed (right).

128

The effect of surface contact width and the percentage of material cracked were intended to be measured on a series of blocks that had the complex shape shown in Figure 3. These test blocks presented notches in groups of 9, 6 and 3 mm, in regions that were 20, 16 and 12 mm thick. Additionally, the blocks were slotted to create contact widths of 15, 10 and 5 mm. The blocks were machined from aluminum (measured conductivity ~43% IACS), zinc (measured conductivity ~29% IACS) and A-36 steel (conductivity from literature ~3% IACS).

Figure 3. Notched test block used for ACDP crack depth studies in this work.

In addition, cracks found on the surface of the nodular iron cylinder head from a large engine using magnetic particle inspection were studied. Located between the various intake and exhaust ports on this component, as shown in Figure 4, the apparent crack depths were measured using ACPD and the cracks then broken open destructively to permit a visual reading of the crack depth based on surface discoloration.

Figure 4. Cracks detected on a cylinder head using MPI, then measured with ACPD and broken open.

129

RESULTS We can first examine the effect of sample conductivity on the accuracy one might expect for using COTS devices to measure cracks via the ACPD method. Figure 5 shows test results on our three sample materials, for a test geometry that represented the thickest sample regions with the widest test probe contact areas. One sees very good correlation between actual and measured crack depths, particularly for low conductivity material.

measured crack depth (mm)

ACPD crack depth measured on materials of varying conductivity 12 10 8 6

Steel

4

Zinc Aluminum

2 0 0

2

4

6

8

10

12

actual crack depth (mm) Figure 5. Correlation between measured and actual crack depth on samples of various electrical conductivity. Error increases slightly for deeper cracks on more conductive material.

The next figure shows the correlation between measured and actual crack depths for the same three materials, this time where the samples were thinnest. This situation, shown in Figure 6, would correspond to the measurement of cracks that were an increasing percentage of sample thickness. Effectively, the increasing proximity of the backwall can be seen to adversely affect crack depth measurement. Figure 7 shows the correlation between measured and actual crack depths on the three test materials, this time on the region of the test blocks that presented the most narrow contact regions. It was this test condition, where the test probe made contact with sample surfaces in regions that had the closest boundaries parallel to the main plane of the conceptual schematic shown in Figure 1, that the accuracy of the ACPD method was poorest. It may also be influenced by the finite thickness of the test pieces used in this study, but cracks of 9 mm in depth were nearly doubled in their estimated size for steel; this measurement was even more inaccurate for more highly conductive material.

130

ACPD crack depth measured on thinnest material - crack is largest percentage of wall thickness measured crack depth (mm)

14 12 10 8 6 Steel

4

Zinc

2

Aluminum

0 0

2

4

6 8 10 actual crack depth (mm)

12

14

measured crack depth (mm)

Figure 6. Correlation between measured and actual crack depth on samples at thinnest region of test blocks. Error increases as cracks become greater percentage of sample thickness.

ACPD crack depth measured on narrow material - crack is on slender ridge -

50 40 30 20

Steel Zinc

10

Aluminum

0 0

10

20

30

40

50

actual crack depth (mm) Figure 7. Correlation between measured and actual crack depth on samples at most narrow region of test probe contact. Error increased dramatically on all samples.

131

The effects of proper calibration procedure for use of this ACPD device are shown in Figure 8, along with some observation into the influence of surface preparation. When using such COTS devices, one may calibrate the device solely on a crack free region, only performing a “material calibration.” For best practices, one should also machine notches of known depth into the test material and include these readings into the device’s calibration. Knowledge of the tapered notch geometry on the 4340 steel clock sample permitted such readings. This figure shows quantitatively the error that would be encountered in the measurement of increasingly large cracks if calibration procedures were minimized and improved device accuracy was not achieved by using practice notches.

measured crack depth (mm)

ACPD crack depth measured on 4340 steel - ground smooth, wire brushed surfaces 14 12 10 8 6 4

ground smooth, mat'l cal ground smooth, mat'l & notch cal wire brushed, mat'l cal wire brushed, mat'l & notch cal

2 0 0

2

4

6

8

10

12

14

actual crack depth (mm) Figure 8. Correlation between measured and actual crack depth for 4340 steel, showing the benefit of using proper material and notch calibration. Wire brushing the test surface as opposed to machine grinding smooth was not seen to significantly affect ACPD readings. Finally, the result of using ACPD to measure crack depth on the complex geometries of the bridges between valve ports in cylinder heads is shown in Figure 9. Nondestructive results were obtainable after adequate mechanical cleaning of the cylinder head surface to remove fuel combustion residue; this was necessary to permit electrical contact for the probe to make readings. The destructive measurement of crack depth based on visual measurement of discolored crack surfaces was a somewhat subjective measurement, given that the crack front was often not a clearly defined edge that was parallel to the component surface. This notwithstanding, a very good correlation was noted for cracks depths measured using ACPD and visual means. The agreement deteriorated when cracks were found to be greater than about 8 mm deep. But it was also found that most of these deeper cracks occurred in regions where an interior surface of the cast component made them an increasing percentage of casting wall thickness.

132

Crack Depth on Cylinder Heads correlation between ACPD and visual results

ACPD crack depth (mm)

20

15

Glowplug E-E E-I I-I

10

5

Bridges E-E have internal surface near 0 0

5

10

visual crack depth (mm)

15

20

Figure 9. Correlation between crack depths measured on a cylinder head using ACPD and as determined by visual discoloration upon destructively breaking open. Agreement is quite good to about 8mm crack depth, and outliers are in predominantly thinner areas of the casting.

CONCLUSIONS A commercially available ACPD device was used to measure crack depths (simulated by thin notches) in various test blocks and a real industrial part. The electrical conductivity of the test material was varied by machining the blocks from steel, zinc and aluminium. The geometry of these blocks presented cracks that varied in percentages of part thickness and proximity to edges perpendicular to the crack planes. The industrial component was a cylinder head that had cracks on bridges between valve ports, where these regions had complex shapes. An initial study into the effect of surface condition was also performed. While it is known that test samples for ACPD studies should be clean enough to provide electrical contact between the probe and test piece, it may be anticipated that industrial application of the measurement will be attempted in regions that see rough cleaning. It was seen that samples with a higher electrical conductivity can show increasing measurement error as the cracks get deeper. The level of concern for such error will be dependent on the need for precision in the crack depth

133

measurement, but for cracks of 9 mm deep, the largest deviation was approximately 2 mm. Shallower cracks were more accurately measured. The error in crack depth measurement was slightly larger when the cracks became a larger percentage of wall thickness. This was expected and cited in literature found on this topic. The most dramatic effect on crack measurement accuracy was found when cracks were measured on thinner regions, with material edges perpendicular to the crack plane got closer together. This error was small when the cracks were small, but became significantly greater for deeper cracks in more conductive material. While a 6 mm crack in steel might be measured as being 10 mm deep, a 9 mm crack in aluminum could give a reading of nearly 50 mm. Obviously, when measuring cracks on relatively thick projections of a component, an observation of crack depth using ACPD could tend to be unreasonably large. This should be a sign that constraints on the geometry of the test region make successful crack depth readings invalid. Surface preparation prior to ACPD tests in a production facility may be expected to be rough and pressured by time concerns. The results on wire brushed steel samples proved quite similar to those for the same steel that had been machine ground smooth. There was some slightly increased inaccuracy in apparent crack depth for a wire brushed sample than for a ground sample, but that may be partly due to surface roughness affecting the crack depth reading to a small degree. The most important lesson learned in subjecting these steel samples to ACPD measurement was to strictly adhere to best practices as defined in the ACPD device user manual. Both an accounting for material variation as well as calibration on practice notches are needed for optimized results. Finally, the challenging geometry of cracks on bridges between valve ports on a cylinder head was not an insurmountable obstacle for successful ACPD crack depth measurements. Placement of the probe could indeed be challenging, and it was seen that thin contact surfaces may be problematic. This notwithstanding, crack depths measured nondestructively agreed very well with observations made after breaking open these cracks and deducing the crack length based on surface face discoloration. This agreement degraded when the interior surface of the casting got closer to the front surface, effectively increasing the percentage of wall thickness that was cracked. Future work will focus on ACPD readings along penny-shaped cracks in low conductivity samples to define crack size and shape. A wider range of surface preparation techniques will also be examined. These will be chosen such that they would conceivably lend speed and easier sample preparation to manufacturers, while not preventing the location of the cracks in question to be hidden to the magnetic particle techniques that will be used to locate them.

REFERENCES 1. 2. 3. 4.

Lugg, M. C., “An Introduction to ACPD,” Technical Software Consultants Ltd., TSC/MCL/1146, United Kingdom, 20 February 2002. Saguy, H. and Rittel, D., “Flaw detection in metal by the ACPD technique: Theory and experiments,” NDT&E International 40, pp. 505-509, Elsevier Ltd., 2007. Lugg, M.C., “Data interpretation in ACPD crack inspection,” NDT International, Vol. 22, No. 3, June 1989. “The Potential Drop Technique & Its Use In Fatigue Testing,” Matelect Application Note Reference FPD04, London.

134

Microwave Inspection of Fiber Reinforced Plastic Products for Absolute Thickness and Remaining Wall Microwave Inspection of Fiber Reinforced Plastic Products for Absolute 1 2 Thickness and Remaining Wall Robert J. Woodward and Karl Schmidt URS Energy & Construction, Inc. 1 and Karl Schmidt2 Robert J. Woodward (803) 634-7075; e-mail [email protected] 1 URS Energy & Construction, Inc. 2 (803) 634-7075; e-mail [email protected] Evisive Inc. (215) 962-0658; e-mail [email protected] 2 Evisive Inc. (215) 962-0658; e-mail [email protected] 1

SUMMARY

The need to detect and characterize wall thickness of fiberglass pipes and tanks, while only having access to only the one surface is increasing in many industries. This paper describes the scanning microwave technique which is used to detect thickness and, especially variations in thickness, caused by mechanical damage, such as erosion; or chemical attack, including osmotic blistering. The interferometric scanning microwave technique described here has the ability to detect small changes in wall thickness, which is often associated with the very earliest stages of chemical attack. This microwave technique inspection technique has been developed to be applicable to inspection whether the inspected vessel is in service or out of service, empty of filled with its associated process fluid. Proper selection of beam power and frequency makes it possible to inspect vessel with wall thickness in excess of 3 inches in some cases.

Standing Microwave Interferometric Microwave Basic Theory

As illustrated in Figure 1 the object being examined is bathed in microwave energy. The energy is reflected at all interfaces which have a change in dielectric constant. Both the transmitted and reflected microwaves impinge on the detectors. The combined signal, interference signal, is captured with position information to create an image. Since the voltage at the detector is the result of the interaction between a beam and many reflections, which differ in both phase and amplitude, information about the entire volume of the specimen is superimposed to create a single, 3-dimensional image through the thickness of the part.

Figure 1

Wall Thickness Evaluation Basic Theory

Figure 2 displays how the interference voltage is plotted on both the “A” and “B” channels. A change in depth correlates to a change in the phase relationship between the two channels. The change in phase relationship can be unambiguously correlated to known depths for thickness changes less than λ/2. Depths between the known points can be interpolated.

135

Tra ns mi tte

0

r

-2

Tra ns mi tte

+0

B A

¼ λ

+1

r

B A

¼ λ

Figure 2 As the thickness increases the phase relationship between the “A” and “B” channels is plotted on a phase space graph as can be seen in Figure 3(a), where the phase vector angle increases linearly with depth, and the phase vector amplitude decreases by attenuation. Expressed as ATAN2 (A, B) the relationship with depth is linear over λ/2, and wraps at that point. This creates an image that resembles a saw-tooth pattern as shown in Figure 3(b). Software which recognizes the 2π radian step “unwraps” the data to create an image that can be calibrated in thickness value over several λ/2 thicknesses as shown in Figure 3(c).

Figure 3(a)

Figure 3(b)

Data Collection Methods

Figure 3(c)

Capturing scanning microwave data in the time domain only, “Line Scanning” provides a 2 dimensional image with thickness as the ordinate and time as the abscissa. The time value can be correlated to position, as shown in Figure 4. Line scanning is typically used for sampling and indications are further evaluated by imaging with position information as shown in Figure 5.

136

Figure 4

Wall loss = .150”

Wall Loss = .155” Wall Loss = .145” Wall Loss = .168”

Wall Loss = .060”

Wall Loss=.050” Wall Loss = .074” Wall Loss = .060”

REFERENCES 1.

2.

S. S. Udpa (Editor) and Patrick O. Moore (Editor). Nondestructive Testing Handbook, Third Edition: Volume 5, Electromagnetic Testing: May 1, 2004 Schmidt, Karl and Little, Jack, Evisive Scan Technical Overview and Theory Training, Copyright Evisive Inc.

137

Application of Flash Thermography in Automotive Carbon Fiber Application of Flash Thermography in Automotive Carbon Fiber Composites Composites Haibo Zhao and Patrick Blanchard Ford Motor Company Haibo Zhao and Patrick Blanchard Research and InnovationFord Center, MD 3135 RIC, Dearborn, MI 48121 Motor Company (313) 390-0620; fax (313) 390-0514; e-mail [email protected] Research and Innovation Center, MD 3135 RIC, Dearborn, MI 48121 (313) 390-6230; fax (313) 390-0514; e-mail [email protected] (313) 390-0620; fax (313) 390-0514; e-mail [email protected] (313) 390-6230; fax (313) 390-0514; e-mail [email protected]

ABSTRACT

To achieve continuous improvement in vehicle fuel economy and emissions, automotive OEMs are examining the benefits of advanced lightweight composites based upon carbon fiber. The potential for mass reduction is substantial. However, to date, application of carbon fiber composites has been limited in the automotive market to only low volume niche market vehicles. Lack of adoption is often due to issues related to processing cycle time and material cost, but the absence of a rapid low cost quality control method is also considered a key impediment to widespread proliferation. That is, current quality systems are either time consuming or destructive which results in additional cost burden to new application development. Hence, this undermines the business case for carbon composites when compared to alternate lightweight material systems. This has prompted a need to develop low cost, low cycle time inspection methods that can be deployed in an industrial environment. One path to reducing inspection costs is the use of non-destructive test methods. As non-destructive testing enables parts to be used after testing, the cost associated with scrapping parts is eliminated. Flash thermography, a full-field NDT technology, was evaluated in order to determine the capability of the technique as a potential means of quality control and in-service damage assessment. Carbon fiber composite panels in the form of fabrics were tested to identify the capability of thermography in detecting potential manufacturing flaws. Results from testing show that flash thermography can be used to detect a wide array of potential defects. However, considerable development is still required to establish a quality control system that would be robust to a manufacturing plant environment.

INTRODUCTION

Carbon fiber based components often consist of multiple layers of oriented fibers held together by a matrix resin. Each layer is engineered so that the laminate, as a whole, meets the performance requirements for the structure. The anisotropic properties of each lamina influence the damage tolerance of the composite and overall behavior when subjected to high energy impacts. Unlike metals, global deformation or localized impacts can result in damage such as fiber damage and intra-ply delamination. Often these and other manufacturing defects are difficult to detect visually and repair of broken fibers is not usually practical. In order to confirm the integrity of a composite part, it is necessary to examine the part to determine characteristics such as fiber density, fiber orientation, void content, and ply boundaries. NDT (NonDestructive Testing) methods exist today and are applied in the aerospace industry on carbon fiber composite parts for this exact purpose. However, the methodologies used are scaled to the inspection of very large parts with low production volumes, and often require highly skilled technicians. For the successful integration of NDT into an automotive manufacturing environment, the same technologies applied for aerospace need to be scaled towards high volume production where a multitude of different parts can be programmed for automated inspection across the same machine at high frequencies and analysis of the results enables operators to readily identify quality concerns.

138

Thermography is a non-contact optical imaging technique for detecting invisible infrared radiation. It has the advantages of being a low cost full-field method that is conducive to short cycle times. For this reason, thermography was selected for investigation to further understand its capabilities and limitations. In pulsed thermography, a heat source such as a brief pulse of light is used to heat the surface of a sample while an infrared camera records changes in the surface temperature. As the sample cools, the surface temperature is affected by internal flaws such as disbonds, voids or inclusions, which obstruct the flow of heat into the sample. An infrared camera is used to capture the changes in surface temperature after being thermally excited. Using time lapse images, internal defects can be detected based upon changes to the surface temperature map of the sample[1]. In this paper, the capability of flash thermography for detection of potential composite manufacturing flaws is evaluated. The materials under evaluation were carbon fiber epoxy panels molded into 305mm by 305mm flat panels. The part structural integrity inspection was conducted at Thermal Wave Imaging Inc. using ThermoScope® II.

EXPERIMENTAL Flat Fabric Panel Preparation

Seven layers of carbon fiber prepreg were stacked together and compression molded into a 305mm by 305mm panels with a thickness of around 4.7 mm. Defects were intentionally placed at different ply layers with different sizes. Table 1 is a summary of defect layout. To create fiber misalignment, a circular disk was cut out from the prepreg and rotated 90° before placing it back. Teflon inserts with thickness of 1mm were sandwiched between ply layers as foreign objects to be detected. No fiber zone was created by simply cut out an area from the prepreg. Dry fiber was prepared by first cutting out a piece from the prepreg and then burning out the resin in an oven. Then the dry fibers were placed back to the prepreg.

Flash Thermography System

The samples were tested using a commercially available flash thermography system with 2 quartz-xenon linear flash lamps, housed in a reflective hood. Each lamp is mounted in a reflector and driven by a 6 kJ power supply that provides a truncated rectangular optical pulse with variable duration ranging from 10 msec to 200 microseconds. The system employs a 640 x 480 pixel InSb camera operating in the 2-5 micron spectral range at a maximum full frame rate of 130 Hz (higher frame rates are attained by reducing the frame size). Digital data from the camera was transferred to a PC in real-time, and the temperature-time sequence before, during and after flash heating was processed using Thermographic Signal Reconstruction (TSR). A reconstructed signal was generated for each pixel as well as its 1st and 2nd logarithmic derivatives and various attributes of the derivatives [2]. Samples were placed 300mm from the camera lens and lamps and heated with a 3 msec duration rectangular flash pulse. Data was acquired for 33 msec prior to the flash pulse and for 30 seconds after, at a frame rate of 30 Hz. The sample was inspected in an “as is” condition, i.e. no surface preparation was applied.

139

Table 1: Defect layout in carbon fiber fabric panel. Defect Type

Size and shape

Location

Fiber misalignment A1

50 mm diameter circular disk

Ply 2

Fiber misalignment A2

25 mm diameter circular disk

Ply 2

Fiber misalignment B1

50 mm diameter circular disk

Ply 4

Fiber misalignment B2

25 mm diameter circular disk

Ply 4

Fiber misalignment C1

50 mm diameter circular disk

Ply 6

Fiber misalignment C2

25 mm diameter circular disk

Ply 6

Missing Fiber A

50 mm diameter circular disk

Ply 6

Missing Fiber B

50 mm diameter circular disk

Ply 4

Missing Fiber C

50 mm diameter circular disk

Ply 2

Dry fiber A1

50 mm diameter circular disk

Ply 2

Dry fiber A2

25 mm diameter circular disk

Ply 2

Dry fiber B1

50 mm diameter circular disk

Ply 4

Dry fiber B2

25 mm diameter circular disk

Ply 4

Dry fiber C1

50 mm diameter circular disk

Ply 6

Dry fiber C2

25 mm diameter circular disk

Ply 6

Foreign Object A

30 mm by 30 mm square

Between ply 1 and ply 2

Foreign Object B

15 mm by 15 mm square

Between ply 3 and ply 4

Foreign Object C

10 mm by 10 mm square

Between ply 6 and ply 7

RESULTS AND DISCUSSION

Figure 1 contains images captured by a flash camera showing views from both sides of the carbon fiber composite panel. Using the visible spectrum, no defects could be detected. However, by switching to an infrared camera, defects introduced into the panel are clearly shown within the surface temperature map. As shown in Figure 2, all the foreign objects (10mm x 10mm, 15mm x15mm and 30mm x 30mm), embedded at three different depths are clearly identified (shown as dark squares in Figure 2). From Side 1, the sizes of Square A and B measured in the images are the same size as the actual Teflon inserts. However, the edge definition of Square C is noticeably blurred resulting in an inability to quantify the defect accurately. The loss of resolution is attributed to the depth of the defect in the panel. This is corroborated when viewing the panel from the opposite side. Under these circumstances, Square C was closer to the testing surface and is clearly defined in both size and shape on the IR image. Although Square A became the furthest from the inspection surface when inspected from side 2, the size and shape are still detectable. Hence, the resolution of flash thermography appears to be a function of both defect size and depth. In summary small defects at deeper location are difficult to detect. However, since this panel could be examined from both sides, thermography is capable of accurately reflecting the size and shape of foreign objects in this carbon fiber composite with a thickness of 4.7 mm.

140

Side 2 Side 1 Figure 1: Images of carbon fiber fabric panels molded with seven prepreg layers.

Figure 2: TSR images revealing foreign objects embedded between prepreg layers. Figure 3 demonstrates the capability and limits of thermography in detecting missing fibers. Carbon fibers have a higher heat conductivity compared with the resin matrix. Hence, it is proposed that the absence of fiber in a prepreg layer will cause a lower heat transfer rate compared with surroundings, resulting in a hot spot on the surface. As shown in Figure 3, fiber free zones on the second ply and the fourth ply were accurately portrayed. However, the shape and size of the fiber free zone on the six ply was not truly representative of the defect. This again suggests that the resolution of flash thermograph diminishes with the depth of the defects. Based upon this observation, it is not recommended to apply thermography to detect missing fibers on parts with a thickness greater than 3 mm.

141

Figure 3: TSR images reveal no fiber zones. Figure 4 is the TSR image of areas where dry fibers were placed within plies of the laminate. On a practical note, it is not feasible to prevent wet out of these fibers during the molding process. However, there should still be areas in the panel with a relative increase in fiber volume fraction. For dry fiber regions in ply 2, defects of both sizes were visible. However, the same defects placed on the fourth and sixth ply were largely undetectable.

Figure 4: TSR images of no resin zones. Figure 5 is the TSR image highlighting defects that included misaligned fibers. Due to the tow count used in the weave pattern, fiber orientation in each ply appears difficult to detect and so a measure of misalignment with plies was not feasible. This is attributed to an insignificant change in through thickness heat transfer that would be associated with a rotation of individual plies.

142

Figure 5: Thermographic Signal Reconstruction images of misaligned fiber zones. A summary of the observations from the thermography evaluation of a fabric carbon fiber panel are contained in Table 2. Table 2: Results of defect detection in a carbon fiber fabric panel. Defect Type

Location

Detectability

Fiber misalignment A1

Ply 2

No

Fiber misalignment B1

Ply 4

No

Fiber misalignment B2

Ply 4

No

Fiber misalignment C1

Ply 6

No

Fiber misalignment C2

Ply 6

No

Missing Fiber A

Ply 6

No

Missing Fiber B

Ply 4

Yes

Missing Fiber C

Ply 2

Yes

Dry fiber A1

Ply 2

No

Dry fiber A2

Ply 2

No

Dry fiber B1

Ply 4

No

Dry fiber B2

Ply 4

No

Dry fiber C1

Ply 6

Yes

Dry fiber C2

Ply 6

Yes

Foreign Object A

Between ply 1 and ply 2

Yes

Foreign Object B

Between ply 3 and ply 4

Yes

Foreign Object C

Between ply 6 and ply 7

Yes

Fiber misalignment A2

Ply 2

143

No

CONCLUSION

The capability of flash thermography was evaluated for detection of an intentional set of flaws in a sample carbon fiber composite panel. Findings and recommendations are summarized below based upon the data. 

Thermography is capable of detecting foreign objects through 4.7 mm thickness when two sided inspection is feasible. The size and shape of foreign objects can be accurately detected under these circumstances. However, the size of the defects that could be detected in 1-sided inspection appeared to be limited the 10 mm by 10 mm square Teflon insert at a depth of 4.0 mm.



Fiber orientation could not be discerned possibly due to the high frequency weave pattern of fabrics. Distinguishing individual carbon fiber is challenging due to the resolution limitation of camera lens.



The absence of fiber for a 50 mm defect could be detected at a depth of 2.7 mm from the test surface. Beyond 2.7 mm, the shape and size of the resin rich area are not distinguishable.



Further investigation is recommended for flash thermography be applied to a target application to determine suitability for common automotive components. In addition no correlation of mechanical properties to defects detected is currently feasible. Therefore, threshold criteria will need to be developed to determine if components are of acceptable quality.

REFERENCES 1.

2.

Moore, P., Nondestructive Testing Handbook, third edition: Volume 3, Infrared and Thermal Testing, Columbus, OH, American Society for Nondestructive Testing, 2014. Steven M.S., “Thermographic Characterization of Composites”, SAMPE 2013 Proceedings: Education & Green Sky-Materials Technology for a Better World, Long Beach, CA, May 6-9, 2013.

144

ABSTRACTS

145

Health Assessment of Heat Exchanger Tubes through Eddy Current Testing (ECT) and Internal Rotary Inspection System (IRIS) and Their Comparative Study Performed on Various Tube Materials

What’s Missing in NDE Capability Evaluation? Charles Annis ABSTRACT We consider the wide-spread practice in engineering of creating home-made analysis procedures when superior, established statistical methods are widely available. This appears to result from the idea that the current problem is unique, thus never before encountered, and so requires an ad hoc solution. Unfortunately many of these faulty ad hoc procedures have been promulgated in industry publications as being problem-solving exemplars, or worse, presented as industry standards. We consider several examples and call attention to established methods for dealing with them.

Abdul Razzaq Al-Shamari Mohammad Al-Shaiji G. Santhosh Kumar Arnab Gupta Inspection & Corrosion Team Kuwait Oil Company, Ahmadi, Kuwait

ABSTRACT Heat exchangers are an integral part of process industry and failure of same during operation can prove very costly from HSE and business perspective. During service, heat exchanger tubes comes in contact with corrosive media like H2S, CO2, water and low pH fluids which contribute to degradation of tubes in form of pitting and uniform corrosion from both sides. Hence, timely & precise detection of defects in tubes is always a challenge to the personnel associated with integrity assessment of these equipment. Being an oil & gas major, Kuwait Oil Company has large number of tubular heat exchangers of varied construction, such as Shell & Tube type, Fin-Fan cooler type and U-tube type, as necessitated per engineering and process requirement. Intricate design, variation in tube metallurgy, tube layout and adherent process deposits on tubewall; constantly add to the existing NDT difficulties of defect detection. Deployment of advanced NDT techniques like ECT (Eddy Current Testing) /RFET (Remote Field Eddy Current Testing) and IRIS (Internal Rotary Inspection System) increase the probability of detection (PoD) of defects like miniscule isolated pitting/s and tube wall loss. In this paper, authors share how supplemental techniques like ECT and IRIS are utilized in KOC as effective tools for assessing the defects and make a comparative study from end user’s viewpoint. Measures adopted to overcome various challenges common in oil and gas industry, prior to deployment of particular inspection technique for giving the best intended result, is also highlighted.

146

Inspection of Welds From One Side Only Using Phased Arrays

Chemical and Petroleum EMI Inspection Standards

Tim Armitt and David Miller Lavender International NDT USA, LLC

WD Averitt New Tech Systems USA PO Box 1560 Brenham, TX 77834 (281) 330-9957 e-mail [email protected]

ABSTRACT Welds connecting pipes to flanges and pipes to elbows, more often than not, suffer from severe restrictions in terms of accessibility to scan the weld from both sides. There are hundreds of thousands of these joints across the petrochemical industry. The majority of inspections are being conducted in full knowledge that the NDT test is not fully code compliant. A justifiable concern exists that many single sided weld inspections miss reportable defects. This presentation addresses the issues and goes through ways of optimising defect detection capability, characterisation and sizing.

ABSTRACT Presentation to address recent improvements in qualifications of inspection standards. Specifically, inspection standards used for the calibration of electromagnetic inspection (EMI) systems used in the pipe inspection industry. The presentation will address the fabrication of the standards and common field use issues that affect performance. General calibration procedures will be reviewed as well as recent industry concerns that have brought about changes that have improved the process. Several brief case histories will be presented for reference. Keywords: pipe inspection, EMI, standards, OCTG

147

What to do When ASME Code Ultrasonic Exam Requirements are Not Enough to Find Critical Flaws?

Verifying Magnetic Field Strength for Magnetic Particle Examination of Some Complex Parts

David Bajula   CEng Acuren Inspection GM - USA Adv. NDT Services ASNT/ACCP Level III UT, RT, MT, PT, VT, ET, LT, IR, ML AWS CWI, API-QUTE, QUPA, QUSE, 510

John Brunk 10541 Walmer Street Overland Park, KS 66212 (913) 991-0213 e-mail [email protected] ABSTRACT ASTM E 1444E1444M requires the use known real or artificial defects, notched shims or Hall effect probe gaussmeter readings to determine that relevant discontinuities of any orientation in the area of interest can be detected. It is not hard to find test objects for which none of these options appear to be easy, or even possible. Some examples of examining parts where everything was tried to determine the most effective verification technique are illustrated and discussed.

ABSTRACT The requirements for performing ultrasonic examinations on new welds in accordance with ASME are covered in various documents including but not limited to ASME V; Article 4, ASME Section VIII; xxxxxx (pressure vessels), ASME B31.3; xxxxxx (process piping) and ASME B31.1; xxxxxxx (power piping). This presentation will provide details on the difficulty for the detection and evaluation of transverse flaws in welds, especially when the welds are not flat topped or ground flush. Even with state-of-the-art ultrasonic techniques such as Phased Array and TOFD, transverse cracking can be easily missed unless supplemental techniques, that is techniques above and beyond code requirments, are performed. We will present data that clearly shows that the code needs to address the defficiency in detection of transverse flaws in the “as welded” condition and provide recommendations on techniques that can overcome this gap.

Keywords: MT, Field strength verification

148

Further Development of an Optimised Array Wheel Probe for Inspection of Fibre Glass Composites

Load-Enhanced Methods for Lamb Wave in situ NDE of Complex Components

Joe Buckley Sonatest Ltd Dickens Road Old Wolverton Milton Keynes Buckinghamshire MK12 5QQ United Kingdom 441908316345 441908321323 e-mail [email protected]

Xin Chen, Jennifer Michaels, and Thomas Michaels Georgia Institute of Technology 771 Lindbergh Dr NE Apt 5210 Atlanta, GA 30324 (678) 986-9936 e-mail [email protected] ABSTRACT Ultrasonic guided waves, called Lamb waves in plates, provide a promising tool for in situ monitoring of defects. It is well-known that varying environmental conditions such as temperature and loads significantly change guided wave signals and thus can obscure the detection of damage. However, applied tensile loads open cracks and enhance their detectability. Previous research by the authors demonstrated a load-differential method for detection and localization of fatigue cracks in simple structures that leveraged the effect of loading on crack opening. Here, load-enhanced concepts are further extended to components with complex geometries. Guided wave signals at multiple frequencies and loads are used to improve crack detection and localization. Discrimination of cracks from benign scatterers is obtained by incorporating expected scattering information into analysis algorithms. The efficacy of these methods is demonstrated using data from fatigue tests performed on aluminum plate specimens with different structural complexities such as bonded doublers, multiple fastener holes, and installed fasteners.

ABSTRACT With growing maturity in the Wind power industry has come the need to maximise the efficiency and reliability of turbine equipment. One key aspect of this is weight reduction, both to reduce manufacturing cost and to reduce loading on gearboxes, bearings and associated equipment. A previous paper detailed the initial development stages of a Specialised array wheel probe, optimised for Thick Glass fibre or Glass Carbon composite materials as used in Wind and Tidal power applications. This paper extends the story, showing results from practical use of this technology. Keywords: Wind power, Composite UT.

Keywords: Guided Waves,Load-enhanced,Scattering

149

Digital Image Correlation Techniques for Aerospace Applications

Small, Flexible and Advanced Phased Array Module for Customizing NDT Applications

Tsuchin Philip Chu, Anish Poudel, and Thomas Heller Southern Illinois University 1230 Lincoln Dr MC 6603 Carbondale, IL 62901 (618) 453-7003 (618) 453-7658 e-mail [email protected]

Gavin Dao Advanced OEM Solutions, Cincinnati, USA e-mail [email protected] ABSTRACT Having a small, advanced and open platform phased array (PA) ultrasonic module provides new opportunities for both industry and research. Industrial applications can benefit by creating custom solutions that fit their particular need. A smaller form factor advanced PA module allows mounting the instrument on a scanner or robotic arm, thus shortening the transducer cable, which provides better signal quality, better integration, avoids cable failures and reduces cost. An open instrument allows customization of the software to create a dedicated software that can simplify user operation. Researchers require instrumentation that allows access to low-level parameters, raw waveforms, Full-Matrix Capture (FMC) capability, and the freedom to control and interface to instrumentation with their choice software language. However, key characteristics are essential such as a complete phased array feature set, excellent signal-to-noise ratio, fast data throughput, ruggedness, compact form factor and cost effectiveness.

ABSTRACT This paper discusses on the review of digital image correlation (DIC) techniques for aerospace applications. This paper looks at several examples where DIC technique has been successfully applied. Some of these include strain measurement in Friction Stir weld (FSW) samples, thin-film coatings at micro-level, and particulate composites. DIC is now a proven technique both in terms of the technical advancement and in terms of the applications for which it is applied. By combining the DIC method with existing NDE methods, it can be a powerful quantitative tool for monitoring damages in composites. This paper also discusses on the in-plane bulk displacement and deformation measurements by using ultrasonic C-scans. The possibility of integrating DIC technique with the state-of-art Acoustography NDE technique is also explored. Keywords: Digital Image Correlation, Acoustography, UT, C-Scan, deformation, strain.

Keywords: Non Destructive Testing; Phased Array; Ultrasound; Full –Matrix Caputre

150

When We Get it Wrong: Why NDT Matters

The focus of more and more university engineering programs is on computational design with extremely limited exposure to actual material behavior when subjected to loads, and essentially no significant exposure to deterioration associated with corrosion, creep and fatigue. The designer is unaware of the need to assess when during service and where in the system or structure this deterioration might occur in order to inform the maintenance engineers and to adequately provide access through design for the inspection. Of course this task can further be complicated by the unpredictability of various forms of deterioration. It is essential that provisions are made during planning and design to instrument the system or structure to assess the condition of significant components during fabrication and to allow for assessing whether the final system or structure performs as intended. It is also important for the inspection needs to be brought to the attention of the design team by a member of the team that understands the inspection capabilities as well as the potential deterioration mechanisms likely to occur.

Douglas Davies Alcoa Fastening Systems 800 S State College Blvd Fullerton, CA 92831 (714) 871-1550 x228 e-mail [email protected] ABSTRACT An update of the presentation I did at ASNT Vegas, I will cover some well known incidents having NDT related causes (UAL flight 232 Delta FL 1288) the Alaska pipeline failures and address the disturbing trend of NDT systems being audited by people with very little NDT knowledge. Keywords: NDT

NDE Engineering: SHM and NDE during Planning and Design of Critical Assets John C. Duke, Jr. Virginia Tech Federal Highway Administration TurnerFairbank Highway Research Center e-mail [email protected] ABSTRACT Traditionally nondestructive inspection has been relied on to assure that systems and structures are fabricated without critical defects. However for many complex systems and structures computational systems are being relied on more heavily to achieve efficient designs. Reasonable design assumptions are used, including details regarding connections, material properties, material conditions associated with fabrication processes such as welding and fit up stresses, etc. Often performance specifications serve as the basis for material selection and issues associated with deterioration are not a primary focus. At times actual systems and structures once fabricated may perform different than the actual design, but it may not be obvious until problems develop after a period of use. In addition, when deterioration occurs which may affect the integrity of the system or structure it is important that it be detected and assessed.

Fig. 1. Schematic of Life Cycle Critical Asset Management

Points of Emphasis: • Maintenance and inspection consideration during planning and design, – Where, when, how inspect – How to repair • Skill set needed by the NDE Subject Matter Expert – Deterioration Processes/Damage Mechanism – Nondestructive Technologies • Detect damage • Track damage • Characterize materials properties • Characterize deterioration rate

151

Film vs. Digital Radiography on Thin Walled Orbital Arc Tube Welds

Gap in NDT Engineering Mark Gehlen Uniwest 122 South 4th Ave., Pasco, WA 99301 (509) 544-0720; fax (509) 544-0868; e-mail mark. [email protected]

Jeffrey Elston, Bence Bartha, Paul Vona and Wenche Cheng PaR Systems Mail Code PaR Hangar N Kennedy Space Center, FL 32899 (321) 258-2064 e-mail [email protected]

ABSTRACT The role of NDT Engineering will be explored and contrasted to the roles played out by technicians and researchers.  The notion that an NDT Engineering “Gap” exists will be discussed.  Ideas will be offered on how ASNT might play a more meaningful role in providing resources and support to the NDT Engineering field.

ABSTRACT Thin wall orbital arc tube welds during space vehicle processing require Non-Destructive Evaluation, i.e., radiography. Film radiography is currently the best technique for inspecting these welds, however recent improvements in digital radiography may allow for the detection of small size flaws that need to be resolved for this application. Digital radiography has significant advantages compared to conventional film radiography in terms of environmental safety (no chemicals), shorter exposure time, processing time and archival time. In In an effort to develop comparable results using methods of digital radiography a series of tests were conducted to develop inspection techniques and compare the results to film radiography. Several weld samples of varying sizes and materials were created and examined with film, Computed Radiography (CR) and Digital Radiography (DR). This paper details the equipment, techniques and settings used for the study and the resulting comparison of each technique. The project focused on the relative resolution and image quality of thin walled tube welds using various CR scanners and image plates. The outcome of the results are being used to formulate a path toward development and qualification of digital radiographic methods for inspecting thin walled orbital arc tube welds on Space Vehicles. Keywords: RT, Film, CR, DR, orbital arc tube weld

152

New Known Defect Standard for the Verification of Penetrant System Performance

What Makes a Great NDE Career? Gary Georgeson Boeing Research & Technology (206) 662-3847; e-mail [email protected]

David Geis Magnaflux 3624 W Lake Ave, Glenview, IL 60025 (847) 657-5300 fax (847) 657-5388 e-mail [email protected]

ABSTRACT Dr. Georgeson has learned a lot about what makes a great NDE career after almost 26 years at Boeing. Much of what he has learned was learned through his mistakes. The worst moments turned out to be the greatest blessings. And what he thought was most important, turned out to be the least. He will share briefly from his own experience on what makes a great career.

ABSTRACT One of the primary requirements for quality assurance of fluorescent penetrant systems is the daily System Performance Check. The overall performance of the entire system (materials, processing parameters, and equipment condition) is verified by processing a Known Defect Standard and comparing the results against an established baseline. The requirements for this Known Defect Standard are detailed in ASTM E1417 and AMS 2647, and NADCAP includes checklist requirements for both the Known Defect Standard and the daily System Performance Check. There are several commercially available Known Defect Standards, the most common being the TAM 146040 Test Panel, the PSM-5 Test Panel, and KDS Panels. These test panels use five starburst stress crack patterns in chrome plating, which have extremely high variability and essentially no reproducibility between panels. In addition, stress cracks propagate with time and process cycling, so panels must be recertified on an annual basis. The deficiencies of the TAM 146040 Test Panel are well known and documented (See “Misconceptions Within Liquid Penetrant Inspection”, George Hopman, ASNT Annual Conference 2013), but until now there has not been an alternative that satisfies the needs of the industry. Magnaflux® has developed a new test panel that fits the needs of the Known Defect Standard while avoiding the problems of the TAM 146040 Test Panel. This new test panel – the Verifi™ Panel – has repeatable defect sizes and depths, providing the sensitivity required to demonstrate deviations in the materials, processing parameters, and equipment condition of a penetrant system. The patented design of the panel is unique and provides repeatability verification within the panel itself, and provides measurable reproducibility from panel to panel. Keywords: PT, System Performance Check, Known Defect Standard, TAM 146040

153

Comparison of Industrial Process Control CT Systems VTomex C and Speed/Scan

Developing New Generation ECT Flaw Detectors to Meet Client Needs John Hansen ETher NDE Endeavour House 3 Roundwood Lane Harpenden Herts AL5 3BW United Kingdom 441582767912 e-mail [email protected]

Juan Mario Gomez GE Measurement & Control Lewistown, PA (717) 447-1226 e-mail [email protected] ABSTRACT Industrial computed tomography (CT) enables the nondestructive and 3-dimensional capturing and analysis of the complete geometric structure of the inspected part. This facilitates processes such as the analysis of faults (pores and material inclusions), material and density analysis or the examination of completeness and dimensional control (coordinate metrology) in 3D helping CT users to early detect deviations and optimize their production parameters. Two new CT systems specially engineered for reliable 3D statistical production process control will be compared regarding benefits for different industrial production surveillance tasks: the compact phoenix v|tome|x c high energy CT scanner and the high throughput speed|scan CT 64 helix CT system (both GE Measurement & Control). The fast CT system can scan large castings like cylinderheads of approx. 500 mm in diameter x 900 mm in length within 15 seconds, with a typical scan, reconstruction and evaluation cycle speed of ~1 min. per part.

ABSTRACT In NDT there is often a need to design a Flaw Detector that is scientifically at the cutting edge of new design, but can be lacking the ability to appeal to the end users specific needs; be it their ECT knowledge or application specifications. In the age of technology that is so simple to use that many small children can quickly grasp the concepts of a mobile phone by the age of three, the need to design flaw detectors with the same predictive ability and able to appeal to the next generation has become of paramount importance. Up until now Eddy Current Flaw Detectors have generally emulated previous generation analogue instruments in a digital way, but have missed the mark when it comes to the balance of cost and ease-of-use for the end user. Digital technology offers the ability to make the inspectors job easier and in this paper we look to explore how the designing of a new Eddy Current Flaw Detector strikes the right balance between state-of-the-art technology, scientific excellence and the ability to adapt to its user regardless of their ECT knowledge level. Keywords: ECT, Flaw Detection, Designing detectors for the next generation

154

Evaluating Dual-Mode Pulse Reflectometry

3D Eddy-Current Based Inspection of Lightweight Components – From Idea into Realization

Yoav Harel AcousticEye 12 Greenway Plaza Suite 1100 Houston, TX 77046 (408) 933-8658 e-mail [email protected]

Martin Schulze, Matthias Pooch and Henning Heuer Fraunhofer IKTSMD MariaReicheStr 2 Dresden Saxony 01109 Dresden 004935188815628 e-mail [email protected]

ABSTRACT Recently introduced dual mode pulse reflectometry combines Acoustic Pulse Reflectometry (APR) and Ultrasonic Pulse Reflectometry (UPR, known also as Guided Waves). This method is a non-traversing technology for inspection of heat exchanger and boiler tubes. Further advances in transducer technology and signal processing, enable increases in speed of inspection, resolution and thresholds of detections. To assess the capabilities of this method, several heat exchanger mockups comprising a variety of tubes were manufactured, incorporating a large series of defects. In this presentation we will present measurements taken on these tubes, demonstrating the system’s typical detection capabilities.

ABSTRACT With the increased use of Carbon Fiber Reinforced Platics (CFRP) in the automotive and aerospace industries, and the production using RTM - (Resin Transfer Moulding) processes today’s lightweight components are available in high-volume, reproducible and also cost-effectively. The development and practical use of eddy current testing along the supply chain of CFRP´s was started in 2008 at the Fraunhofer Institute for Non-Destructive Testing (Dresden branch) IZFP-D. The system was previously developed for the inspection of planar components such as non-crimp fabrics, prepregs and CFRP plate materials. The high frequency eddy current inspection system EddyCus® was designed for industrial use in harsh environment for production monitoring and non-destructive testing of delaminations, fiber accumulations, dry spot inspection, angular deviation detection and foreign material inclusions. Customers of this system wanted increased automation and enhanced testability of 3D structures . In order to meet these requirements it has been installed on a robot at IKTS-MD. Using a light stripe sensor unknown component geometry can be digitalized easily. Then, the scan paths are generated in space by means of a path planning software and a mandatory collision check is performed. The EddyCus® high frequency eddy current system has been extended to a measure with a timetriggered data acquisition setup with subsequent coordinate recalculation algorithms. The individual components are already tested in detail. In the presentation the actual state of the art of the overall system is presented.

Keywords: Pulse Reflectometry,UT,AT, Heat Exchanger

Keywords: 3D-Eddy Current Inspection

155

New Ultrasonic Phased Array Method for Characterizing Circumferential Welds at Thin-walled Pipes

Designing an Integrated, Affordable and Versatile Ultrasonic Solution using Advanced Technologies

Susanne Hillmann1, Frank Uhlemann2, David-M. Schiller-Bechert1, and Zsolt Bor1 1 Fraunhoefr IKTS-MD Dresden, Germany

Deborah Hopkins1 and Michel Brassard2 1 BERCLI PhasedArray Solutions 2614 McGee Ave Berkeley, CA 94703 (510) 717-8859 e-mail [email protected]

Ingenieurbüro Prüfdienst Uhlemann, Peitz, Germany

2

ABSTRACT Described is the development and the potentials of a new ultrasonic phased array method, which can be used in power plants and chemical industry. The advantages are, that this method is much faster and cheaper then the previously used radiography method. Using optimized operation schedules of the testing staff, power plants can finish their maintenance period after half of the inspection time then before. The method is developed, tested, optimized and validated and starts to establish in Europe at this time.

Techno Diffusion NDE

2

ABSTRACT Automation used to be the domain of large-scale manufacturing and typically means multi-million-dollar ultrasonic inspection systems using proprietary software for motion control and data acquisitionprocessing. These systems are typically not well integrated with the manufacturing process and are not easily modified to accommodate other components. The UT solution presented here is based on a versatile platform that is designed to integrate advanced technologies, for both ultrasonics and motion (path generation). The design is modular to allow upgrades and integration of new techniques as required. It is therefore an investment in capability that is not tied to a single part or to technology available today. As advances in processing power enable techniques that can be performed in near real time including multiple probes used in parallel, total focusing methods, and surface adaptive UT, it makes automation more cost effective on a smaller scale. For example, the advent of composites in aerospace has resulted in hundreds of complex parts that are well suited for automated inspection in a small tank, which is much more cost effective when a wide variety of parts can be inspected at the same station. Examples are presented to demonstrate how to specify an integrated system and evaluate the tradeoffs between cost and capability that can be optimized to meet performance and price targets. Keywords: Automated UT integrated inspection system, phased array, total focusing method, full matrix capture, surface adaptive, parallel probes,scan path, composites, aerospace, complex parts

156

A Brief History of the Magnetic Particle Inspection Method

Inspection Range of Guided Wave Testing of Pipeline

George Hopman Quality Control Council of the United States 911 E. Camelback Rd. Unit #3086 Phoenix, AZ 85014 (602) 321-2950 e-mail [email protected]

Sang Kim and Heui Kim Guided Wave Analysis LLC 7139 Callaghan Rd San Antonio, TX 78229 (210) 842-7635 e-mail [email protected]

ABSTRACT The magnetic particle method has come to be a fairly developed method of inspection. Many dedicated individuals have contributed to this inspection method that started out as an art form and eventually became a science. This paper will document the progress of the magnetic particle inspection method over the years, from its inception in 1929 to the current state of the method. The paper will acknowledge the primary people and companies that contributed to this essential primary inspection method.

ABSTRACT Guided wave testing (GWT) has been widely used to detect corrosion or crack defect for long-range inspection in many pipelines as a rapid screening tool. The inspection range of guided wave testing depends on defect size, operating frequencies, and pipe conditions. The presentation shows several inspection reports, how to decide on the inspection range with the acquired data, and what frequency should be used for guided wave testing depending on defect size, coating, and generalized pipe surface corrosion. Keywords: GWT, inspection range, frequency

157

DAC Recording Using Phased Array Probes Made Easy

X-Ray Backscatter Tomography for Large Scale Constructions

Wolf Kleinert and York Oberdoerfer GE Sensing & Inspection Technologies RobertBoschStr 3 Huerth NRW 50354 Germany +491704534678 e-mail [email protected]

Peter Krueger Fraunhofer IKTSMD MariaReicheStr 2 Dresden Saxony 01109 Germany 004935188815513 e-mail [email protected]

ABSTRACT Defect sizing using phased array probes requires recording of Distance Amplitude Correction (DAC) curves for each angle to be applied for testing which is very time consuming. Using a new probe technology (trueDGS) enables calculation of DAC curves for all angles. Therefore it is sufficient to record one single DAC curve for one single angle. Using the recorded echoes the material characteristics, such as the sound attenuation in the calibration block, are derived. After recording the single DAC curve the curves for all other angles are calculated considering the different probe parameters for each angle, such as delay line length, near field length and sensitivity difference. A brief introduction to the new trueDGS technology will be given. The method for deriving DAC curves for phased array probes will be presented. The accuracy of this new method will be proven by taking the recorded echoes from other angles and comparing these measurement values with the model. By using this approach defect sizing using phased array probes is as easy as using a single element angle beam probe. This new approach results in high accuracy and a huge productivity gain.

ABSTRACT Large constructions like airplane wings or wind turbine blades create their stability from internal delicate constructions. These structures need to be tested on their integrity during build up and while lifetime of the product. For many of those constructions a handy to use testing procedure is missing. Ultrasound testing is often not possible on complicated compositions - as in airplane wings - or porous media – as used in wind turbine blades. Thermography may give some information on near-surface defects but cannot penetrate deeply. X-ray transmission radiography and tomography is usually not possible due to very low transmission and because of bad double sided access on large systems. Backscatter technologies can circumvent some of these problems. In the old days, an x-ray backscatter technology known as CompScan was used to image tree-dimensionally regions in the wider vicinity of the surface. This method had the disadvantage that it was – due to the scanning measurement - very slow. We want to show in this contribution a novel approach to the backscatter technology. This setup features a parallel recording of many backscatter images. The images are produced by a series of pinhole cameras focusing on a single region illuminated from a powerful x-ray source. The images are reconstructed to the volume using algorithms very similar to those used in computed tomography. We want to present some case studies here recorded with a sample implementation of the method and derived basic parameters for a dedicated machine.

Keywords: UT, Phased Array, DAC, Accuracy, Simplification, Productivity Gain

Keywords: X-Ray, Tomography, accessibility from one side

158

Ultrasonic Measurement of Residual Stresses in Welded Elements and Structures

UT Camera Imaging: An Alternative to Phased Array for Aerospace & Petrochemical Applications

Yuriy Kudryavtsev and Jacob Kleiman Structural Integrity Technologies Inc 80 Esna Park Drive Units 79 Markham Ontario L4J 0A1 Canada (416) 917-1519 e-mail [email protected]

Bob Lasser, Randy Scheib, David Rich, and Ola Mallaug Imperium, Inc. Beltsville, MD ABSTRACT Imperium will report on a study comparing the performance of the Company’s real time 2D ultrasound camera technology to phased array ultrasonic imaging. Imperium’s 2D technology uses a 120 row x 120 column ultrasound detector array to generate real time C-scan images. This is a very different technique than that used by phased array ultrasound testing (PAUT) systems, the currently prevailing ultrasonic imaging inspection technology. The differences in ultrasound physics suggest differences in performance, and one important performance metric is resolution. For a given operating frequency, the 2D camera system will be compared to a commercially available PAUT system. Our results will show that for a given ultrasound frequency, the 2D camera provides superior resolution. This paper will explain how this improved performance is a result of the detector array construction, image orientation, lack of geometric distortion, and lack of speckle. Additional metrics including ease of use, training, and cost comparisons will be outlined. We will also describe the industrial applications that have benefited from the technology as well as additional opportunities for the technique, resulting in a simpler, quicker alternative to phased array systems.

ABSTRACT The application of an ultrasonic non-destructive method for residual stress (RS) measurements has shown that, in many cases, this technique is very efficient and allows measuring the RS both in laboratory conditions and in real structures in field for a wide range of materials. Using this technique, one can measure the RS at the same points many times, studying for instance, the changes of RS under the action of service loading or effectiveness of stress-relieving techniques. An ultrasonic computerized complex (UCC) for nondestructive measurement of residual and applied stresses was developed recently. The complex includes a measurement unit with transducers, basic supporting software, an advanced database and an Expert System, housed in a laptop, for analysis of the influence of RS on the fatigue life of welded elements. In general, the ultrasonic method allows one to measure the RS in both cases: averaged through thickness or in surface layers. The present version of UCC allows measuring the averaged through thickness biaxial RS in plates 2 - 150 mm thick and surfacesubsurface RS at the predetermined depth. The results of ultrasonic RS measurement in large scale welded specimens and structures are also discussed in this paper. Keywords: UT, residual stresses, welded elements

159

Differentiation of 3D Scanners and Their Positioning Method When Applied to Pipeline Integrity

The Art of Technical Writing Oscar Lewis West Penn Testing Group 1010 Industrial Blvd New Kensington, PA 15068 (724) 339-1900 e-mail [email protected]

JeromeAlexandre Lavoie, Jerome Beaumont, and Mark Maizonnasse Creaform 3D 5825 rue SaintGeorges Levis G6V G6V 4L2 Canada (418) 833-4446 e-mail [email protected]

ABSTRACT Throughout history, procedures have been written to convey information for various reasons. Whether it is something as simple as directions on the side of a frozen pizza box to a manual on how to rebuild a transmission, the goal of a procedure is the same; to document a process so the end result is repeatable. How those instructions are written, however, is just as varied as what the instructions are for. Very few people have said “When I grow up, I want to write technical procedures.” Yet many of us end up being responsible for writing procedures as part of our job description. So, many of us are either not prepared to write effective procedures, or revert to what we have been taught from “traditional” English classes. The problem is “traditional” English classes teach what I call “literature” writing. Meaning, there are rules concerning sentence structure, punctuation, and even guidelines for styles of writing to keep the reader’s interest, but those rules do not completely transfer well into technical writing. I will discuss the goals of technical writing, many pitfalls of using traditional “literature” writing, and give some suggestions on how to write a more effective procedure.

ABSTRACT In the world of 3D scanning, the right scanner depends on the application, but also on the main goal of the people who will use it. Each method has its benefits but also its tradeoffs. This paper aims at helping service providers and asset owners select the most suitable 3D scanner solution for their inspection needs. 3D scanners are differentiated according to one of their main features: the positioning method they use. The measuring arm, the tracked 3D scanner, the structured light, and the portable 3D scanner categories will be investigated. More specifically, the two main positioning methods used by portable 3D scanners will be discussed: positioning through targets, and positioning through natural features. A third method called hybrid consists in combining the two. The positioning method is defined as the way a system captures the 3D space and then aligns the data collected during the scanning phase. 3D scanners are used for pipeline fitness-for-service evaluation in replacement of conventional methods such as pit gauge and ultrasound probes. Corrosion and mechanical damage can now be characterized with very high accuracy and repeatability. Each scanner category has been tested for corrosion assessment on a pipeline. We will see how they perform against each other and the importance of a proper positioning method.

Keywords: Technical writing, procedures, English, instructions

Keywords: 3D scanners, Pipeline Inspection, Corrosion, Mechanical Damages, pipeline integrity, direct assessment

160

Lessons Learned from Failed Radiographic Qualifications

The purpose of the Course is to certify and re-certify Non Destructive Technology (NDT) inspectors, in understanding and applying API RP 578 through an approved API-U Training Course that will qualify personnel in proper Guidelines and Application procedures utilizing XRF and OES technologies for PMI. The Certification course is covered in 2-day sessions and instruction on both, classroom theory with field testing procedures as a requirement. The need and now requirement for Positive Material Identification (PMI) has dramatically grown in the past few years in refinery and petrochemical plant operations to 100% alloy material verification in today’s risk-based quality control (QC) and quality assurance (QA) environment. This paper will explain why, in the United States of America (USA), the Occupational Safety and Health Administration (OSHA) implemented the “National Emphasis Program” (NEP) in June of 2007. This instruction provides guidance to Occupational Safety and Health Administration (OSHA) national, regional, and area offices and state programs which choose to implement a similar program concerning OSHA’s policy and procedures for implementing a National Emphasis Program (NEP) to reduce or eliminate workplace hazards associated with the catastrophic release of highly hazardous chemicals (HHC) at petroleum refineries. In July 27, 2009 OSHA implemented a National Emphasis Program (NEP) for the Petrochemical industry as a pilot program and then, implemented the Chemical National Emphasis Program ( CHEMNEP) (CPL 03-00-014)-Nationwide in 11/29/2011. Because of the corporation between different countries and the global nature of Occupational Safety and Health Administration (OSHA), instruction, we believe that this information and instruction is of great global interest. It is very apparent that “Mechanical Integrity” (MI) in the Oil & Gas Petrochemical process facilities and infrastructures are of the up most importance in today’s inspection activities.

Scott McClain ARDEC Picatinny Building 908 Picatinny Arsenal, NJ 07806 (973) 724-8428 e-mail [email protected] ABSTRACT This presentation details, with specific examples, symptons and root causes, for common problems encountered in qualifying and auditing various Radiographic inspections at a variety of manufacturing facilities. Keywords: Radiography, Digital Radiography, Qualification, Audit

Codes and Specifications for Minimizing Corporate Risk using Proper PMI Practices Don Mears Analytical Training Consultants Kingwood, Texas (281) 684-8881 e-mail [email protected] ABSTRACT Analytical Training Consultants has written and submitted to the American Petroleum Institute (API) a training course on API-RP 578 “Material Verification Program for New and Existing Alloy Piping Systems”, and was approved by the American Petroleum Institute (API) as a Certified Training Provider, (TPCP-#0118) in January 2008. Starting in 2014 this now operates under the API-U organization. We are the only API-U Company that is approved to train and certify Positive Material Identification (PMI) technicians, globally. This approved API-U PMI certification course is the topic of my presentation and will offer the reasons, purpose and criteria for this huge global need to be properly trained to perform Positive Material Identification (PMI) in today’s Petrochemical Oil & Gas Industry. We will explain the need and requirements for “Positive Material Identification” (PMI) using Portable X-Ray Fluorescent (XRF) and Mobil Optical Emission Spectroscopy (OES) Technologies.

161

Advancements in Digital Radiography for the Inspection of Welds CR, DR and DICONDE

applications. Advanced flat panel arrays (DR) are also now available and suited for filed applications. These detectors eliminate the intermediate step of secondary processing and the images are acquired in near real time. Both CR and DR technologies will be reviewed and analyzed. DICONDE compliancy is also a key component of the inspection of welds. DICONDE ensures that operators are not limited by current proprietary formats, eliminating the need for future data conversion and simplifying data integration from other NDT information sources, such as pipe management databases. This ensures that customers can choose the best-in-breed hardware and software platforms, while always ensuring the reliability of their data and its format. DICONDE based software platforms also allows one to manage all inspection data; Radiographic, Visual, Ultrasonics, Eddy Current, etc. on one platform and allows quick efficient sharing of inspection data and information. The phrase a picture is worth a thousand words comes true with DICONDE; the image and all the key information about that asset and inspection are stored right with the image. This provides a standard structure for querying on images and opens up the opportunity for advanced trending and data analysis based on inspection history ultimately leading to better asset management and improved asset uptime.

Johnny Gibson and Richard Mills Digital X-Ray Products GE Inspection Technologies 50 Industrial Park Rd. Lewistown, PA 17044 ABSTRACT Imaging in today’s digital age requires knowledge of the various detectors and the effect the detector’s attributes have upon image quality. There are numerous different modalities and many of these are best suited for a particular application. The application will determine which digital technology should be chosen. Basic questions, such as the need for portability, throughput, desired image quality and cost, must be answered. The scope of this presentation is intended to provide an overview of a few selected case studies where digital technology has proven to be cost effective and very successful from an image quality and productivity prospective for the inspection of in-process welds in various O&G applications. Digital radiography is now applicable in most areas where traditional film-based radiography is successful. Digital radiographic systems include Computed Radiography (CR), which uses a flexible, reusable, phosphor-based media that can be implemented in the same manner as film. The phosphor imaging plate can be carried into the field, exposed and returned to a location for field

162

Monitoring of Time-Dependent Degradation in Pipelines with Ultrasonic Guided Waves using Permanently Installed Sensors

Inspection of Composite Stingers using the SAUL (Surface Adapted Ultrasound) Method With Circular Arrays Guillaume Neau1 and Igor Ivakhnenko2

Peter Mudge1, Paul Jackson2, and Keith Thornicroft2

M2M 1 rue de terreneuve miniparc du verger batH les ulis idf 91940 France +33665559386 e-mail [email protected] 1

1 TWI Ltd Granta Park Great Abington Cambridge Cambridgeshire CB21 6AL UK +44 (0) 1223 899000 e-mail [email protected]

ATK Aerospace Systems

2

Plant Integrity Ltd

2

ABSTRACT In the aerospace industry, most composite structures undergoing NDT have complex and variable geometries. Traditional ultrasound NDT requires either specific probes, specific coupling tools for each geometry reference, andor high-resolution profile learning features, making the whole process difficult to automate. Surface-Adapted Ultrasound (SAUL) is a new technique developed by M2M that adds adaptability to any given probe. Moreover, this process does not require the knowledge of the precise geometrical and acoustical properties of the component undergoing inspection. In this article, the inspection of composite radii using curved array is studied with and without the SAUL process. Results show substantial advantages in terms of data quality and reliability.

ABSTRACT The determination of the condition of high pressure pipelines which are not suitable for internal in-line inspection (ILI) is problematic. Such transmission and distribution lines tend to be buried and have protective coatings which make them inaccessible for easy external inspection access. This is particularly relevant for gas lines. The situation is most significant for sections, where the pipeline crosses roads, railways or rivers, where a protective casing is present around the pipe for the length of the crossing. For these sections the outside of the pipe is inaccessible for direct assessment, even when it is excavated, and the costs of exposing the ends of the crossing may be excessive. Consequently, there is a need for a method of determining the conditionintegrity of such pipelines in general and that of cased crossings in particular. The PHMSA guidelines state that where it is not possible to gather sufficient information to determine the integrity of a cased section, excavations and direct examinations are required (unless a hydrotest is performed). Further, the Guidelines state that long range ultrasonic testing using ultrasonic guided waves (GWUT or GWT) is the only ‘Other Technology’ which may be used to provide information about the condition of cased crossings where ILI cannot be applied. This paper will describe developments by TWI and Plant Integrity Ltd, which address the need for improved performance of ultrasonic guided wave testing for examination of cased crossings and other inaccessible sections of pipeline, in particular the use of permanently installed sensors to permit high sensitivity assessment of time-dependent changes in condition of the line through, for example corrosion. The stability requirements for the sensors and novel methods of analysis and display of the test data to reveal time-dependent changes will be presented.

Keywords: phased-array, composites, aerospace

Keywords: GWT, permanently installed sensors, cased crossings, corrosion monitoring

163

Shearography NDT of Aerospace Composites

Laser Ultrasound: Inspecting Next Generation CFRP Structures

John Newman Laser Technology Inc 1055 West Germantown Pike Norristown, PA 19403 (610) 631-5043 e-mail [email protected]

Mark Osterkamp, Kenneth Yawn, and David Kaiser PaR Systems 1 Lockheed Blvd MZ 6852 4122C Fort Worth, TX 76108 (817) 777-7820 e-mail [email protected]

ABSTRACT Shearography nondestructive testing has gained rapid acceptance during the last decade as a highly cost effective means for the production inspection of aerospace composites and sandwich structures and is now included in NAS410. NDT was first introduced on the B-2 program in 1986 for inspecting composite materials and structures. Development of digital CCD cameras, the PC and small high power solidstate lasers have led to dramatic performance improvements.

ABSTRACT Since the introduction of carbon fiber reinforced polymer (CFRP) materials, their use has grown with each successive generation of new aircraft. Their strength-to-weight ratio and built-to-shape characteristic makes CFRP an attractive choice for design engineers. As a larger percentage of aircraft is made from CFRP, the complexity of these shapes continues to increase. Soon to be gone are the days of simple flat or rotationally symmetric composites (aka “black aluminum”) that you can test in an ultrasonic immersion tank or simply fixture in a water squirter. Design engineers are creating the next generation of composites – the integrated stiffener, fastener-less structures whose varying thickness and complex shape pose a nightmare for conventional ultrasonic systems. Laser Ultrasonics is a non-contact inspection methodology that uses lasers to induce and detect ultrasound in material structures. By eliminating the need for a coupling medium such as water, it is capable of rapidly inspecting large part areas without the specialized tooling or prolonged setup times common with conventional ultrasonic systems. The full-waveform data is readily interpretable by operators trained on conventional UT systems, simplifying integration of the technology into existing production lines.

Keywords: Shearography, NDT Testing, alternative NDT,

Keywords: Laser Ultrasound LaserUT UT

164

From the Aerospace to the Petroleum Industry: Fourier Transform Infrared Spectroscopy as a New NDE Tool

suggestion, Agilent started developing a handheld portable FTIR spectrometer. The rationale for this approach was to enable the FTIR analysis to be brought directly to the site of the sample, enabling non-destructive assessment of the part on location. In this presentation, the presenter will show how mobile FTIR can assist inspectors in qualifying parts. In particular, the presenter will illustrate the use of FTIR as a new NDE through the detection of incipient heat damage by FTIR, a technique developed by Boeing, the FAA, AMTAS and the University of Delaware and in use today for maintenance of the Boeing’s 787 Airliner. Here, FTIR can detect changes in the molecular structure of the composite induced by environmental stresses such as UV or lightning strikes, and correlate them to variations in physical strength of the composite (or shearing). To the same token, the author will demonstrate FTIR’s applicability for polymer aging assessment using an example in the power generating industry. Subsequently, the audience will be guided through the use of FTIR based chemometrics for the determination of curing time andor level (linked with the variation of the Tg) of various composites and polymer matrices, for the determination of composites surfaces preparation readiness for bonding purposes (Abrasion level, type of peel ply used, contamination level) , factors widely affecting quality indices. The presenter will then shift gear to address surface preparation (like coating) and contamination detection on various surfaces like metals and polymers. Finally, art conservation will be used as a conclusion to illustrate the wide reach of FTIR as a non-destructive analytical tool.

Frederic Prulliere, John Seelenbinder, Frank Higgins, Graham Miller, Christopher Sasso, James Fitzpatrick, and Leung Tang Agilent Technologies 5 Rowe Avenue Rockport, MA 01966 (864) 320-8625 e-mail [email protected] ABSTRACT Over time, NDE techniques like Eddy Current, X-Ray have focused on a physical rather than a holistic approach to part inspection. In many instances, the physical defects captured are merely symptoms of a deeper issue linked with the chemistry of the material. Complementary techniques like Fourier Transform Infrared Spectroscopy (FTIR) providing an insight into the chemical makeup of the inspected material can be very desirable for inspectors. For example, for materials containing a mixture of carbon and hydrogen (and other elements like Oxygen), FTIR looks at the chemical fingerprint of the part to infer identity andor possible chemical changes affecting parts specifications. For example, elastomer or composite based parts can show early failure due to insufficient curing not detectable by conventional methods. In a recent past, FTIR was mostly considered to be a lab bound technique where samples were brought to the lab for analysis. This procedure is hardly applicable to large composite parts, like aircraft wings which cannot be taken apart andor to non-movable parts like reinforced composites pipe in an upstream oil and gas application. At Boeing’s

Keywords: FTIR, NDE, Composites, Polymers, Chemical Fingerprint, Complementary Technique to Classical NDE, Curing, Heat Damage Assessment, Surface Preparation, Coating, Tg, Contamination Detection, Polymer Aging

165

Eddy Current Inspection of Twisted Tube Heat Exchangers / Field Trial Test Results

Optimization of Image Processing with Cone Bean CT Software Tools for 2D and 3D Digital Radiography

Tim Rush1, Skip Hoyt2, and Olivier Lavoie3 1 Mistras Group, Inc.

Daniel Shedlock, Martin Hu, David Nisius, and Josh Star-Lack Varian Medical Systems 8712 Barbee Ln Knoxville, TN 37923 (865) 560-6654 e-mail [email protected]

2

Marathon GBR

Koch Heat Transfer Company

3

ABSTRACT Over the last 20 years, twisted tubes have made their way into the Petrochemical and Chemical industries. In addition, twisted tubes are free of baffles or supports, which allow them to have the highest heat transfer coefficient of all tubular heat exchangers (up to 40% higher). Twisted tubes come in a variety of sizes from .625” up to 1” inch in diameter with material types such as stainless, titanium, brass, monel, carbon steel, duplex, and nickel (to name a few). Because of their unique enhancement design and geometric shape it’s been brought to the industries attention how to inspect twisted tubes? There are some existing applications such as long range guide wave ultrasonic and acoustic eye. However, the applications have shown little success thus far. This paper is a field case study of the application and the test results performed at the Marathon GBR turnaround. Also with the collaboration of Eddyfi, and Mistras recent innovation and development of new technology (Twisted Probe) that will have the capability of detecting, sizing and characterizing defects both internally and externally for the inspection of twisted tubes. Also, review of field data collection (titanium twisted tubes) will be presented to show its inspection speed, durability, and detection capabilities.

ABSTRACT CST (Cone-Beam CT Software Tools) is a software product for users of Flat Panel X-Ray detectors that provides an image processing pipeline for manufacturers of 2-D and 3-D systems. The toolkit contains a library of dll’s that can be used by integrators to quickly develop and deploy cone beam CT systems or to enhance projection radiographs (2-D imaging). For 3-D reconstruction, the product contains plugins that allow for CPU or GPU implementation of the FDK algorithm for arbitrary geometries. A ring correction algorithm is also included. For 2-D imaging, the package contains a resolution enhancement algorithm (REA) that corrects the light spread from thicker scintillator screens to allow the user to take advantage of higher efficiency screens without sacrificing spatial resolution. Applicable to both 2-D and 3-D imaging are plugins for lag correction, scatter correction, and beam hardening correction. Lag is a well-known issue associated with amorphous silicon flat panels. Scatter and beam hardening artifacts arise from the interaction of the interrogating x-ray beam with the object being scanned, and the plugins use physics-based models to correct for the resulting signal errors. CST has been established in the medical area and work is now beginning to bring it to bear on industrial applications. This paper will discuss the physical processes and methodologies for correcting these artifacts in flat panel imaging for industrial applications and will provide imaging examples demonstrating the effectiveness of each algorithm. Keywords: CT DR Cone Beam Scatter Correction Resolution Enhancement Lag 3-D GPU CPU

Keywords: TWISTED TUBES, heat transfer, titanium, corrosion types, etc. Applications used ECT twisted tube probe of titanium condenser bundles. Acknowledgements: Marathon GBR, Eddyfi, and Koch Heat Transfer Company

166

Optical Coherence Tomography using On-Chip Spectrometers

Microwave to Detect Cold Fusion Joints in High Density Polyethylene Pipe

Arthur Nitkowski, Kyle Preston, Nicolás Sherwood-Droz, Bradley Schmidt, and Arsen Hajian Tornado Spectral Systems 2359 N Triphammer Rd Ithaca, New York 14850 (607) 3796185 e-mail [email protected]

Robert Stackenborghs1, Ken Murphy2, and Brian Gray3 1 Evisive Inc 1925 Ryder Drive Baton Rouge, LA 70808 (225) 769-2780 e-mail [email protected] Exova

2

Spectrum NDT

3

ABSTRACT Tornado Spectral Systems has developed a chip-based spectrometer called OCTANE, the Optical Coherence Tomography Advanced Nanophotonic Engine, built using a planar lightwave circuit with integrated waveguides fabricated on a silicon wafer. Intended to support low-cost, highvolume applications, these spectrometers are well-suited for spectral-domain optical coherence tomography (SD-OCT) imaging for both biological and industrial non-destructive testing applications. The field of integrated optics enables the design of complex optical systems which are monolithically integrated on silicon chips. The form factors of these systems can be significantly smaller, more robust and less expensive than their equivalent free-space counterparts. Fabrication techniques developed by the microelectronics industry have previously been adapted for the telecom industry, and more recently for sensing applications. The chip-based spectrometer operates around a center wavelength of 860 nm with 70 nm spectral bandwidth and is able to record both polarizations of light on a 2048 pixel, 80 kHz line scan rate detector array. These specifications enable OCT images to be acquired with 2.7 mm imaging depth with 9.2 um axial resolution. OCT systems with spectrometers using integrated optics have large advantages in applications where size, robustness and cost matter: field-deployable devices, assembly line or factory installations, handheld scanning and more.

ABSTRACT This presentation describes an innovative apparatus and method that uses electromagnetic energy in the microwave frequency range to volumetrically examine dielectric materials, including high density polyethylene piping fusion joints. This presentation describes the theory of use and presents several HDPE inspection case studies. Specifically, this presentation describes the mechanics of cold fusion joint detection and in several cases the inspection results are compared to mechanical test results that confirm the accuracy of the examination. Keywords: HDPE, Microwave (MWT)

Keywords: OCT, NDT, spectrometer, nanophotonics, integrated optics, planar lightwave circuit, microfabrication

167

3D Reality Capture for Asset Management

Inservice Inspection for Enhanced RBI of Aboveground Storage Tanks

Tom Taylor Acuren 101 Old Underwood Rd La Porte, Texas 77571 (281) 228-0000 e-mail [email protected]

Sam Ternowchek Mistras Group Inc 195 Clarksville Rd Princeton Junction, NJ 08550 (609) 915-2798 e-mail [email protected]

ABSTRACT 3D reality capture technologies are evolving quickly. Quality is improving while costs are dropping. Learn how laser scanners achieve sub millimetre accuracy and capture the as built condition of facilities and equipment very quickly. This data can be leveraged using powerful visualization and simulation tools for optimizing asset integrity management.

ABSTRACT Typical Risk Based Inspection (RBI) programs rely on historical data and past inspections to evaluate an above ground storage tank. In many case this data can be several years old and not present an accurate picture of the tank’s current condition. In service inspections, that are conducted while the tank is on line, allow for a more accurate assessment of the tank’s condition, hence, a better assessment of the tanks overall condition. Using advanced NDT methods such as AE for assessing the condition of the tank floor, AUT for corrosion mapping of the shell and guide bulk wave UT for the annular ring, a current assessment of the tank’s condition is available for use with the RBI evaluation. These inspections are performed while the tank is in service so there is little interruption to tank operation. An example of how this assessment is performed will also be presented.

Keywords: 3D Mapping

Keywords: Risk Based Inspection, AUT, AE

168

Visualizing Internal Health of Frame Turbines Remote 3D Surface Scanning & Analysis Verify & Validate Before it Fails

Evaluation of Friction Stir Welds with Advanced Ultrasonic and Eddy Current Techniques

Paul Thompson North America GE Inspection Technologies Everest Vit (912) 335-8969 mobile (512) 213-8369 fax (866) 899-4184 e-mail [email protected] www.Geit.Com 622 Maupas Avenue Savannah, GA 31401

Evgueni Todorov, Roger Spencer, and Harvey Castner EWI 1250 Arthur E Adams Dr Columbus, Ohio 43221 (614) 688-5268 (614) 6885001 e-mail [email protected] ABSTRACT Friction stir welding (FSW) is used to join aluminum extrusions to form lightweight panels used in the construction of Navy ships. The National Shipbuilding Research Program—Advanced Shipbuilding Enterprise (NSRP- ASE) sponsored a large study to investigate technologies that could be used to improve the quality assurance capabilities for inspection of these FSW panels. To address the challenging FSW quality specifications, phased array ultrasonic (PAUT) and array eddy current (AEC) nondestructive evaluation (NDE) techniques were selected after an extensive literature review. Computer modeling of typical PAUT and AEC techniques justified the technique selection. High sensitivity and resolution PAUT and AEC field setups were developed. Statistically significant number of FSW specimens containing kissing bonds, incomplete penetration (IP) and incomplete consolidation (IC) were fabricated and tested with both techniques. Cross sections for metallographic analysis at selected areas were performed to objectively evaluate the NDE performance. The detection rate or score and probability of detection were obtained at several detection thresholds. While baseline FSW specimens (no flaw) were correctly classified with very low false call probability, the detection pattern on specimens with flaws was intermittent and required deviation from standard examination practices and procedures. Both advanced techniques demonstrated good detection capabilities. The NDE results were well correlated to other mechanical and fatigue tests conducted under the same project.

ABSTRACT 3D surface scanning for indication detection and analysis on remote visible surfaces improves safety & productivity. Optical metrology allows 3d visualization of internal surfaces in reciprocating & rotating equipment, piping, tubing, weld quality, and vessels used in aerospace, automotive, oil & gas and power generation. 3D surface scanning enables viewing an xyz surface map of an indication to enable the analysis process with high resolution. Benefit to inspectors & assets owners is ability to make better decisions faster about the health and viability of bringing the equipment back on line. Having this level of probability of detection and precision of analysis helps visual internal configurations and health of the asset and assists to quickly, accurately remove any doubt about the status of an indication.

Keywords: Friction stir welding, UT, ET, NDT modeling, POD, shipbuilding

169

Pulsed Terahertz Methods for Non-contact Inspection

Numerical Analysis of Acoustic Spark Source Focused by an Ellipsoidal Reflector for Air-Coupled Ultrasonic Excitation

Thomas Tongue and Brian Schulkin Zomega Terahertz Corp 15 Tech Valley Dr Suite 102 East Greenbush NY 12061 (518) 833-0577 e-mail [email protected]

Yi-Te Tsai, Jinying Zhu1, and Michael R. Haberman2 1 Department of Civil, Architectural and Environmental Engineering The University of Texas at Austin, Austin, Texas 78712

ABSTRACT The Terahertz (THz) band of the electromagnetic spectrum (0.1 – 10 THz) is very attractive for non-destructive testing applications in industrial processes. Pulsed Terahertz inspection methods combine many of the benefits of more common NDT techniques such as microwaves and ultrasonics to provide novel inspection capabilities. Both timeof-flight and spectroscopic information are measured by pulsed terahertz waves in non-conductive materials such as plastics, ceramics, paper, foam and certain composites. Similar to ultrasound, the position, amplitude and polarity of pulses contained in a terahertz waveform can be used to measure coating and layer thicknesses, as well as reveal cracks, voids, and gaps between layers. Until recently, pulsed terahertz systems were too slow to provide on-line inspection, but new techniques have unlocked high-speed process control. We present results using commercialized terahertz systems targeting the inspection market to measure layer thicknesses, bonds and cracks in plastic and ceramic materials unavailable by other methods.

2 Applied Research Laboratories and Department of Mechanical Engineering The University of Texas at Austin, P.O. Box 8029, Austin, Texas 78713-8029

ABSTRACT A spark source focused by an ellipsoidal reflector has been proposed as an excitation source for air-coupled nondestructive testing (NDT) of solids. Similar to an extracorporeal shock wave lithotripter which is used to break apart kidney stones, this device uses a high- amplitude spark source located at one focal point of an ellipsoidal reflector to generate an outgoing acoustic wave that is subsequently redirected to the other focal point of the ellipsoid where the target is located. Proof-of-concept operation of this device has been shown in an experimental study by the authors where Lamb waves in a concrete plate were excited. However, the proof-of-concept device suffers from significant energy loss at the air-solid interface due to the very large acoustic impedance mismatch between air and concrete. This paper presents numerical analysis and optimization of the ellipsoidal reflector in order to maximize excitability of the impact-echo mode of a solid elastic plate.

Keywords: Terahertz, pulsed, cracks, coatings, layers

170

Rapid Screening of Insulated Pipes for Corrosion under Insulation using Advanced Electromagnetic Technique

Review of Magnetostrictive Sensors for  Guided Wave Screening of Heat Exchanger Tubing

Ankit Vajpayee and David Russell Russell NDE System Inc 4909 75 Avenue Edmonton Alberta T6B 2S3 Canada (780) 468-6800 (780) 462-9378 e-mail [email protected]

Sergey Vinogradov, Charles Duffer, and Glenn Light Southwest Research Institute 6220 Culebra Road, San Antonio, TX 78238 (210) 522-3342; fax (210) 684-4822 email [email protected] ABSTRACT Magnetostrictive sensing (MsS) technology is widely used for guided wave testing of piping and also has been developed for screening of small bore heat exchanger (HX) tubing. Using guided waves for the screening of HX tubing is a very challenging application of guided wave testing because of the presence of multiple geometry features such as tube support plates and also of chemical deposits in the tube. Overcoming these difficulties required development of a number of different types of probes. One MsS probe type utilizes EMAT magnetostriction to generate guided waves directly in the tube wall (called electromagnetic coupling) used on ferromagnetic materials with high magnetostriction coefficient. Another type of probe is the “dry” coupled probe utilizing pressure and/or shear wave couplant. A review of current status of this work together with the results of field trials will be presented.

ABSTRACT Corrosion under Insulation (CUI) is one of the most expensive issues our industry is facing today. For a reliability specialist in a hydrocarbon processing environment, the issue has the potential to be catastrophic. For example: refinery’s steel piping is subject to temperature fluctuations. Thermal insulation applied to the pipe or vessel mitigates the effects, but the presence of seams, gaps or other discontinuities in the insulation layer makes them susceptible to infiltration by outside moisture or from the process environment. The result of infiltration is moisture held in contact with the pipe – resulting in CUI. Its occurrence can be unpredictable and undetectable based on visual examination. Traditional methods of addressing this issue involve selective removal of insulation for visual inspection; radiography or spot thickness measurements with PEC (Pulsed Eddy Current). This paper discusses the development and deployment of an Advanced Electromagnetic Rapid Inspection Technique on insulated pipe without the need to remove the insulation. Keywords: Corrosion Under Insulation (CUI), Electromagnetic Technique, Wall Thickness Measurement

171

Examination of Tube-to-Header Welds with Flexible Eddy Current Probes

Using Guided Waves to Monitor for Fatigue Damage on Electrical Transmission Tower Line Hangers

Jonathan Bartlett 1, Gary Burkhardt1, and Stanley Walker2 1 Southwest Research Institute 6220 Culebra Rd San Antonio, TX 78238 (210) 522-6099 e-mail [email protected]

Glenn Light, PhD and Sergey Vinogradov, PhD Southwest Research Institute 6220 Culebra Road San Antonio, TX 78238 ABSTRACT The number of green energy generation sites such as wind, solar, and other types of green energy being developed across the US and the world is increasing daily. These sites are usually in low population areas where the power is not needed. To get the power where is it needed, transmission lines are being installed to transport the power to the high population areas. Transmission lines including towers are expensive and often require a large amount of land on which the tower is located that has to be purchased or leased. Some of the newer transmission line systems utilize a monopole tower due to reduce the amount of land needed for the tower and to minimize the manufacturing and installation costs. For the monopole configuration, the electrical lines hang from a branch unit that is welded to the central monopole. This branch is subjected to a wide range of wind and temperature loading that can potentially cause weld fatigue cracking. Based on this potential failure mechanism, a nondestructive monitoring technology is needed to monitor the weld for cracking. SwRI has conducted a laboratory study to evaluate magnetostrictive guided wave technology for monitoring the weld for simulated cracking. The purpose of this paper is to present data collected on the laboratory mockup of an electrical tower branch hanger unit and to demonstrate that the guided wave monitoring approach can be used to provide early defect detection capability that might be indicative of potential weld failure. Data were collected for various defect sizes and over a temperature range of -30o F to 130o F (estimated to be the temperature range experienced by the transmission towers).

EPRI

2

ABSTRACT Heat recovery steam generators (HRSGs) in gas-turbine combined-cycle power plants contain numerous steel tubes that are welded to headers. Electric Power Research Institute (EPRI) has interest in the inspection of the tubeto-header joint welds to detect the presence of cracks. The geometry of the weld area is complex and is characterized by weld beads located at the intersection of a large diameter cylindrical header with a small diameter cylindrical tube. The surfaces of the HRSG tube-to-header welds are difficult to inspect with conventional eddy current testing (ET) probes due to the complex weld intersection and irregularities of the weld surface. In this project, SwRI developed two flexible ET array probes. One probe incorporates a compliant suspension that can conform to the weld geometry. Although this probe could be used with a suitable fixture for manual scanning, it is better suited to a mechanical scanner that can scan circumferentially and follow the geometry of the tube-toheader geometry. The second array probe was designed for manual scanning and was incorporated into a glove worn by the inspector. Finger pressure allows the probe to conform to the weld surface as it is scanned. Laboratory testing on welded mockups supplied by EPRI demonstrated detection of implanted cracks in the welds. The glove probe was evaluated at HRSGs at four different power plants. Overall, the use of ET array probes was shown to be effective for inspecting HRSG tube-to-header welds. Keywords: Heat Recovery Steam Generator, Flexible Eddy Current Array Probe

172

Inservice Composite Inspection, Adhering to the Level II Certification Requirements

The Role of NDT in Risk-based Inspection for Highway Bridges Glenn Washer and Massoud Nasrollahi University of Missouri Department of Civil Engineerin Lafferre HAll Columbia, MO 65203 (573) 239-1368 e-mail [email protected]

Rusty Waldrop USCG NDI Program Manager GS12 Elizabeth City, NC e-mail [email protected] ABSTRACT I am in the process of bringing up to speed aviation inspectors in composite material construction inspections. The document which is adopted for the Level II certification program is NAS-410. The requirement for the certification is basically: Two parts per method one part per technique. In my interpretation I am designing the process to have in the ultrasonic method the following applications in regards to aircraft composite structures. UT longitudinal wave using a delay line tip for thin laminate ply structures. These laminate structures can be carbon fibers, fiberglass, Kevlar ect. I consider this a technique. Another technique which is administered is a bonded material test using the resonance method for thin laminate plies. Can this technique be as reliable as UT with a delay line tip? It appears to be less difficult to set up and simpler to teach. Is there any pit falls? I consider this a technique within the ultrasonic realm. Pitch-Catch RF using the bonded material tester is considered an ultrasonic technique in the program and it is used to inspect sandwich construction materials to detect skin to core separations. Do other entities consider these methods techniques under the realm of a method? The program would require at a minimum two parts for UT a compressional wave application, and one part for the following techniques shear wave application, pitch catch RF and resonance and a delay line tip UT applications. At a minimum this would be 6 applications for one certification. Throw into the logistics of a UT practical examination a minimum of a ten point check list plus the requirement for a known condition to be detected for each application. Can a level III use one of the techniques to validate the finding of another in a practical examination? How do other entities prepare for this? How are components and or test samples developed or acquired for composite testing? The purpose of a practical is to show proficiency in the inspection of components. It is incumbent of the level III to ensure the practical requirements are adhered to and that the level III is confident of the individuals’ proficiency.

ABSTRACT The National Bridge Inspection Standards currently mandate a uniform inspection interval of 24 months for most highway bridges. This presentation will describe a new, riskbased methodology for inspection planning that anticipates a non-uniform inspection interval, ranging from 12 to 72 months,for highway bridges based on a risk analysis. This risk analysis identifies and assesses the likelihood of specific damage modes for bridges, and anticipated consequences. The role of NDT in condition assessments supporting the analysis will be discussed. Keywords: Risk-based inspection, bridges, RBI, infrastructure, concrete, steel bridges, decks

173

Imaging Plate Life for Industrial Field Radiography Applications

NDT Joins the IoT Revolution!  Marieke Wijtkamp

Brian White Carestream NDT 1049 Ridge Road Rochester, NY 14615 (585) 627-8017 e-mail [email protected]

DESCRIPTION: Join this session to find out how the Internet of Things (IoT) is positively impacting the remote visual inspection process. There are over 200,000 unconnected NDT devices deployed today. You will hear about a new innovation that significantly expands the value of these devices by adding live collaboration with remote experts to the inspection process. Unveiled at this event, you will see how you can immediately stream live visuals from these legacy NDT devices to the desktops, tablets and smartphones of remote collaborators in real-time.  Whether teams are inspecting aircraft, assets at an oil platform, or equipment in a power plant, learn how they can collaborate on live video and images to make better, faster decisions.  See this new live video inspection capability in action and hear examples of how it saves NDT service companies and enterprises time and money.

ABSTRACT Imaging plate life is determined by the operating environment and by a user’s specific image acceptance criteria. Use conditions for field radiography can be extreme. Radiographers work with a wide range of shot techniques, and in conditions with variable temperatures and humidity. For pipeline gamma radiography, imaging plates are repeatedly inserted and removed from flexible cassettes in contact with lead screens and are wrapped around a radius. Imaging plates get handled for multiple cycles, whereas film is only handled once. Customers expect an imaging plate to last for hundreds of cycles in order to realize a cost savings relative to film. Therefore, imaging plates must have sufficient durability to allow for repeated use without producing detrimental imaging artifacts. Imaging plate mean time to failure for various use conditions has been investigated to identify opportunities for computed radiography technology adoption in field radiography applications. Keywords: Digital imaging, computed radiography, industrial field radiography, imaging plate, durability, customer use conditions, pipeline, welding, RT

174

TDI (Time Delayed Integration) Technique and Device in NDT of Digital Radiography (DR)

A Reconfigurable System Design for Pipeline Inspection Using Guided Waves Ultrasound

LINBO YANG, Chinlee WANG, Nguyen Luu, and Dongri Meng XScan Imaging Corporation 107 Bonaventura Drive San Jose, CA 95134 (408) 432-9888 e-mail [email protected]

Lei Zhang University of Regina 3737 Wascana Parkway Regina SK S4S0A2 Canada (306) 337-2588 e-mail [email protected]

ABSTRACT There are basically two types of DR devices in current RT application. One is 2D device called flat panel, the other is 1D device called linear diode array (LDA). Both have advantages and disadvantages. For flat panel, it is difficult to make a large sized panel and to avoid object X-ray scatter during NDT operation especially at higher energy. To combine the advantages from both types, TDI technique is then introduced. A TDI device is between flat panel and LDA. This device has multiple diode lines and the signal for each line can be passed to the next line. As the object passes over each line, each line collects signal and then passes the signal to the following line. After the object passes the final line, the full integrated signal is read out. As a result signalto-noise ratio in TDI camera is much higher in other DRs. Esentially, a TDI device starts with 1D and then performs a limited 2D operation. Therefore, TDI device can be very large sized and has much less scatter problem. Some TDI device images are presented and results are discussed.

ABSTRACT Ultrasonic Guided Waves Testing (UGWT) is an efficient Non-destructive Testing (NDT) method used for quick long range defect scanning. Increasing numbers of requirements for quick long range testing have led to urgent need for improvement of testing methods and the development of new testing equipment to help the researchers in laboratory and help technicians in field inspection. The market for multi-channel ultrasonic instruments is characterized by small volumes of sales and a high ratio of development costs to sale price. The consequence is that the investment required to develop an instrument is significant and upgrades and other changes to instrument specification or configuration are difficult for many manufacturers to justify. These factors are particularly relevant for long range guided wave technology, where the technology is still relatively new in the market place and the instrumentation has different characteristics from other type of ultrasonic systems. In order to support the design and manufacture of new testing system, this study used a fully programmable gate array (FPGA) based control circuits for the electronics, which allows the component count to be decreased and also permits the system to support future functions to be modified via flexible reconfiguration of the FPGA. This has benefits of reduction of size, weight and power consumption of the electronics and provides a means of upgrading the product without expensive re-design of hardware.

Keywords: RT, DR, TDI, High Energy X-ray, Scatter, Large Sized DR device, signal-to-noise ratio

Keywords: NDT, Guided Waves, FPGA, pipeline inspections

175

Application of Flash Thermography in Automotive Carbon Fiber Composites

Reducing Arc Flash Risks with Electrical Maintenance Safety Devices

Haibo Zhao and Patrick Blanchard Ford Motor Company 2101 Village Road Dearborn, MI 48121 (313) 390-0620 e-mail [email protected]

Martin Robinson RISS Group 10306 Technology Terrace Bradenton, FL 34211 (941) 907-9128 e-mail [email protected]

ABSTRACT With the increasing usage of composite materials on automotive vehicles, non destructive testing of composites becomes a necessary area. Especially with the fast development of automotive carbon fiber composite manufacturing technologies, more metal components now can be replaced by composites to reduce the weight. That being the case, a suitable non destructive testing approach is needed in the plant to assure high quality products. In this paper, flash thermography was evaluated as a non-contact, full-field approach for non destructive testing of automotive carbon fiber composites. Two types of carbon fiber structures were studied and the capability of flash thermography in detecting the manufacturing defects was discussed.

ABSTRACT Electrical accidents, such as arc flashes, happen daily; however, there are ways companies and individuals can reduce the occurrence of these accidents and protect everybody concerned from the physical, financial, and statutory consequences. The National Fire Protection Association (NFPA) regulation 70E provides a reference for facilities to meet the requirements of electrical workplace safety while regulation 70B outlines the best practices for setting up and maintaining and Electrical Preventive Maintenance (EPM) program. VALUE OF NFPA At the heart of NFPA 70E and OSHA initiatives is the hierarchy of control. This concept attempts to mitigate arc flash risk wherever possible. In order of preference, the hierarchy of control prioritizes: 1. Risk Elimination 2. Substitution (with lower risk) 3. Engineering Controls (such as arc resistant switchgear) 4. Safe Work Practices 5. PPE

Keywords: Automotive, Composite, NDE, Flash Thermography

NFPA 70E and OSHA state that electrical equipment should be de-energized prior to opening. Some maintenance tasks have to be completed while the switchgear is loaded and energized, rapidly causing electrical maintenance safety devices (EMSDs) to be a hot topic. Their popularity is growing as companies strive to improve profitability, uptime and safety. Those who are implementing EMSD based programs are reaping significant benefits in terms of efficiency gains, cost control and fire prevention. Because lower PPE levels are required, inspections are quicker. EMSDs in Action One of the tasks that need to be completed on electrical equipment whilst it is energized and under load is infrared (IR) scanning. IR cameras can only measure what they can see and cannot see through glass or plastic viewing windows commonly fitted in switchgear. To allow the inspections to be completed under load, we use an IR window, an EMSD, that allows an IR camera to see the energized loaded connections through a special lens materials in the IR windows.

176

Allowing switchgear to remain closed and in a safe and guarded condition thus ensuring that the inspector is never exposed to the dangers of arc flash or electrocution. Much of the recent acceptance of IR windows has coincided with the increase in level of awareness regarding electrical safety, risk reduction and arc flash. Organizations such as the IEEE have been at the vanguard of this movement with its “Safer by Design” campaign. In response, switchgear manufacturers are increasingly installing IR window at the point of manufacture. Other tasks where EMSDs are used to eliminate a potential arc flash include: • Airborne Ultrasound (EMSD - Ultrasound Ports) • Voltage Detection (EMSD - External Voltage detection ports) • Motor Current Analysis (EMSD – Voltage Tap Off Connections)

standardized ensuring that any trend analysis data is accurate. Other benefits include: • Maintain switchgear in an enclosed and guarded condition • Remove risk of electrocution and possible triggers of an arc flash incident • Removal of high risk behaviors • Conduct valuable, fully loaded online inspections • Access Inaccessible equipment • Because there is no panel removal required: o inspections require less manpower o inspections require lower Personal Protection Equipment (PPE) levels o inspections are faster and more efficient o More inspections are completed due to ease of operation Summary It is significant that most electrical maintenance and safety standards value the use of Condition Based Maintenance (CBM) inspections such as IR surveys, ultrasound inspections, vibration analysis, MCA, and partial discharge testing as a critical part of an electrical preventative maintenance program (EPM). EMSDs, such as IR windows, have now provided a way for companies to comply with the recommendations for inspection processes, while complying with the mandate for arc flash avoidance. Most if not all of these organizations agree that electrical equipment should not be opened unless it is de-energized. EMSDs provide a way for companies to comply with recommendations for inspection and safety standardsguidelines while protecting personnel, equipment and profits.

Other EMSD strategies include the use of online monitoring systems that transmit data back directly to the client utilizing either wired or wireless sensor systems. These systems include: • Temperature measurement (Contact and non-contact systems) • Vibration Analysis (Rotating UPS and generator Systems) • Power Quality (Online and fixed data collection systems) • Partial Discharge (Online and fixed data collection systems) Benefits of EMSDs The benefit of using ESMDs is that they standardize the inspection routes as they become data collection points for the test equipment. They also ensure that all the inspection parameters are fixed and that all collected data is

Keywords: NFPA - National Fire Protection Assoc., Condition-Based Monitoring - CBM, Infrared - IR, PPE Personal Protection Equipment

177

Author index Abdelrahman, Marwa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Al-Shamari, Abdul Razzaq . . . . . . . . . . . . . . . . . . . . . 146 Al-Shaiji, Mohammad . . . . . . . . . . . . . . . . . . . . . . . . . 146 Alward, Kersten . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Annis, Charles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 Armitt, Tim. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 Austin, Russell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Averitt, WD. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 Bajula, David . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 Bartlett, Jonathan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Bartha, Bence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 Beaumont, Jerome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Bhardwaj, Anuj. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Bhardwaj, Mahesh C.. . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Blanchard, Patrick. . . . . . . . . . . . . . . . . . . . . . . . . . 138, 176 Bor, Zsolt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Brassard, Michel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Brunk, John. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 Buckley, Joe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Burkhardt, Gary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Calzada, Juan G. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Carden, Daniel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Chen, Xin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Cheng, Wenche. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 Chu, Tsuchin Philip. . . . . . . . . . . . . . . . . . . . 37, 72, 96, 150 Dao, Gavin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 Davies, Douglas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 Davis, Russell A. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Donikowski, Greg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Duffer, Charles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Duke, John C., Jr. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 ElBatanouny, Mohamed K. . . . . . . . . . . . . . . . . . . . . . . . . 9 Elston, Jeffrey. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 Enyart, Darrel A. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Fetfatsidis, Konstantine A. . . . . . . . . . . . . . . . . . . . . . . . . 17 Fitzpatrick, James. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Foos, Bryan C. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Furr, Parrish A. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Gehlen, Mark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 Geis, David. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Georgeson, Gary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Ghaziary, Hormoz. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

Gibson, Johnny. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 Gomez, Juan Mario. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 Gostautas, Richard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Gray, Brian. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Greene, Kenneth J. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Gupta, Arnab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 Haberman, Michael R. . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Hajian, Arsen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Hansen, John. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 Harel, Yoav. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Harvey Castner. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Hassen, Ahmed Arabi. . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Haszler, Alfred . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Heller, Thomas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 Heuer, Henning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88, 155 Higgins, Frank. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Hillmann, Susanne . . . . . . . . . . . . . . . . . . . . . . . . . . 88, 156 Hoa, Suong Van. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Hopkins, Deborah. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Hopman, George. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Hoyt, Skip. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Hu, Martin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Ivakhnenko, Igor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Jackson, Paul. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Jinhu, Wu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Johnson, Michael . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Johnson, Wayne . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Kaiser, David . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Kim, Heui. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Kim, Sang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Kleiman, Jacob. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Kleinert, Wolf. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Krueger, Peter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Kudryavtsev, Yuriy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Kumar, G. Santhosh . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 Lasser, Bob. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Lavoie, JeromeAlexandre. . . . . . . . . . . . . . . . . . . . . . . . 160 Lavoie, Olivier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Lei Zhang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Lessmann , Karsten . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Lewis, Oscar. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Light, Glenn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171, 172

178

Author index Lixiang, Hong. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Luu, Nguyen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Maizonnasse, Mark. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Mallaug, Ola. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 McClain, Scott . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Mears, Don. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 Meng, Dongri. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Michaels, Jennifer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Michaels, Thomas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Miller, David. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 Miller, Graham. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Mills, Richard. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 Mudge, Peter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Murphy, Ken. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Naghashpou, Ali. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Nasrollahi, Massoud. . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Neau, Guillaume. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Newman, John. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Nisius, David . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Nitkowski, Arthur . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Oberdoerfer, York. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Osterkamp, Mark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Patel, Kashyap. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Patsora, Iryna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Pergantis, Charles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Pooch, Matthias. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Poudel, Anish . . . . . . . . . . . . . . . . . . . . . . . . 37, 72, 96, 150 Preston, Kyle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Prulliere, Frederic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Rich, David. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Robinson, Martin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 Rush, Tim. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Russell, David. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Sandhu, Jaswinder Singh . . . . . . . . . . . . . . . . . . . . . . . . . 96 Sasso, Christopher. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Scheib, Randy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Schiller-Bechert, David-M. . . . . . . . . . . . . . . . . . . . . . . 156 Schmidt, Bradley. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Schmidt, Karl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Schulkin, Brian. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Schulze, Martin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

Seelenbinder, John . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Shedlock, Daniel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Sherwood-Droz, Nicolás. . . . . . . . . . . . . . . . . . . . . . . . . 167 Shrestha, Shashi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Spencer, Roger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Stackenborghs, Robert . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Stanley, Roderic K. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Star-Lack, Josh. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Tamutus, Terry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Tang, Leung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Taylor, John G. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 Taylor, Tom. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 Ternowchek, Sam. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 Thompson, Paul . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Thornicroft, Keith. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Todorov, Evgueni . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Tongue, Thomas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Trach, Phillip W. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Tsai, Yi-Te. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Uhlemann, Frank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Utrata, David. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Vaidya, Uday K. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Vajpayee, Ankit. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Vinogradov, Sergey. . . . . . . . . . . . . . . . . . . . . . . . . 171, 172 Vona, Paul . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 Waldrop, Rusty. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Walker, Stanley. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 WANG, Chinlee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Washer, Glenn. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Watson, Jon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 White, Brian. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Wijtkamp, Marieke. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Woodward, Robert J.. . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Yanchun, Zheng . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 YANG, LINBO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Yawn, Kenneth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Yester, Michael. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Zhao, Haibo. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138, 176 Zhifa, Chen. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Zhu, Jinying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Ziehl, Paul H. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

179