a knowledge-based framework for construction methods - cIRcle

A KNOWLEDGE-BASED FRAMEWORK FOR CONSTRUCTION METHODS SELECTION by Ibrahim A. Al-Hammad B.A.Sc, King Fahad University of Petroleum and Minerals, Saudia Arabia, 1981 M.A.Sc, The University of Colorado, U.S.A., 1985

A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T O F T H E REQUIREMENTS F O R T H E D E G R E E O F D O C T O R O F PHILOSOPHY

in T H E F A C U L T Y O F G R A D U A T E STUDIES D E P A R T M E N T O F CIVIL ENGINNERING

We accept this thesis as conforming to the required standard

T H E UNIVERSITY O F BRITISH COLUMBIA

April 1991 © Ibrahim A. Al-Hammad, 1991

In presenting this thesis in partial fulfilment

of the requirements for an advanced

degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department

or

by his or

her

representatives.

It

is understood that

copying or

publication of this thesis for financial gain shall not be allowed without my written permission.

Department of The University of British Columbia Vancouver, Canada

DE-6 (2/88)

Abstract

The objectives of t h i s t h e s i s are to investigate,

formulate,

and structure the problem of methods s e l e c t i o n , and apply a Knowledge-Based Expert System (KBES) approach. conceptual is

A complete,

KBES framework for the methods s e l e c t i o n problem

proposed

and

selected

aspects

of

i t were

implemented

using NExpert Object. Defined consists

hierarchically, of

the

construction

following

strategy,

are

to

versus others. at

two

first

attributes:

The

method design

frame

element,

resources,

and

roles of the KBES control

specify a method and

then

rank i t

In so doing, the control strategy i s applied

levels:

detailed

conceptual

construction

construction process model. strategy

a

a

feasibility

preliminary level.

feasibility

The

level,

and

a

former i s used to reduce

the number of available methods and rank them f o r processing by the l a t t e r .

The preliminary f e a s i b i l i t y part constitutes

declarative knowledge with high l e v e l premises. The of

detailed f e a s i b i l i t y

the

method.

This

level,

develops the

component

contains

attributes empirical,

a n a l y t i c a l , and procedural knowledge that draws on the engineering construction.

knowledge

domains

of

design,

civil

analysis

Because the notion of a frame i s a useful

of i d e n t i f y i n g the

and way

a t t r i b u t e s of a construction method, a

conceptual

frame

i s used

throughout

t o demonstrate the

build-up of the method attributes through preliminary, then detailed f e a s i b i l i t y . An

expert

system

called

CMSA

(Construction

Methods

Selection Assistant) was developed to implement a subset of the proposed solution approach with Cut-and-Cover tunnelling as the problem domain. methods

selection

domains.

shell

CMSA, as designed, constitutes a that

can be

applied

t o other

I t e n t a i l s a solution paradigm of Suggest, Design,

Predict, and Analyze operators. CMSA

incorporates

knowledge)

as

well

previous as

experience

algorithmic

(shallow

procedures

(deep

knowledge). Key elements central to CMSA knowledge base include r i s k , design technical f e a s i b i l i t y ,

resources compatibility,

and time performance measures,

and regulatory

cost

constraints.

Allowance i s made f o r modelling project context variables. A

range

of geotechnical

conditions

were treated

for

the

example problem domain. The

KBES

problem problem,

framework proposed

shows

promise

helping

to

f o r the methods

for tackling organize

this

site

contributing to productivity improvement.

iii

selection

ill-structured

experience,

and

Contents Abstract Contents Figures Tables Screens Listings Acknowledgement Acronyms

i i iv vii viii ix ix x xi

1. Introduction 1.1 1.2 1.3 1.4

Background Research Objectives and Methodology Problem Domain Organization of the Thesis

2. Literature Survey for Methods Selection Problem 2.1 2.2

1

1 2 4 5

9

Introduction 9 Construction Methods 9 2.2.1 D e f i n i t i o n of Construction Methods 12 2.2.2 Terminology Used i n the Thesis 16 2.3 Decision Making Model f o r Method Selection 20 2.3.1 Background 20 2.3.2 Simulation Techniques 20 2.3.3 Decision Analysis 22 2.3.4 Decision Support Systems (DSS) 23 2.4 Knowledge-Based Expert Systems 28 2.4.1 KBES Components 29 2.4.2 Expert Systems f o r Construction Management 33 2.4.3 KBES f o r Construction Methods Selection 35

3. Cut-and-Cover Methods in Soft Ground 3.1 3.2 3.3

3.4 3.5 3.6 3.7

Introduction Tunnelling Background Cut-and-Cover Tunnelling A l t e r n a t i v e s 3.3.1. Background 3.3.2 T r a d i t i o n a l Cut-and-Cover Tunnelling 3.3.3 Milano Cut-and-Cover Tunnelling 3.3.4 Major Operations Common to Cut-and-Cover Tunnelling GWSS Alternatives 3.4.1 Common Types of GWSSs Excavation Operations Factors A f f e c t i n g Methods Selection and Design Cut and Cover Tunnelling Project Example 3.7.1 Background 3.7.2 Lagging and Excavation Construction Cycle

iv

45

45 46 48 48 50 50 51 54 57 67 70 73 73 77

Contents 4. A KBES Framework for Methods Selection and Design 4.1 4.2

Introduction A KBES framework for Method Selection 4.2.1 General 4.2.2 Methods Selection Defined 4.2.3 Methods Shell 4.2.4 Sketch of System Features and Operation 4.3 CMSA Development 4.3.1 Overview 4.3.2 Context Modelling 4.3.3 Preliminary F e a s i b i l i t y 4.3.4 Detailed F e a s i b i l i t y Level 4.4 CMSA Risk Component Development and Evaluation

5. CMSA Implementation 5.1 5.2

5.3

Introduction NExpert Object Overview 5.2.1 Major NExpert Object Modules 5.2.2 NExpert Primitives 5.2.3 Viewing Knowledge Structure 5.2.4 The Inference Process CMSA Implementation 5.3.1. CMSA Overview 5.3.2 Solution Paradigm and Knowledge Base 5.3.3 Knowledge Representation 5.3.4 Technical F e a s i b i l i t y Part 5.3.5 CMSA Chaining and Reasoning (Control Strategy)

6. The Prototype Example 6.1 6.2

Introduction Example Problem Description 6.2.1 Session Start 6.2.2 Problem Context S p e c i f i c a t i o n 6.2.3 Modified Example 6.3 Risk Component Assessment Implemented 6.3.1 Introduction 6.3.2 NExpert Risk Implementation 6.3.3 Risk Routine

7. Conclusions and Recommendations for Further 7.1 7.2 7.3

Summary Contribution of The Thesis Further Research

Bibliography

81

81 84 84 84 89 92 106 106 110 115 130 149

154

154 156 156 160 167 168 174 174 176 187 207 214

219

219 219 220 224 238 239 239 246

250

250 251 252 255

v

Contents Appendix A.l A.2 A. 3

A: Pressures and Moments Computation Introduction L a t e r a l Pressure Calculations Design P r i n c i p l e s f o r S t r u c t u r a l Members

262 262 262 267

Appendix B. l B.2 B. 3

B: P i l e Driving Production Rate Derivation Introduction S o i l / P i l e F r i c t i o n Calculations P i l e Driving Production Rate Estimation

272 272 272 278

Appendix C. l C.2 C.3 C. 4

C: Interviews Introduction Minutes of Meeting with Dillingham Contractors Minutes of Meeting with Quadra Construction Project Site V i s i t

297 297 297 306 308

Appendix D. l D.2 D.3 D.4 D.5

D: CMSA P a r t i a l L i s t i n g and Miscellany Introduction P a r t i a l L i s t i n g of CMSA Knowledge Base Vibratory Hammer Selection Knowledge Unit Cost Quotations Sample Data Base F i l e s

310 310 311 331 333 335

vi

Figures Figure 2.1 Figure 2.2 Figure 2.3 Figure 2.4 Figure 2.5 Figure Figure Figure Figure Figure Figure

2.6 2.7 2.8 2.9 2.10 3.1

Figure 3.2 Figure 3.3 Figure 4.1 Figure 4.2 Figure 4.3 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure

4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 5.1

Figure Figure Figure Figure Figure Figure

5.2 5.3 5.4 5.5 5.6 5.7

Figure Figure Figure Figure Figure Figure

5.8 5.9 5.10 5.11 5.12 5.13

Design and Construction Interaction 11 Construction Model Process 13 Overview of C l a s s i f i c a t i o n System f o r Construction Technology 14 Example of Element, A t t r i b u t e , 14 Suggested Data Structure f o r Selected Technology 28 Sample Element A c t i v i t y Frame 38 Example of Knowledge Source 40 Labor Component Frame 42 Equipment Component Frame 43 Process Component Frame 43 Cost Comparison f o r Tunnelling A l t e r n a t i v e versus Cut-and-Cover A l t e r n a t i v e 48 Cut and Cover Tunnel Project Isometric 75 Barchart and Time-Space Diagram f o r the Seattle Project 76 Hierarchy of Construction Method Frame Attributes 85 Construction Methods Selection S h e l l 90 Construction Methods Selection System Process 94 Detailed F e a s i b i l i t y (Phase 2) 95 Steel Sheet P i l e (SSP) Method Frame 107 Prototype Model 109 S o i l P r o f i l e Scenarios 112 GWSS Frame Synthesis 117 CMSA Rule Execution Loop 119 Drive.c Routine Interface with CMSA 145 States of Nature for Methods Selection 150 Risk Assessment Tree Diagram 153 NExpert Object Open A l Environment Framework 159 NExpert Rule Construct 162 The Class and Object Hierarchy 166 Rules Perpendicular to Frames 169 Backward Chaining for Inference 171 NExpert Inference Framework 173 Knowledge Base Organization and Control Strategy 177 Implementation Solution Paradigm 178 Design Element Class Hierarchy 190 Design Element Instance Frame 191 Steel Sheet P i l e Class i n NExpert 192 Steel Sheet P i l e Selection Rule 193 Steel Sheet P i l e s Database (SSP.NXP) 195

vii

Figures Figure Figure Figure Figure Figure Figure Figure Figure Figure

5.14 5-15 5.16 5.17 5.18 5.19 5.20 5.21 5.22

Figure Figure Figure Figure Figure Figure

5.23 5.24 5.25 5.26 6.1 6.2

Figure Figure Figure Figure Figure Figure Figure

6.3 A.l A.2 B.l B.2 B.3 B.4

Construction Resource Class Hierarchy Impact Hammer Element Vibratory P i l e Driver Element Double Acting Hammer Database (DAAH.NXP) Vibratory Hammer Database (VIBRO.NXP) Impact Hammer Class i n NExpert Hammer Selection Rule i n NExpert Construction Strategy Class Hierarchy Construction Process Model Class Hierarchy Technical F e a s i b i l i t y Rule Method Technical F e a s i b i l i t y i s True Technical F e a s i b i l i t y Diagnostic Rule CMSA Model of Chaining and Reasoning Instantiation Tree Risk Assessment Decision Tree f o r Steel Sheet P i l e Risk Framework Assessment Flow Chart Pressure and Moments Envelopes S o i l P r o f i l e f o r Two S o i l s Scenario Hammer Blow Count versus S o i l Resistance Hammer Blow Count versus Driving Depth Drive.c Routine Flow Chart Drive.c Routine Development Flow Chart

196 197 198 199 199 200 201 205 206 208 212 213 217 237 242 248 264 267 281 285 287 288

Tables Table 3.1 Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table

3.2 4.1 4.2 4.3 4.4 4.6 5.1 5.2 6.1 6.2 A.l A.2 B.l B.2 B.3 C.l

P a r t i a l Space of Design/Construction Elements f o r Cut-and-Cover Tunnel Seattle Project General Information Methods Selection Space f o r GWSS GWSS Project Context Data S o i l P r o f i l e Input-Format 1 S o i l P r o f i l e Input-Format 2 Hammers for Different S o i l s Truth Matrix for NExpert Knowledge Base S t a t i s t i c s Risk Assessment Data Input Summary "SSP_Risk.nxp" F i l e for SSP A l t e r n a t i v e S o i l Types Properties Employed i n CMSA Lagging Members Values for Angle of Internal F r i c t i o n Ultimate Skin F r i c t i o n f o r Sands "Out.out" Sample Output Sample P i l e Driving Resources Unit Cost viii

53 74 87 97 113 114 137 164 186 247 249 263 271 274 274 284 307

Tables Table D.l Table D.2 Table D.3

Vibratory P i l e Drivers S i z i n g Crane Selection Format Representative Resources Unit Costs

332 333 334

Screens Screen Screen Screen Screen Screen Screen Screen Screen Screen Screen Screen Screen

6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 6.10 6.11 6.12

Screen 6.13 Screen 6.14 Screen 6.15

Knowcess Hypothesis Command Menu CMSA Overview Rule Network Window CMSA Rule Network Window GWSS Feasible Alternatives S o i l P r o f i l e S p e c i f i c a t i o n (1) S o i l P r o f i l e S p e c i f i c a t i o n (2) S o i l P r o f i l e S p e c i f i c a t i o n (3) S o i l P r o f i l e S p e c i f i c a t i o n (4) Water Table Level Input Hypothesis Select_Suitable_Sheet_Pile" Hypothesis Select_Suitable__Hammer Selected_Hammers Class and i t s Dynamic Objects Hammer E f f i c i e n c y Input "Drive.txt" Explanatory F i l e "Resultsl.nxp" F i l e ll

n

M

221 221 222 225 225 225 226 226 226 228 230 230 232 232 235

Listings Listing Listing Listing Listing Listing Listing Listing Listing Listing Listing

B.l C.l C.l D.l D.2 D.3 D.4 D.5 D.6 D.7

Drive.c Routine f o r P i l e Driving Extracted Rules of Thumb Extracted Rules of Thumb (continued) P a r t i a l L i s t i n g of CMSA Steel Sheet Data Base "SSP.nxp" Soldier P i l e s Data Base "HP_Pile.nxp" Struts Data Base "Strut.nxp" Lagging Data Base "Lag.nxp" Impact Hammer Data Base "Hammer.nxp" Vibratory Hammers Sample Data Base

ix

289 3 04 305 311 336 337 338 339 340 341

Acknowledgements

I am

greatly indebted

advisor

Professor

incisive Russell

input

Alan

and

provided

research.

My

i n large part of t h i s t h e s i s to Russell

for

constructive priceless

his

invaluable

criticism.

guidance

and

Professor

throughout

gratitude extends to my

my

this

t h e s i s supervisory

committee members of Dr. Caselton, Dr. Sassani, and Dr.

S.O.

Russell. Special

recognition

extends

to

the

Saudi

Arabian

Educational Mission and B r i t i s h Columbia Science Council for purchasing NExpert Object program and IBM 386 to conduct the research work.

Special thanks to Stuart Brown of Dillingham

Co.

Simonett

and

John

of

Quadra

interviews and exposing t h e i r Reflecting

back

at

my

Inc.

for

me

experience.

residence

in

campus,

benefited greatly from a number of people. thank Ronald Yaworsky and

granting

have

I would l i k e to

Leon Phem for t h e i r

Along with other friends, I extend my

I

assistance.

appreciation to Saad

Al-Mubiyedh, Mohammad Al-Robesh, Abdul-Aziz A l - J a l l a l , Tariq Al-Faris, Ranasinghe

Bernardo —

for

de

Castello, Rachid their

support,

Nakeeb, and encouragement,

Malik and

friendship that I value. Everything I have achieved can be a t t r i b u t e d to the love i and caring that my mother and family have provided. x

Acronyms

BC CMSA CS CPM CR DA DMM DSS PC GWSS IE KB KBES lbc lbpr PZ SPL SPT SSP STW ubc ubpr

Backward Chaining Construction Methods Selection Assistant Construction Strategy Construction Process Model Construction Resource Design A l t e r n a t i v e Decision Making Models Decisions Support Systems Forward Chaining Ground Wall Support System Inference Engine Knowledge Base Knowledge Based Expert Systems, Lower Bound Cost for a Project or A c t i v i t y Lower Bound Production Rate f o r a Project or Activity Heavy Steel Sheet P i l e s Soldier P i l e s and Lagging Standard Penetration Test Steel Sheet P i l e Slurry Trench Wall Upper Bound Cost f o r a Project or A c t i v i t y e Upper Bound Production Rate f o r a Project or Activity.

xi

1. Introduction

1.1 Background Construction methods selection i s a challenging general,

there

are

numerous

alternative

for each a c t i v i t y , and

The

In

practice,

decision

for

number of

the p o t e n t i a l

i n t e r a c t i o n among them, makes methods s e l e c t i o n process.

In

methods

performing each major a c t i v i t y i n a project. methods available

problem.

makers

a

rely

for

complex on

past

experience from s i m i l a r projects to provide solutions to the current ones. Given the

selection of a method, t r a d i t i o n a l techniques,

such as network analysis,

simulation

can be used to predict time and these techniques are evaluative, incorporate

or decision

analysis,

cost performance.

However,

not generative, and

h e u r i s t i c knowledge e x p l i c i t l y .

The

do

strength

not of

quantitative modelling techniques l i e s i n t h e i r lower l e v e l prediction and A

optimization.

combination

knowledge

is

Consequently,

of

essential a

both to

computerized

descriptive effective

and

methods

decision-making

procedural selection. tool

which

embodies a Knowledge Based Expert System (KBES) i s worthy of investigation.

Such a t o o l

should

integrate

quantitative

and q u a l i t a t i v e assessments i n order to produce and acceptable solutions.

analyze

Chapter 1. Introduction

The focus

of t h i s

2

thesis i s on developing

a

conceptual

framework f o r describing the methods s e l e c t i o n problem and on i d e n t i f y i n g the roles that

can be played

by knowledge

based systems.

1.2 Research Objectives and Methodology An

extensive

general

review

of the l i t e r a t u r e

statement and formulation

problem has been developed.

revealed

that

no

of the methods s e l e c t i o n

In fact,

few researchers

ever

address the issue, p r e f e r r i n g to focus on s p e c i f i c problems. In t r e a t i n g s p e c i f i c problems, research has been directed at applying Little

operations

attempt

research

has been

and systems

analysis

tools.

made to t r e a t the problem

as a

design s i m i l e as opposed to an analysis one and incorporate construction

knowledge

in

Complicating

the problem

construction

method

in

the

cognitive

process.

i s the dimensionality terms

of

a

large

of

number

a of

quantitative and q u a l i t a t i v e a t t r i b u t e s , the combinatorial problem of combining methods and the multiple c r i t e r i a used for evaluating a method. The main goals of t h i s thesis are twofold: generalized

statement

and

structure

to develop a

f o r the

methods

s e l e c t i o n problem, and to demonstrate the a p p l i c a b i l i t y of a knowledge based approach to t h i s problem. prototype

expert

system

called

CMSA

For the l a t t e r , a

(Construction

Selection Assistant) has been developed.

Methods

Chapter 1. Introduction

This

3

t h e s i s addresses the

methods s e l e c t i o n problem i n

broad terms of organizing, structuring and and to propose and

formulating i t ,

implement a KBES framework approach.

An

appropriate domain example of Cut-and-Cover t u n n e l l i n g w i l l be

used.

A

central research

objective

i s to

contribute

toward the representation of a construction methods design environment

that

can

handle

a

methods/strategy s e l e c t i o n problems.

wide The

variety

of

main function of

the environment i s to provide the construction user with subset

of

feasible

including preliminary

methods, values

given

a

for design

project

context,

parameters of

short l i s t e d methods. S p e c i f i c research objectives are as follows: 1.

Develop a generalized description and structure for the construction methods selection problem;

2.

Identify s p e c i f i c roles based system can f u l f i l l ;

a

knowledge

3.

Develop a detailed representation for describing i n d i v i d u a l methods;

structure

4.

Formulate a process of traversing alternate methods and pruning a l t e r n a t i v e s ;

. 5.

that

Treat multiple decision c r i t e r i a , time, cost, and r i s k ;

a

including

6.

Consider both h e u r i s t i c and knowledge; and

procedural

7.

Develop a prototype system, using the context of Cut-and-Cover tunneling, to demonstrate and partially validate findings from objectives 1 through 6.

the

Chapter 1.

Introduction

4

The research methodology employed consists of a number of parts.

First,

an

extensive

conducted to i d e n t i f y the useful

approaches to

specific

project

tunnelling) provide

to

a

base

context

the

was

specificity

which

of

literature

state-of-the-art and

each of

bring

review

could

be

to i d e n t i f y

objectives. selected to

the

Second,

a

(Cut-and-Cover process

further

was

and

to

generalized

by

examining other methods s e l e c t i o n problems. S p e c i f i c research methodologies are as follows: 1.

Extensive l i t e r a t u r e review of previous approaches to the problem as well as conducting field interviews;

2.

S p e c i f i c project context s e l e c t i o n GWSS alternative for the tunnelling);

3.

Devising a wholistic definition for the construction method a t t r i b u t e s which serves as a basis f o r structuring the problem;

4.

Employing KBES Techniques, including knowledge a c q u i s i t i o n , knowledge representation schemes, and control strategy design; and

5.

U t i l i z i n g an Expert Systems s h e l l (NExpert Object) as a basis for a working prototype to demonstrate v i a b i l i t y of suggested approach.

(sheet p i l i n g Cut-and-Cover

13 Problem Domain Cut-and-Cover tunnelling has explore richness

and is

construction among

structure the in

terms

been chosen as

methods s e l e c t i o n problem. of

the

number

a l t e r n a t i v e s available,

construction

a vehicle

activities,

the

and

of high its

design

to Its and

interaction distinctive,

Chapter 1.

Introduction

5

d i s c r e t e , r e p e t i t i v e construction a c t i v i t i e s .

Further,

the

experiential

spans

the

fields,

which

knowledge

geotechnical,

associated

structural

and

with

construction

makes i t a good candidate f o r exploring solving

framework.

The

it

poorly

a general

structured

problem

nature

of

the

problem domain makes i t amendable to a KBES approach. Problem requires

domain

solving,

knowledge

about

within

the

thesis

soil-structure

context,

interaction

to

design the ground wall support system (GWSS); Cut-and-Cover construction

techniques;

sequencing

and

scheduling;

optimizing resources; and general project management.

1.4 Organization of the Thesis The

remainder

Chapter

2

of

this

thesis

examines previous

selection.

A

start

is

organized

work on

as

construction

i s made toward s e t t i n g out

d e f i n i t i o n of the methods s e l e c t i o n problem. approaches simulation applied

to

reviewed.

(decision and

so

analysis,

forth),

methods The

follows.

and

goal

of

this

a

operations

problem chapter

research, approaches

modelling, is

to

are

provide

comprehensive view of the s t a t e - o f - a r t of previous at modelling

general

Conventional

knowledge based

selection

methods

a

attempts

the construction methods s e l e c t i o n problem.

Chapter 3 examines Cut-and-Cover tunnelling, the selected problem

domain,

emphasis placed

and on

related

the

construction

ground wall

support

methods, system

with (GWSS)

Chapter 1.

design

Introduction

and

6

installation

alternatives.

Analytical

h e u r i s t i c design and construction procedures are

and

discussed.

Cut-and-Cover tunnelling i s a knowledge-rich problem domain that i s used as a vehicle to explore the methods s e l e c t i o n problem through hypothetical used to expose how

examples.

These examples

are

excavation operations are interwoven with

GWSS retaining system design a t t r i b u t e s . Chapter

4

sets

out

a

framework

methods s e l e c t i o n problem.

for

Design and

structuring

construction

the

tasks

are analyzed for methods s e l e c t i o n ; a comprehensive method frame

definition

comprehensive prescribed

KBES

The

or

system

control

alternatives

level,

introduced;

approach

including

description. method

is

by

for

features

operating

knowledge;

costs

are

used

levels.

Key

rejected

at the

and

simplified

rank

elements two

declarative

compatibility,

to

to

two

and

a

synthesizes a

high

constitutes

feasibility

that

method

at

both

accepted

or

decomposed into procedural

categories measures,

risk

strategy

levels:

that

is

Evaluation c r i t e r i a based

make

l e v e l s are

performance

selection

f e a s i b l e alternatives

that

knowledge

approach

ranks at

conceptual

control

detailed

constitutes procedural knowledge. on

and

feasibility and

a

method

strategy

preliminary

declarative

and

of and

assessment

design, regulatory. is

risk, A

presented.

Chapter 1. Introduction

Treatment

of project

7

context

variables

(state v a r i a b l e s ) ,

p a r t i c u l a r l y of s o i l conditions, i s also discussed. Implementation

of the model

features

described

l a s t part of chapter 4 i n the form of an expert described The

i n chapter 5.

main features

in

the

system, i s

The s h e l l used i s NExpert Object.

of t h i s

shell

are b r i e f l y

highlighted.

Knowledge structure and control strategy for CMSA prototype development follow the conceptual represent

method a t t r i b u t e s of design

strategy,

construction

process models. and

to screen

Samples

model.

of

related

Frames are used to

element, construction

resources

and construction

Rules are used for the control alternatives

knowledge

and represent

constructs

soil

utilized

strategy, profiles.

i n CMSA are

provided. Chapter

6

presents

a

detailed

example

of the CMSA

prototype synthesizes a Cut-and-Cover tunnelling a l t e r n a t i v e of s t e e l sheet p i l e s .

Conclusions

further research are contained

and recommendations f o r

i n chapter 7.

A number of Appendices contain d e t a i l s of the algorithms and

CMSA knowledge base.

pressure

Appendix A contains

and moment c a l c u l a t i o n s

ground wall

support system

formulas f o r

f o r the design

(GWSS).

of the

Appendix B covers the

mathematical derivations for s o i l resistance for impact p i l e d r i v i n g hammers. duration

of

pile

They serve as an algorithm t o p r e d i c t the driving

for a

single

pile.

This

Chapter l . Introduction

8

information i s used i n turn by the CMSA to p r e d i c t a t o t a l production

rate.

Appendix

knowledge a c q u i s i t i o n formal that

and informal contributed

C provides

process interviews

insights into the

followed. conducted

to the knowledge

Covered and s i t e

acquisition

are

the

visits process.

Appendix D supplies a l i s t i n g of the CMSA knowledge base and the data bases used.

2. Literature Survey for Methods Selection Problem

2.1 Introduction The l i t e r a t u r e review i n t h i s chapter i s divided into three parts.

First,

problem

attributes

Subsequently, general

instances

and

selection

of methods

definitions

the attributes

methods

thesis.

previous

are

and d e f i n i t i o n

problem

summarized. f o r a formal

are set out f o r t h i s

Second, operations research and systems

approaches examined.

to

specific

methods

selection

analysis

problems

are

Third, recent work involving the application of

expert systems to t h i s problem i s explored. approach,

selection

i t i s assumed that the reader

For the l a t t e r

i s familiar

with

KBES concepts and terminology.

2.2 Construction Methods To r e s t r i c t the scope of the thesis problem, we assume that the design of the permanent f a c i l i t i e s the

kind

of methods selection

tool

i s fixed.

Clearly,

investigated

i n this

thesis could be used by designers to better l i n k the design and

construction processes, as well as by contractors who

seek to optimize t h e i r decisions given a design. Gray (1986) has argued asserts

that the concept

the case f o r the former where he of b u i l d a b i l i t y

suggests

it

is

advisable to involve the contractor i n the early stages of

Chapter 2. Method Selection Problem and i t s Approaches

the

design

process.

In essence,

i t means that

10

a deeper

understanding of the methods used by contractors to analyze the

problems

and

risks

inherent

i n a design

has to be

achieved. Few

construction

involvement

in

Traditionally, facilitate

contract

methods

to

to

frequently,

contracts

documents used

with

submit

methods

allow

planning

competition

project

contractor

the

most

more

be

contracts

and

inhibit

among may

this

as

stages.

i n order

bidders.

provisions

seen

contractor

design

specify

h i s methods are

early

Occasionally,

the that

construction permit

the

f o r approval. the

to

purview

More of

the

contractor and contract documents are s i l e n t on the methods to be employed. In

a

conventional

system,

preliminary decisions regarding

the

contractor

makes

construction methods based

on the information available during b i d preparation.

I f the

contractor i s successful i n being awarded the job, previous decisions are reviewed and further decisions are taken i n l i g h t of more complete information. In

an

idealized

system,

the contractor

i s allowed

to

bring h i s construction methods knowledge to the early design stage.

Figure 2.1 contrasts conventional

systems

(Gray 1985).

This

figure

among relevant decision categories.

versus

idealized

illustrates

interaction

For the

conventional

Chapter 2. Method Selection Problem and i t s Approaches

11

system,

of

the

previous

arrows

decisions

modification.

show on

a

unidirectional

more

recent

ones

influence with

little

For the i d e a l i z e d case, the arrows are b i -

d i r e c t i o n a l between the three decision categories.

Parties

Approach Conventional

Idealized

Planning and Design

Feasibility Study Master Plan Design Alternatives Work Volume Contract Preparation

Owner, A/E, Creditor,

Construction Planning

Resource Allocation Productivity Planning Methods Selection Detailed Scheduling Procurement Subcontract Evaluation

CM, Contractors, Subconts., Sureties, Material Suppliers, Traders,..

Different Site Conditions Change Orders Actual Productivity

CM, Contractor, Owner, A/E, Subcontractors, Sureties, Material Suppliers, Trades

Construction Execution

CM, Contractor*

* involved in idealized system

Figure 2.1

Design and Construction Interaction [Adapted from Gray 1985]

Chapter 2. Method Selection Problem and i t s Approaches

12

The decision making process of methods s e l e c t i o n tends to be evolutionary i n nature, i n which each decision i s l i m i t e d by

decisions made at e a r l i e r

approach, or dynamically,

2.2.1 No

stages,

as

traditional

as i n an i d e a l i s t i c approach.

D e f i n i t i o n of Construction Methods (Previous Work) universal

definition

s e l e c t i o n problem has

of

the

emerged.

construction

terminology

methods

Most d e f i n i t i o n s given

the l i t e r a t u r e are context s e n s i t i v e .

We

in a

commonly found i n the

in

In t h i s section, the

literature

i s reviewed.

incorporate i t i n our general d e f i n i t i o n s on the methods

s e l e c t i o n problem as appropriate. Construction technology can be defined as the science of construction materials,

involving

the

methods, and

judicious

equipment

planning, preparation, and execution Tatum described

use

of

available

including the

necessary

(Merritt 1976).

a construction technology

classification

system that includes a hierarchy of four parts: components, elements, Figure

a t t r i b u t e s , and

2.2

presents

hierarchical (1987).

values

a sketch

construction Further,

components: material

1987

and

1988).

of a proposed model for

process

as

defined

2.3

shows

the

figure and

(Tatum

equipment resources,

the

by

Tatum

four

major

construction

applied resources, project requirements and constraints, and construction

processes

(Tatum

1988).

A

synthesis

of

the

Chapter 2. Method Selection Problem and i t s Approaches

first model

three components, Construction process, by

which

performance

measures

13

represents a

under

different

scenarios are derived. The a t t r i b u t e s of a construction method

can be further

elaborated upon as shown i n figure 2.4.

^

Owiwr

^

( CapLl ")

/ArehltocV ~N \jnflln««r J

I -^Coratniotor^"

Supply Resource* (Elements) Place Tools People

Information Energy Materials Foreman Crew

J

Method Product

Figure 2.2

Construction Model Process [from Tatum 1987]

Chapter 2. Method Selection Problem and i t s Approaches

14

Materials and Permanent Equipment Resources Construction Applied Resources

+ Information + Skills + Equipment + Tools + General Conditions + Space + Energy + Time

Construction Processes

+ Methods + Tasks

Project Requirements and Constraints

Constructed Product

Legend: • component of technology + element of technology

Figure 2.3 Overview of C l a s s i f i c a t i o n system f o r Construction Technology [from Tatum 1988]

Possible Text 1 Construction Methods + + + + + + + + +

primary location degree of automation' degree of complexity experience available degree of interdependancy^ point of origination fundamental process basic type degree of uncertainty

Values of Attributes: fab shop, offsite, staging area, workface, yard

3

3

3

crew, designer, planner, superintendent, vendor batch flow, Job shop, worker-paced assembly line

Notes 1. element of the construction process component 2. text values for attributes that do not allow quantitative values 3. attribute has quantitative value ranging from 1 (minor) to 6 (extreme) + Indicates an attribute of the element

hand tools, heavy equipment manual, precision

Figure 2.4 Example of Element, A t t r i b u t e , and Value [from Tatum 1988]

Chapter 2. Method Selection Problem and i t s Approaches

Construction

method has been defined

which resources

15

as the manner i n

on s i t e are used to achieve s p e c i f i e d forms

of construction

(Mansero 1987).

elements i d e n t i f i e d as being

In t h i s reference,

major

part of the description of a

construction method include: 1.

a precise sequence of operations;

2.

the r e l a t i v e operations;

3.

i n t e r a c t i o n patterns with other

4.

construction plant;

5.

expendable material and temporary works;

6.

temporary services; and

7.

craft s k i l l s .

Although selection, confine

many there

this

pace

factors

of composite

must

be

parts

of

operations;

considered

in

plant

are basic p r i n c i p l e s that can be used to problem.

For

example,

in

building

construction, a crane i s considered to be a key resource for materials handling f o r major operations such as transporting forms, concrete,

and other

materials.

(1986) developed a systematic a

suitable

crane

several f a c t o r s . tower

cranes

For instance,

Gray

approach f o r the s e l e c t i o n of

f o r a high-rise

building

that

embraces

Among these factors are number and type of

versus

mobile

cranes,

work

load,

productivity rate, and crane work space and reach, f i n i t e number of available cranes.

required given a

This provides a tangible

Chapter 2. Method Selection Problem and i t s Approaches

16

example of s t r u c t u r i n g the knowledge pertaining to a complex problem of multi-task resource s e l e c t i o n .

2.2.2

Terminology Used i n the Thesis

This section presents the d e f i n i t i o n s of a number of terms that form part of the o v e r a l l construction methods s e l e c t i o n problem.

They are elaborated upon i n subsequent

chapters.

We take a r e s t r i c t i v e view that the design of the permanent facility

i s fixed.

provisions may engineering

Later,

we

acknowledge that

contract

consider other design a l t e r n a t i v e s or

proposals.

In addition, the

value

requirements for

temporary works, such as shoring and formwork, may

become a

s i g n i f i c a n t design component.

Design Approach: its

elements,

response

to

Includes

and

dimensioning

specifying the

loads,

required

the

structure

and

of

materials

in

types

functions,

site

features,

regulations, and so on.

Construction

Plan/Strategy:

Represents

the

high

level

abstraction of major aggregate a c t i v i t i e s that are sequenced i n a pre-determined l o g i c a l manner to r e a l i z e a design.

The

estimate

the

of

an

activity

duration

selected construction method and

is

process.

dependant

on

Chapter 2. Method Selection Problem and i t s Approaches

An

example

building

is

simultaneously

of

a

construction

to

proceed

strategy

excavation

in

a

17

high-rise

activity

downward,

with constructing the super-structure upward,

i n contrast to the t r a d i t i o n a l bottom up construction.

Construction Process Model: equipment, project

labor

context

operation

material,

constraints

or construction

construction concrete

and

process.

high-rise

superstructure

Involves the i n t e r a c t i o n among

can

under

which

i s used i n

instance,

building,

concrete

be modelled using

in

other

a

major

a

analyzing reinforced

placement

for

the

CYCLONE (construction

process simulation program, see Halpin 1976) progress rate which i n d i r e c t l y

and

characterize

cycle that For

physical

to measure i t s

indicates the whole project

progress pace. The and

construction

process model combines design

construction resources

according

to a set of

elements criteria.

Thus, i t could be used as a decision-making t o o l to rank, and

accept or

reject

for the model may system), t o t a l may

include

construction.

Quantifiable measures

include progress rate (productivity of the

cost, the

alternatives.

and

duration.

quality

of

work

Q u a l i t a t i v e variables and

safety

during

Chapter 2. Method Selection Problem and i t s Approaches

Construction Method: design,

strategy,

Consists of a h i e r a r c h i c a l assembly of

resource,

and

process

characterize a s p e c i f i c operation. trench

wall

following

(STW)

method

components.

for

The

and/or

a

GWSS

design

permanent

strategy i s bottom-up. those

involved

concrete process

in

The

each

placement, and includes

the

specifies a

physical

and

resources

The

the

include

concrete mix

facility).

operation,

design

construction

such

as

excavation,

r e t a i n i n g system installment.

sequencing

of

a

employed include a l l

arrangement of

resource

that

encompasses

elements

The

resource i n t e r a c t i o n ,

which constitutes a construction cycle. cycle

components

For instance, a s l u r r y

s p e c i f i e d s l u r r y type and density, and (temporary

18

the

The

construction

activities

interactions

as

well

based as

on

other

requirements imposed by the project context.

Construction

Methods Selection:

methods s e l e c t i o n consists of: method

which

strategy, process;

and

consists

construction

Designing

attributes;

of

design

A

w h o l i s t i c paradigm

Suggesting elements,

resources,

and

a

of

preliminary construction construction

a the method i n terms of specifying i t s

Synthesizing

a method by

means of a model to

predict i t s performance measures for the suggested method; Analyzing thereby

the

performance measures versus

accepting

or

rejecting

a

expected

method

goals,

alternative.

Chapter 2. Method Selection Problem and i t s Approaches

19

"Recommendation" f o r a change i n the method a t t r i b u t e values to

alter

a method that

design effectiveness,

was rejected

i s included

or to enhance method

i n the Analyze operator.

In essence, each element of the method i s instantiated, and the

resultant

attributes

values

define

the most

method based on project context variables.

suitable

Chapter 2. Method Selection Problem and i t s Approaches

20

23 Decision Making Model for Method Selection 2.3.1

Background

Considerable adapting

effort

various

has

been

operations

expended

research

in

and

applying

systems

analysis

t o o l s to the problem of construction methods s e l e c t i o n . general,

the

problem

might

Scarpa

1980,

dynamic

approach be and

analyzed 1984),

programming

simulation, (Halpin

has

been using

queuing

(Gaarslev

1976,

to

show

how

a

optimization

and

In

specific

(Gates

and

theory

(Ringwald

1987),

1977,

Selinger

1980),

Ashley 1980), and so f o r t h : i . e .

given a construction related problem, here i s how

i t might

be analyzed. Generalized

approaches

and

definitions

of

methods

s e l e c t i o n which can be applied to a large class of problems, are the exception, work of Halpin 1986),

not the r u l e .

A notable exception

(Halpin and Woodhead 1976, Halpin and

i n which he has attempted to develop a

i s the Bernold

simulation

approach f o r t r e a t i n g a broad range of problems.

In t h i s

section we review previous work directed at improving one's a b i l i t y to model and r e f i n e construction methods s e l e c t i o n .

2.3.2

Simulation Techniques

Simulation

can be used

i n planning

and

scheduling

r e p e t i t i v e cycles i n a construction project. oriented simulation

c a l l e d CYCLONE was

highly

A construction

developed by

Halpin

Chapter 2. Method Selection Problem and i t s Approaches

(Halpin and

Woodhead 1976), and

l a t e r refined

21

(Riggs

1980,

Halpin and Bernold 1986), extended and integrated with other systems such

as

INSIGHT

(Paulson

et

a l . 1981

and

1987).

CYCLONE i s used for modelling at the construction

operation

l e v e l and

behavior

of

a

i s e s p e c i a l l y useful for predicting the

construction

cycle

design,

data of durations and

methods

analysis

to

can

be

used

given

the

(Halpin

show the

required

1976).

impact

input

Sensitivity

of

each

major

v a r i a b l e on productivity. Simulation

applications

i n tunnelling are mostly used to

predict the tunnelling advance rate, and cost breakdowns for major equipment items.

Both deterministic

approaches have been used Performance

measures

for

(Miller those

1987

and

and

methods

stochastic

Touran

are

in

1987).

terms

of

tunnel advance rate (feet/day) and t o t a l costs i n general. Knowledge Based Expert Systems (KBES) and be

combined to form a computer aided

for modelling processes.

O Keefe x

simulation

decision making t o o l

(1986) has and

its

explored

the

Expert

Systems-simulation

areas.

Suitable applications include i n t e l l i g e n t front ends

for simulation

taxonomy

can

application

packages which provide advice on how

best to

formulate and interpret the r e s u l t s from a simulation model. For example, Bernold be used i n conjunction process

scenarios.

(1987) showed how

h e u r i s t i c rules

with CYCLONE to evaluate

may

construction

Chapter 2. Method S e l e c t i o n Problem and

2.3.3

Decision

Decision

and

illumination decision

These

term

three

used

professional of

decision

of

follows

repeated

logical

1983).

three and

The

sequential

informational.

until

i s less than the

body

the

(Howard

stochastic,

are

information

for

a

the

cost

value

of

of

obtaining

Decision analysis has been applied to methods s e l e c t i o n

(Gaarslev

1977,

b a s i c a l l y an

Ashley

et

analysis, as

al.

manager

relying

to

select

primarily

methodology

on

the

his

is built

on

1979

and

1983).

It

opposed to design t o o l .

(1983) proposed c r i s i s decision the

describe

practice

procedure

phases

to

problems

deterministic,

additional it.

is a

modelling

phases:

22

Analysis

analysis

knowledge

i t s Approaches

analysis

most

and

bases:

Ashley

a t o o l to aid

appropriate

experience two

as

is

alternative

intuition.

a decision

tree

This and

p o l i t i c a l c o n f l i c t resolution.

This model i s supposed to be

utilized

at

level.

during The

criterion,

construction

tool, has

using

been

project

applied

to

the

strategic

profit

as

select

construction method for a hypothetical

an

the

planning decision

alternative

sewer tunnel during a

c r i s i s dealing with encountered surface settlement. Ayyub framework cost,

and

Haider

which

benefits,

(1985)

considers and

strategy a l t e r n a t i v e .

proposed

a

information

consequences

of

decision on

analysis

relative

each

risk,

construction

The decision c r i t e r i o n , safety of the

Chapter 2. Method S e l e c t i o n Problem and

construction

operations

as

a

23

i t s Approaches

function

of

a

construction

strategy, i s measured i n terms of the completed structure's consequent p r o b a b i l i t y of affect

safety

states

of

theory best

are

identified

these

factors

with

cost

strength

including

of t h i s

initial

single

can

objective

factors

one

cost

using

function

(cost)

The

fuzzy

set

failures.

The

with the minimum of

the

structure

expected cost of structure approach l i e s

that

qualitative.

quantified

terms. However, i t s weakness l i e s

only.

main

being

a l t e r n a t i v e i s the

under construction and The

are

as

The

to estimate the r i s k of construction

construction

cost,

damage.

failure.

in treating linguistic i n i t s dependance on

based

on

safety

a

factors

Besides a decision analysis approach, decision making employ

other

relatively

new

techniques

including

computer aided t o o l s , such as decision support system (DSS). Since there i s some overlap between DSS

and ES, t h i s subject

w i l l be investigated next.

2.3.4

DSS

D e c i s i o n Support Systems

can

provides

be

defined

a computer-based decision aid

that

convenient access to decision models dealing with

production, (Blanning

distribution, 1984) .

management research

as

(DSS)

(OR)

Both DSS

informations techniques.

financial and

ES

systems

analysis, and incorporate (MIS)

and

so forth

features

of

operations

Chapter 2. Method S e l e c t i o n Problem and

Turban

and

approach

Watkins

between

connections

DSS

between

t h e i r integration.

(1985) and

the

contrasted

ES, two

and

decisions

and

and

model rather than a causal model. along

and

can

in

possible

benefits

with

a

of

judgmental

Furthermore, an ES o f f e r s

supporting

justification

or

using

the

user's

loosely think of a DSS

an

A DSS helps a user

choose among alternatives based on the

model, mainly One

the

the

contains

explanation using transferred expertise. evaluate

differences

examined

systems

24

A major difference i s that an ES makes,

rather than supports,

conclusions

i t s Approaches

judgement

and

system

discretion.

as a quantitative

(causal)

modeling approach to a problem, whereas an ES i s regarded as a

qualitative

(judgmental)

and

quantitative

modeling

approach. DSS

for

industry,

Methods

Selection:

Mansero

and

Within

Chapman

(1987)

the

construction

argued

that

DSS

provides the best means for methods s e l e c t i o n for reinforced concrete structures rather than an ES, the output from which would be model

alternative

concrete selecting planning Kim

too p r e s c r i p t i v e .

framed

ways

of

providing

buildings,

suitable

with

formwork

in-situ

particular

methods

from

a

system to reinforced

emphasis

on

construction

perspective.

(1984) proposed and

adaptable

They proposed a DSS

tunnelling

implemented a DSS

method

(which

adapts

to

select

an

construction

Chapter 2. Method Selection Problem and i t s Approaches

methods the

to encountered geological conditions)

design/construction

rock.

methods

selection

to

for

and

equipment.

wall

support,

including

optimize

tunnels

In h i s work, he deals with construction

excavation

analytical

in

methods for

selection

He proposes a framework for generating

support information

25

of

decision

i n adaptable tunnelling and then derives

methodologies

for

the

proposed

stochastic dynamic programming (DP)

DSS,

algorithms.

employing A relative

confidence l e v e l i s used as a measure for ranking competing alternatives. and

Decision variables considered are

support

conditions,

methods. and

advance rate, costs). terms

objective stochastic

conditions. with The

variables

cost/time factors equipment and

The of

State

costs,

selected

variables

objective

of the

and

describing

in

geological deterministic

from previous tunnelling optimization

overhead

i s expressed

Cost/time factors are treated as

constants derived

geological

(productivity i n terms of

material

function

are

excavation

projects.

framework, i n adaptable

tunnelling, i s to i d e n t i f y the most c o s t - e f f e c t i v e chain

of

excavation and support methods, each of which i s t e c h n i c a l l y feasible.

At

the

feasible

for the

subsequent tunnel

same time,

anticipated

each

must

geological

be

economically

conditions

segments ( i . e . the tunnel

i s divided

equal segments to r e f l e c t the changing geological along the

tunnel).

in i t s into

conditions

Chapter 2. Method Selection Problem and i t s Approaches

The

objective

subject

to

feasibility, method

function

two

(total

constraints.

The

method

criteria.

cost), i s

first,

technical

r e f e r s to the a b i l i t y to employ an

suited to the encountered

support

construction

The

that

satisfies

second

ground

constraint,

excavation

conditions

s t r u c t u r a l and economic

26

and a

functional

optimization,

r e f e r s to employing the most c o s t - e f f e c t i v e combination of different

methods

along

the

tunnel,

given

geological

variations. The features

of Kim's work that are of d i r e c t

relevance

to the t h e s i s work described herein are: 1.

Planning decisions are divided into construction and a construction phase.

a

pre-

2.

Output decision variables, for the preconstruction and construction phases, include f e a s i b l e types of combined construction methods, l e v e l of confidence i n s e l e c t i n g a method (the major criterion f o r ranking alternatives), expected loss f o r each method as an upper l i m i t for the additional geotechnical exploration expense, and t o t a l cost and time f o r each alternative.

3.

Input state variables encompass geological conditions and cost/time data f o r each method. Costs are defined as cost of equipment, material, labor, and mobilization/demobilization charges f o r methods changes. Time data include the productivity of each method versus a ground c l a s s , method change duration, and lead time as predetermined data.

4.

Variations of a construction method are treated as discrete methods. For instance, the model has been applied to s e l e c t combined methods among f i v e method variations of drill and blast excavation/support a l t e r n a t i v e s , as opposed to

Chapter 2. Method S e l e c t i o n Problem and i t s Approaches

27

competing with other major methods such as tunnel boring machine, or p a r t i a l face boring. 5.

The model does not consider important project a t t r i b u t e s such as physical constraints, material handling, ground subsidence, and ground water control.

Law

(1987) proposed

a conceptual

design

of construction

activities

DSS

f o r the d e t a i l e d

associated

with projects

characterized by s i g n i f i c a n t r e p e t i t i o n . High-rise b u i l d i n g was

selected

Structure smaller

as the problem

(WBS) work

was

used

operations

domain.

A

to decompose

i n order

construction technologies.

Work

Breakdown

activities

to s e l e c t

into

appropriate

The a t t r i b u t e s of construction

a l t e r n a t i v e s consist of productivity, equipment and material unit costs, and crew makeup. see

figure

2.5,

attributes methods

such in

flexibility confidence

does as

terms of

not

incorporate

the technical of

the

The suggested data structure,

project

equipment

some

important

feasibility physical

used,

of

those

constraints,

and

the

level

of

associated with each method.

Law indicated, i n the problem recognition section, that i n the design have

of the construction a c t i v i t i e s ,

to draw

projects

on t h e i r

to apply

(CM) selections.

their

experience

approach

quantitative

similar

Such knowledge resides

for aspects

methods of

engineers

knowledge of construction

personnel and i s r a r e l y documented. DSS

with

field

with

previous method

a few key

I t i s obvious that the

selection

the problem.

addresses

Law

stressed

the the

Chapter 2. Method Selection Problem and i t s Approaches

28

importance of modelling the q u a l i t a t i v e or judgmental part of construction methods assessment and s e l e c t i o n .

EFCO Floating Slab Formwork System Work Tasks: Literature File No.: SlbFm-012 Slab Formwork Dimension: 20 M by 10 M Bay ** INSTALL AND DISMANTLE FORM * "

No of Units



01 02 03 04 05 06 07 08 09 10

Set roller support bracket ahead Lower support brackets & form Strip slab edge hand rail (1 end) Attach tugger winch Position rolling scaffold Roll slab form out to pick-points Crane hookup 4 points Slab swing-out and reset Clean and oil form Install filler panel

Figure 2.5

'84sf

Unit Time

Total Time

40 15 30 30 15 30 5

320 120

20 6/100sf 90

Comments

30 30 15 30 20 80 46 180

4 Men and Crane 2 Men

Suggested Data Structure f o r Selected Technology [from Law 1987]

2.4 Knowledge-Based Expert Systems In

this

section,

we

identify

the main

components

of a

Knowledge Based Expert System and i d e n t i f y relevant work i n the construction domain, e s p e c i a l l y methods s e l e c t i o n , that relates to each component. Modelling explored

of —

uncertainty existing

i n knowledge

uncertainty

has

methods

not

been

include

p r o b a b i l i s t i c methods (Bonissone 1985, Duda et a l . 1979),

Chapter 2. Method Selection Problem and i t s Approaches

29

confirmation theory (Buchanan and S h o r t l i f f 1985), fuzzy set theory

(Zadeh

1976).

While relevant to construction management problems,

it

i s outside

uncertainty

in

1975)

the

and

scope

terms

Dempster-Shafer

of

of

this

theory

thesis.

outcomes

of

(Shafer

However, site

the

conditions

(especially of geological conditions) was treated. Gashing

(1985) defines KBES as

an

interactive

computer

program incorporating judgement, experience, rules of thumb, intuition,

and

other

expertise, to

provide

knowledgeable

advice about a v a r i e t y of tasks.

2.4.1

KBES Components

A t y p i c a l KBES has four major components; a Knowledge Base, consisting of Knowledge Representation

and A c q u i s i t i o n ;

an

Inference Engine; a Context; and an Explanation F a c i l i t y . 1.

Knowledge Base

(KB)

The knowledge base (KB)

contains a symbolic representation

of expert rules of judgement and experience i n a form that enables the inference engine upon i t . the

to perform

logical

deductions

Such facts and rules are s p e c i f i c to the domain of

problem.

attributed

to

Difficulties knowledge

in

developing

representation

a and

KBES

are

knowledge

acquisition. Knowledge Representation :

Knowledge Representation

(KR) i s

the set of syntactic and semantic conventions used to encode

Chapter 2. Method Selection Problem and i t s Approaches

30

the facts and relationships that constitute knowledge

in a

knowledge

based

developing knowledge

system

KBES

has

(Winston

shown

representation

that

is

1986). a

often

Experience i n

robust, the

key

yet

precise

to

avoiding

s u p e r f i c i a l i t y or shallowness i n the solution of r e a l i s t i c problems

(Jackson 1986).

Selection of a KR technique i s a

fundamental step i n the application of ES to a problem.

The

KR process i s concerned with the problem of encoding the knowledge

so

computer.

that In

i t can

general

be

the

easily following

manipulated by elements

must

the be

represented:

domain terms which deal with the language or

jargon

by

used

the

expert

in

the

field;

structural

relationships which treat the interconnections of compound entities;

and

causal relationships which deal with cause-

e f f e c t r e l a t i o n s between components. KR techniques may Shortliffe

1976),

include production rules predicate

logic

(Clocksin

(Buchanan and and

Mellish

1981), semantic nets (Minsky 1968, Duda et a l . 1978), frames (Minsky 1975), and object oriented programming

(Bobrow and

S t e f i k 1983, Goldberg 1981). Knowledge A c q u i s i t i o n

:

transfer,

transformation,

and

the

Knowledge A c q u i s i t i o n of

(KA) i s the

problem-solving

techniques from some knowledge source to a program and S h o r t l i f f e 1985).

(Buchanan

The major bottleneck i n building an

Chapter 2. Method Selection Problem and i t s Approaches

ES

i s the s c a r c i t y

of knowledge

engineering

31

s k i l l s to

i n t e r a c t with one or more human experts. Several methods are used i n the KA process. the use of unstructured prototype and

interviews,

They include

structured

system development, r u l e induction,

finally

1987).

machine

learning of rules

These methods embody theories

interviews, observation,

(Gruber

and Cohen

and knowledge from

computer science, psychology, l i n g u i s t i c s , and sociology, i n addition to technological expertise. 2.

Inference Engine

The

inference

contains

the

characterized control

engine

(IE) i s the part

general by

problem-solving

strategies

the reasoning

which

process

of a

KBES

knowledge draw

(Mikroudis

that

and i s

inferences 1986).

guides the development of a solution using

and

The IE

the f a c t s and

rules stored i n i t s KB and the information i t acquires from the user.

Thus, the IE i s used to derive new facts from

known facts and to regulate occurs. are

not

i n which

reasoning

IE strategies used to make inferences include, but limited

inheritance. forward

the order

to,

modus

ponens,

resolution,

and

Control strategies include backward chaining,

chaining,

agenda

control,

(Charniak and McDermott 1985).

mixed,

and

others

Chapter 2. Method Selection Problem and i t s Approaches

3. The

32

Context context

i s a temporary data storage

i n which known and

deduced facts are stored during a consultation session. The context builds up dynamically during the s o l u t i o n process of a p a r t i c u l a r problem.

I t i s used by the inference engine to

determine the next step i n the process.

Data may come from,

or go t o , an extended data base, analysis/design programs, or

even

provides

data

acquisition

devices.

The

inference

a further mechanism f o r representing

relationships and f o r assigning values

tree

hierarchical

to object s l o t s by

instantiating. 4.

Explanation F a c i l i t y and Others

The explanation f a c i l i t y

(EF) component serves to p a r t i a l l y

trace

process

the ES reasoning

conclusions The

i n order

to j u s t i f y the

made during a consultation.

two widely

used commands are HOW and WHY.

For a

network of goals, rules, and hypotheses, HOW asks what rules were involved

i n solving the problem.

reasons some information

i s requested

WHY

states

f o r the

by the system. In a

goal driven ES, the HOW rule propagation goals to the i n i t i a l

asks

d i r e c t i o n i s from

(backward chaining).

d i r e c t i o n i s v i c e versa (forward chaining).

The WHY

33

Chapter 2. Method Selection Problem and i t s Approaches

2.4.2

Expert Systems for Construction Management

Construction complex

engineering

decision-making

allocation,

planning

and

management

problems,

involves

such

and scheduling,

as

safety,

many

resource

analysis of

construction

processes,

improvement.

The solution of these are highly dependant on

engineering

and

and productivity measurement and

trade

judgement,

rules

of

thumb,

and

subjective evaluations. As

stated

in

the

management decision-making

previous

section,

construction

tools have t r a d i t i o n a l l y employed

quantifiable models (networks, OR techniques, e t c . ) . strength l i e s

Their

i n t h e i r rigorous analysis of the a v a i l a b l e

data culminating i n an optimal, or near optimal, solution to the problem. the

Their main weakness i s the t o t a l dependance on

quantitative data

necessary

to represent

the various

relationships that describe the problem, many of which are imperfectly understood When

construction

qualitative judgement,

(Warszwaski 1986). management

information and

or

intuition),

decisions

involve

relationships or

when

more

(experience,

multiple

decision

c r i t e r i a are present, the t r a d i t i o n a l approach i s of limited use, being more at the t a c t i c a l than s t r a t e g i c l e v e l . limitations

can

be

overcome

incorporating the experience,

to

a

certain

heuristics,

acknowledged experts into an ES.

degree

Such by

and judgement of

Chapter 2. Method Selection Problem and i t s Approaches

Since

1984

several

papers,

articles,

and

34

conference

proceedings have been published

that provide an overview of

ES applications i n Construction

Management (Wager 1985,

86

1990),

1987,

Levitt

1987,

Mohan

applications i n t h i s f i e l d Reviews recently

of

current

(Wager 1986,

ES

(Chin 1987, applications

L e v i t t 1987).

and

suggest

CIB-

further

Mohan 1990). have been

Applications

reported identified

include: 1.

construction project Switzerland);

organization

design

2.

time estimating systems ( C i v i l & C i v i c construction firm);

3.

r e p e t i t i v e construction of Texas);

4.

decision making and risk I n s t i t u t e of Technology);

5.

intelligent construction risk systems (University of Texas);

6.

layout of temporary (Stanford U n i v e r s i t y ) ;

7.

evaluation of project personnel based on progress data from project time/cost monitoring systems (MIT);

8.

vertical construction (University of I l l i n o i s , and

9.

project planning and control (Stanford L e v i t t et a l 1988);

10.

construction project monitoring (CMU);

11.

maintenance advisor (PTY Ltd, Australian elevator construction and maintenance contractor);

r i s k analysis analysis

(Zurich,

Australian (University (Georgia

identification

construction

facilities

planning/scheduling CMU); University,

Chapter 2. Method Selection Problem and i t s Approaches

12.

equipment and plant s e l e c t i o n Technology, Loughborough, U.K.).

35

(University

of

Other applications have been suggested by Trimble (1987), Warszwaski (1985), Chin (1987), Mohan (1990) as follows: 1.

design synthesis and i n t e r p r e t a t i o n of b u i l d i n g code regulations;

2.

estimating procedures and cost c o n t r o l ;

3.

the analysis scheduling;

4.

s e l e c t i o n of appropriate plant and equipment;

5.

site

6.

construction financing;

7.

design and construction planning buildings;

8.

q u a l i t y control;

9.

safety practices;

10.

contractual claims analysis; and

11.

evaluation of a l t e r n a t i v e construction methods at early design stages.

and

evaluation

of

construction

planning;

of prefabricated

2.4.3 KBES for Construction Methods Selection Research

work on the use of expert been

tailored

to

systems

specific

f o r methods

selection

has

applications.

Generally,

such systems can be c l a s s i f i e d as r u l e based, or

frame and/or object oriented. Within the f i r s t category, some applications are directed at s e l e c t i n g a key resource, s p e c i f i c job. of

craneage

particularly

equipment f o r a

Gray and L i t t l e (1985) examined the influence resources

required

to l i f t

large

units

in a

Chapter 2. Method Selection Problem and i t s Approaches

high-rise

building

resources

on

and

the

the

effect

activity

Subsequently,

an

expert

analysis

was

developed

desirable

crane on

the

site

multiple

duration

system to

of

for

select

crane

calculations.

craneage

and

36

resource

locate

for a high-rise

the

and

most

low-rise

j building

construction.

imported

into

a

Later,

more

this

comprehensive

expert system directed

at

expert

system

planning

was

rule-based

determining a l l work

activities

i m p l i c i t i n the design of a high-rise b u i l d i n g (Gray 1986). Activities

were

defined

(resource l a b e l l e d : material,

according trade,

significant

function

(direction

horizontal),

and

operationally

to

type

plant),

of

of

work

operationally

movement: v e r t i c a l significant

or

location

(grouping a c t i v i t i e s of d i f f e r e n t sequence and s i z e ) . Components of an a c t i v i t y ' s duration resource

level.

The

work volume

Resource l e v e l i s variable.

are work volume

i s set

by

Resources are

the

set

and

design.

in

fairly

coarse groups, either gang or piece of plant, at the minimum l e v e l consistent with normal practice. were

used

to

(precedence, and processes.

select

an

time l i n k s ) ,

Rules and h e u r i s t i c s

activity, and

link

activities

perform problem

This application shows how

solution

expert systems can

be

used to encode algorithms (network a n a l y s i s ) , and h e u r i s t i c s for a c t i v i t y selections.

Chapter 2. Method Selection Problem and i t s Approaches

In

the

same

category,

other

applications

37

include

s e l e c t i n g a crane type (tower crane versus mobile crane) and size

for

high-rise

Wijestundera

building

1987),

and

construction

scraper

equipment

earthmoving, given the

s p e c i f i c project

1988).

authors

The

former

construction

equipment

be

formulated.

Moreover,

applications

should

information

pertaining

evaluation.

for

conditions

concluded

i s l a r g e l y based

i n t u i t i v e knowledge, allowing

(Harris

that on

road (Harris

selecting

uncertain

and

only broad rules of thumb to they

include to

suggested

output

data

plant

and

that and

further

production

labor

resource

In both examples, the knowledge base i s l a r g e l y

h e u r i s t i c i n nature, and knowledge a c q u i s i t i o n was for deriving the those

and

inference

applications,

essential

for problem solution.

methods were represented

as

Also,

in

equipment

for c a p i t a l intensive projects. The

second

application

category,

frame

based

systems, includes work done by Logcher and Nay

expert

(1985), Kunz

et a l . (1986), and Hendrickson et a l . (1988). Chief

among

those

applications

(Hendrickson et a l . 1988). planning,

is

including construction.

used

to

excavation,

plan

is

Construction

Planex, a KBES for modular

high-rise

foundation,

and

Planex

construction buildings, structure

Planex s t a r t s with a design a l t e r n a t i v e

as

input which consists of several design elements (a footing,

Chapter 2. Method Selection Problem and i t s Approaches

column,

and

building).

beam The

element a c t i v i t y etc).

Figure

for a design

frames

modular

element

reinforced-concrete

generates

( i . e excavation,

2.6 shows a sample

The element

activity

construction

pouring

element

concrete,

activity

created t o describe the excavation a c t i v i t y footing.

38

required f o r a

i s identified

number using the extended MASTERFORMAT code.

frame

by a code

The f i r s t s i x

s l o t s define i t s designation and relevant parents of design elements slots

and project

activities.

These

are followed by

f o r amount of work, unit-of-measure, crew, material-

package, duration,

and successor

element a c t i v i t i e s .

The

crew a t t r i b u t e has been evaluated to excavation-foundation05. Element-Activity p01 -•OO-bOO-1 O0-ca-02-220-10-01

SLOT

VALUE

is-a

ea

ea-name ea-code ea-of-DE parent-EA

axcavatlon-column-fbotIng-01 01-220-1^01 p01 -sOO-MO-fOO-de-60-01 -01

ea-of-PA amount-of-work unit-of-measure crew material-package duration successors

p01-sCO-b00-f00-ea-01-220-10 p01-200-b00-(00-pa-10-60 24.0 cu-yd excavation-fbundation-05 none 16 hours p01-s00-b00-f00-ea-02-220-10-02

Figure 2.6 Sample Element A c t i v i t y Frame [from Hendrickson et a l 1988] After grouped

element into

activities project

have

been

activities

created,

based

on

they are selected

Chapter 2. Method Selection Problem and i t s Approaches

technologies.

39

The a c t i v i t i e s are then sequenced and t h e i r

duration estimated i n order to develop the schedule. The

knowledge base i s organized into a set of knowledge

sources that represents rules, h e u r i s t i c s , functions.

and c a l c u l a t i o n

Decisions and computations undertaken during the

planning process can be stored i n any of the frames i n the Planex hierarchy of frames and can be inherited upward and downward activity

between element

operator

design frames.

modules,

applied

project

activity,

and

When frames are created by the

relevant

knowledge base w i l l operations

element,

knowledge

be evaluated. to create

a

sources

Within

i n the

a sequence of

construction

plan, the

selected technology operator uses h e u r i s t i c s related to s o i l and s i t e information, resource productivity information and other factors, by a c t i v a t i n g relevant knowledge sources (KS) designated

as KS-technology-xx-xx,

to group

element and

project a c t i v i t i e s under an a u x i l i a r y group object that i s used

to store the common technology

shows

an

example

of

a

choice.

knowledge

source,

Figure 2.7 namely

KS-

Technology, f o r selecting excavating equipment pictured as a decision table, whereas i t i s a c t u a l l y encoded as frames and production conditions, condition source

rules.

This

three

rules,

of

KS contains and three

"KS-water-level"

two project actions.

i s an

that has to be evaluated

embedded

first.

context

The

second

knowledge

The t h i r d

rule

Chapter 2. Method Selection Problem and i t s Approaches

40

indicates that i f none of the previous two rules were f i r e d , the appropriate

technology i s "special machine".

KS-Technology-Example Object

Slot Op

soil-characterlstlcs

Value

RULES

soiltype

Is

hard

true falsenotfireo

KS-waterlevel

Is

wet

false ture notfirea

T

•

T

then then then selected selected selected

power-shovel clamshell special-machine

Figure 2.7 Example of Knowledge Source [from Hendrickson et a l . 1988] According

to

Planex, element are

the

inference

and project

supposedly selected

selected

from

figure

2.7.

tables

and

the Task

and

activities assembled

KS-technology durations

c a l c u l a t i n g rules

strategy

implemented

in

f o r the excavation based

on the

instantiating

shown

are estimated from

in

decision

i n a manner s i m i l a r to

used i n MASON system (Hendrickson et a l . 1987).

plant

that

Precedences

among element a c t i v i t i e s are also determined and recorded i n s l o t s of the element a c t i v i t y frames. be of two types: physical or

These precedences can

resource-related.

Chapter 2. Method Selection Problem and i t s Approaches

For

the

Planex

system,

the

authors

41

indicated

that

determining the equipment to be used, the number of crews or pieces

of

equipment,

inter-task

precedence

and

task

duration, involve diagnosis and p r e d i c t i o n as contrasted synthesis, i n a c t i v i t y d e f i n i t i o n . the way

to

This example shows that

i n which a method, mimicked as equipment selected

based on

project

context,

i s used to combine element

and

project a c t i v i t i e s among numerous a l t e r n a t i v e s , i s e s s e n t i a l i n constructing the planning Logcher system

and for

Hierarchical work

(1985)

analyzing

described

risks,

conceptual

expert

project

risks.

represent

resources,

information.

a

construction

frames were used to

packages,

relevant

Nay

schedule.

Figures

and

2.8,

2.9,

project

tasks,

additional

site-

and

2.10

show a

sample of the labor, equipment, and process frames that are to

be

created

during

a

session.

Their

values

will

be

inherited by Work Package and Review Data frames f o r further manipulation

of

the

presented show how represented

project

risk

construction

analysis.

The

frames

related concepts could

be

by describing and d e t a i l i n g t h e i r a t t r i b u t e s .

In t h i s application, frames were found to be a knowledge representation

strategy

capable

of

capturing

relevant

problem c h a r a c t e r i s t i c s , while rule-directed inference used to associate project r i s k s with work packages.

was

Chapter 2. Method Selection Problem and i t s Approaches

42

Labor type:

union: cost: quality: ( s k i l l and manpower required by work package) Productivity: quantity: productivity l e v e l : (output/unit time) schedule: (regular and overtime hours/week) production rate: (output/unit input) morale: Safety: accidents: shutdowns:

Figure 2.8 Labor Component Frame [from Logcher and Nay 1985] The foregoing diversity

expert systems sample applications show the

of the methods, s e l e c t i o n l i t e r a t u r e .

thesis viewpoint, each example addresses s p e c i f i c of methods s e l e c t i o n .

From the instances

There i s a lack of a w h o l i s t i c scheme

for specifying and analyzing

a method.

The l i t e r a t u r e has

showed how some expert systems have been incorporated and/or evolved into a larger i n t e l l i g e n t system f o r planning, layout,

etc.

In

envision a generic

the

same vein,

i t i s conceivable

site to

t o o l that consists of a series of ESs,

small and big, that are t i e d together to s e l e c t and specify the most suitable method.

Chapter 2. Method Selection Problem and i t s Approaches

Equipment

type:

43

general s p e c i f i c a t i o n : description: date information supplied: supplied by: equip name: rated capacity: alternative equip type: operating hrs. u n t i l mainten.: source: supplier: 33,000 lb. ft. SSPDesignation Is PZ27

Context, User Context, User Context, User

THEN

(Hammer Size + SSP Size + Soil Profile) Is Technically Feasible HammerandSSP Are Compatible

User, Context

AND

Source

{Control Clauses if Accepted} AND AND

Check DetailedFeasibility for ResourceCompatibility Check Performance_Measures_Feasibility {Control Clauses if NotAcceptable}

AND AND

Eliminate SSP Go to Next GWSS Alternative

Resources,

including

compatible. constructed

For with

design

instance, a

substituted

to

suit

some

subset

equipment, or v i c e versa,

materials,

of

materials

existing

resources,

be

are

only

tools

where some materials

existing

must

and/or

have to be

such

as

crew

i n the above

rule

expertise and available equipment. In

this

vein,

pronounces the s o i l The

the f i r s t context

premise

condition to be cohesive

soil.

second premise dictates a lower bound f o r the hammer

power.

This i s provided

lower l e v e l greater defining

which draws from a

r u l e that r e l a t e s the l i k e l i h o o d of SPT bring

than a threshold a

by the context

feasible

and minimum hammer energy,

hammer

subset.

The t h i r d

thus

premise

Chapter 4. A KBES Framework for Methods Selection

126

i d e n t i f i e s the type of SSP section as PZ sections, which are heavy and have a high section modulus. context or user volunteered implied

This i s based on the

information.

r e l a t i o n s h i p between the

pile

For the former, an driving

conditions,

and/or hammer s i z e , versus SSP minimum s i z e , i s established, reducing the set of f e a s i b l e SSP. The object i s to reduce set of available l i s t s of hammers and SSPs into reduced f e a s i b l e subsets that s a t i s f y expected goals, e.g. cost and time. Control

passes

horizontally

measures f e a s i b i l i t y .

to

check

the

performance

I f the horizontal inference

carries

on, the next performance f e a s i b i l i t y w i l l be checked at t h i s stage.

This

for production cost

data

component establishes upper and

lower bounds

and costs from previous projects, and/or unit

manuals,

given

including s o i l p r o f i l e and

for

a

project

context

scenario

equipment/material spread.

following r u l e mirrors t h i s component function.

The

Chapter 4. A KBES Framework f o r Methods Selection

127

Performance Measures F e a s i b i l i t y

4.

RULE

Performance Measures Are Feasible

IF

Lower_Prod._Bounds < ProductionRate < Upper_Prod._Bounds LowerCostBounds < ProductionCost < Upper_Cost_Bounds

AND

Source User User

AND THEN

Steel Sheet Piles.Performance Measures Are Feasible {Control Clauses if Accepted} Check Detailed_Feasibility for Other Components Check Detailed_Feasibility for PerformanceMeasures

AND AND/OR

{Control Clauses if NotAcceptable} Eliminate SSP Go to Next GWSS Alternative

AND AND

These upper and lower contract as unit user

to

for a

estimate

are

i n the

given

the Production_Rate design

construction resources and processes. estimates

(stated

cost and construction duration) , and the

i s supposed

Production_Cost

bounds are goals

conceptual,

a

rule

alternative, At t h i s

of

thumb

and and

l e v e l , the based

on

experience from a project context, whereas further p r e c i s i o n of estimates could be done at the lower l e v e l . Control then passes to the regulatory component to check compliance

with

regulatory and safety requirements.

example i s presented i n the next rule.

An

Chapter 4. A KBES Framework f o r Methods Selection

Regulations

5.

128

Are S a t i s f i e d

RULE

Regulation Is Satisfied

Source

IF AND AND AND

OSHA and Local Safety Regulations Are Satisfied Environmental Hazards Are Acceptable Pile Driving Level of Noise Is Acceptable Hammer Vibrations Are Acceptable

User User User, Context Context, User

THEN

Regulatory Conditions Are Satisfied {Control Clauses if Accepted} Evaluate Another GWSS Alternative Rank Preliminary Feasible Alternatives Check Detailed_Feasibility Regulations

AND/OR AND/OR AND/OR

{Control Clauses if NotAcceptable} Eliminate SSP Go to Next GWSS Alternative

AND AND

The

first

premise

ensures

that

safety

regulations

relevant to working conditions f o r labor are s a t i s f i e d . This means that proper labor a l l o c a t i o n , labor protection, and so forth, have to be met f o r the major operations

(excavation,

p i l e d r i v i n g , muck removal, etc) of a proposed method, where some methods require more consideration

than others.

The

user has to ensure the v a l i d i t y of t h i s clause since

there

are

numerous

provisions

verifies

acceptance

method.

For example,

possible

i n some areas,

dismissed.

to meet.

The

second

f o r the environmental i f diesel

material

the d i e s e l

premise

hazards

disposal

of a i s not

hammers class w i l l be

This condition must be v e r i f i e d by the user.

Chapter 4. A KBES Framework f o r Methods Selection

The

third

noise

premise

emanating

sets

from

a lower

pile

bound

driving,

129

f o r an accepted e.g. noise

from

construction i s being severely l i m i t e d with an objective of not

more than 85 dB at 50 f t created

hammers and vibros checked driver

(Hunt 1979).

by the derived components

This

context

l i m i t a t i o n could be

where properties

are retrieved

confirmed by the system.

by compressors f o r

from

i t s data

of p i l e base and

The user may exercise h i s judgment

as to whether further r e s t r i c t i o n are warranted.

The fourth

premise ensures that vibrations emanating from p i l e d r i v i n g do

not d e s t a b i l i z e surrounding

cases,

surrounding

precautions user

structures

are warranted.

or derived

structures,

context

are

where, i n some

braced

or

other

Similar to the previous premise, could

be used

to approve

this

condition. Once t h i s rule category i s s a t i s f i e d , and by implication, others as well, then a GWSS method a l t e r n a t i v e i s considered to be p r e l i m i n a r i l y f e a s i b l e . pile, is

i s regarded as p r e l i m i n a r i l y f e a s i b l e .

successfully selected,

list

In t h i s example, s t e e l sheet

of the available

preliminary

feasibility

A f t e r an SSP

another GWSS a l t e r n a t i v e i n the alternatives

by repeating

will

be tested f o r

the same cycle shown

above. The

foregoing

control

structure

may

allow

solution

synthesis t o continue even i f a high l e v e l condition

failed

Chapter 4. A KBES Framework for Methods Selection

(say regulatory The

system may

condition) , or was

130

v i o l a t e d , within

limits.

allow the user to further pursue and

explore

p a r t i a l solutions for the components that remain f e a s i b l e . Therefore, alternatives

a

tentative

is

list

available

of

for

preliminarily

further

feasible

synthesis

at

the

detailed level.

4.3.4 The

Detailed F e a s i b i l i t y

Level

d e t a i l e d f e a s i b i l i t y l e v e l contains detailed design

and

analysis knowledge which i s organized into several knowledge bases as shown i n figure 4.6. are to confirm the

The

objectives

of t h i s l e v e l

f e a s i b i l i t y of alternatives produced at

the preliminary

f e a s i b i l i t y stage, and to complete the frame

description

each method which survives

of

checks.

The

with the

design component knowledge base

At

first,

through

control strategy

technical

For instance, data

base,

message

is

guides t h i s process beginning

feasibility

s i m p l i f i e d analysis

and

for an

(KB

3,4,

and

5).

a t t r i b u t e i s sought

design procedures

(KB-3).

i f the available sheet p i l e s , represented by a do

not

sent

contain

to

the

the

user

the next preliminary the

Control

section, of

the

then

a

technical

i s then passed to

GWSS a l t e r n a t i v e .

steel

design element was

required advising

i n f e a s i b i l i t y of the a t t r i b u t e .

Assuming

a l l feasibility

sheet

pile

(propped

sheet

piles)

f e a s i b l y sized, control i s passed to

the

Chapter 4. A KBES Framework for Methods Selection

pile

driver

Based

selection

on the s o i l

and s i z i n g

knowledge

base

p r o f i l e and conditions,

design

element

of the s t e e l

system

attempts to pick

sheet

pile

131

(KB-4).

and s p e c i f i e d a t t r i b u t e s , the

the most suitable

and productive

hammer type and s i z e from a hammer data base.

The hammer

s e l e c t i o n must s a t i s f y the technical f e a s i b i l i t y

conditions.

On the other hand, i f the hammer type and s i z e i s selected f i r s t , because of a v a i l a b i l i t y , t h i s may d i c t a t e the s i z e of the

sheet p i l e .

be

reversed.

Thus, s t e e l sheet and hammer s i z i n g could For CMSA, the former

knowledge

processing

approach design element s i z i n g to hammer s i z i n g i s adopted. Included i n the technical f e a s i b i l i t y t e s t for the hammer type and s i z i n g , i s the p i l e d r i v i n g strategy drive p i l e s i n singles or i n p a i r s . include other wave patterns). i s dependant on the s o i l and

so f o r t h .

strategy predict

After

the

conditions,

alternative

(KB-6).

i s attained,

assessment routine

or

not

pile,

strategy energy,

hammer, and

module i s applied to

(time and cost)

f o r the

After the cost estimate f o r control

i s passed t o the r i s k

(KB-2), which uses the r i s k assessment

framework described i n section 4.4. whether

p i l e driving

sheet

a prediction

the performance measures

candidate a l t e r n a t i v e

(Other strategies may

The choice of either

a steel

are s p e c i f i e d ,

(KB-5), i . e . ,

the

risks

I t i s used t o determine

associated

with

a l t e r n a t i v e , when priced out, exceeds some maximum

the GWSS threshold

Chapter 4. A KBES Framework f o r Methods Selection

value.

I f they

infeasible

and

do,

the a l t e r n a t i v e

the control

strategy

132

i s deemed moves

to

to be

the

next

candidate. I f the alternative's r i s k costs

are acceptable, then i t

passes control to the diagnosis or analysis component (KB7) .

I f the method synthesis

satisfied, construction

then

KB-7

strategy,

(technical f e a s i b i l i t y ) i s not

recommends

Currently,

automated, with the alternatives singles or i n p a i r s .

declared

change

in

either

hammer type or s i z e , or s t e e l sheet

p i l e type, s i z e , and grade.

infeasible

a

being

only the f i r s t i s to drive

piles in

I f the recommendations s t i l l y i e l d an

SSP GWSS alternative,

to be i n f e a s i b l e .

then

the a l t e r n a t i v e i s

I f i t i s technically feasible,

then production rate and unit cost are determined using the prediction module of KB-6. At

the end, an evaluation

criterion

will

be chosen to

rate the f e a s i b l e SSP a l t e r n a t i v e i n order to rank i t with other

successful

candidates.

Section

4 . 5 elaborates

on

several c r i t e r i a schemes. Other issues relevant when a solution remedy

to

fails

explore,

to t h i s section concern what to do within

what

remembered, where to track stops and declares system.

the system

attributes

—

values

what kind have

to

of be

back to, and when the system

an a l t e r n a t i v e i s not f e a s i b l e within the

Chapter 4. A KBES Framework f o r Methods Selection

133

In the discussion that follows, a step-by-step i s presented,

approach

along with i l l u s t r a t i v e rules to demonstrate

the approach just outlined.

First,

we examine the design

element synthesis. 1.

Design Element Synthesis (KB-3)

Attention i s directed at s i z i n g the design element (in t h i s case

the s t e e l

sheet

piling)

feasibility

requirement.

feasibility

means

support

system,

deflection

that

piles

criteria

For given

used

this

sized

technical

attributes

so that

(not implemented)

to determine

the technical

example,

default

can be

three part process i s adopted. is

to s a t i s f y

for a

stress

and

are s a t i s f i e d .

A

F i r s t , a rule-based approach

pressures

and moments.

Second, a

search i s made f o r a sheet p i l e that s a t i s f i e s the allowable stress

criteria.

retaining

strategy

Third,

i f a pile

i s altered

can't

(spacing

be found, the

of struts

and/or

wales i s involved) and control i s passed back to the second step,

with

iterations

occurring u n t i l

either

a

feasible

s t e e l sheet p i l e design i s confirmed or no f e a s i b l e solution exits.

This t h i r d

prototype.

step has not been

implemented

Example rules f o r the f i r s t

i n the

two parts of t h i s

process are now described. 1.1 Rules

Pressure and Moment Calculations (KB-3-1) are

corresponding

used to

to

retrieve

the

soil

the strata

required scenario

properties (context

Chapter 4. A KBES Framework for Methods Selection

134

information) , to assign the default horizontal and v e r t i c a l spacing

for

the

retaining

spacing

(struts

spacing

h o r i z o n t a l l y 12 f t and v e r t i c a l l y 15 f t ) , and to perform the calculations

of pressure

and moments.

Details

of the

calculations are given i n Appendix A. An example of the rule format follows. Rule:

Compute Earth Pressure and Moments for the two Soil Layers Scenario

IF AND AND THEN

Source

Soil_Layers.Number is Two Soil_layer_l is Loose_Sand Soil_Layer_2 is Stiff_Clay Maximum_Lateral_Pressure = (Loose Sand.Unitweight * LooseSand.Depth * (K(a) for LS) + StiffClay Unit Weight * Stiff_Clay_Layer.Depth) * (K(a) for SC) Moments for Steel Sheet Piles = Maximum_lateral_Pressure *

AND

User User User

L (vertical spacing) ~2 / 8 {Control Strategy} AND

The

Specify the Steel Sheet Pile Section

first

profile.

three

premises

the user

The system retrieves relevant

compute the l a t e r a l etc.

query

earth

In the concluding

pressure part,

—

f o r the

soil

s o i l properties to unit

pressure

weight,

K(a) ,

and moments are

computed (See Appendix A) , and control i s passed to specify the sheet p i l e . 1.2

Selecting and S i z i n g Design Elements [KB-3-2]

Moment

information

i s passed

from the f i r s t

step

to the

second, and combined with an allowable stress condition to determine required

the section exceeds

modulus required.

the maximum

size

I f the modulus

available

i n the data

Chapter 4. A KBES Framework f o r Methods Selection

base,

the

current

design

is

technically

135

infeasible.

F e a s i b i l i t y may be achieved by modifying the spacing of the retaining system. The KB-3 design component contains mainly factual applied

rules

f o r sheet p i l e s , s o l d i e r p i l e s , lagging, wales, and

struts.

I t currently

strength.

Other

uses

design

a single

criteria

c r i t e r i o n based on such

as

deflection,

allowable settlement, and so forth, have not been

treated.

A t y p i c a l rule employed i s : RULE

Select a Steel Sheet Pile

IF

SSP.Section_Modulus = Maximum_Moments / (Fb * Fs) 38.3 in~3 < SteelPileSection Modules < = 46.8 i n ^ SSP.designation Is PZ_27 Retrieve PZ27 Properties Database Calculate the Quantity Take-Off for Sheet Piles

AND THEN AND AND

Source Context (step 1) Context

{Control Strategy} Select a Hammer Class

AND

This

rule

i s i n t e r n a l , where derived

context from the

previous rule i s used to specify a SSP of P Z _ 2 7 .

Control i s

passed to hammer selection next. KB-3-3,

although not implemented,

would

permit changing

the retaining system i n an attempt to f i n d a f e a s i b l e s t e e l sheet p i l e s i z e .

Chapter 4. A KBES Framework for Methods Selection

2.

Specifying A Construction

2.1

Selecting a Hammer (KB-4-1)

136

Resource (Resource Level)

The main types of hammers include: 1.

Drop hammers;

2.

Steam hammers (Single and Double Acting),

3.

A i r hammers (Single and Double Acting);

4.

Diesel Hammers (Single, Double, and D i f f e r e n t i a l Acting hammers;

5.

Hydraulic hammers;

6.

Vibratory Hammers; and

7.

others.

Selecting

the

most

suitable

pile

project involves the consideration

hammer

for

a

given

of several factors, such

as s i z e and p i l e s types, number of p i l e s , c h a r a c t e r i s t i c s of the s o i l , type of

location of the project, r i g available, and

owned by A

the

topography of the

types and

site,

sizes of hammer

contractor.

pile

driving

contractor

usually

is

concerned

with

s e l e c t i n g the hammer that w i l l drive the p i l e s for a project at

the

lowest

cost

within

the

required

production

rate.

Broad brush rules found i n the l i t e r a t u r e are s i m i l a r to the one

shown i n table

suitable Such

a

for

different

table

representation;

4.6,

is

recommend

homogeneous

convenient

however,

complicated when two

which

the

soil

for

selection

the

hammer most

classifications. a

One_Soil_Layer process

is

more

or more s o i l layers are present.

The

Chapter 4. A KBES Framework f o r Methods Selection

s e l e c t i o n depends on the ordering-of the s o i l

137

layers, the

depth Of each layer, and the SPT p r o f i l e .

For example,

the

layer

top layer

i s soft s o i l

and the lower

if

i s dense

s o i l , a vibratory, l i g h t impact hammer i s used to drive them to

the dense layer,

and then

another,

heavier hammer i s

u t i l i z e d to drive them t o r e f u s a l .

8AND8 (N0N-C0HE8IVE 80IL8) H- 8heet ConePipe Pipe Wood Open Ctoeet Beam Pile orvte Very L O O M

DA

vgjB, VjNB) VgjB, VgjB)

\ QQM9

DA

V£NB)

DA

Medium

SA

VjNB)

DA

Dent*

SA

Very Dense

SA

V

£A'

SA

SA

SA

V(NB)

v , r

VjNB) VgjB) VgjB) VgjB)

SA

VgjB,

DA DA

SA SA SA

(a)

CLAYS (COHESIVE 8OILS) H- Sheet ConePipe Pipe Wood Open Cloeec Beam Pile crete Very8oft

DA

VgjB)

DA

VgjB)

V

SA

Medium

DA

V^NB)

SA

V(NB) DA

V DA

SA

Stiff

SA

DA

SA

DA

DA

SA

VeryStltf

SA

SA

SA

SA

SA

SA

Hard

SA

SA

SA

SA

SA

SA

W DA • Double Acting (Diem or Mr/Steam) SA - Slngl* Acting (Oleeri or Air/Steam) V-Vlbratory NB • No Bearing Formula Required

Table 4.6

Hammers f o r Different S o i l s [from Barber 1987]

Chapter 4. A KBES Framework for Methods Selection

138

A t y p i c a l rule format for the single s o i l layer case i s : RULE

Select A Single Acting Air Hammer

Source

IF AND THEN AND OR

Soil_Layers.Number = One Soil Layer Type of Soil_Profile = Cohesive Soil Select An Impact Hammer Choose A Single Acting Air Hammer Choose Double Acting Air Hammer

User User, Context Context (Experiential)

3

Control Strategy AND

The

Specify a Hammer Size

f i r s t and second premises query the user for the s o i l

profile.

The l a t t e r inquires about the cohesiveness of the

layer although the system i s already aware of the s o i l properties.

layer

The reason for t h i s i s to allow the user t o use

h i s judgement i n determining t h i s quality since t h i s clause, i f true, excludes the vibratory class from consideration and thus focuses on the impact hammer c l a s s . For

two s o i l

layers,

hammer alternatives —

there

could

be several

feasible

e.g. use a single hammer (impact or

v i b r a t o r y ) , use a combination of hammer types, or a range of sizes of the same hammer

(use a l i g h t e r one t o drive the

top, s o f t layers and the heavier one to drive p i l e s to t h e i r r e f u s a l depth).

Vibratory p i l e driver use i s not recommended f o r a s o i l profile with a sizable cohesive layer (e.g. c l a y ) . Therefore, t h i s premise eliminates the vibratory p i l e d r i v e r subset and examines impact hammers only.

Chapter 4. A KBES Framework f o r Methods Selection

139

A t y p i c a l rule format f o r the two s o i l layers case i s : RULE

Select A Single Acting Air Hammer

Source

IF AND AND THEN AND

Top_Soil_Layer is LooseSand Lower Soil Layer Is Stiff Clay Stiff_Clay_Layer.Depth > 15 ft Choose A Single Acting Air Hammer Size the Hammer

Context Context Context

{Control Strategy} AND

Specify a Hammer Size

The above rule checks the cohesive s o i l to

a threshold which

clay

depth

excludes vibratory

to the sand

layer

ratio

the impact

hammer,

single

hammers.

I f the

i s very high then a

vibratory p i l e d r i v e r could be favorable. rule,

layer thickness

acting

Thus,

for this

a i r hammer, i s

selected with reference to table 4 . 5 .

2.2 The

S i z i n g The Hammer [KB-4-2] s i z i n g process s t a r t s by selecting the highest hammer

t h e o r e t i c a l energy.

This

i s consistent with the "greedy"

algorithm, described previously, i n which maximum production rates

and reserve capacity

are sought.

conditions are examined however. p i l e must be avoided. driving

the p i l e

Two

additional

F i r s t , damage t o the sheet

Second, the hammer must be capable of

to refusal

depth . 4

I f either

of these

R e f u s a l i s the depth to which p i l e s have to be driven, to a t t a i n t h e i r designed resistance strength through skin and end bearing. For non-displacement p i l e s (see figure 4 . 7 ) , sheet p i l e s and s o l d i e r p i l e s , t h e i r depth extends below the tunnel bottom (excavation depth) by 5 f t to 15 f t (Winterkorn and Fang 1 9 7 5 ) . 4

Chapter 4. A KBES Framework for Methods Selection

conditions

140

cannot be met, then either the hammer energy i s

decreased and/or the sheet p i l e

s i z e i s increased.

I f no

s a t i s f a c t o r y solution can be found, the GWSS a l t e r n a t i v e of SSP

i s considered to be i n f e a s i b l e .

A rule format for determining Hammer Size according to SSP

s i z e , using empirical knowledge, i s shown below. RULE

Size the Hammer

IF AND THEN

SSPCrossSectionArea is Ap Context (SSP Database) Hammer_Type Is SingleActingAirHammer Context HammerSize (Hammer_Rated_Energy) Context = < 3,000 * Ap (lb-ft) Single_Acting_Air_Hammer.Size > (DAAHDatabase), Context

AND

'

Source

= RequiredHammer.Size {Control Strategy} AND

The

Do Pile_Driving_Strategy

above

rule,

based

determines the maximum multiplying prevent

3,000

pile

successfully, driving

control

contractors'

magnitude of the hammer

lb/in 2

damage.

on

A

by

experience, energy by

SSP_Cross_Section_Area

After i s passed

executing to s e l e c t i n g

this

to rule

the

pile

strategy.

3. P i l e Driving Strategy fKB-5") Soil

conditions

strategy

and hammer power dictate the p i l e

i n terms of d r i v i n g i n singles or p a i r s .

driving

Secondary

factors relevant to t h i s strategy are SSP s i z e and length of

Chapter 4. A KBES Framework f o r Methods Selection

pile

segment.

To prevent buckling,

141

the maximum

allowable

d r i v i n g depth with respect to adjacent p i l e s i s < = 13 f t . As

a

greedy

contractors pairs

approach

is

have suggested

that

i s a preferred

In_Pairs

favoured,

first.

r e s t r i c t e d to two states:

as

other

t r y i n g to drive p i l e s i n

approach, the p i l e

i s selected

and

Driving

driving

strategy

conditions

are

s o f t and hard d r i v i n g Conditions.

They are i n f e r r e d from the s o i l conditions using s i m p l i f i e d rules.

For the former, soft d r i v i n g conditions, the d r i v i n g

strategy i s In_Pairs; while the l a t t e r d r i v i n g strategy i s InSingles.

Other pertinent factors such as hammer type and

s i z e , type and grade of sheet p i l e s , number of p i l e d r i v e r s and

complexity

of

the project

were

not

considered

explicitly. The

following

driving

rules

In_Singles

exemplify

the conditions

and In_Pairs.

Pile

driving

for pile In_Singles

rule (experiential) i s : RULE

Drive in Singles

Source

IF AND AND THEN

PileDriving.Conditions Are Hard Refusal.depth > 30 ft Hammer.size < 22,000 Ib-ft Drive Piles In_Singles

Context, User Context Context

{Control Strategy} AND

Do Performance Measures

Chapter 4. A KBES Framework f o r Methods Selection

142

P i l e d r i v i n g In_Pairs rule (experiential) i s : RULE

Drive Piles in Pairs

Source

IF AND AND THEN

Pile Driving.Conditions Are Soft RefusaLdepth < 90 ft Hammer.size > 10,000 Ib-ft Drive Piles In Pairs

Context, User Context Context

{Control Strategy} AND

Do Performance_Measures

Predict (KB-6) The steps involved are: 1.

Determine production rate and cost.

2.

During p i l e d r i v i n g production analysis, a p i l e damage check i s involved (blow counts) If

blow count exceeded, then backtrack and change one or more choices

Otherwise

determine production + cost. If not acceptable, then . .

In choosing between p i l e d r i v i n g ln_Pairs or In_Singles, consideration must be given t o both set up time and d r i v i n g time.

Production Time and Cost Performance Given a f e a s i b l e SSP size, hammer type and s i z e , and d r i v i n g strategy,

then

performance,

i t remains

to determine

and perform other checks,

time

and

cost

such as regulatory

considerations on noise l e v e l . 5

S a f e t y and regulation factors are not implemented at t h i s level. The noise l e v e l i s i m p l i c i t l y s a t i s f i e d as affirmed by the preliminary regulation f e a s i b i l i t y check. 5

Chapter 4. A KBES Framework f o r Methods Selection

Using a hammer dynamic energy

143

(modified Engineering News

formula ) production routine written i n C (Drive.c), control 6

i s passed to t h i s routine f o r purposes of computing

a pile

d r i v i n g production rate excluding fixed set up time.

Output

information i s passed back to the control strategy which i n turns

interprets

the

routine

results.

A

constraint

is

included dealing with the maximum number of blows per foot, beyond which p i l e damage i s l i k e l y . blow count

i s reached,

the routine

In the event that t h i s stops summing up

the

incremental production time and returns a message to CMSA i n d i c a t i n g t h i s v i o l a t i o n and where i t happened —

i . e . the

depth of p i l e where i t interrupted d r i v i n g . The

hammer blow count

different

i s empirical

and

soil/pile/hammer where technical

can

vary with

feasibility

is

monitored by observation. For CMSA, the system takes action based on a t h e o r e t i c a l blow count from the model derived i n Appendix

B.

T h i s formula, among numerous of hammer energy formulas, i s applied only to some types of impact hammers. For vibratory p i l e d r i v e r s , r u l e of thumbs are u t i l i z e d to estimate t h e i r productivity. b

Chapter 4. A KBES Framework f o r Methods Selection

144

The r u l e format f o r running the "Drive.c" routine i s : RULE

Do Performance Measures

Source

IF AND AND AND AND

Selected_SSP_Type Is PZ27 Selected Hammer = Single Acting Air_Hammer (SAAH) SAAH.Rated_Delivered_Energy = 15000 Ib-ft Pile Driving Conditions Is Soft PilesDrivingStrategy Is "InPairs"

Context Context Context Context Context

{Control Structure if Accepted} THEN AND

Compute ProductionRate DoSSP.Risk Assessment {Control Structure if Unaccepted}

ELSE

Analyze Technical Feasibility

8

The above r u l e pools s p e c i f i e d method a t t r i b u t e s — pile,

hammer and p i l e d r i v i n g strategy,

input parameters to the Drive.c Figure

4.10

shows

input

sheet

and sends them as

routine. parameters

f o r the

Drive.c

routine passed by CMSA v i a an input text f i l e which then i s processed routine,

by the routine.

After

executing

the numerical

output parameters are passed back t o the control

strategy for interpretation and further manipulation. Output values

variables

f o r each

include of

skin

incremental friction,

and

end

cumulative

bearing

soil

'Blow count s a t i s f a c t i o n i s i m p l i c i t i n t h i s clause. I f the threshold blow count per foot i s v i o l a t e d , then control i s passed t o the technical f e a s i b i l i t y analysis to investigate a remedy. Technical f e a s i b i l i t y at t h i s stage refers to whether a method attributes combination (SSP + P i l e Driver + Construction Strategy) achieves i t s goals of production, cost, damage free d r i v i n g , and so f o r t h .

Chapter 4. A KBES Framework f o r Methods Selection

resistance,

blow

count,

and

production

145

progress

(relationships and runs are detailed i n Appendix B).

Drlve.c Input File

/Git • Soil Profile SSP properties Hammer Properties. /

Compute: Skin Friction End Bearing Friction Hammer Blows Rate Incremental Pile Penetration Rate

NO

/ Is Average _ Set Satisfied? YES

Compute Pile Driving Cycle Duration

Figure 4.10 A criterion

Drive.c Routine Interface with CMSA

of maximum acceptable hammer blow count of

150 blows/ft (a bench-mark from f i e l d engineers t o interrupt i f r e f u s a l i s reached) i s set as the threshold. count

exceeds t h i s

limit,

I f the blow

then the Drive.c routine stops

Chapter 4. A KBES Framework f o r Methods Selection

146

computation, flags the depth where i t happened and sends a message to the control

strategy

that the method combination

which

i n turn

interprets

(technical f e a s i b i l i t y )

i s not

f e a s i b l e , or else, i t i s f e a s i b l e and the production rate i s passed by the routine to the control strategy. A f t e r the SSP a l t e r n a t i v e passed t h i s t e s t successfully, the quantity take-off and cost estimate computation f o r the whole project follow.

Once d e t a i l e d method cost i s known, a

r i s k assessment i s ensued subsequently.

5. Risk Assessment (KB-2) The r i s k

factor has a s i g n i f i c a n t

construction

methods

i n general

selection i n particular.

impact on the choice of and

on

The considerable

shoring

method

emphasis placed

on an informal r i s k assessment, p a r t i c u l a r l y with respect to the l i k e l i h o o d of catastrophic r i s k , was highlighted i n an interview

with

a

seasoned

construction

engineer

(see

Appendix C). Based

on

discussions

with

construction

personnel,

a

review of the l i t e r a t u r e and an analysis of the amount of data

likely

construction

to be available when making decisions methods,

a

s i m p l i f i e d a n a l y t i c a l CMSA

assessment framework was developed as described 4.4.

about risk

i n section

What i s important i s the r o l e of r i s k i n the control

strategy.

I f the p o t e n t i a l f o r catastrophic or unacceptable

Chapter 4. A KBES Framework f o r Methods Selection

risks

are high,

withstanding

then

a

i t s appeal

method because

will

be

of time

147

dropped, or unit

not cost

performance. As

described

three

states

design

i n section

4.4, the r i s k

of nature of geological

alternative.

model

involves

conditions,

given

The three states correspond to better

than, equal to, and worse than expected conditions. cost

categories

nature.

a

describe

the outcome

f o r each

Several state of

An example rule f o r the case when the state of

nature i s AsExpected, i s as follows: "As-Expected" State of Nature

RULE

Compute Risk for

IF AND AND AND Context

GWSS = Steel_Sheet_Pile (SSP) SoilConditions Is AsExpected As_Expected_Conditions.Likelihood = Pp Consequence Costs (Dp) = Sum (Dp(i))

THEN AND

Compute AsExpected RiskCostComponent AsExpectedConditions.RiskComponent = (Pp * Dp)

Source

Preliminary Feasibility Context User User,

Control Strategy AND

The

first

Compute Other Risk Component Costs

premise

identifies

the candidate

evoke r i s k relevant s l o t s from the preliminary

i n order to level.

The

second premise affirms the state of nature by context.

The

third

premise queries

the user

about the i n d i v i d u a l cost

item estimates (Dp(i)), while context conclusion, as

adds them up.

the As_Expected r i s k element w i l l be

the product

of i t s l i k e l i h o o d

In the

evaluated

by the sum of i t s cost

Chapter 4. A KBES Framework f o r Methods Selection

items. nature

148

Next, control i s passed t o compute other states of risk

components

—

i . e . better

than

expected and

worse than expected. Next, we examine the SSP method analysis component.

6.

Analyze (KB-7)

The following

control

strategy

rule

examples are used to

d i r e c t the search f o r changes to the construction method i n order to achieve f e a s i b i l i t y . RULE (1) IF AND

Change Pile Driving Strategy from "In_Pairs" to "InjSingies" Source PueDrivingStrategy Is "In_Pairs" Context, User TechnicalJ'easibility.State is False Context (blow count < = threshold) {Control Strategy}

THEN AND

Change PileDrivingStrategy to InSingles Do Performance Measures

RULE (2)

Increase Hammer Energy

Source

IF AND

Pile_Driving_Strategy Is InSingles TechnicalFeasibility Is False

Context Context

{Control Strategy} THEN AND

Increase the Hammer Delivered Energy Do Performance Measures

Appendix B contains the l o g i c and relevant dynamic formulas derivation f o r the technical feasibility test under pile/hammer/soil scenarios.

Chapter 4. A KBES Framework for Methods Selection

149

RULE (3)

Change GWSS Alternative

Source

IF AND

Driving_Strategy Is In_Singles Hammer/Pile Are Not_Compatible

Context Context

{Control Strategy} THEN AND

4.4

Change SSPAlternative to SPLAlternative Do Design SPLAlternative

CMSA Risk Component Development and

Interviews regarding

and

discussions

with

contractors'

personnel

the decision making process dealing with methods

s e l e c t i o n have highlighted t h e i r process,

Evaluation

with r i s k s .

concern, early on

i n the

P a r t i c u l a r emphasis i s placed

on

the

p o t e n t i a l f o r large/catastrophic r i s k s which often accompany A method which i s

underground work, work i n water, etc.

more l i k e l y to be subject to such r i s k s tends to be shunned even i f there are s i g n i f i c a n t cost/time with i t . as

benefits

associated

Contractors tend to seek a s o l u t i o n that a c t i v e l y ,

opposed to passively,

controls r i s k

(e.g.

H piles

and

lagging with s t r u t s rather than shotcrete). Based on the above, the use of a r i s k c r i t e r i o n to alternatives framework intensive. may

early on

should

be

i s important. simple

use

and

risk not

assessment

overly

portray

method, e.g. see figure effort

and

the

operating

environment

for

for a

4.11.

data required

to

specify each r i s k

and corresponding states simply are not a v a i l a b l e . assessment

data

For example, a complex set of states of nature

realistically

The

to

The

screen

Cut-and-Cover

tunnelling

So

alternatives

type risk for

Chapter 4. A KBES Framework f o r Methods Selection

this

thesis,

must

be

simplified

to

treat

150

only

those

conditions that could lead to unacceptable r i s k s . In

this

section,

illustrate

the

the

context

approach.

Two

of

GWSS

risk

is

used

categories

to are

considered:

Figure 4.11 1.

Normal

conditions influence

States of Nature f o r Methods Selection

Risks:

Normal

those

site

(access, weather, ground, management, etc.)

that

productivity

and

risks

deals

with

other variables,

thus

creating

uncertainty i n time and cost estimates. 2. Large/Catastrophic

Risks:

These are treated

explicitly

i n the decision making process through a s i m p l i f i e d decision tree shown i n figure 4.12.

The basis of t h i s decision tree

Chapter 4. A KBES Framework for Methods Selection

this

thesis,

must

be

simplified

to

treat

150

only

those

conditions that could lead to unacceptable r i s k s . In

this

section,

illustrate

the

the

context

approach.

Two

of

GWSS

risk

is

used

categories

to are

considered:

Figure 4.11 1.

Normal

conditions influence

States of Nature f o r Methods Selection

Risks:

Normal

those

site

(access, weather, ground, management, etc.)

that

productivity

and

risks

deals

with

other variables,

thus

creating

uncertainty i n time and cost estimates. 2. Large/Catastrophic

Risks:

These are treated

explicitly

i n the decision making process through a s i m p l i f i e d decision tree shown i n figure 4.12.

The basis of t h i s decision tree

Chapter 4. A KBES Framework f o r Methods Selection

i s as follows. a l t e r n a t i v e s , we

F i r s t , since we

151

are using r i s k to preserve

examine alternatives i n d i r e c t l y .

Second,

three basic state of nature are treated: 1.

conditions better than expected;

2.

conditions as expected; and

3.

conditions worse than expected.

Conditions

here r e f e r s to that condition most l i k e l y

lead to unacceptably large or catastrophic r i s k s . each

state

of

nature

treat

three

to

Then, for

more conditions

—

no

f a i l u r e , minor f a i l u r e , major f a i l u r e ; or no damage, element damage, system damage, where: 1.

minor f a i l u r e r e f e r s to p a r t i a l damages for the wall structures and/or the surroundings; and

2.

major f a i l u r e r e f e r s to GWSS collapse, or retaining system collapse, and/or other major surrounding damages.

The

user

i s required

to

branch i n the decision tree. vector

of

incremental

assign

p r o b a b i l i t i e s to

At the end

costs

each

of each path i s a

(positive/negative).

This

vector of costs i s : Labor Equipment Materials Loss of L i f e Loss of Reputation Subsurface Subsidence Season Loss The

user i s asked to estimate the costs associated

t h i s vector as a f r a c t i o n or percentage of t o t a l

with

estimate,

Chapter 4. A KBES Framework f o r Methods Selection

given

the states of nature.

contractors In

152

This corresponds to the way

estimate.

terms

of

computation,

the

incremental

costs

summed, and then discounted by the p r o b a b i l i t i e s . are

deemed

assign

to be unacceptable,

an

infinite

cost

to

the user a

cost

are

If risks

i s required to

category.

This

e f f e c t i v e l y eliminates the a l t e r n a t i v e . Then, Total cost of a l t e r n a t i v e = For

base cost + expected value of incremental further

challenge

development

i s to i d e n t i f y ,

governing

risk

of

the

expert

costs

system,

the

f o r each method a l t e r n a t i v e , the

considerations

(e.g.

ground

condition

variable, flood p o t e n t i a l , etc.) We also need to allow the facility process tree.

f o r the user that

leads

to describe

and record

to the s p e c i f i c a t i o n

the thought

of the decision

Shoring Alternatives

Event Chances

Consquences Costs (Outcomes)

Encountered Geological Conditions More Favorable Than Expected (PI)

CD-

Risk Category

Steel Sheel Pile

Equipment Loss

Risk Cost

%ot Total Cost %of Labor Loss Total Cost %of Material Loss Total Cost %ot Life Loss Total Cost Subsurface Sub- %of Total Cost sidence Loss %of Season Loss Total Cost Other Losses Catastrophic Damage (Not Acceptable)

x

%of Total Cost

154

5. C M S A Implementation

5.1 Introduction The

primary

objective of t h i s

chapter

issues involved i n the implementation system. Object,

The f i r s t an expert

objective

part

covers

system

used

i s to f a m i l i a r i z e

environment:

knowledge

i s to explore the of a prototype CMSA

an overview

of NExpert

t o implement

the reader

constructs,

CMSA.

with

The

the NExpert

syntax,

operators,

inference mechanism, and so forth. The

second

prototype

part

covers

development.

selected

details

The h e u r i s t i c

problem

paradigm (figure 4 . 6 ) consists of Suggest, and Analyze the

operators.

preliminary

The f i r s t

knowledge

that

of the CMSA solution

Design,

Predict,

operator, Suggest, suggests

maps

a preliminary

f e a s i b l e GWSS a l t e r n a t i v e f o r further d e t a i l i n g by the low level

feasibility

part.

The

other

correspond to the low l e v e l f e a s i b i l i t y specifies

the design

construction

the

operators

component:

Design

element, construction resources, and

strategy; Predict

formula t o predict

three

applies the hammer dynamic

i t s performance; and Analyze

synthesized method to t e s t

diagnoses

i t s feasibility.

If

the

assembled method meets i t s goals, then i t accepts the GWSS alternative

(SSP) .

I f i t does not meet i t s goal, then i t

suggests recommendations f o r re-designing the method.

Chapter 5. CMSA Implementation

The

construction

elements,

and

frames.

Rules

are

attributes resources

used

to

for

resource

selection

strategy

(pile

alternatives.

of

thumb

literature

review

appendices

were

interviews.

and

in

engineering

(sheet

analyzing

feasibility),

and

of

and

frames,

extracted

as

2,3,

from

describes

foremen, and

from a

interviews.

the

and

GWSS

databases

presented and

4

texts,

others.

s e l e c t i o n process

varying degrees. experts

method

provided.

algorithms chapters

ranking

rules,

process

and

in

the

in

the

journals,

and

Knowledge a c q u i s i t i o n for methods s e l e c t i o n and

superintendents, method

piles),

the

analysis i s the domain of a number of experts,

the

specific

construction

formulas),

written i n NExpert Object w i l l be Rules

other

as

drivers s e l e c t i o n ) , construction

(dynamic

Examples

design

represented

s t r u c t u r a l members

(pile

(technical

are

the

These include

d r i v i n g pattern),

model evaluation

of

represent

control strategy.

knowledge

synthesis

method

construction

knowledge and design

155

one site

at

engineers,

Each contributes

different

times,

Knowledge a c q u i s i t i o n from s i t e and project visit

and

Such interviews

was

undertaken.

provides

the

contributed

and

v a l i d a t i n g the problem solving knowledge base.

to

office

Appendix

r e s u l t s of to

to

improving

C two and

156

Chapter 5. CMSA Implementation

5.2 NExpert Object Overview An

Al toolkit

specialized should

i s ideal

tools.

offer

programming,

f o r problems

I t i s desirable

a

hybrid

and

access

rule to

that that

system,

a

general

need

a mix

such a

of

toolkit

object

oriented

purpose

language.

A d d i t i o n a l l y , the t o o l k i t should interface with conventional software such as databases, spreadsheets, graphics packages, and word processors.

Neuron Data's NExpert Object exhibits

several of these features. NExpert i s a powerful, hybrid,

rule and object based

expert system s h e l l that speeds up the prototyping process for expert systems f o r non-programmers.

I t i s mainly a rule

based system f i t t e d with object oriented features, such as f i r i n g a routine from a premise i f c e r t a i n conditions have changed. lack

of

The version used by the author suffered good

documentation

and

examples

to

from a

explain

and

demonstrate a l l of NExpert's features.

5.2.1

Major NExpert Object Modules

NExpert Object consists of the following modules: 1. It

NExpert Development contains

pop-up

Package, i s the core of the system.

windows

for editing

text,

database,

rules, objects, classes, etc; v i s u a l display f o r rules and objects networks; a reasoning kernel providing control,

backward

chaining,

forward

inheritance

chaining,

pattern

157

Chapter 5. CMSA Implementation

matching c a l l s to external routines, and others; and trace f a c i l i t i e s (Transcript, Encyclopedia, 2.

Reports).

The C a l l a b l e Interface, i s a l i b r a r y of C routines and

function kernels used to embed NExpert within a conventional programming language.

I t consists of C functions that make

up the NExpert Object

Development

and Runtime Environments.

It can be used to e s t a b l i s h communications application advantage

programs

and NExpert

of structures

between external

applications

and functions

that

by

taking

NExpert

uses

internally. There

are at

interface. external "trap"

NExpert

code

three

functions

as exported

specific

replacement, other

least

NExpert

ways

to use

the

can be c a l l e d

callable

directly

C functions; external functions

or a standard

and declare

by

code can their

own

message can be passed between

windows applications and NExpert

using

Microsoft's

dynamic exchange protocol (a system f o r s e t t i n g up standard messages).

The c a l l a b l e interface includes C functions to

initialize,

start,

change

knowledge

stop,

and

resume

structures;

and

sessions;

change

query

or

list

of

the

hypotheses or agendas. In embed

order

to i n s t a l l

NExpert

additional

within

function

handlers

another

application,

software requirements such

i n NExpert, or there

as the MS

are

Windows

158

Chapter 5. CMSA Implementation

Software

Development

K i t , version

2.03 or l a t e r ,

and a

Microsoft C Compiler version 5.0 or l a t e r . 3. NExpert Object Runtime, i s a run time package that i s used to run an application without access to the knowledge base

(a development

mechanisms).

package

stripped

The developer can define,

of

i t s debugging

in detail,

how an

application w i l l run and design an interface that i s v i s i b l e to the c l i e n t . An overview of the NExpert development framework i s shown i n figure 5.1. 4. Hardware and Software Requirements — and

software

requirements

for

Minimum hardware

NExpert

Object

under

MS/Windows are: IBM PC or compatible with 64OK of conventional memory, plus 1 MByte of expanded memory on a 286 machine 1 MByte of expanded or extended memory on a 386 machine (2M i f you are using Windows 386) Enhanced Graphics Adapter (EGA) or VGA. Color EGA requires a video board with at l e a s t 64K of memory. -

1.44 M floppy disk drive hard disk with at l e a s t 3 MByte a v a i l a b l e Microsoft Windows Runtime 286 or 386, version 2.03 or l a t e r . Mouse (compatible with MS-Windows: Bus or S e r i a l , Microsoft, LogiMouse, Mouse Systems, etc.), P a r a l l e l port.

Chapter 5. CMSA Implementation

Recommended

hardware

159

and software

i n addition

to the

above: 2 additional MBytes of Extended memory. This allows you to store NExpert Object i n RAM drive, 386-based machine (for development)

External devices/computers multi-process

Networking Ethernet, DecNet.

real-time data Input Into NEXPERT event-driven architecture

Inter-process Communication Vax Calling Conventions Dynamic Data Exchagne

Multi-tasking

NEXPERT!

DataBases DB III. Lotus 123, ORACLE, SOL, RDP, EXECL...

o

Retrieve

• Show

graphics text

data storage read/write

Knowledge Bases dynamic access to knowledgebases Load/Unload

Figure 5.1

explanations graphics text focus of attention conclusions reasoning trace active values reports what If

NExpert Object Open A l Environment Framework [from Neuron Data 1989]

Chapter 5. CMSA Implementation

160

5.2.2 NExpert Primitives - Building Blocks and Operations The basic b u i l d i n g blocks of NExpert Object are described i n this

subsection.

1.

Rules Rules are the preferred way to process objects i n either

forward

or backward

chaining.

In NExpert,

rules are

expressed i n the following form (see figure 5.2 f o r NExpert Rule Syntax) IP AND

Conditionl Condition2

AND

ConditionM

THEN

Conclusion

AND AND

Actionl Action2

AND

ActionM

The

l e f t hand side

property

(LHS) of a r u l e t e s t s the value of a

f o r some object or c l a s s .

existential instance

(Hypothesis)

operator

(test

Mixed q u a l i f i e r s such as

the condition:

...) and universal operator

are a l l instances

(test

i s there

any

the condition:

...) can be used to lend expressive power

to the LHS conditions. The more antecedent

LHS of the rule i s composed of one or

( i f ) clauses

which are c a l l e d

conditions.

Operators and t h e i r purposes on t h i s side include:

Chapter 5. CMSA Implementation

Yes and No [>/ = (SectionModulus) (1.9)) /v

(< (Section_Modulus) (2.4))

(Is (< |Selected_SSP| > .designation) ("PSA32"))

) (@HYPO= SelectSSP) (@RHS = (Do ( < | Selected_SSP | > .designation) (Selected_Steel_Pile)) (CreateObject ( < | Selected_SSP | > ) (| Matched_SSP |)) (Do (SelectPileDriver) (SelectPileDriver)) ~

) Figure 5.12

Steel Sheet P i l e Selection Rule

This r u l e i s interpreted clause by clause as follows. 1.

L e f t hand Side

The f i r s t and second clause say that i f the required section modulus (which i s calculated by a p r i o r rule) p i l e i s between 1.9 i n says there

3

f o r a sheet

and 2.4 i n , and i f the t h i r d clause 3

i s a sheet p i l e object i n the Selected_SSP that

s a t i s f i e s those section modulus l i m i t s , then "PSA_32" of A36 ASTM i s selected. The

notation implies a pattern matching

operator

f o r sheet p i l e s which are treated as objects i n a

l i s t i n order to match whatever property i s s p e c i f i e d i n the LHS.

194

Chapter 5. CMSA Implementation

2.

Hypothesis

The hypothesis ("Hypo") name c a l l e d Select_SSP i s invoked by "backward chaining" v i a the control strategy.

I f the Left

Hand Side conditions were s a t i s f i e d , then t h i s hypothesis i s evaluated as TRUE:

otherwise i t i s FALSE or NOTKNOWN i f one

of the conditions was f a l s e or not known respectively. 3.

Right Hand side

The pile

"Do"

operator assigns the designation of a s t e e l

to the variable

"Selected_Steel_Pile",

and

sheet

then

uses

the "CreateObject" operator to l i n k the selected sheet p i l e (Selected_SSP)

to a new

class or l i s t

of "Matched_SSP", to

separate i t from the rest of the sheet p i l e s f o r subsequent use

i n hammer

selection

and

technical

feasibility.

Matched_SSP object i n h e r i t s i t s attributes and values the Selected_SSP

class

l a s t clause of the "Do"

(pattern matched sheet p i l e ) .

The from The

operator transfers control i n order

to invoke the "Select_Pile_Driver" hypothesis using backward chaining. that

This hypothesis i s then used to s e l e c t a hammer

satisfies

soil

and

pile

conditions

including

the

Matched_SSP properties which are treated as constraints. NExpert text database Format (*.nxp), f l a t database, f o r the s t e e l sheet p i l e s i s shown i n figure

5.13.

195

Chapter 5. CMSA Implementation

\SSP_l.Designation\ = "PZ38" \SSP-l.Weight_per_foot\="57.00" \SSP_l.Cross section_area\="16.77" \SSP l.Driving_width\ = "18" \SSP_l.Surface_area\="5.52" \SSP_l.Section_modulus\="46.8" \SSP_2.Designation\="PZ32" \SSP-2.Weight_per foot\="56.00" \SSP_2.Cross section_area\="16.47" \SSP 2.Driviifg_width\="21" \SSP_2.Surface_area\="5.52" \SSP_2.Section_modulus\="383"

Figure 5.13 Steel Sheet F i l e s Database (SSP.NXP) 2.2

Construction Resource Class

Figure

5.14 shows a construction

Hierarchy resource hierarchy

which

divides Cut-and-Cover tunnelling c a p i t a l intensive resources into classes and objects. slots

for activity

and task

cost,

productivity,

etc.

A resource class has universal type and i d e n t i f i c a t i o n , The subclasses

unit

of bulldozers,

cranes, and hammers have further s l o t s to characterize them in

terms

properties, piles,

of

their

etc.

bulldozers

a c t i v i t i e s ; and

functionality,

For instance, are used

size,

operational

f o r a GWSS of s t e e l sheet

f o r " c l e a r i n g " and excavation

cranes are used f o r handling

and h o i s t i n g

materials, muck removal, and carrying a hammer f o r the p i l e driving activity.

Chapter 5. CMSA Implementation

196

Construction Resource Class Slots Activity: Task:

Excavation, Pile Driving,.. Hoisting, Pile Driving, Excavation,..

Resource Type:

L a Intensive, Capital Intensive. D O r

Bulldozer Subclass Hydraulic Hammer Subclass Vibratory Subclass Matched_Hammer Subclass

Drop Hammer

Slots Hammer_Mode! Theortlcal_Energy -|

| Stroke_per_MIn Length_of_Stroke Ram_Welght

Vlbro

Manufacturer Selected SAAH: YES

(^)

Figure 5.14

Class, Subclass

A

Ob|ect

•

Slot

construction Resource c l a s s Hierarchy

Chapter 5. CMSA Implementation

The

hammer

vibratory

subclass

197

was divided

hammers and hydraulic

were further c l a s s i f i e d

into

impact

hammers.

hammers,

Impact hammers

i n accordance with t h e i r

operating

mode as single acting a i r hammer, double acting a i r hammer, d i e s e l hammer, d i f f e r e n t i a l acting hammer and drop hammer. Each hammer type i s represented by a frame, an example of which i s shown i n figure 5.15. Single Acting Air Hammer SAAH2 None

Class Objects SubObjects Slots Hammer Model Ram Weight Strokes Per Minute Length of Stroke Theoretical Energy HammerManufacturer Efficiency

Figure 5.15 Suppose

that

availability (SAAH). retrieved

the s o i l suggest

"S 20" "20,000", lb "60" "36", in "60,000", lb-ft Vulcan" "87%"

Impact Hammer Element context

variables

use of a single

and

acting

a i r hammer

Then, a database f o r the SAAH subclass and required

record

a t t r i b u t e s would

into

the corresponding

object

and properties

The

construction

knowledge

f o r hammer

resource

would be be mapped

i n NExpert.

selection

and

performance was encoded i n a rule, which involves r e t r i e v i n g s o i l s t r a t i f i c a t i o n information and p i l e properties from the

Chapter 5. CMSA Implementation

198

p i l e frame v i a a message which sends the selected p i l e

cross

section t o a h e u r i s t i c rule which l i m i t s the maximum s i z e of the hammer. In addition t o developing a data base f o r single a i r hammers, data bases were created hammers

as well

as vibratory

ones.

acting

f o r double acting a i r This

permits

one to

explore the e f f i c i e n c y of the prototype system to s e l e c t the most appropriate example

frame

resource as a function of s o i l context. f o r a vibratory

An

hammer i s shown i n figure

5.16. Knowledge

f o r s e l e c t i n g and s i z i n g a vibratory hammer i s

described

i n Appendix D.

requires

heuristic

The prediction of production rates

knowledge,

and

i s done

vibratory hammer type and size versus s o i l

Class Objects SubObjects Slots Vibratory Model DynamicForce Horse_Power Frequency Amplitude MaximumPull SuspendedWeight ShippingWeight VibratoryManuf. Efficiency

by

mapping

stratification.

Vibratory Vibrol None "1412" "20,000", lb "650", HP "400", Vibration per min "1.5", in "80", Tons "1020", lb "20.5", lb "ICE" "93%"

Figure 5.16 vibratory P i l e Driver Element

Chapter 5. CMSA Implementation

199

The data bases developed f o r the double acting a i r hammer types and vibratory hammer types are shown i n figures 5.17 and 5.18 respectively. \Hammer_01.Hammer_Model\="2" \Hammer_01.Ram_Weight\="3000" \Hammer_01.Strokes_per_Min\ = "70" \Hammer_01.Length_of_Stroke\="29" \Hammer_01.Thero_Energy\ = "7260" \Hammer_02.Hammer_Model\="1" \Hammer_02.Ram_Weight\="5000" \Hammer_02.Strokes_per_Min\ = "60" \Hammer_02.Length_of_Stroke\ = "36" \Hammer_02.Theor_Energy\="15000"

Figure 5.17 Double Acting Hammer Database (DAAH.NXP) \Vibratory_l.Dynamic_Force\ = "204" \Vibratory_l.Model \ = "1412" \Vibratory_l.Manufacturer\ = "ICE" \Vibratory_l.Frequency\ = "1200" \Vibratory_lAmplitude\ = "1" \Vibratory_l.Horse_Power \ = "650" \Vibratory_l.Max_Pull_Extract\ = "80" \Vibratory_l.Pile_Clamp_Force \ = "250" \Vibratory_l.Suspended_Weight\ = "10.20" \Vibratory_l.Shipping_Weight\ = "20.5" \Vibratory_l.Dynamic_Force\ = "204" \Vibratory_l.Model \ = "1412" \Vibratory_l.Manufacturer\ = "ICE" \Vibratory_l.Frequency\ = "400" \Vibratory_lAmplitude\ = "1.50" \Vibratory_l.Horse_Power \ = "650" \Vibratory_l.Max_Pull_Extract\ = "80" \Vibratory_l.Pile_Clamp_Force \ = "250" \Vibratory_l.Suspended_Weight\ = "10.20" \Vibratory_l.Shipping_Weight\ = "20.5"

Figure 5.18 Vibratory Hammer Database (VIBRO.NXP)

Chapter 5. CMSA Implementation

200

The NExpert text format f o r the selected class i s shown i n figure 5.19.

impact hammer

(@CLASS= Selected Hammer (©PROPERTIES = HammerModel Length_of_Stroke Ram_Weight Strokes_per_Min Theor Energy

)

) Figure 5.19

Impact Hammer Class i n NExpert

Shown i n figure 5.20 i s an impact hammer s e l e c t i o n r u l e . This r u l e i s f i r e d i f previous rules have indicated single selected

acting

a i r hammer

may

( i f one i s feasible)

Class f o r l a t e r use.

be

suitable.

i s linked

that a

The hammer

to Matched_Hammer

The MatchedJHammer i n h e r i t s the same

a t t r i b u t e s and values of the "pattern"

.

Chapter 5. CMSA Implementation

201

(@RULE = Select_Vibratory_PD_or_DA_Hammer_PD_for_LS_ON_STCL ©COMMENTS = "Select a hammer type based on soil profile and conditions (Using Hunt 1979 Table)";@WHY="Inference category for this rule is set to 1 since double acting air hammer (DAAH-Hammer) overrides the vibratory selection under this rule condition (assumption)"; (@LHS = (Is (Soil_Profile_Scenario) ("Loose_Sand_ON_Stiff_Clay")) (Retrieve ("daah.nxp") (@TYPE = NXP;@FILL=ADD;@CREATE= | SelectedHammer | ;\ )) (< = (< | SelectedHammer | > .Theor_Energy (|Matched_SSP|.Cross_Section_Area * 3000) (Name (MAX(< |Selected_Hammer| > .Theor_Energy)) (Max_Energy)) (= (< | Selected_Hammer | > .Theor_Energy-Max Energy) (0))

) (@HYPO= SelectPileDriver) (@RHS = (Let (Hammer_Type) ("Double_Acting_Air_Hammer")) (CreateObject ( < | Selected_Hammer | > ) (| Matched_Hammer |)) (DeleteObject ( < | Selected_Hammer | >) (| Selected_Hammer |)) (Do (< | Matched_Hammer | > .Hammer_Model) (Hammer_Model)) (Do (SPProduction) (SPProduction)) ) )

Figure 5.20 The

rule

Hammer Selection Rule i n NExpert

i n figure 5.20 can be interpreted clause by

clause as follows: 1.

Left Hand Side

1.1

The f i r s t that

condition checks the s o i l

profile

scenario

d i c t a t e s the type and/or c l a s s of hammers that

should

be

used

(Impact

Double_Acting_Air_Hammer

(DAAH),

Suppose, based on experience, loose impact

sand

on s t i f f

hammer,

versus

clay

and within

. Vibratory,

Diesel_Hammer, e t c .

that a s o i l

i s best this

handled

subclass,

acting a i r hammer i s the preferred choice.

p r o f i l e of using an a

double

Chapter 5. CMSA Implementation

1.2

202

I f the

first

condition

i s successfully

(returns

True), the second condition

satisfied

assumes that the

DAAH subclass best matches the s o i l p r o f i l e .

Therefore

the database of DAAH w i l l be retrieved and attached to Selected_Hammer Class. 1.3

The

third

condition

reduces

the

list

of

Selected_Hammers by retaining only those hammers which have a t h e o r e t i c a l energy ( l b - f t ) equal to or less than 3,000

times

the

Matched_SSP.

cross

sectional

area

of

the

This rule of thumb i s used i n the f i e l d

by p i l e d r i v i n g contractors to prevent s t e e l sheet p i l e damage caused The 11

driving

by over-sized strategy

In_Singles".

state

pile

driving

i s assumed

equipment. here

However, i f i t i s "In_Pairs",

upper bound i s doubled.

and hammer s i z e .

hammer

dictate

weight

may

then the

There i s a tradeoff

the d r i v i n g strategy

to be

between

Furthermore, the

the weight

of the p i l e

segment (length and weight), e s p e c i a l l y for the case of a vibratory p i l e driver. 1.4

One hammer

selection

productive, theoretical level)

or

efficiency. subclass

largest,

energy

assuming

have

c r i t e r i o n i s to pick

(cost

those

hammer

i n terms

i s not considered

hammers have

Different different

the most

models

at this-

the same hammer

within

hammer

of i t s

same

hammer

efficiencies for

Chapter 5. CMSA Implementation

pile/soil/hammer

203

scenarios.

Therefore, condition four

picks the largest hammer from the Selected_Hammer

list

using NExpert operator "MAX". 2.

The hypothesis through

named

Select_Pile_Driver

i s triggered

"Forward chaining" v i a the control strategy.

I f the Left Hand Side conditions were s a t i s f i e d ,

then

t h i s hypothesis i s evaluated as "True"; otherwise i t i s "False" or "Notknown" when one of the conditions was Notknown. 3.

Right Hand Side The "Let" operator assigns the hammer subclass type as a s t r i n g , Double_Acting_Air_Hammer,

to the Hammer_Type

variable. The

"CreateObject"

operator

links

the hammer

which

s a t i s f i e d the previous conditions to the Matched_Hammer class.

The

"DeleteObject"

Matched_Hammer

from

Matched_Hammer

failed

production

rate,

as

operator

deletes

the

if

this

Selected_Hammer to

produce

inferred

from

the

required

the

technical

f e a s i b i l i t y diagnosis. The

"Do"

operator

designation

to

a

subsequent treatment. control

to

then

assigns

variable

of

the

hammer

Hammer_Model

model for

The l a s t "Do" operator transfers

the SP_Production

hypothesis

where the

Chapter 5. CMSA Implementation

204

pile/soil/hammer combination i s examined f o r t e c h n i c a l feasibility.

3.

Construction Strategy Class

For the

Hierarchy

Cut-and-Coyer tunnelling problem, the

construction

strategy class can be described i n terms of a hierarchy figure 5.21).

At

the higher

levels,

o v e r a l l construction

approaches, such as top-down or bottom-up method), are treated. influence

lower

These high

level

(see

(e.g. the Milano

l e v e l strategies greatly

strategies,

activities

and

their

sequencing. For instance, for a top-down strategy using a s t e e l sheet p i l e GWSS, strategies at the p i l e d r i v i n g a c t i v i t y l e v e l are drive "In_Singles lower l e v e l

or "In_Pairs".

11

strategy,

whereas the

CMSA deals only with the higher

level

strategies

are assumed to be fixed.

4.

Construction Process Model Class

The

construction

process

design and resource certain

model

(CPM)

draws

on

selected

frames i n a process aimed at s a t i s f y i n g

constraints

performance measures.

and

quantifying

construction

method

Chapter 5. CMSA Implementation

205

GWSS Class

Figure 5.21 Figure 5.22 representative

Construction Strategy Class Hierarchy

shows, at the top l e v e l , a CPM slots.

Class

subobjects

element with

(model b u i l d i n g

elements) include crews, equipment, layout, etc.

One,

or a

construction of more than one,

of these e n t i t i e s emulates a

systematic

operations,

representation

of

which

in

turn

y i e l d s q u a n t i t a t i v e / q u a l i t a t i v e performance measures. For CMSA, i t i s assumed that q u a l i t a t i v e v a r i a b l e s are dealt with quantifiable

prior

to using the

performance.

a t t r i b u t e s need to be phase.

process

Thus,

a

model to subset

of

accessible to the process

determine resource modelling

For instance, at the p i l e d r i v i n g a c t i v i t y

level,

Chapter 5. CMSA Implementation

206

three object instances, from two resource and p i l e s ) , and the s o i l p r o f i l e ,

classes (hammers

have t o be bound i n the

process model.

Construction Process Model Class Slots

SSP Subclass

Subobjects:

Crews, Equipments, Layout,.

Model Criteria:

Analytical, Approximate.,

Heuristics,.. Activities Involved: Pile Driving, Excavation,.. Model Performance Attributes: Time, Cost,.. Safety,

Soil Profile Subclass' Pile Driving Process Model SAAH Hammers

Slots • Soll.Pro (Proflle-1

Pile Driving Strategy: ln_Pairs, ln_Slngles Model Criteria: Dynamic Formula, or WEAP.

Pilejnstance (SSP-3)

Technical Feasibility. True, False

(Pile Driving Process Subob|ects)

Q

Class, Subclass

/ \ Object, Subobject •

Figure 5.22

Slot

Construction Process Model Class Hierarchy

At the process model l e v e l , the s l o t s t r e a t p i l e strategy,

process

model

type

and t e c h n i c a l

driving

feasibility.

Chapter 5. CMSA Implementation

Process

model

type

207

deals

with

the type

of a n a l y t i c or

numerical algorithm used. In

CMSA, f o r p i l e

d r i v i n g , the solution model used i s

based on dynamic formulas which combine the hammer and s o i l properties t o provide an approximate s o l u t i o n . The

third

slot

i s of the technical

feasibility

state

which has a boolean value of either true or f a l s e .

5.3.4

Technical F e a s i b i l i t y Fart

Given the s e l e c t i o n of a p i l e

type,

a hammer and a p i l e

d r i v i n g strategy, the next step i s to check the f e a s i b i l i t y of

the combination

This

check

using

involves

the construction

assessing

technical

p r e d i c t i n g time and cost performance. involves determining i f the p i l e

process

model.

f e a s i b i l i t y and

Technical f e a s i b i l i t y

can be driven to r e f u s a l

without damaging i t and whether or not the rate of d r i v i n g can

satisfy

production

rate

constraints

discussion below i s given i n the context

or targets.

The

of the PREDICT and

ANALYZE operators of section 4.3.4.

1.

Predict:

appropriate

The

construction

"Predict"

operator

selects

the

process model f o r p r e d i c t i n g the

method performance a f t e r "Design" has been done.

The rule

shown i n figure 5.23 combines the pre-selected matched sheet pile

section

and matched

hammer

properties

i n order

to

Chapter 5. CMSA Implementation

predict

pile

driving

progress

208

rate

based

on the dynamic

formula, as derived i n Appendix B.

(@RULE = Single_Pile_Variable_Production_Time @INFCAT=5; ©COMMENTS = "This rule computes the variable component of driving a single pile based on Dynamic Formula. Note that effective energy consideration is not included - e.g. hammer efficiency as function of the type of the hammer, and pile group effect on pile driving are not treated."; (@LHS = (Name (Tunnel.depth + 5) (L)) (Show ("Drive.txt") (@KEEP = FALSE; ©WAIT = TRUE;)) (Is (Driving_Conditions) ("Soft")) (Name("In_Pairs") (PilesDrivingPattern)) (Name ( < | Matched_Hammer | >. Strokes_per_Min) (F)) (Name( < | Matched_Hammer | > .Theor_Energy*Hammer. Total_Efficiency) (E)) (Name( < | Matched_SSP | > .Surface_Area*2) (SA)) (Name(< | Matched_SSP | > .Cross_Section_Area * 2) (Ap)) ) (@HYPO= SPProduction) (@RHS = (Write ("Hammer.nxp") (©TYPE = NXP; ©FILL=NEW; ©ATOMS = L,F,E,SA,Hammer_Type;\ )) (Write ("soil.nxp") (©TYPE = NXP; ©FILL = NEW; ©ATOMS = Soil_Type_l,\Start_l,Finish_l,Soil_Type_2, Start_2,Finish_2;\)) (Execute("Drive.exe") (©TYPE=EXE;@WAIT=TRUE;)) (Retrieve("out.nxp") (©TYPE = NXP; ©FILL=ADD; ©FWRD = TRUE; ©CREATE = |Var_Time|; \@ATOMS= Variable_Time.amount;)) (Retrieve ("out.nxp") (@TYPE=NXP; @FTLL=ADD; ©FWRD = TRUE; ©CREATE = | Feasibility | ;\@ATOMS=TechnicaI_Feasibility.State;)) (Do (CheckFeasibility) (CheckFeasibility))))

Figure 5.23 1.

Technical F e a s i b i l i t y Rule

Left Hand Side

The rule shown i n figure 5.23 s t a r t s with the depth of the tunnel, or excavation, from the context information

(tunnel

depth) and adds 5 feet to i t , as a default, to determine the

Chapter 5. CMSA Implementation

pile

length.

prevent,

The

soil

209

extra

boiling

displays a text f i l e ,

length

or

i s to

heave.

minimize,

The

second

if

not

condition

"Drive.txt", which b r i e f s the user

on

the questions to be asked by the system, what t h i s r u l e w i l l do and The

what i s expected to happen a f t e r f i r i n g t h i s

third

condition

checks

that

the

driving

rule.

conditions

correspond to Soft, as i n f e r r e d from the strategy component based

on

string

soil

value

strategy

conditions. of

,,

variable.

"Matched_Hammer" variables F and surface

In_Pairs

and

,,

next

for

the

Conditions frequency

E.

cross

The

four

and

Condition

condition

and

The

five,

s i x assigns

hypothesis,

assign

the

energy

to

the Matched_SSP

variables SA

These variables are used as part of the formula routine.

a

Pile_Driving_Pattern

theoretical

section areas to

assigns

and

Ap.

input f o r dynamic

SP_Production, r e f e r s to

whether or not the sheet p i l e production rate i s acceptable (True or F a l s e ) .

2.

Right Hand Side

The

input required

two

files

for p i l e d r i v i n g routine i s written i n operator.

The

i s "hammer.nxp", which includes hammer and

pile

data, and the second i s "soil.nxp", which includes the

soil

first

one

(*.nxp format) using the

p r o f i l e input.

"Write"

210

Chapter 5. CMSA Implementation

The

"Execute"

"Drive.exe",

operator

and

runs

calls

it

the

in

the

executable DOS

file,

environment.

"Drive.exe" i s a compiled program f o r predicting the speed of p i l e

d r i v i n g given a soil/hammer/pile combination.

It

f a i l s i f a constraint such as the allowable number of blows per foot run i s v i o l a t e d . the

database

running

file

"Out.nxp", which contains

"Drive.exe"

contained

results.

i n the f i l e

returns two values.

The "Retrieve" operator r e t r i e v e s

More

"Out.out".

a summary of

detailed

output i s

The "Retrieve"

operator

The f i r s t i s the variable time f o r p i l e

d r i v i n g , which indicates the speed of d r i v i n g , and which can be tested against the p i l e d r i v i n g productivity. variable

i s the state

of technical

True i f the p i l e d r i v i n g operation

The second

f e a s i b i l i t y , which i s

i s successful, otherwise

i t i s False i f the p i l e i s not driven to i t s r e f u s a l or the damageability bound has been v i o l a t e d . The

next

diagnosing

hypothesis, the cause

feasibility. strategy low,

"Do

Check

of a False

For example, a driven

Feasibility", response

involves

f o r technical

"In_Pairs"

construction

i s i n f e a s i b l e because the production rate

i s too

or the maximum number of blows before damage w i l l occur

i s exceeded.

Chapter 5. CMSA Implementation

2.

211

Analyze (Diagnose):

Here, we examine how the CMSA can diagnose a f a i l u r e of a technical f e a s i b i l i t y t e s t and suggest a remedy. the hypothetical s t i f f clay

Consider

example of a s o i l p r o f i l e of loose sand on

(given a high loose sand/Stiff clay depth ratio)

which implies s o f t d r i v i n g conditions which i n turn suggests driving

i n pairs.

Suppose

that

f e a s i b i l i t y returns the value False.

the state

of technical

Thus:

The "In_Pairs" p i l e d r i v i n g strategy may be i n f e a s i b l e and p i l e d r i v i n g should be done In_Singles, even though the s o i l scenario suggested s o f t d r i v i n g conditions. (Note that the p i l e d r i v i n g strategy also a f f e c t s the rate of production as well as the fixed time f o r p i l e driving.) The rationale behind t h i s i s that when p i l e s are driven "In_Singles", s o i l resistance, mainly skin f r i c t i o n and secondary end bearing, w i l l be decreased by h a l f . The number of blows per foot exceeds the allowable limit which indicates high soil resistance, or i n s u f f i c i e n t hammer energy. One. way t o remedy t h i s s i t u a t i o n i s to pick a larger hammer. However, the hammer s e l e c t i o n c r i t e r i o n already considered picking the largest hammer that s a t i s f i e d the s t e e l sheet p i l e constraints. Therefore, i f a bigger hammer i s to be chosen, then the strength of the sheet p i l e must be increased. In other words, a heavier s t e e l sheet p i l e section i s required. The production rate did meet the required progress rate v a r i a b l e plus fixed time. As a r e s u l t d r i v i n g "In_Singles", and/or use of a larger hammer and p i l e section may be suggested.

Chapter 5. CMSA Implementation

212

A text f i l e of "Tech_Fea.txt" i s displayed t o the user to explain t h i s r u l e , as shown i n figure 5.24, along with the expected actions. true,

then

I f the technical f e a s i b i l i t y condition i s

the "Do" operator

invokes

the hypothesis of

Compute_Production_of_SSP i n order to c a l c u l a t e p i l e d r i v i n g costs. (@RULE= Technically Feasible Alternative ©COMMENTS = "If pile driving conditions state (based on soil profile and conditions) is soft, then the pile driving pattern will be In_Pairs, else will be In_Singles. The selection of either driving strategy will be reflected in pile driving rate where fixed and variable time computation will be different for each.; (@LHS = (Yes (Technical_Feasibility.State)) (Is (DrivingConditions) ("Hard")) (Is (PilesDrivingPattern) ("In_Pairs")) (Show ("Tech_Fea.txt") (@KEEP = FALSE;@WAIT = TRUE;)) ) (@HYPO= CheckFeasibility) (@RHS = (Do (Compute_Production_of_SSP) (ComputeProductionofSSP)) )

Figure 5.24 A r u l e directed condition

Method Technical F e a s i b i l i t y at reversing

the technical

i s True feasibility

from False to True by changing the p i l e

strategy from "In_Pairs" to "In_Singles"

driving

i s shown i n figure

Chapter 5. CMSA Implementation

213

(@TRUE= TechnicallyFeasible @INFCAT=3; ©COMMENTS = "If the driving conditions is soft, pile driving pattern was In_pairs, and combination is not technically feasible then reset the pile driving pattern status into In_Singles. This requires setting the hypothesis SPProduction to Unknown to reevaluate the technical feasibility with the new strategy; (@LHS = (No (Technical Feasibility.State)) (Is (DrivingConditions) ("Soft")) (Is (Piles_Driving_Pattern) ("InPairs")) (Show ("Tech_Fea.txt") (©KEEP=FALSE;@WAIT=TRUE;)) ) (@HYPO= CheckFeasibility) (@RHS = (DO ("InSingles") (Piles_Driving_Pattern)) (Reset (SP_Production)) (Do (SPProduction) (SPProduction)) )

Figure 5.25 The

first

technical diagnosis.

Technical F e a s i b i l i t y Diagnostic Rule

condition

feasibility

checks

to see i f the state

i s False

i n order

The second condition

of

to perform

ensures that

a

the d r i v i n g

condition i s " s o f t " which commonly suggests the use of the "In_Pairs" d r i v i n g strategy. checks the v a r i a b l e

The t h i r d condition e x p l i c i t l y

"Pile_Driving_Conditions"

value.

The

"Show" operator displays the text f i l e of "Tech_Fea.txt" to explain the process to the user. The

hypothesis

"Check_Feasibility"

i s the same as the

previous one with a d i f f e r e n t rule (in NExpert, a hypothesis embraces one or more rules) .

In order to f i r e

diagnostic

category

figure

rule,

i t s inference

5.25) i s set to three

the f i r s t

(QINFCAT shown i n

which i s NExpert's

priority

Chapter 5. CMSA Implementation

mechanism that

fires

214

rules within

the same hypothesis i n

ascending order of t h e i r inference category. To continue, and

that

driving

assume t h i s rule's conditions were s a t i s f i e d

the system strategy

control

from

d r i v i n g "In_Singles". SP_Production strategy routine

driving

altered

the p i l e

"In_Pairs"

to

pile

Then, the "Reset" operator resets the

hypothesis

backtracks with

pile

strategy

to Unknown

and re-runs

changed

input

such

that

control

the "Drive.c"

variables.

numerical

The "Do"

operator

then forces the re-evaluation of the Unknown hypothesis.

5.3.5

CMSA Chaining and Reasoning (Control

Strategy)

In t h i s section, we explain the CMSA chaining and reasoning features

f o r each

major

operator:

Design,

Predict and

Analyze. Currently, CMSA i s implemented using backward chaining i n the

NExpert

solution

Object

sense

propagates

in

of chaining a

forward

definitions. reasoning

Its

fashion.

NExpert, o r i g i n a l l y a r u l e based system, i s a hybrid system which makes i t n o n - t r i v i a l to r a t i o n a l i z e i t s chaining and reasoning

approach

with

conventional

instance,

technical f e a s i b i l i t y ,

employs

a

trial

Design/Predict/Analysis attribute

value,

and cycle

terminology.

For

covered i n section

5.3.4,

error

procedure

i n order

i . e . a design

for

to modify a

alternative,

the

design

construction

Chapter 5. CMSA Implementation

resource,

215

or construction

strategy.

This

facility

is

l a b e l l e d as non-monotonic reasoning i n NExpert Object, i . e . , making assumptions and r e t r a c t i o n s . What

follows

is a

description

reasoning process i n CMSA. of

hypotheses

—

of the chaining

Figure 5.26 shows a CMSA network

rules grouped

i n categories

connected f o r chaining and inference propagation Shown i n t h i s the

methods

and

and i n t e r purposes.

figure are control strategy clauses used i n synthesis

process.

CMSA

accomplishes the

control strategy task using the RHS "Do" operator a f t e r an hypothesis has proved to be true. The

Do operator

performs two operations.

I t triggers a

s p e c i f i c knowledge evaluation and then passes control to the next operation.

For example, the following clause from the

Design box of figure 5.26,

Do

(Select_Pile Driver)

i s extracted

(Select P i l e Driver)

from figure 5.20, and i s interpreted as f i r e

the (Select_Pile_Driver) hypothesis.

I t i s triggered i f the

previous control clause, extracted from figure 5.12,

Do

(Select SSP)

(Select SSP)

216

Chapter 5. CMSA Implementation

i s successfully f i r e d —

i . e . select s t e e l sheet p i l e .

Once

the p i l e d r i v e r selection i s done successfully, then control passes

t o applying the hammer dynamic formula

the hypothesis

embedded i n

(SP_Production), i . e . predict a productivity

rate and check i f the p i l e reaches i t s r e f u s a l depth.

The

clause that t r i g g e r s t h i s hypothesis i s found i n the Predict box i n figure 5.26, i . e . Do

(SP_Production)

(SP_Production)

CMSA s t a r t s with the Suggest operator which currently i s a surrogate f o r the preliminary f e a s i b i l i t y knowledge base. The user i s prompted with a choice of GWSS a l t e r n a t i v e s . The Design task follows r i s k evaluation i f the l e v e l of risk

i s acceptable.

The design

element

forward reasoning and chaining modes. task i s more complicated. profile.

exhibit

The hammer s e l e c t i o n

The hammer type i s based on s o i l

Hammer size i s based

on compatibility conditions

with the design elements of s t e e l and goal requirements,

subtasks

sheet and s o l d i e r

(e.g. production rate),

piles

although i t

i s assumed with CMSA that the maximum energy hammer w i l l be the default

choice.

Thus hammer s e l e c t i o n

modes of chaining with

a forward

involves mixed

mode of reasoning.

By

contrast, " R l " an expert system that configures VAX computer systems, exhibits forward (McDermott 1984).

chaining with backward reasoning

Chapter 5. CMSA Implementation

Figure 5.26

CMSA Model of Chaining and Reasoning

218

Chapter 5. CMSA Implementation

Once

operator

method

Predict

operator

selects

attributes a

are generated,

suitable

method

the

(procedural

routines) for measuring the method performance for the given project

context.

Next,

the Analyze

operator

applies

analysis and/or i n t e r p r e t a t i o n routines t o the r e s u l t s from the Predict operator.

This, i n general

involves backward

modes of chaining and reasoning. During

the Predict/Analyze

i s based on minimum costs.

process, Further,

a l t e r n a t i v e ranking the Analyze

operator

may e x h i b i t non-monotonic reasoning by making and r e t r a c t i n g assumptions.

For instance, i f the p i l e d r i v i n g construction

strategy state of In_Pairs i s "True", but the method d i d not s a t i s f y a goal c r i t e r i o n ,

(say production

r a t e ) , or method

elements compatibility resulted i n a f a i l e d

solution, the

o r i g i n a l plan could be altered by the CMSA control structure by r e t r a c t i n g the p i l e d r i v i n g strategy state of In_Pairs to be "False" and assessing the value "True" to the In_Singles state

value.

Thus,

a mix forward

mode of chaining and

reasoning are used for the Analyze operator.

6. The Prototype Example

6.1 Introduction Features of the prototype CMSA implemented are described i n this

chapter.

Input/output

data,

solution

strategy

processing, s e n s i t i v i t y of decisions to input changes, and the explanation

facility

are described

first.

An example

problem i s demonstrated using a step by step approach. example consists of two Cut-and-Cover parallel —

methods i n

s t e e l sheet p i l e s and s o l d i e r p i l e s and lagging.

The second part of t h i s chapter example

shoring

The

of the r i s k

assessment

i s devoted t o a d e t a i l e d process,

implemented i n

NExpert Object as an independent module.

6.2 Example Problem Description The

example problem i s a proposed tunnel, 1000 f t long, 60

f t deep and 20 f t wide. layers: below

The s o i l

p r o f i l e consists of two

a 40 f t top layer of loose

that.

The contract

sand, and s t i f f

duration

for this

clay

project i s

estimated to be a maximum of 240 days (or 5 ft/day), with a unit cost of $2,800 per foot ± $500.

I t i s assumed that

upper and lower bounds for u n i t costs and production

rates

are given for the Cut-and-Cover tunnel a l t e r n a t i v e s . A record of a session i s provided, CMSA

system

responses.

Example

with

screens

user

input and

are shown as

Chapter 6. Prototype Example

appropriate. facilities,

Due

220

to the lack

explanation

files

of NExpert's

with an extension

explanation " t x t " were

used extensively to explain some of the CMSA operations and query processes to the user.

6.2.1

Session Start

A session commences by suggesting either a "datum" f o r input variables or suggesting screen 6.1.

a NExpert "hypothesis"

as shown i n

This window i s invoked by the command "Suggest"

followed by the command "Knowcess" from the Expert Command Menu.

From t h i s window, any hypothesis could be highlighted

and put i n the Suggest/Keep corner before NExpert s t a r t s the session,

e.g., the hypothesis

t r i g g e r the CMSA session. any datum

"Select_A_GWSS"

Also t h i s window i s used to l i s t

(a premise i n a rule) and t r i g g e r s the beginning

of the session by f i r s t evaluating the premise to the hypothesis). rule.

i s used to

Once

a

(as opposed

Control i s then passed to evaluate the

hypothesis

or

datum

i s placed

i n the

Suggest/Keep corner, "OK Knowcess" command i s selected, the session s t a r t s .

Chapter 6. Prototype Example

Screen 6.2

CMSA Overview Rule Network window

221

Chapter 6. Prototype Example

222

IRULE NETWOR

^

7 Name 10 (1) a Select_GWSS_of_SPL

(114)

9 Show "Textl .txt" @KEE JB(1)

GWSS Is "Soldiet_P

9 =>Do (1) Select_GWSS Name 10 (1) a [1] Select_GWSS_of_SSP

(113)

Show "Textl txt" @KEE (1) GWSS Is "Steel_Sh« =>Do (1) Select_GWSS

Screen 6.3 After

firing

CMSA Rule Network Window

"Select_A_GWSS",

a

text

file

(CMSA.txt)

appears on the screen, introducing the user to the Cut-andCover tunnelling problem. chaining" mode. 6.2

Thus, CMSA i s set i n a "backward-

For the loaded CMSA knowledge base, screen

shows the o v e r a l l rule network using the Overview Rule

Network rules

(ORN) that

window. are

fired

The highlighted successfully

NExpert Command Menus are shown ( F i l e ,

branches represent

during

the

session.

Edit, Expert, etc.)

i n the NExpert Environment Screen. By zooming i n on the dotted box of ORN, the "Rule Network Window" focuses v i s u a l l y on a i n d i v i d u a l rule or a group of rules i n the CMSA knowledge base, as shown i n screen The Rule Network Window (RNW)

6.3.

i s enlarged to the s i z e of the

Chapter 6. Prototype Example

screen.

223

In the upper r i g h t hand side of the s l i d e , the Rule

Network Overview (RNO) i s shown.

In the middle of the RNW,

the r u l e "Select_GWSS_of_SSP" i s shown f i r e d

(a condition or

hypothesis i s indicated by the following icons:

"True" by a

check mark, "False" by a highlighted check mark, by

a question

"Unknown"

mark, "Not Known" by an empty box, "Being

Currently Investigated:" by a target, and "Being Evoked:" by an a s t e r i s k . Continuing with the session, CMSA provides the user with the

following

p o t e n t i a l f e a s i b l e a l t e r n a t i v e s as shown i n

screen 6 . 4 . »

Select A GWSS ?

The

1.

Steel Sheet P i l e (SSP).

2.

Soldier P i l e and Lagging (SPL). system asks the user to choose one GWSS a l t e r n a t i v e

to s t a r t the detailed KB part.

For example purposes, the

two a l t e r n a t i v e s of Steel Sheet P i l e s and Soldier P i l e s and Lagging screening

are assumed

to have

survived

the preliminary

process.

Assume the user chooses the s t e e l sheet p i l e a l t e r n a t i v e . The

next part

context.

of the session

involves

specifying the s o i l

Chapter 6. Prototype Example

6.2.2

224

Problem Context S p e c i f i c a t i o n

What i s the Number of S o i l L a y e r s ? 1. One 2. Two »

TWO

The user i s prompted f o r either a single s o i l layer or a two s o i l layer scenario. i n screen

The l a t t e r has been selected as shown

6 . 5 . .

CMSA asks f o r the s o i l type f o r the top layer as shown i n screen 6.6.

A l t e r n a t i v e l y , a second input format i s based

on a Standard Penetration Test (SPT) P r o f i l e . If the f i r s t input format i s adopted Loose_Sand as the top layer. the

, assume you select

Screen 6.7 shows the input f o r

40 foot loose sand layer depth.

Screen 6.8 shows that

s t i f f clay i s chosen as the second s o i l

layer.

Screen 6.9

shows the water l e v e l input. Other queries include the tunnel depth i n feet the

(60), and

tunnel length i n feet (1000).

This input format i s selected f o r s i m p l i c i t y . I f the input f o r the SPT p r o f i l e i s selected, the user w i l l be required to input SPT readings at 5 foot i n t e r v a l s f o r the depth of the tunnel or excavation.

225

Chapter 6. Prototype Example

SESSION CONTROL What i s t h ^ Value o f GWSS ?

Screen 6.4

GWSS Feasible A l t e r n a t i v e s

SESSION CONTROL! What i s t h e Ualue o F Number_of_Soil_Layers. ? TuoSoilLayers OneSoilLayer

Screen 6.5

Soil Profile Specification

(1)

SESSION CONTROL What i s - t h e Ualue» of- SoiI_Type_1 ?

Hoose-Saridl

OK

G

Dense_Sand

m

S o f t Clay S t i f f Clau

~ NO TK NO UN'

Screen 6.6

Soil Profile Specification ( 2 )

Chapter 6. Prototype Example

Screen 6.7

(3)

Soil Profile Specification

gsESSION CONTROL] What i s the Ualue of S o i l T y p *> 2 ?

•fstiff ClauB

OK

DenseSand Soft Clay

k

NOTKNOWN

Screen 6.8

S o i l P r o f i l e S p e c i f i c a t i o n (4)

•SESSION C O N T R O L H H B H B I What i s the depth of Water Table ? OK 20;

Screen 6.9

Water Table Level Input

Chapter 6. Prototype Example

GWSS

227

Technical F e a s i b i l i t y Assessment:

The CMSA System then

processes the following operations: 1.

2.

i? Calculate pressure and moments ': A rule i s used to do the computations f o r t h i s scenario; see Appendix A f o r the computation procedures. The hypothesis "Calculate Pressures and Moments f o r "Two_Soil_Layers Scenario" i s fired. -1

Select a suitable sheet p i l e i f i t e x i s t s with the given data base of "ssp.nxp" (see screen 6.10). The following operations are c a r r i e d out by CMSA. Retrieve Steel Sheet P i l e database "ssp.nxp"; F i r e Section Modulus Rule which incorporates moments based on Retaining System Spacing; Select PZ_27;

ASTM

Steel

Sheet

Pile

Section of

Attach Selected PZ_27 to Matched_SSP; Inherit PZ_27 properties t o Matched_SSP; I f t h i s f i r e s successfully, then pass control to Select P i l e Driver.

Pressure and moments calculations are based on the "default" spacing of the retaining system (15 f t v e r t i c a l l y and 12 f t h o r i z o n t a l l y ) .

Chapter 6. Prototype Example

228

9 (1) Section_Modulus > = Ki') 9

9Bf1 ] Select_SSP_of_PSA_32 (13S

SectionModulus < .desic

9 =>Do 9 =>CieateOb|ect Do (1) Select_Pile_D

I-

(1) Section_Modulus > =

I STOP J

(1) SectionModulus < .desic II] Select_SSP_of_PZ_27 (138)

Show "hammei.txt* @KI =>Do =>CreateObject Do (1) Select_Pile_D

Screen 6.10

Hypothesis "Select_Suitable_Sheet_Pile" i s F i r e d Successfully

After

successfully

(SSP.nxp), Hammer"

control

selecting

i s passed

hypothesis.

After

a suitable

SSP data base

to the "Select a

pile

a

section

Suitable has

been

successfully selected, the system selects a suitable hammer as

follows.

Based

upon

the s o i l

stratification

sequencing, a generic type of hammer i s selected. example, the impact vibratory one.

"Select

the type Pile

For t h i s

i s more s u i t a b l e

than a

A Double Acting A i r Hammer (DAAH) i s found

to be an appropriate Once

hammer type

and

impact hammer.

of hammer

Driver"

i s specified

(the hypothesis

i s fired) , i t i s then

sized

using

Chapter 6. Prototype Example

heuristic

rules

229

(the

"Selec^Vibratory^rJDAAH^amme^PD" 6.11).

hypothesis

i s fired,

see screen

The CMSA proceeds as follows: Retrieve DAAH database "DAAH.nxp"; Based on experiential rules, the DAAH i s sized i n terms of "Rated Delivered Energy". I f the search i s not successful and no such s i z e exists within the database, a message appears offering the alternatives of abandoning the search and q u i t t i n g ; selecting another hammer; or selecting another shoring a l t e r n a t i v e . The default strategy i s to pick the hammer with the highest delivered energy; If a DAAH hammer of the required s i z e i s found, i t will be attached to the "Matched^ammer" object and i t s properties (model, delivered energy, unit cost, etc) are inherited from the "Selected_Hammer" l i s t . Cost i s not considered as part of the c r i t e r i o n at t h i s stage; Control strategy

The

i s then passed knowledge base.

to

the

driving

selected hammers class and dynamic objects created

a f t e r f i r i n g the above rule are shown i n screen 6.12.

Chapter 6. Prototype Example

(1) Soil_Type_1 Is "Loo (1) Soil_Type_2 Is "Sofi Retrieve 'dan nxp' @T" Name MAX(Let (1) Pile_D nver "D =>Let (1J Harnmei_Type =>CieateObjecl DeleteOh|t:i:t Do .driving_width)) (NumberofSSPUnitWidth)) (Do (Number_of_SSP_Umt_Width*Nujnber_of_PUesj)er_SSP_Umt_Width) (Number_of_Piles)) (Do(Number_of_SSP_Unit_Width*Total_Dr^^^ (Total_Production_Time_in_Days)) (Do (NumberofPiles/TotalProductionJTime_in_Days) (Productivity_in_Number_of_Piles_Per_Day)) (Do (CEIL(NumberofPiles)) (Number_of_Piles)) (Do (hypo) (hypo))

)

)

(@RULE = Production_Measures_of_SSP_for_Driving_Single_Pile ©COMMENTS = "Total time consists of two components: variable component which depends on soil/pile/hammer scenario, the fixed which is dependant on pile welding and positioning, and movement of pile driver. Theoretically the number of sheets used is the same, however, the delivery could be "In-Pairs" requiring lessfixedtime for the crane. The computation of the total driving time is adjusted by identifying PileDrivingConditions state (InPairs, In_Singles). For the former, driving sheets in pairs, the total driving variable time is perhaps less than the latter, however, the total drivingfixedtime (as number of the setups) is less than the former. The "PLength" text file explains how pile segment length effects the total driving time; ©WHY="This is to measure the productivity performance of sheet pile and a given pile driver."; (@LHS = (Is (Pile_Driving_Conditions) ("InPairs")) (Name (Fixed_Time_per_given_Hammer) (Fixed_Time_per_Pair_of_piles.amount)

) (Show ("PJLength") (©KEEP = FALSE;@WAIT = TRUE;)) (Name (CEIL(Tunnel.depth/SSP.StandardJL£ngth)*Fked_Timejer_Pairs_of_piles.amount) (Fixed_Time_per_Pair_of_SSP_Unit_Width.amount)) (Name (CEIL(Tunnel.depth/SSP.Standard_Length)*2) (Number_of_Piles_per_Pair_of_SSP_Unit_Width)) (Name (Fixed_Time_per_Pair_of_SSP_Unit_Width.amount+VariableTime.amount) ( Total_Driving_Time_per_Pair_of_SSP_Unit_Width)) ) (©HYPO = ComputeProductionofSSP) (@RHS = (Do (2*12*(Tunnel.Length)/(< | Selected_SSP | > .drivingwidth)) (Number_of_Pair_of_SSP_Unit_Width))

L i s t i n g D.l P a r t i a l L i s t i n g of CMSA (continued)

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

318

(Do(Number_of_Pair_of_SSP_Urut_Width*Numte ofSSPJJrntWidth) (Number_of_Pairs_of_PUes)) (Do (Number_of_Pair_of_SSP_Umt_W Pair_of_SSP_Unit_Width/(\8*60)) (TotalProductionTimeinDays)) (Do (2*Niunber_of_Pair_of_PUes/Total_Production_Tmie_in_Days) (ProductivitymNumberofPuesPerDay)) (Do (CEIL(Number_of_Pair_of_Piles)) (Number_of_Piles)) (Do (hypo) (hypo))

) (@RULE = WriteDesiredOutputResults (@LHS = (Name (1) (a)) ) (@HYPO= hypo) (@RHS = (Write ("Results.nxp") (@TYPE = NXP;@FILL=NEW;@UNKNOWN = TRUE;@ATOMS = Techmcal_FeasibUity,\Number_of_PUes,Total_Production_Time_in_Days,\ Productivity_in_Number_of_Piles_Per_Day,Selected_Steel_Pile,\ Pile_Driver,Hammer_Model;)) (Show ("Results.nxp") (@KEEP = FALSE;@WAIT=TRUE;)) \(Do(Done) (Done))

)

)

(@RULE= Select_GWSS_of_SSP (@LHS = (Name (10) (a)) (Show ("Textl-txt") (@KEEP=FALSE;@WAIT=TRUE;)) (Is (GWSS) ("Steel_Sheet_Pile"))

) (@HYPO= SelectAGWSS) (@RHS = (Do (Select_GWSS_of_SSP)

(Select_GWSS_of_SSP))

) ) (@RULE= SelectGWSSofSPL (@LHS = (Name (10) (a)) (Show ("Textl.txt") (@KEEP=FALSE;@WAIT=TRUE;)) (Is (GWSS) ("Soldier_Piles_and_Lagging"))

)

L i s t i n g D.l P a r t i a l L i s t i n g of CMSA (continued)

Appendix D.

)

CMSA P a r t i a l L i s t i n g and Miscellany

(@HYPO= SelectAGWSS) (@RHS = (Do (SelectGWSSofSPL) )

(SelectGWSSofSPL))

(@RULE= Soldier_Piles_is_Selected (@LHS= (Name (10) (a)) (Is (Number_of_Soil_Layers) ("Two_Soil_Layers"))

) (@HYPO= SelectGWSSofSPL) (@RHS = (Do (Calculate_Pressure_and_Moments_for_Two_Soil_Layers_of_SPL) (CalculatePressureandMomentsfor_Two_Soil_Layers_of_SPL)) ) (@RULE= Soldier_Piles_is_Selected (@LHS = (Name (10) (a)) (Is (NumberofSoilLayers) ("One_Soil_Layer")) ) (@HYPO= SelectGWSSofSPL) (@RHS = (Do (Calcmate_Pressure_and_Moments_for_Single_Soil_Layer_of_SPL) (Calculate_Pressure_and_Moments_for_Single_Soil_Layer_of_SPL))

)

)

(@RULE = Steel_Sheet_Pae_is_Selected (@LHS = (Name (10) (a)) (Is (Number_of_Soil_Layers) ("One_Soil_Layer")) ) (@HYPO= Select_GWSS_of_SSP) (@RHS = (Do(Cal_Pres_and_Mom_for_SSL_of_SSP) (Cal_Pre_and_Mom_for_SSL_of_SSP))

)

)

L i s t i n g D.l P a r t i a l L i s t i n g of CMSA (continued)

319

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

320

(@RULE = SteelSheetPileisSelected (@LHS = (Name (10) (a)) (Is (NumberofSoilLayers) ("TwoSoilLayers"))

) (@HYPO= SelectGWSSofSSP) (@RHS = (Do(Cal_Pres_and_Mom_for_TSL_of_SSP) (Cal_Pres_and_Mom_for_TSL_of_SSP))

) ) (@RULE = SelectVibratoryPDorDAHammerPDforVeryLooseSand @INFCAT=0; ©COMMENTS="Select type of pile driver (mainly hammers) based on soil conditions (heuristics found in Hunt 1979)"; (@LHS = (Is (SoiLtype) ("Very_Loose_Sand")) ) (@HYPO= SelectPileDriver) (@RHS = (Let (Pile_Driver) ("Vibratory")) (Do (RetHammerEnergy) (Ret_Hammer_Energy))

)

)

(@RULE = Select_Vibratory_PD_or_DA_Hammer_PD_for_Very_Loose_Sand ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics found on Hunt 1979)"; (@LHS = (Is (Soil.type) ("Very_Loose_Sand"))

) (@HYPO= Select_Pile_Driver) (@RHS = (Let (PileDriver) ("DoubleActingHammer")) (Do (RetHammerEnergy) (Ret_Hammer_Energy)) )

) (@RULE = Select_Vibratory_PD_or_DA_Hammer_PD_for_Very_Loose_Sand @INFCAT=0; ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics found in Hunt 1979)";@WHY="Inference category (INFCAT = 0) is set to 0 in order to set a low priority for selecting a vibratory pile driver as opposed to impact hammers."; (@LHS =

L i s t i n g D.l

P a r t i a l L i s t i n g of CMSA (continued)

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

321

(Is (Soil.type) ("Very_Loose_Sand")) ) (@HYPO= Select_Pile_Driver) (@RHS = (Let (Pile_Driver) ("Vibratory")) (Do (RetHammerEnergy) (Ret_Hammer_Energy)) ) (@RULE = Select_Vibratory_PD_or_DA_Hammer_PD_for_Very_Dense_Sand ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics found in Hunt 1979)"; (@LHS = (Is (SoiLtype) ("Very_Dense_Sand"))

) (@HYPO= SelectPileDriver) (@RHS = (Let (PileDriver) ("Double_Acting_Hammer")) (Do (RetHammerEnergy) (RetHammerEnergy))

) (@RULE = Select_Vibratory_PD_or_DA_Hammer_PD_for_Very_Dense_Sand @INFCAT = 0; ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics found in Hunt 1979)"; (@LHS = (Is (SoiLtype) ("Very_Dense_Sand"))

) (@HYPO= SelectPileDriver) (@RHS = (Let (PileDriver) ("Vibratory")) (Do (RetHammerEnergy) (RetHammerEnergy)) )

)

(@RULE = Select_Vibratory_PD_or_DA_Hammer_PD_for_Medium_Sand ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics found in Hunt 1979)"; (@LHS = (Is (SoiLtype) ("Medium_Sand"))

) (@HYPO= SelectPileDriver) (@RHS = (Let (Pile_Driver) ("Double_Acting_Hammer"))

L i s t i n g D.l

P a r t i a l L i s t i n g of CMSA (continued)

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

322

(Do (RetHammerEnergy) (RetHammerEnergy)) ) ) (@RULE = Select_Vibratory_PD_or_DA_Hammer_PD_for_Medium_Sand @INFCAT=0; ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics based on Hunt, Hall)"; (@LHS = (Is (SoiLtype) ("MediumSand"))

) (@HYPO= SelectJPile_Driver) (@RHS = (Let (Pile_Driver) ("Vibratory")) (Do (RetHammerEnergy) (Ret_Hammer_Energy))

) (@RULE = Select_Vibratory_PD_or_DA_Hammer_PD_for_Loose_Sand ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics based on Hunt, Hall)";@WHY = "Inference category is set to 1 for DA-hammer which indicates that it has higher priority than vibratory pile driver"; (@LHS = (Is (SoiLtype) ("Loose_Sand")) (Retrieve ("daah.nxp") (@TYPE=NXP;@FILL=ADD;@CREATE = | SelectedHammer | ;\

)) (Name (MAX( < | Selected_Hammer | > .Theor_Energy)) (Max_Energy)) (= (< | Selected_Hammer | > .Theor_Energy-Max_Energy) (0))

) (@HYPO= SelectPileDriver) (@RHS = (Let (PileDriver) ("DoubleActingHammer")) (Let (HammerType) ("SASH")) (CreateObject ( < | SelectedHammer | > ) (| Suitable_Hammer |)) (DeleteObject ( < | Selected_Hammer | > ) (| Selected_Hammer |)) (Do (< | SuitableHammer | > .HammerModel) (Hammer_Model)) (Do (SPProduction) (SPProduction)) )

) (@RULE = Select_Vibratory_PD_or_DA_Hammer_PD_for_Dense_Sand @INFCAT = 0; ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics based on Hunt, Hall)";

L i s t i n g D.l

P a r t i a l L i s t i n g of CMSA (continued)

Appendix D.

)

CMSA P a r t i a l L i s t i n g and Miscellany

323

(@LHS = (Is (SoiLtype) ("Dense_Sand")) ) (@HYPO= SelectPileDriver) (@RHS = (Let (Pile_Driver) ("Vibratory")) (Do (Ret_Hammer_Energy) (Ret_Hammer_Energy)) )

(@RULE = Select_Vibratory_PD_or_DA_Hammer_PD_for_Dense_Sand ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics based on Hunt, Hall)"; (@LHS = (Is (SoiLtype) ("Dense_Sand"))

) (@HYPO= SelectPileDriver) (@RHS = (Let (PileDriver) ("Double_Acting_Hammer")) (Do (Ret_Hammer_Energy) (RetHammerEnergy))

) (@RULE = Select_Vibratory_PD_for_Very_Stiff_Clay ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics based on Hunt, Hall)"; (@LHS = (Is (SoiLtype) ("Very_Stiff_Clay"))

) (@HYPO= Select_Pile_Driver) (@RHS = (Let (PileDriver) ("Single_Acting_Hammer")) (Do (Ret_Hammer_Energy) (Ret_Hammer_Energy))

)

)

(©RULE = Select_Vibratory_PD_for_Very_Soft_Clay @INFCAT=0; ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics based on Hunt, Hall)"; (@LHS = (Is (SoiLtype) ("Very_Soft_Clay")) )

L i s t i n g D.l

P a r t i a l L i s t i n g of CMSA (continued)

Appendix D.

)

CMSA P a r t i a l L i s t i n g and Miscellany

324

(@HYPO= SelectPileDriver) (@RHS = (Let (Pile_Driver) ("Vibratory")) (Do (RetHammerEnergy) (RetHammerEnergy)) )

(@RULE= Select_Vibratory_PD_for_Stiff_Clay ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics based on Hunt, Hall)"; (@LHS = (Is (Soil.type) ("StiffClay"))

) (@HYPO= SelectJ>ile_Driver) (@RHS = (Let (PileDriver) ("DoubleActingHammer")) (Do (RetHammerEnergy) (Ret_Hammer_Energy)) ) (@RULE = Select_Vibratory_PD_for_Medium_Clay ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics based on Hunt, Hall)"; (@LHS = (Is (Soil.type) ("Medium_Clay"))

) (@HYPO= SelectPileDriver) (@RHS = (Let (PileDriver) ("DoubleActingHammer")) (Do (Ret_Hammer_Energy) (RetHammerEnergy))

)

) (@RULE = Select_Vibratory_PD_for_Medium_Clay @INFCAT = 0; ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics based on Hunt, Hall)"; (@LHS = (Is (SoiLtype) ("Medium_Clay")) ) (@HYPO= SelectPileDriver)

L i s t i n g D.l

P a r t i a l L i s t i n g of CMSA (continued)

Appendix D. CMSA P a r t i a l L i s t i n g and Miscellany

325

(@RHS = (Let (Pile_Driver) ("Vibratory")) (Do (RetHammerEnergy) (Ret_Hammer_Energy)) ) ) (@RULE = Select_Vibratory_PD_for_Hard_Clay ©COMMENTS = "Select type of pile driver (mainly hammers) based on soil conditions (heuristics based on Hunt, Hall)"; (@LHS = (Is (Soil.type) ("HardClay"))

) (@HYPO= SelectPileDriver) (@RHS = (Let (PileDriver) ("Single_Acting_Hammer")) (Do (RetHammerEnergy) (RetHammerEnergy))

)

)

(@RULE= Select_SSP_of_PZ_38 ©COMMENTS = "Select SSP designation of PZ_38, Section_Modulus is in in 3";@WHY="This is used to determine the type of sheet pile from which other properties can be deduced from the SSP data base (SSP.NXP) for further treatment."; (@LHS = (>= (SectionModulus) (38.3)) (< (SectionModulus) (46.8)) (Is ( < | Selected_SSP | > .designation) ("PZ_38")) A

) (@HYPO= SelectSSP) (@RHS = (Do ( < | SelectedSSP | > .designation) (Selected_Steel_Pile)) (Do (SelectPileDriver) (SelectPileDriver))

) ) (@RULE= Select_SSP_of_PZ_32 ©COMMENTS = "Select SSP designation of PZ_38, Section_Modulus is in in~3";@WHY="This is used to determine the type of sheet pile from which other properties can be deduced from the SSP data base (SSP.NXP) for further treatment."; (@LHS = (>= (SectionModulus) (30.2)) (< (SectionModulus) (38.3)) (Is (< |Selected_SSP| > .designation) ("PZ_32"))

) (@HYPO= SelectSSP)

L i s t i n g D.l

P a r t i a l L i s t i n g of CMSA (continued)

Appendix D. CMSA P a r t i a l L i s t i n g and Miscellany

326

(@RHS = (Do (< |Selected_SSP| > .designation) (SelectedSteelPile)) (Do (SelectPileDriver) (Select_Pile_Driver)) )

) (@RULE= Select_SSP_of_PZ_27 ©COMMENTS = "Select SSP designation of PZ_38, SectionModulus is in in 3";@WHY="This is used to determine the type of sheet pile from which other properties can be deduced from the SSP data base (SSP.NXP) for further treatment."; (@LHS = (>= (Section_Modulus) (10.7)) (< (SectionModulus) (30.2)) (Is ( < | Selected_SSP | > .designation) ("PZ_27")) A

) (@HYPO= SelectSSP) (@RHS = (Do (< |Selected_SSP| > .designation) (Selected_Steel_Pile)) (CreateObject ( < | Selected_SSP | > ) (| Matched_SSP |)) (Do (SelectPileDriver) (Select_Pile_Driver))

) (@RULE= Select_SSP_of_PSA_32 ©COMMENTS = "Select SSP designation of PZ_38, SectionModulus is in in 3";@WHY="This is used to determine the type of sheet pile from which other properties can be deduced from the SSP data base (SSP.NXP) for further treatment."; (@LHS = (>= (SectionModulus) (1.9)) (< (SectionModulus) (2.4)) (Is (< |Selected_SSP| > .designation) ("PSA32")) /v

) (@HYPO= SelectSSP) (@RHS = (Do ( < | Selected_SSP | > .designation) (Selected_Steel_Pile)) (CreateObject (< |Selected_SSP| >) (|Matched_SSP|)) (Do (SelectPileDriver) (SelectPileDriver))

) ) (@RULE= Select_SSP_of_PSA_28 ©COMMENTS = "Select SSP designation of PZ_38, Section_Modulus is in in"3";@WHY = "This is used to determine the type of sheet pile from which other properties can be deduced from the SSP data base (SSP.NXP) for further treatment.";

L i s t i n g D.l

P a r t i a l L i s t i n g of CMSA (continued)

Appendix D. CMSA P a r t i a l L i s t i n g and Miscellany

327

(@LHS = (>= (Section_Modulus) (2.4)) (< (Section_Modulus) (2.5)) (Is (< | SelectedSSP | >.designation) ("PSA28")) ) (@HYPO= SelectSSP) (@RHS = (Do (< |Selected_SSP| > .designation) (SelectedSteelPile)) (CreateObject ( < | SelectedSSP | > ) (| MatchedSSP |)) (Do (SelectPileDriver) (SelectPileDriver)) ) ) (@RULE= SelectJSSP_of_PMA_22 ©COMMENTS = "Select SSP designation of PZ38, Section_Modulus is in in 3";@WHY="This is used to determine the type of sheet pile from which other properties can be deduced from the SSP data base (SSP.NXP) for further treatment."; (@LHS = (>= (Section_Modulus) (2.5)) (< (Section_Modulus) (5.4)) (Is (< |Selected_SSP| > .designation) ("PMA22")) /v

) (@HYPO= SelectSSP) (@RHS = (Do (< |Selected_SSP| > .designation) (Selected_Steel_Pile)) (CreateObject ( < | Selected_SSP | > ) (| Matched_SSP |)) (Do (SelectPileDriver) (SelectPileDriver))

)

)

(@RULE= Select_SSP_of_PMA22 ©COMMENTS = "Select SSP designation of PZ38, Section_Modulus is in in^3";@WHY="This is used to determine the type of sheet pile from which other properties can be deduced from the SSP data base (SSP.NXP) for further treatment."; (@LHS = (>= (Section_Modulus) (2.5)) (< (Section_Modulus) (5.4))

) (@HYPO= Select_SSP) (@RHS = (Let (SSP) ("PMA22")) (Do (SelectPileDriver) (SelectPileDriver))

) )

L i s t i n g D.l

P a r t i a l L i s t i n g of CMSA (continued)

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

328

(@RULE= Select_SSP_of_PDA27 ©COMMENTS = "Select SSP designation of PZ38, SectionModulus is in in 3";@WHY="This is used to determine the type of sheet pile from which other properties can be deduced from the SSP data base (SSP.NXP) for further treatment."; (@LHS = (>= (SectionModulus) (5.4)) (< (SectionModulus) (10.7)) /v

) (@HYPO= SelectSSP) (@RHS = (Let (SSP) ("PDA27")) (Do (SelectJPileJDriver) (Select_Pile_Driver)) ) (©RULE = Single_Pile_Variable_Production_Time ©COMMENTS="This rule computes the variable driving time for a pile series under soft driving conditions and "InSingles" pile driving patters. The rest of the rule premisses assign variables to selected hammer and pile properties. These variables are then written to the input files "Hammer.nxp" and "Soil.nxp" for the Drive.c routine (see chapter 5)."; @WHY="This rule assumes that soil conditions state is at best and thus pile driving time will be shorter than other soil scenario."; (@LHS = (Name (Tunnel.depth+5) (L)) (Show ("Drive.txt") (©KEEP = FALSE;@WAIT=TRUE;)) (Is (DrivingConditions) ("Soft")) (Is (PilesDrivingPattern) ("InSingles")) (Name (< | SuitableHammer | > .Strokes_per_Min) (F)) (Name (< |Suitable_Hammer| >.Theor_Energy*Hammer.Total_Efficiency) (E)) (Name ( < | Selected_SSP | > .Surface_Area) (SA)) (Name ( < j Selected_SSP j > .Cross_Section_Area) (Ap))

) (@HYPO= SPProduction) (@RHS = (Write ("Hammer.nxp") (©TYPE=NXP;@FILL=NEW;@ATOMS=L,F,E,SA,Hammer_Type;\ )) (Write ("soil.nxp") (©TYPE = NXP;@FILL=NEW;@UNKNOWN=TRUE;@ATOMS = SoilTypel A Start_l,Finish_l,Soil_Type_2,Start_2,Finish_2;\ )) (Execute ("Drive.exe") (©TYPE = EXE;@WAIT = TRUE;)) (Retrieve ("out.nxp") (©TYPE = NXP;@FILL=ADD;@FWRD = TRUE;@CREATE = | VarTime | ;\

L i s t i n g D.l

P a r t i a l L i s t i n g of CMSA (continued)

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

©ATOMS = VariableTime.amount;)) (Retrieve ("out.nxp") (©TYPE=NXP;@FILL=ADD;@FWRD=TRUE;@CREATE = | Feasibility | ;\

©ATOMS=Technical_Feasibility.State;)) (Do (CheckFeasibility) (Check_Feasibility))

)

)

(©RULE = Single_Pile_Variable_Production_Time ©COMMENTS = "This rule computes the variable driving time component for a pile series."; (@LHS = (Name (Tunnel.depth+5) (L)) (Show ("Drive.txt") (©KEEP = FALSE;@WAIT=TRUE;)) (Is (DrivingConditions) ("Hard")) (Name (< | SuitableHammer | > .Strokes_per_Min) (F)) (Name (< | SuitableHammer | > .Theor_Energy*Hammer.Total_Efficiency) (E)) (Name ( < | Selected_SSP | > .Surface_Area) (SA)) (Name (< |Selected_SSP| >.Cross_Section_Area) (Ap))

) (@HYPO= SPProduction) (@RHS = (Write ("Hammer.nxp") (©TYPE = NXP;@FILL=NEW;@UNKNOWN = TRUE;@ATOMS = L,\ F,E,SA,Hammer_Type;)) (Write ("soil.nxp") (@TYPE = NXP;@FILL=NEW;@ATOMS = Soil_Type_l,\ Start_l,Finish_l,SoU_Type_2,Start_2,Finish_2;\ )) (Execute ("Drive.exe") (©TYPE = EXE;@WAIT = TRUE;)) (Retrieve ("out.nxp") (©TYPE = NXP;@F1LL=ADD;@FWRD = TRUE;@CREATE = | VarTime | ;\ ©ATOMS = Variable_Time.amount;)) (Retrieve ("out.nxp") (©TYPE=NXP;@FILL=ADD;@FWRD=TRUE;@CREATE = | Feasibility | ;\

©ATOMS=Technical_Feasibility.State;)) (Do (Check_Feasibility) (Check_Feasibility))

) ) (©RULE = Single_Pile_Variable_Production_Time @INFCAT = 5; ©COMMENTS = "This rule computes the variable driving time component for a pile series.";

L i s t i n g D.l P a r t i a l L i s t i n g of CMSA (continued)

329

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

(@LHS = (Name (Tunnel.depth+5) (L)) (Show ("Drive.txt") (@KEEP = FALSE;@WAIT=TRUE;)) (Is (Driving_Conditions) ("Soft")) (Name ("InPairs") (Piles_Driving_Pattern)) (Name (< |Suitable_Hammer| >.Strokes_per_Min) (F)) (Name (< | SuitableHammer | >.Theor_Energy*Hammer.Total_Efficiency) (E)) (Name (< |Selected_SSP| >.Surface_Area*2) (SA)) (Name ( < j Selected_SSP | > .Cross_Section_Area*2) (Ap)) ) (@HYPO= SPProduction) (@RHS = (Write ("Hammer.nxp") (@TYPE=NXP;@FILL=NEW;@ATOMS=L,F,E,SA,Hammer_Type;\ )) (Write ("soiLnxp") (@TYPE = NXP;@FILL=NEW;@ATOMS = SoU_Type_l,\ Start_l,Finish_l,Soil_Type_2,Start_2,Finish_2;\ )) (Execute ("Drive.exe") (@TYPE = EXE;@WAIT=TRUE;)) (Retrieve ("out.nxp") (@TYPE = NXP;@FILL=ADD;@FWRD = TRUE;@CREATE = | VarTime | ;\ ©ATOMS = Variable_Time.amount;)) (Retrieve ("out.nxp") (@TYPE = NXP;@FILL=ADD;@FWRD = TRUE;@CREATE = | Feasibility | ;\

©ATOMS=TechnicalFeasibility.State;)) (Do (CheckFeasibility) (CheckFeasibility))

)

)

(@GLOBALS = ©INHVALUP = FALSE; ©INH VALDOWN=TRUE;

L i s t i n g D.l P a r t i a l L i s t i n g of CMSA (continued)

330

Appendix D .

CMSA P a r t i a l L i s t i n g and Miscellany

331

D.3 Vibratory Hammer Selection Knowledge Base Development Vibratory p i l e driver s e l e c t i o n knowledge was extracted from the l i t e r a t u r e . This component serves t o select and s i z e a vibratory hammer for a given s o i l and p i l e scenario. What i s missing, i s a model which can predict p i l e d r i v i n g penetration rate as opposed to the dynamic formulas applied to the impact hammers. Furthermore, there i s no i n d i c a t i o n what kind of vibratory p i l e d r i v e r s u i t s a s o i l p r o f i l e (as opposed to impact hammers types such as SAAH, DAAH, e t c . ) . The interview with Quadra Construction Co. (see Appendix C) revealed that vibratory p i l e driver type s e l e c t i o n i s based on experience. In summary, there i s no mechanism to check whether a p i l e driven by a vibratory p i l e d r i v e r could reach i t s r e f u s a l depth nor how long would i t take t o drive a p i l e to i t s r e f u s a l depth. Therefore, t h i s part i s l i m i t e d to vibratory pile driver selection without a technical f e a s i b i l i t y test. The vibratory p i l e driver s e l e c t i o n i s based on empirical formula and charts found i n Barber (1987) — where a vibratory hammer can be s p e c i f i e d at t h i s l e v e l by i t s dynamic force and amplitude. From graph 1 of the previous reference, for a given s o i l p r o f i l e SPT and d i f f e r e n t s t e e l p i l e unit weights, one can f i n d a vibratory dynamic force. Table D.l presents vibratory dynamic force ( i n tons) computed i n terms of SPT (substituted as N) against d i f f e r e n t p i l e unit weights as derived from Graph 1 (Barber 1979). For instance, the f i r s t r e l a t i o n s h i p implies the following governing equation. For p i l e unit weight of 10 lb/ft, Vibratory

Dynamic Force = 0.21 * N

(D.l)

where N i s the standard penetration t e s t (SPT). I f the p i l e unit weight i s s p e c i f i e d (as CMSA prototype s p e c i f i e s s t r u c t u r a l member of SSP/SPL), then the dynamic force for a suitable vibratory hammer can be determined, and v i c e versa.

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

Line Number

Pile Unit Weight (lb/ft)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

10 20 30 40 50 60 70 80 90 100 110 120 130 140 150

Table D.l

332

Vibratory Dynamic Forces (Tons) 0.21* N 0.50 * N 0.75 * N 1.00 * N 1.20 * N 1.49* N 1.72 * N 2.00* N 2.33 * N 2.50 * N 2.70 * N 3.03 * N 3.23 * N 3.45 * N 3.85 * N

Vibratory P i l e Drivers S i z i n g

Graph 2 of Barber (1979) presents a l i n e a r r e l a t i o n s h i p between the p i l e length and vibratory amplitude. This i s transformed into the following governing equation: Amplitude

(in)

= P i l e Length

(ft) * .016 + 0.12

(D.2)

Using equation D.2, either the amplitude can be determined given a p i l e length, or vice versa. XMAS t r e a t s p i l e length ( p i l e segment) as a variable volunteered by the user. For the CMSA prototype, equation D.l r e s u l t s i n dynamic force designation of a vibratory s i z e (where p i l e unit weight i s predetermined by the hypotheses Select_SSP and/or Select_SPL). Equation 2 s p e c i f i e s the vibratory amplitude based on user p i l e segment s i z e . C r a n e Selection: Cranes are considered as secondary resources for p i l e d r i v i n g . Although crane s e l e c t i o n depends on several factors, f o r CMSA prototype development, crane s e l e c t i o n i s determined by the weight of the hammer ram weight. Table D.2 presents rules of thumb f o r s e l e c t i n g cranes designated by t h e i r carrying capacity versus the t h e o r e t i c a l energy of the impact hammer. For instance, a 35 ton crane may be used to carry a hammer with an upper bound

Appendix D.

of 8750 (8750 -

CMSA P a r t i a l L i s t i n g and Miscellany

ft-lb,

15000)

40 ton crane f o r hammers within f t - l b , and so f o r t h .

Cranes (Ton)

333

the range

Impact Hammer Energy (lb-ft)

25 40 60 100

8750 15000 25000 > 25000

Table D.2 Crane Selection Format (From Means Heavy Construction Cost Data 1987) "Crane.nxp" i s the f i l e that contains the crane data base that are used t o represent cranes i n CMSA prototype. D.4 Unit Cost Quotations Cost quotation sources f o r the CMSA prototype include cost data manuals, previous projects, interviews, and l o c a l vendors. What follow are comments about cost estimates and t h e i r break down. Contractor experience from previous jobs was used to set upper and lower bounds for unit cost rates (x $/ft run), or unit cost per surface area (x $ per square foot of s t e e l sheet p i l i n g ) , and production rate i n (day/ft). -

The interview with Quadra Construction Co. (see Appendix C) provided experienced based estimates for the crews and equipment involved i n the p i l e driving activity. For instance, from the project site visit, the following unit costs were obtained:

For sheet piling d r i v i n g , the cost of material i s approximately 3 / 4 of the t o t a l cost of p i l e d r i v i n g operation, whereas 1/4 i s labor cost and equipment (labor cost almost i s almost equal t o the equipment cost f o r a vibratory hammer). Representative table D . 3 .

costs

employed by CMSA are shown below i n

Appendix D .

1.

2.

3.

4.

•5. 6.

CMSA P a r t i a l L i s t i n g and Miscellany

Steel Sheet Piles and H - Piles Grades (40 and 50 ksi)

Unit Cost

$ 930/Ton

Lumber Size

Unit Cost

Available Length

2 3 4 4 6

$ 0.27/ft $ 0.50/ft $ 0.75/ft $ 1.35/ft $ 2.30/ft

6 to 20 ft 6 to 20 ft 6 to 20 ft 6 to 20 ft 8 to 20 ft

Impact Hammer

Theoretical Energy (lb-ft)

Rent

SAAH, DAAH SAAH, DAAH SAAH, DAAH

8750 15,000 25,000

$ 4000/month $ 5000/month $ 6000/month

4 4 4 6 6

Crane 25 Ton 40 Ton 60 Ton 100 Ton

Rent $5000/month $6500/month $9000/month $11000/month

Crew (Pile Driving) 4 men / day

Cost $22000/month

Vibrator Model Model ICE 812 Model ICE 216

334

Dynamic Force (Tons) 145.5 36.4

Rent $11000/month $7000/month

Table D.3 Representative Resources Unit Costs i n Vancouver, B.C, [Quadra Construction Co Ltd] In comparison with other shoring methods, f o r s t e e l sheet p i l i n g , material cost savings are of major concern. Unit prices are as quoted from the l o c a l market (1990) . These prices vary with the amount of material quantity, length and s i z e of the piles/lumber segments, material quantity, and so forth.

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

335

D.5 Sample Data Base Files As indicated e a r l i e r , these include the design elements f o r the construction method (structural members (steel sheet piles, soldier piles, s t r u t s , wales, and laggings) + construction resources (hammers + cranes). The data bases contain a sample of what could be used. CMSA uses a subset of these data bases. 1.

Steel Sheet P i l e s

The following properties and dimensions were taken from Winterkorn and Fang (1975) and are found i n database "SSP.nxp". The SSP are divided into three groups according t o t h e i r section modulus around the X-axis (the assumption used i s that the section modulus rather than i n t e r l o c k strength constitutes the basis for SSP s e l e c t i o n including shape. A p a r t i a l l i s t i n g of those used i n CMSA i s show i n table D.3. Group_l : This i s mainly Z-section with Section Modulus (S): 46.8

in 3 A

«

S «

30.2 i n 3 A

There are 3 SSP sections i n t h i s group. Group_2:

This i s mainly invert U-section with Section Modulus (S): 10.7 i n 3 A

«

S

«

2.4 i n 3 A

There are 2 SSP sections i n t h i s group. Group_3:

This i s mainly PSA and PSA (straight) sections with Section Modulus (S): 1.9

in 3 A

«

S

«

2.4 i n 3 A

There are 4 SSP sections i n t h i s group which depend on using them i n applications involving i n t e r l o c k strength rather than section modulus. Properties are shown f o r information purposes. Group_4: Miscellaneous

sheet p i l i n g are not included.

Properties treated i n the data base are as follows D.2) : 1. 2.

Designation; Weight_per_foot i n ( l b ) ;

(listing

Appendix D.

3. 4. 5. 6.

CMSA P a r t i a l L i s t i n g and Miscellany

336

Cross_Section_Area (a) i n ( i n 2 ) ; Surface_area i n ( f t 2 / f t ) {excludes i n t e r l o c k area}; Driving_width i n ( i n ) ; and Section_Modulus i n ( i n 3 ) . A

A

A

Group_l: Z-Section \SSP_l.Designation\ = "PZ38" \SSP-l.Weight_per_foot\ = "57.00" \SSP_l.Cross_section_area\ = "16.77" \SSP_l.Driving_width\ = "18" \SSP_l.Surface_area\="5.52" \SSP_l.Section_modulus\ = "46.8"

\SSP_2.Designation\ = "PZ32" \SSP-2.Weight_per_foot\ = "56.00" \SSP_2.Cross_section_area\ = "16.47" \SSP_2.Driving_width\ = "21" \SSP_2.Surface_area\ = "5.52" \SSP_2.Section_modulus\ = "38.3" \SSP_3.Designation\ = "PZ727" \SSP_3.Weight_per_foot\ = "40.5" \SSP_3.Cross_section_area\="11.91" \SSP_3.Driving_width\="18" \SSP_3.Surface_area\="4.98" \SSP_3.Section_modulus\ = "30.2"

L i s t i n g D.2 2.

Sample of Steel Sheet P i l e Data Base Used i n CMSA "SSP.nxp"

Soldiers P i l e s

Structural properties are given f o r use when Standard HPP i l e s are u t i l i z e d as rakes, wales or as other s t r u c t u r a l members (AISC 1978). The database "HP_Pile.nxp" contains selected s o l d i e r p i l e members. The normal Material S p e c i f i c a t i o n i s : ASTM A36, ASTM A572 grades 42 through 60 (HP 14 * 117 i s not available i n grade 60). Those H-piles are available i n welded form from Kaiser Steel Corporation.

Appendix D .

CMSA P a r t i a l L i s t i n g and Miscellany

337

Note that other s t e e l s t r u c t u r a l members such as WF-sections could be used. WF-sections have more flange width and thus are more r e s i s t a n t to l a t e r a l pressure. For the prototype application, attention was l i m i t e d to H-piles. Properties treated i n the data base are as follows: 1. 2. 3. 4. 5. 6. 7.

Designation; Weight_per_foot i n (Lb); Cross_Section_Area (a) i n ( i n 2 ) ; Surface_area i n ( f t 2 / f t ) {excluding i n t e r l o c k area}; Driving_width i n ( i n ) ; Driving_depth i n ( i n ) ; and Section_Modulus i n ( i n 3 ) . A

A

A

\HP-l.Designation\ = "14_117" \HP-l.Weight_per_foot\="117" \HP_l.Cross_section_area\="34.4" \HP_l.Driving_width\ = "14.89" \HP_l.Driving_Depth\ = "14.23" \HP_l.Surface_area\ = "7.11" \HP_l.Section_modulus\ = "173" \HP-2.Designation\ = "14_102" \HP-2.Weight_per_foot\ = "102" \HP_2.Cross_section_area\ = "30.0" \HP_2.Driving_width\ = "14.78"

L i s t i n g D.3

S o l d i e r P i l e s Sample Data Base "HPPile.nxp"

3. Struts Properties and dimensions of American-Produced Standard (W) Shapes f o r Columns (Struts) and Beams (Wales) f o r Internal Bracing Retaining System (Table 1-22 Manual of Steel Construction, AISC) were used. The W14 series was adopted for the prototype because of the v a r i e t y a v a i l a b l e i n t h i s s i z e range. Further, Fy = 36 k s i , and Fa = 19 k s i (Winterkorn and Fang 1975). The properties i n the data base are as follows. 1. 2. 3. 4.

Designation; Weight_per_foot i n (Lb); Cross_Section_Area (a) i n ( i n 2 ) ; and Radius_of_Gyration r(y) i n ( i n ) . A

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

338

\Strut_l.Designation\="W_14_132" \Strut_l.Weight_per_foot\="132.00" \Strut_l.Cross_section_area\="38.8" \Strut_l.Radius_of_Gyration\ = "3.76" \Strut_2.Designation\ = "W14120" \Strut_2.Weightjper_foot\="120.00" \Strut_2.Cross_section_area\="35.3" \Strut_2.Radius_of_Gyration\="3.74" \Strut_3.Designation\ = "W14109" \Strut_3.Weight_per_foot\="109.00" \Strut_3.Cross_section_area\ = "32.0" \Strut_3.Radius_of_Gyration\="3.73"

L i s t i n g D.4 4.

Struts Sample Data Base "Strut.nxp

11

Lagging:

Structural properties are given f o r use when timber i s u t i l i z e d as lagging, rakers, wales or as other s t r u c t u r a l members (CSA Standard 1976). Several nominal timber sizes are available such as 2_4, 3_4, 4_4, 4_2, 4_3, 4_6, and 6_6 are stored i n "lagging.nxp" database. Properties pertinent limited to: 1. 2. 3. 4. 5.

t o lagging

design, within

Designation: Lag_?_?; Lagging_Width, i n ; Lagging_Thickness, i n ; Lagging_Section_Modulus, i n 3 ; and Lagging_Unit_Cost_per_foot, i n Canadian Dollars/foot. A

CMSA, are

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

339

\Lagging_l.Designation\="Lagging_2_4" \Lagging_l.Width\ = "4" \Lagging_l.Thickness\="2" \Lagging_l.Section_modulus\="3.06" \Lagging_l.Unit_Cost_per_Foot\ = "0.27" \Lagging_2.Designation\ = "Lagging_3_4" \Lagging_2.Width\ = "4" \Lagging_2.Thickness\="3" \Laggmg_2.Section_modulus\="5.10" \Lagging_2.Unit_Cost_per_Foot\ = "0.50"

L i s t i n g D.5 5.

Lagging Sample Data Base "Lag.nxp

11

Hammers Sample

Hammer databases are named according to the c l a s s of the hammer. For instance, "DAAH" stands f o r Double Acting A i r Hammer; "SAAH" stands f o r Single Acting A i r Hammer, e t c . L i s t i n g D.6 shows a sample of impact hammers which i s implemented i n CMSA. Table 17-16 of Peurifoy (1970) data on p i l e d r i v i n g hammers has been adopted f o r the impact hammer data base. Data f i e l d s used are:: 1. 2. 3. 4.

Ram_Weight i n ( l b ) ; Stroke_per_minute i n (no u n i t s ) ; Length of stroke i n ( i n ) ; and Theoretical_Energy (ft-lb) per blow.

This database Vulcan.

contains single

acting

a i r hammers of type

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

340

\Hammer_01.Hammer_Model\ = "2" \Hammer_01.Ram_Weight\="3000" \Hammer_01.Strokes_per_Min\="70" \Hammer_01.Length_of_Stroke\ = "29" \Hammer_01.Thero_Energy\="7260" \Hammer_02.Hammer_Model\ = "1" \Hammer_02.Ram_Weight\="5000" \Hammer_02.Strokes_per_Min\ = "60" \Hammer_02.Length_of_Stroke\ = "36" \Hammer_02.Thero_Energy\ = "15000" \Hammer_03.Hammer_Model\="0" \Hammer_03.Ram_Weight\="7500" \Hammer_03.Strokes_per_Min\ = "50" \Hammer_03.Length_of_Stroke\ = "39" \Hammer_03.Thero_Energy\="24375"

L i s t i n g D.6

Impact Hammer Sample Data Base "Hammer.nxp"

6. Vibratory Hammers Sample "Vibro.nxp" i s the database f o r v a r i e t y of vibratory p i l e drivers. The following d i f f e r e n t models of vibratory hammers are adopted from Peurifoy (1970) f o r Foster Vibro driver/extractor ( l i s t i n g D.7). Properties of i n t e r e s t are: 1. 2. 3. 4. 5. 6. 7. 8.

Maximum Energy delivered, f t - l b per sec; Vibration frequency, rpm, min; Vibration frequency, rpm, max; Total horsepower; Voltage; Maximum Amplitude; Cycles per sec, and; and Approximate weight, l b .

Appendix D.

CMSA P a r t i a l L i s t i n g and Miscellany

341

\Vibratory_l.Model\="2-17" \Vibratory_l.Max_Energy\="18440" \Vibratory_l.Min_Frequency\ = "1090" \Vibratory_l.Max_Frequency\ = "1290" \Vibratory_l.Tot_Horsepower\="34" \Vibratory_l. Voltage\="440" \Vibratory_l.Max_Amplitude\ = "60" \Vibratory_l.Cycles\ = "60" \Vibratory_l Approx_Weight\="6200" \Vibratory_2.Model\ = "2-35" \ Vibratory_2.Max_Energy\ = "37970" \Vibratory_2.Min_Frequency\ = "890" \Vibratory_2.Max_Frequency\="1120" \Vibratory_2.Tot_Horsepower\="70" \Vibratory_2.Voltage\ = "440" \Vibratory_2.Max_Amplitude\ = "120" \Vibratory_2.Cycles\ = "60" \Vibratory_2 Approx_Weight\ = "9100"

L i s t i n g D.7

Vibratory Hammers Sample Data Base [Vibro.nxp]