abstract introduction

Reprinted by permission from Advances in Powder Metallurgy & Particulate Materials—2012. Copyright © 2012 Metal Powder Industries Federation. All rights reserved.

OVERLAPS IN ROLLED PM GEARS Fritz Klocke, Markus Brumm and Eva Gräser Laboratory for Machine Tools and Production Engineering (WZL) of RWTH Aachen University Steinbachstrasse 19, 52056 Aachen, Germany

ABSTRACT This paper is concerned with a manufacturing defect which occurs during densification rolling of PM gears. This manufacturing defect is an overlap of material which results in a notch just below the gears surface. Prior investigations have shown that this defect occurs at the tooth tip and the tooth root. However, the form and the size of the overlaps differ due to contact conditions. These differences will be demonstrated in micro sections. Therefore, the phenomenon is compared with other forming processes in order to identify important influences. Then the densification process is analyzed and reduced to describable model parameters in order to develop an analogy model for an FE simulation. The use of the analogy model is approved by simulation results, which show the influence of geometry, density and feed.

INTRODUCTION In order to produce powder metal gears, a powder mixture is filled in a cavity. The mixture is densified by high pressure applied through upper and lower punches. The generated green gear is sintered at high temperatures due to which the powder particles bond to a single part. After pressing and sintering the gear is still porous and therefore has a lower strength than gears made of conventional steel. The process chain of a powder metal manufacturing is however considerably cheaper, when gears of the same geometry are to be manufactured in mass production. In order to increase the strength, the surface can be densified locally after sintering. One option is the surface densification by rolling. Due to the contact of roller and workpiece and the reducing distance of rollers, pressure is applied to the surface, which leads to the densification. The quality of a PM gear can be described by the gears geometrical quality, the density distribution and the surface quality. Manufacturing defects can occur, as they also occur in other manufacturing technologies. In densification rolling those are the typical S-shaped profile deviation, a deviating

1

densification and tongue-shaped overlaps of material in the tooth root and tip. Overlaps occur also in other areas of forming. Because they are called overlaps in other areas of forming, this term will be used in the following [BARG08]. During densification rolling of pressed and sintered gears, the kinematics at the tooth flanks is similar to conventional tooth contact of gears, cf. Figure 1. In contrast to conventional tooth contact there is also contact in the tooth root. The position of the pitch point depends on the infeed and is therefore moving towards the tooth root during the process. Comparable to the conventional tooth contact, there is a sliding velocity in the direction towards and away of the pitch point. The sliding velocity reaches its maximum at tooth tip and root of the workpiece. On the right Figure 1 shows a metallographic cut with four details sections (1-4) below. Detail 1 is a picture of the tooth root. Under the surface there is a defect which is composed of two slim, horizontal separations of material, one above and a pore in the middle. The profile of the material between the interlayers allows the conclusion that these are overlaps originated from previous sliding contacts of the tool. Obviously the region around the pore is completely densified. Therefore, the pore cannot be originated form the green part. Detail 2 and 3 show sections of the tooth flank where the structure is densified and tangentially deformed without any surface defects. Detail 4 is a picture of the tooth tip. It shows an overlap in direction of the tooth flank. Thus overlaps occur in areas with high sliding velocity during rolling. If there is a relation between the sliding velocity and occurrence of overlaps, is however unproved. Surface Densification in Simulation Pressing

Sintering

oberflächendefekte Surface Defects

Rolling

1

2 3 4

1

2

Tool

100 µm (3.94 10-3 in) 3

100 µm (3.94 10-3 in) 4

Workpiece Relative Density ρrel

> 98 % > 96% > 94% > 92% > 90%

100 µm (3.94 10-3 in)

-3 in) 10010 µm 100 µm (3.94

© WZL

Figure 1:

Surface Densification of PM Gears

There are multiple studies concerning improvement of profile deviation and influencing the densification. However, up to now there is no study that deals with the occurrence of overlaps during densification rolling. Industry has already established measures to avoid overlap defects based on experience. However, a systematic prevention adapted from scientific research is not feasible yet. Overlaps are shape deviations

2

and accordingly imperfections. So on the one hand they can affect further machining, e.g. by increasing local chipping thickness in hard finishing. On the other hand imperfections pose a risk in operation, when they are located on the highly stressed surface as shown above [NIEM83a]. For these reasons a prevention of overlaps is desirable.

OVERVIEW First, the state of technology concerning occurrence of overlaps during densification rolling and other forming technologies will be discussed. This is the basis for a theoretical analysis which abstracts the rolling contact during densification rolling and points out the main influencing variables. This abstraction will be used to develop an analogy model which will be transformed into a simulation and an experiment. The transformation of the analogy model into the simulation will be introduced with the help of calculated examples. The plausibility will be checked.

STATE OF TECHNOLOGY Up to now there are no studies that investigate the occurrence of overlaps. Overlaps describe overlapping material that occur during manufacturing [BARG08]. However approaches exist which describe the occurrence of overlaps in densification rolling and other areas of deforming technology. In Figure 2 different examples for overlaps in other forming technologies are shown. Overlaps are documented in drawing of steel wire, shaft rolling and screw rolling. Steel Wire

Rolled Shaft

Screws

Bild

Bild

Nomenclature

Causes

 Overlaps occur in several

 Spare material leaves the roll gap and is rolled in during the next

products manufactured by forming  Overlaps are material layers

which are not welded to the part and occur during manufacturing

rolling step  In hot rolling the curled material cools down fast due to high

area-to-volume ratio  Due to oxides the material layers do not cold-weld to part  Overlaps in screws lead to early failures of screw joints

Image source: Pomp (Stahldraht), Gruener (Fehlerquellen beim Walzen), Illgner (Materialtrennungen in Schrauben)

© WZL

Figure 2:

Overlaps as Term in Forming Processes

As reasons for overlaps wrong tool adjustment, temperature control and tool design are mentioned. By these false adjustments spare material gets out of the tools as a bulge and is rolled into the surface in the next rolling step. Beyond that, in hot rolling the bulge cools down quickly due to the temperature gradient between material and environment on the large surface. Because of this, the bulge material has higher

3

strength and is rolled into the softer main material. Especially if scaling occurs in addition to cooling down, material does not reweld and the overlap stays separated from the remaining material as an imperfection in the workpiece. [BAUD04, GRUE51, LANG88, LICK66, NN26, POMP52] Especially bolted connections failed too early due to overlaps in screws [ILLG70]. Imperfections can be generally considered as notches, so that the loss of strength in consequence of is a feasible assumption not only for screws [ISSL03]. Not only the occurrence but also the generation of bulges is documented for flow forming. In this process the workpiece is put on a mandrel and is deformed in axial direction by three rotating rolls in order to reduce wall thickness, cf. Figure 3. The deformation is affected by the selection of material, the diameter of rolls and mandrel, the rotating direction of tool and workpiece, the feed of the rolls and the feeding angle between roll and workpiece. Further influencing variables are the friction between mandrel and workpiece as well as between rolls and workpiece and the reduction of wall thickness overall and per step. [LANG88] Flow Forming Process

Forming of Bulge in Process Process Progress

Roller

vf

Δs

Workpiece

α

Mandrel

µ, m

Influence Size of bulge

Feeding angle α

Feed velocity vf

Friction µ, m

Reduction of wall thickness Δs

Yield point Rp0,2

Source: Lange (Umformtechnik – Massivumformung)

© WZL

Figure 3:

Forming of Bulge in Flow Forming Process

The forming of bulges was investigated in respect of the impact of the influencing variables mentioned above. The feeding angle turned out to be a variable with a lower influence. A high feeding angle leads to a larger bulge. A higher feed velocity increases the size of bulges as well. A good lubrication reduces friction between roll and workpiece. This reduces forming of bulges, so that the size remains uncritical. A critical size is reached, when the bulge cannot be rolled in without generation of imperfections in the following steps. The choice of material is highly relevant, because soft materials tend more to bulge forming. That’s why the choice of material has also an influence on the maximum acceptable decrease of wall thickness ∆s: A previous strain-hardened material can help to achieve a higher decrease of wall thickness without any significant bulge forming. However, generally an increase of wall thickness reduction results in an increase of bulges. In a process with several over-rollings, the first step is to strain-

4

harden the material by performing a low decrease of wall thickness. In the second step the decrease of wall thickness can be heightened significantly. In a study realized at WZL a correlation of shear stress and the occurrence of surface defects during densification rolling of PM gears was shown [FOER10]. The shear stress was determined with the help of an FE simulation. The level of the shear stress has not been validated for this model. But the values can be used for comparison within the simulation. In Figure 4 on the left the shear stress is plotted over the workpiece radius. The value is shown by a colored scale. On the ordinate the number of contacts is plotted. The abscissa shows the radius of the gear. So in the diagram the frequency of occurrence of a specific shear stress spectrum is plotted over the radius. Shear Stress in Simulation

Surface Defects

Occurrence of shear stress spectrum 4

400 300

2

3

1

1

2

200

Radius

100 0 [mm] 28 [in] 11.0

30 11.8

Amount of shaer stress

Usable tip circle 75:100 MPa

32 12.6

34 13.4

Rolling circle (10.9 : 14.5 103 psi)

36 14.2

50 µm (1.97 10-3 in)

Usable root circle 3

50 µm (1.97 10-3 in)

4

100:250 MPa (14.5 : 36.25 103 psi) 50:75 MPa

(7.2 : 10.9 103 psi)

25:50 MPa

(3.6 : 7.2 103 psi)

0.1:25 MPa

(0.01 : 7.2 103 psi)

50 µm (1.97 10-3 in)

50 µm (1.97 10-3 in)

Source: Förster, ROLLAIX

© WZL

Figure 4:

Comparison of Shear Stress and Surface Defects

Below the effective root diameter there are fewer contacts between tool and workpiece. This can be explained by the fact that the tool has contact with the root the first time after several infeed steps. Initially there is just contact at the tooth flanks. However the shear stress level is high above average in the tooth root area. Compared to this there are very few contacts in the high shear stress spectrum between the effective root diameter and the tip diameter. Above the tip diameter many contacts in the highest shear stress spectrum occur. In comparison to the detail views on the right of Figure 4 coherence between the degree of microstructure deformation and the shear stress can be noticed. The analysis has been performed with the same workpiece which is shown in Figure 1. The comparison between Figure 1 and Figure 4 suggests that the occurrence of overlaps can be predicted by regarding the shear stress. But this correlation is not proved yet just as the coherence of overlaps and the sliding velocity.

THEORETICAL ANALYSIS OF THE DEVELOPMENT OF OVERLAPS The rolling contact during the densification rolling process is to be transferred into a description model that illustrates the influencing variables and neglects disturbance values. The geometrical conditions

5

during contact between workpiece and tool can be simplified by the contact radiuses ρWzg und ρWst, cf. Figure 5. It should be considered that convex geometries are also possible. Similar methods are used to describe the flank contact of gears [DIN87].

Rolling Contact

Description Model

ρWzg,Wst EWzg,Wst τSch ρ0 Δs µ, m

Contact radius tool, workpiece Young’s modulus tool, workpiece Shear strength workpiece Density Stock Friction

ω

Sliding velocity Rotational speed

EWzg

ω

ρWzg

Δs

ρWst

ρ0 τSch µ, m

EWst

 Geometry of tool and workpiece

 Variable contact radius

 Variable sliding velocity over the profile

 Rolling or sliding

 Stock for densification on the gear

 Stock

 Material of tool and workpiece

 Reduction on density, Young’s modulus and

shear stress

 Friction in rolling contact

 Coulomb friction, shear friction model © WZL

Figure 5:

Description Model for the Contact in Densification Rolling

The material behavior cannot be described completely. However it is possible to consider material properties within the analysis. Due to the shear stress during the rolling process, the shear strength τSch is considered as an important material constant. The densification affects the volume of material that is going to be deformed and the material properties, cf. Beiss [BEIS03]. That is why the density ρ0 has to be considered as well. In order to achieve a densification the workpiece is provided with a stock of material ∆s. Friction affects the occurrence of overlaps in other forming methods and so it has to be considered here, as well. Therefore the friction coefficient model or Coulomb’s friction is used. Additionally to geometry, material and friction, the contact between tool and workpiece is affected by kinematics. Here the rolling contact has to be described as a composition of rolling and sliding movement. Analyzing the metallographic micrograph and surface images the three-dimensional character of the phenomenon overlaps stands out, cf. Figure 6. The description model is only able to describe a twodimensional imperfection. However for a spur gear, the imperfection is assumed to be constant in the gears width. This model is sufficient to describe the mean size of overlaps. Reasons for a threedimensional character can be abnormalities in the material as pores, grain distribution and nonmetallic inclusions and impurity of the workpiece before the rolling process or impurity of the lubricant. Furthermore, the geometrical quality of the workpiece can have an influence, as waviness on the surface can cause an uneven overlap. In the following the findings are to be analyzed in a view of validity for the occurrence of overlaps during the densification rolling process. As a consequence of a wrong design of tools or an unfavorable position of tools, material gets out on the rolling contact instead of being deformed by tools. In densification

6

rolling the material is either compressed or pressed out of the rolling contact so that a bulge is formed. This suggests that the characteristics of the material flow depend on the force transmission and the contact properties. This correlation is still to be analyzed.

Real Defect

Description Model EWzg

ω

ρWzg

ρWst

Δs

ρ0 τSch µ, m  Real defect is three-dimensional

EWst

Model describes two-dimensional defect

 Reasons for three-dimensionality are unknown. Assumptions:

– Material irregularities (Pores, grain size distribution, non-metallic inclusion) – Contamination on workpiece before rolling and in lubricant – Waviness of workpiece surface, a. s. o.

© WZL

Figure 6:

Limitation of the Description Model

A detailed analysis of the bulge forming during flow forming indicates what influencing variables of the densification rolling process are to be considered. The influence of the feeding angle defines the contact area and the angle of force and can therefore be compared with the contact radiuses and the angle of the force vector during densification rolling. The influence of feed can be compared with the interaction of feed curve, rotational velocity of the workpiece and slip between tool and workpiece. A good lubrication can help to avoid critical bulge forming during roller spinning, so the influence of friction on occurrence of overlaps and the possibility to reduce friction has to be analyzed. The decrease of wall thickness stands for the volume of material that is to be deformed in one step. It is comparable to the oversize in densification rolling, that is to be densified during one rotation. The choice of material has a high influence during flow forming and therefore should also be considered for the occurrence of overlaps. The coherence of areas with high shear stress and areas with high structure deformation suggests that the amount of shear stress in densification rolling has to be compared with material constants like the shear strength to detect a possible correlation. During densification rolling the material is either densified or is forming a bulge. An applied force causes a material deformation. In densification rolling the force is applied on a PM gear by gear-shaped tools. To the resulting compressional force, tangential friction is added. It can be supposed that vectors of applied force and material flow are not parallel. This is due to the densification resistance. The densification resistance is influenced by the original densification, the base material and the deformation history of the workpiece. The prediction of bulge forming as a starting point for development of overlaps can be explained, when the material flow can be predicted.

7

APPROACH TO SIMULATE THE DEVELOPMENT OF OVERLAPS The analysis of development of overlaps will be by means of simulation. For this an analogy model has to be developed, which describes the problem accurately. However, the effort in FE simulation should be low. It should be possible to change the influencing parameters independently in order to identify their influence. The analogy model should be transferable to an experiment in order to validate the simulation. Therefore the description model is to be abstracted to the analogy model. It is assumed that overlaps are constant over the width of the gear. This is accordant to the description model. Two dimensional calculations are less time consummating in the FE simulation than three dimensional calculations. The commercial FE program Deform 2D of SFTC is used. The analogy shows the contact of workpiece and tool, cf. Figure 7. In order to use less complex kinematics and contacts the tool is replaced by a circle-shaped roller and the workpiece is simplified as a block. The roller radius and therefore the contact geometry can be changed. The shear strength and the density of the workpiece are used in a material model of the simulation. The stock of material of the workpiece is used as the feed of the roller is as high. The stock is constant over the length of the workpiece. Requirements on Analogy Models  Investigation of relevant influencing parameters independently from each other  Justifiable simulative and experimental effort with best description of the problem as possible  Transfer in experiment reasonable for validation of model

Description Model

EWzg

ω

ρWst

ρWzg

Δs

ρ0 τSch µ, m

EWst

Analogy Model

ρWzg,Wst EWzg,Wst τSch ρ0 Δs µ, m

Contact radius tool, workpiece Young’s modulus tool, workpiece Shear strength workpiece Density Stock Friction

ω vf

Sliding velocity Rotational speed Feed velocity

FW Roller ω

vf

ρWzg µ, m

Workpiece

Δs ρ0, τSch EWst © WZL

Figure 7:

Entwicklung eines Analogiemodells

The friction has also to be investigated by the analogy model. In FE simulation the law of friction and the accordant friction coefficient has to be chosen. In an experiment friction can be changed by lubrication and surface quality. The sliding velocity between workpiece and tool is substituted by feed of roller. The ratio between rotational velocity and feed is according to the ratio of sliding velocity and rotational velocity during densification rolling. These values are not to be changed during one simulation.

8

The presented analogy model was implemented in a FE simulation in Deform. Here, it is to be proofed, if the analogy can describe the problem. The simulation shows a circle-shaped roller, which rolls over a block-shaped workpiece. The roller of a reference has a radius of r = 10 mm = 0.394 in. The density is given with ρrel = 90 %. The feed, which represents the stock for densification, is ∆s = 0.2 mm = 7.9 10-3 in. The roller has a constant velocity. The rotational velocity is free. The situation is shown during the sixth over-rolling. The sixth over-rolling of the tool was chosen, as in the first steps, no bulge forming was visible. This supports the thesis that overlaps can be explained by the current densification and forming history. In the first step, the potential to densification is the highest. In the further over-rollings this potential is reduced until bulge forming starts.

∆s = 0.2 mm; ρrel = 90 %; rWalze = 10 mm = 0.39 in = 7.9 10-3 in hW = 0.68 mm = 0.026 in

ρrel = 85 %

Reference

In comparison to the reference the parameters roller radius, density, feed and roller tangential velocity are varied. Figure 8 shows the results. The reference is shown on the upper left. The roller deforms a densified block. The densification results from the earlier over-rollings by the roller. In front of the roller a bulge is visible. Its height is hW = 0.68 mm = 0.026 in. The figure on the upper right shows a part whose starting density is reduced to a density of ρrel = 85 %. The densification of the former over-rollings is lower than the one from reference. The bulge has a height of hW = 0.17 mm = 0,006 in and is about a quarter from the reference. A lower density means that the possibility to densify is higher. Therefore the material flow will be more into the part than into forming a bulge. Hence, the simulation result is feasible.

hW = 0.17 mm = 0.006 in

Relative Density ρrel

> 98 %

rR = 20 mm = 0,787 in

∆s = 0.3 mm = 11.8 10-3 in

> 96 % hW = 6.27 mm = 0.247 in

hW = 0.08 mm = 0.003 in

> 94 % > 92 % > 90 % © WZL

Figure 8:

Influence of Density, Stock of Material and Roller Radius

The figure on the lower left shows a simulation, where in comparison to the reference the feed was increased to ∆s = 0.3 mm = 11.8 10-3 in. The densification is very good. The bulge, which is almost chiplike, has a height of hW = 6.27 mm = 0,247 in. Due to the increased feed, the earlier over-rolling lead to a higher material flow into the block as densification. In this shown last step, the possibility to densify is not

9

high. Therefore the material flow will be directed mostly into the front of the roller, which leads to the higher bulge. Hence, this result is feasible, too. The figure on the lower right shows a simulation with increased roller radius of r = 20 mm = 0,787 in. The bulge is very low in comparison to the other results with hW = 0.08 mm = 0.003 in. As the movement of the roller is given, the vector of its force is influenced by contact geometry of tool and workpiece. As the simulation shown on the lower right has a larger, more even contact, the bulge has to be smaller according to Langes results in flow forming [LANG88]. Figure 9 shows again the reference on the left. Here the parameter roller velocity is varied. The roller velocity of the simulation shown on the right is doubled in comparison to the reference. It is visible that the bulge is increased and more similar to the simulation with increased stock of material. The densification is similar to the reference. Therefore not only the potential for densification, but also the deforming velocity is critical to bulge forming.

hW = 0.68 mm = 0.026 in

vf = 2 * vref

Reference

vf = 1*vref

hW = 2.63 mm = 0,103 in

Relative Density ρrel > 98 % > 96 % > 94 % > 92 % > 90 % © WZL

Figure 9:

Influence of Velocity

These simulations were performed, in order to show that a simulation based on an analogy model can describe the development of overlaps. Because of the results of the simulation, the analogy model can be used in the simulation of the development of overlaps. However, more work into reducing the element size and increasing performance of the simulation is required. Afterwards, the shown parameters can be changed separately and be investigated by their single and interaction influences.

SUMMARY AND OUTLOOK The powder-metallurgical production is a cost effective alternative to manufacture gears of the same geometry in a large-scale production. When single pressing and sintering are chosen, the part still has about 10% porosity and thus a lower strength than hobbed gears from conventional steel. In order to increase the strength of PM gears, the surface is densified locally. A local densification is sufficient as failure critical stresses occur only near the surface of gears. One possibility to densify the surface is rolling with gear shaped tools. In this paper, a theoretical analysis is performed to reduce the complex phenomena of overlap development in densification rolling to a describable analogy, which can be used in experiments and simulations. The analysis includes a literature research and an analysis of the surface densification process.

10

The literature research was conducted to obtain information on the formation of overlaps. In other rolling processes, the following reasons for overlaps were found: Due to an incorrect design or a bad position of the rolling tools, material leaves the rolling gap instead of being deformed in the tools. Findings from bulge forming in flow forming processes suggest that these parameters can affect overlaps: the contact radii, the feed curve, the rotational velocity of the tool, slip between the teeth of tool and workpiece as well as the lubrication. Also important is stock for densification per revolution and the choice of material. Based on the theoretical analysis an analogy model was presented, which contains the influencing parameters and allows an implementation in a two dimensional FE simulation and in an experiment. A circle-shaped roller is used, which rolls over a block-shaped workpiece. Several simulations were performed in order to test the capability of the analogy model. The FE simulations show feasible results. The simulations show a high influence of feed on bulge forming and a medium influence of the roller radius and the starting density. The prediction of bulge forming as a starting point for development of overlaps can be explained, when the material flow can be predicted. This is e. g. indicated by the high influence of densification during the process on the bulge forming. Furthermore it is indicated by the high influence of roller velocity. The prediction of material flow will also allow more efficient process planning as it can not only be used for the prediction of overlaps and densification but also for tool development in profile optimization. The next steps are the improvement of the simulation and the implementation of the analogy experiment. This parallel development will end in a simulation with which the investigation can be performed.

ACKNOWLEDGEMENTS The investigations described in the present paper were conducted in the project “High-strength gears by powder metallurgical manufacturing” sponsored by the German Research Community (DFG) as a part of a priority programme (SPP) 1551.

BIBLIOGRAPHY [BARG08]

Bargel, H.-J.; Schulze, G. (Hrsg.): Werkstoffkunde. Springer: Berlin, 2008

[BAUD04]

Bauder H.-J.: Entwicklung eines Hochleistungsharnisches für Luftdüsenwebmaschinen. Cuvillier Verlag: Göttingen, 2004

[BEIS03]

Beiss, P.: Mechanische Eigenschaften von Sinterstählen.Tagungsband zum Symposium für Pulvermetallurgie: Material-Prozess-Anwendung, Hagen, 27.-28. November 2003, Hagen: ISL Verlag, 2003, S. 3-24

[DIN87]

Norm DIN 3960 (März 1987). Begriffe und Bestimmungsgrößen für Stirnräder(Zylinderräder) und Stirnradpaare (Zylinderradpaare) mit Evolventenverzahnung

[FOER10]

Förster, A.; Klocke, F.; Gorgels, C.; Kauffmann, P.; Hauptmann, C.: ROLLAIX – Ein Aachener Tool zur FE-Analyse von Rollierprozessen. In: Tagungsunterlagen zum Abschlusskolloquium BMBF-Projekt: PM-Zahnräder, Aachen, 25. August 2010

[GRUE51]

Gruener, P.; Brüggemann, T.: Fehlerquellen beim Walzen. In: Stahl und Eisen: Zeitschrift für das Deutsche Eisenhüttenwesen. Band 71. Verein Deutscher Eisenhüttenleute, Verein Deutscher Eisen- und Stahlindustrieller. Nordwestliche Gruppe, 1951, S.20-28 und 71-77

11

[ILLG70]

Illgner, K.: Materialtrennungen in Schrauben und Muttern. In: Drahtwelt. Band 56/12, 1952

[ISSL03]

Issler, L.; Häfele, P.; Ruoß, H.: Festigkeitslehre. Grundlagen. Springer, Berlin, 2003

[KOTT03]

Kotthoff, G.: Neue Verfahren zur Tragfähigkeitssteigerung von gesinterten Zahnrädern. Dissertation, RWTH Aachen, 2003

[LANG88]

Lange, K.: Umformtechnik. Handbuch für Industrie und Wissenschaft: Band 2: Massivumformung. Berlin: Springer Verlag, 1988

[LICK66]

Lickteig, E.: Schraubenherstellung. In: Stahleisen-Bücher. Band 4. Verlag Stahleisen M.B.H., Düsseldorf, 1966, S. 61-75

[NIEM83a]

Niemann, G.; Winter, H.: Maschinenelemente Band 2. Getriebe allgemein, Zahnradgetriebe – Grundlagen, Stirnradgetriebe. 2. Aufl. Berlin: Springer, 1983

[NISC11]

Nischwitz, A.; Fischer, M.; Haberäcker, P.; Socher, G.: Computergrafik und Bildverarbeitung: Band II: Bildverarbeitung. 3. Auflage, Vieweg +Teubner, Wiesbaden, 2011

[NN26]

Firmenschrift: Bauer & Schaurte: Rohe Schrauben. Neuss 1926

[POMP52]

Pomp, A.: Stahldraht. In: Stahleisen-Bücher. Band 1. Verlag Stahleisen M.B.H., Düsseldorf, 1952, S. 10-14

12