Wrinkling in Sheet Metal Forming: Experimental Testing vs.

Numerical Analysis
A. Selman 1 *, E. Atzema 2, T. Meinders 3, A.H. van den Boogaard 3 and J. Huetink 3
1

Netherlands Institute for Metals Research,Rotterdamseweg 137 2628 AL Delft The Netherlands

Corus Research Development & Technology, IJTC, Automotive Applications P.O. Box 10000 1970 CA IJmuiden The Netherlands University of Twente, Department of Mechanical Engineering, P.O. Box. 217 7500 AE Enschede The Netherlands
___________________________________________________________________________ ABSTRACT: Following a number of publications on numerical prediction of wrinkling in thin sheet metal forming, the present part of our work is devoted to the comparison of numerical results with those obtained through experimental testing. A number of hemispherical product samples have been used with various blank holder forces and drawn to different depths to capture the onset of wrinkling, its mode and location. KEY WORDS: Numerical Simulation, Finite Elements, Sheet Metal Forming, Error Estimation, Wrinkling Prediction, Adaptive Mesh Refinement and Experimental Results. ___________________________________________________________________________
3

Please send all correspondence to: A. Selman, University of Twente, Mechanical Engineering Department, Po Box 217, 7500 AE Enschede, The Netherlands.

1. Introduction In a numerical simulation, wrinkles can be detected by a visual inspection of the deformed mesh, provided that the finite element discretisation is fine enough to allow a proper capture of the wrinkles. This, in general, makes it cumbersome to proceed with the analysis. Rather, it is desirable to proceed with a selective refinement to keep the computational cost low (acceptable). In this context [SEL 00, 01a, 01b, 02], [BON 92, 94], [ROD 97], [GEL 98], and more generally in finite element simulations [SEL 90, 93, 97], adaptive mesh refinement plays an essential role. However, this implies that some kind of discretisation error estimators and wrinkling indicators are used to direct the refinement process [SEL 00, 01a, 01b, 02], [AMZ 91], [NOR 97, 98], [BRU 97]. In this work, the analysis of Hutchinson and Neale [HUT 85], which consists of formulating the problem within the context of plastic bifurcation theory for thin shell elements and its extension by Neale [NEA 89] to account for more general constitutive models, is used. Under a number of assumptions, limitations and simplifications a simple wrinkling criterion with some restrictive applicability, is obtained. Nonetheless, the results are used to locally define a wrinkling risk factor or simply a wrinkling indicator, which, in turn, is used to detect the zones (elements) to be refined in a subsequent adaptive mesh refinement process. Hutchinson analysis is, unfortunately, limited to regions of the sheet that are free of any contact. When contact is taken into account the problem is further complicated. Furthermore, given that numerical simulations of complex sheet metal forming involve large scale models, it is obvious that global wrinkling indicators found in the literature - mostly based on eigen value analysis of the global tangent stiffness matrix - should not be used because of their high computational cost. This is to avoid over-loading the already time consuming deep drawing simulations. Consequently, an indicator based on the local change of curvatures has been developed [SEL 00, 01a, 01b, 02]. In wrinkling prediction analyses, the local curvature and the thickness, amongst other parameters, play a major role and should, therefore, be properly approximated from the finite element meshes at all stages of the computation. In this context, the incorporation of discretisation error indicators and adaptive mesh refinement in sheet metal forming processes is doubly important in keeping the computation cost low and allowing a comprehensive wrinkling prediction analysis. In recent publications we have presented numerical results on wrinkling prediction with Adaptive Mesh Refinement [SEL 00, 01a, 01b, 02]. However, in a number of occasions a recurrent question kept rising: How do the numerical results compare with those obtained in experimental testing ? To shed some light on such comparison, a hemispherical product has been used as a benchmark and a number of product samples stamped with various blank holder forces (BHF) and drawn to different depths to capture the onset of wrinkling, its mode and location.

In Section 2, the experimental setting is presented. The finite element code used in the numerical simulations as well as the comprehensive algorithm and indicators used are presented in Sections 3. In Section 4, the overall results are given and a comparison made. 2. Experimental procedure To provide data for the validation of the wrinkling model a number of experimental tests were carried out using a hemispherical punch that was available. The tests were held on a triple action 400 tonne hydraulic press (SMG) used in a dual action mode. The die was mounted on the ram and the punch and blank holder mounted on the bed. The blank holder was supported by 8 pins with a load cell on top of each to measure the actual applied blank holder force. Moreover the pins were shimmed to achieve best possible load distribution under the blank holder. From earlier tests experience [MED 00a, 00b], [BOT 97], the draw depth as well as the blank holder force at which local buckling occurs in this product were approximately known. The blanks were cold-rolled galvanised IF steel with 560 mm diameter and 0.80 mm thickness. Only a small amount of lubricant (Quaker N6130) was applied as otherwise the time to pressing is known to become of significant influence [HAA 00] and also disturbs the tests. The procedure was to apply some oil and then rub it of with a cloth until the surface appears matt. Experience of CRD&T at Corus shows that this amounts to about 1.0 to 1.5 g/m2. The buckling waves were measured on a 3D coordinate measuring machine (CMM), Mitutoyo BH506 and corrected for the slight off-centre of the specimen by the proprietary Scan pack software. Then they were filtered, to remove long wavelength shape defects that were not of interest in this research. This filtering was done by means of Fourier analysis and subsequent inverse Fourier analysis on part of the spectrum in a CRD&T written code. The experimental results are recapitulated in Tables 1 to 3 and compared with those obtained through numerical simulations. 3. FEM Simulation The numerical simulation has been carried out using our DiekA code [DIE 00], which has been dedicated to the simulation of forming processes. In these simulations, large deformations, history dependent material behaviour, contact phenomena and numerical techniques play an essential role. It should be stressed

that the contact behaviour is the driving force in most forming processes. However, it is highly non-linear and often leads to numerical difficulties. Therefore, special interface-elements have been developed and implemented in DiekA. Different friction models are also available along with thermal effects on the interface. Moreover, we have recently incorporated a number of routines covering discretisation error estimators as well as wrinkling prediction indicators for a comprehensive study of wrinkling. In the following subsections, the general algorithm for wrinkling prediction with adaptive mesh refinement is presented along with a brief description of the indicators used. For a detailed description refer to [SEL 01a, 01b]. 3.1 General and comprehensive algorithm for mesh adaptation in sheet metal forming The general and comprehensive algorithm for mesh adaptation in sheet metal forming processes comprises the following steps 1. 2. 3. Generate an initial grid to represent the computational domain and to allow an adequate initial solution Advance the solution for a number of steps Use the error indicator and wrinkling indicators to determine whether mesh refinement is necessary. If yes, compute a new mesh distribution and continue - otherwise go to step 2 Proceed with the mesh refinement and obtain the field values of the solution on the new grid by direct interpolation from the previous grid If the desired load interval has elapsed stop otherwise go to step 2

4. 5.

3.2 Discretisation errors estimation The error estimation used is entirely geometrical and is based on the accuracy with which the finite element mesh can describe the continuous exact shape [BON 92], [ROD 97], [SEL 01a, 01b]. In addition to the geometric error estimation, a thickness error, which measures the difference between the finite element solution and a higher order solution obtained by some recovery technique to substitute for the exact solution, is also taken into account. A good approximation of the geometry and thickness distribution is of primary importance in sheet metal forming, as these play an essential role in wrinkling prediction analysis.

3.3 Wrinkling analysis The basic theory of plastic buckling and relevant relations for the DMV shallow shell theory have been developed by Hutchinson [HUT 74]. The application of this theory to sheet wrinkling was first carried out by Hutchinson and Neale [HUT 85]. In this analysis, we consider a contact free sheet element, which, in the current stage of forming, has attained a doubly curved state with principal radii of curvature and thickness assumed to be constant over the region of the sheet being examined for susceptibility to wrinkling. Moreover, the stress state prior to wrinkling is assumed to be a uniform membrane state. Simplifications arise from the fact that the anticipated short-wavelength modes are shallow and can be analysed using DonnellMushtari-Vlasov (DMV) shallow shell theory. To determine the critical values for buckling Hutchinson bifurcation functional [HUT 74] is used. For the pre-wrinkling geometry and stress state considered above, wrinkling will in most cases be aligned with one of the principal curvatures (stresses) directions. In such a case the analysis simplifies considerably and a simple wrinkling criterion can be obtained. Using the wrinkling critical stress values, we define a wrinkling risk factor as the ratio of the principal stress and the critical stress. Therefore, a wrinkling risk exists whenever the wrinkling risk factor is larger than one. The risk is more important with larger values of the risk factors. In the adaptive procedure we need to detect the zones (elements) to be refined and determine a new size for those zones (elements). The zones detection is obtained by the use of the wrinkling risk factors. The new mesh size is obtained by the use of the wavelength of the wrinkles [SEL 00, 01a, 01b, 02]. As already stated above, Hutchinson analysis is limited to contact free wrinkling. Therefore, in the contact zones a different approach (indicator) had to be used. The indicator we have developed is based on the local change of curvatures (during a single deep drawing step) under compressive stresses [SEL 01a, 01b]. This filters out all changes in curvature that are not due to compressive stresses, such as those caused, for example, by the geometry of the tool and the die. Such changes are, however, taken into account by the (geometric) error estimation, if and when necessary. As for Hutchinson approach, the present indicator is local and is of a post-processing type, and is therefore determined at a very low computational cost. 4. Numerical and experimental results The performance of the wrinkling prediction procedure with adaptive mesh refinement briefly described in this work is here demonstrated and compared with experimental data.

A number of hemispherical product samples have been used with various blank holder forces and drawn to different depths to capture the onset of wrinkling, its mode and location. The following dimensions have been used Punch Radius (mm) Die Radius (mm) Die Shoulder Radius (mm) Blank Diameter (mm) Initial Thickness (mm) 146.5 148.0 20.0 560.0 0.8

Hollomons hardening law has been used with the following set of parameters and material data K (MPa) 527.28 0.3 n 0.225 E (MPa) 2.1*10
5

R0 2.19

R45 1.72 0 (MPa) 172

R90 2.50

A punch speed of 25 mm/s has been used. Coulomb friction coefficient is set to 0.1 for low BHFs and 0.16 for high BHFs, as little lubricant is used. A step size of 0.1mm is taken for all numerical examples and DST finite elements have been exclusively used. Before proceeding further, there is a need to define the onset of numerical as well as experimental wrinkling. Experimentally, it is obviously defined as the instant at which wrinkles are visible with a naked eye. Numerically, it is more difficult to define the onset of wrinkling. Nevertheless, we sought to use a definition which best fits experimental results. Consider the maximum curvature plots versus the time step (Figure 1 and 2). The numerical onset of wrinkling is defined as the instant at which the first point reaches the steep and ascending part of the curve with a constant tangent. The plots of the meshes presented in Figure 1 and 2 are made at the defined numerical onset of wrinkling and at the point of maximum curvature versus the time step. Three series of deep drawing experimental testing have been carried out. These are listed in Tables 1, 2 and 3. For the numerical simulation an average value for each group has been considered as shown in Table 4.

BHF (kN) Draw Depth (mm) (Onset)

130 110

130 105

130 100

130 95

110 100

90 100

Table 1. Series of experimental testing for various BHF showing the onset of wrinkling depth.

BHF (kN) Draw Depth (mm) (Onset)

150 100

150 110

150 105

150 107.5

150 115

Table 2. Series of experimental testing with a BHF of 150 kN showing the onset of wrinkling depth.

BHF (kN) Draw Depth (mm) (No Onset)

180 250

190 250

200 250

210 250

220 250

Table 3. Series of experimental testing of products drawn to 250mm depth with different BHF No wrinkling observed.

BHF (kN) Draw Depth (mm)

120 101 (Onset)

150 107 (Onset)

200 Up to 250 (No Onset)

Table 4. Average values of BHF and draw depth for the onset of wrinkling used in numerical simulations.

The number of waves observed in experimental testing and numerical simulation are listed in Table 5. BHF (kN) Experimental Testing Numerical Analysis 120 27 28 150 27 28 200 0 0

Table 5. Number of waves observed in experimental testing and numerical analysis.

BHF (kN) Exp. Onset Draw Depth (mm) Num. Onset Draw Depth (mm)

120 101 96

150 107 104

200 Up to 250 (No Onset) Up to 250 (No Onset)

Table 6. Comparison of numerical and experimental onset of wrinkling.

In the next subsections both wrinkling indicators along with discretisation error estimation are considered in situations where blank holder forces of 120, 150 and 200 kN are used. 4.1 Numerical and experimental results for the 120 kN BHF In this numerical simulation an average BHF value of 120 kN is used. The simulation is started with a coarse mesh comprising 2250 elements. At the numerical onset of wrinkling at step 958 the finite element mesh comprises 5550 elements. At the maximum of wrinkles development at step 1440 the adapted mesh comprises 7000 elements. The first refinements that take place prior to the onset of wrinkling are due to the thickness variation at the bottom of the product and the geometric error estimation in the region of the die shoulder. Thereafter, the wrinkling indicator senses a potential for wrinkling in the wall of the product and refines the mesh. The blank holder forces used in the experimental testing and the corresponding onset of wrinkling depths are shown in Table 1.

Step 958 5550 Elts

Step 1440 7000 Elts

8 7 6 5 4 3 2 1 0 0 500

BHF 120 kN

Max. Curvature * 100

1000
Step Number

1500

2000

Figure 1. Meshes and corresponding 2-principal curvature contour plots and max. curvature vs. step number (BHF 120 kN).

Figure 1 shows, at the onset of wrinkling (Step 958) and maximum development of wrinkling (Step 1440), the adapted meshes along with the corresponding contour plots of the second principal curvature. The plot of the maximum curvature vs. the time step is also reported in the same figure. As can be seen in Table 5 and 6 and the observation of the location of wrinkles, the numerical prediction is in good agreement with experimental results in terms of location and mode. The onset of wrinkling, as defined above, is also in good agreement with experimental testing although in numerical analyses wrinkles start developing slightly before the experimental ones. It has also been observed that the numerical simulation tends to overestimate the amplitude of the wrinkles. These observations are in accordance with other experiment testing [MED 00a, 00b], [BOT 97] with different die shoulders. 4.2 Numerical and experimental results for the 150 kN BHF In this numerical simulation as well as in the experimental testing a BHF of 150 kN is used. The experimental onset of wrinkling depths are shown in Table 2. The simulation is started with a coarse mesh comprising 2250 elements. At the numerical onset of wrinkling at step 1041 the finite element mesh comprises 5700 elements. At the maximum of wrinkles development at step 1450 the adapted mesh comprises 7300 elements. The first refinements that take place prior to the onset of wrinkling are, again, due to the thickness variation and the geometric error estimation. This is followed by a refinement due to the wrinkling indicator. Figure 2 shows, at the onset of wrinkling (Step 1041) and maximum development of wrinkling (Step 1450), the adapted meshes along with the corresponding contour plots of the second principal curvature. The plot of the maximum curvature vs. the time step is also reported in the same figure.

Again, the numerical prediction is in good agreement with experimental results in terms of location and mode. The onset of wrinkling, as defined here, is also in good agreement with experimental testing. Again, it has been observed that the numerical simulation overestimates the amplitude of the wrinkles.

Step 1041 5 700 Elts

Step 1450 7 300 Elts

BHF 150 kN

8 7 6 5 4 3 2 1 0 0 500 1000
Step Number

Max. Curvature * 100

1500

2000

Figure 2. Meshes and corresponding 2-principal curvature contour plots and max. curvature vs. step number (BHF 150 kN).

4.3 Numerical and experimental results for the 200 kN BHF In this numerical simulation an average BHF value of 200 kN as set in table 4, is used. A Coulomb friction coefficient of 0.16 has been used in this case. As expected the product showed no signs of wrinkling as reported in Table 3 and as can be seen in Figure 3, which presents meshes and contour plots at a couple of deep drawing depths. The product has been drawn up to 250 mm both experimentally and numerically and no wrinkles developed during the forming process. The simulation is started with a coarse mesh comprising 2250 elements and ended with an adapted mesh of 7750 elements. This is essentially due to the fact that the wrinkling indicator senses a potential for wrinkling and, mostly in anticipation refines the mesh. In the present case wrinkling does not develop and a mesh de-refinement is certainly a good option to implement in connexion with wrinkling prediction with adaptive mesh refinement. Figure 3 shows meshes at Step 1067 and 1469 and corresponding contour plots of the second principal curvature. Finally, we note that the numerical prediction is in good agreement with experimental results as no wrinkling is reported in both cases.

Step 1067 5 700 Elts

Step 1469 7 750 Elts

Figure 3. Meshes and corresponding 2-principal curvature contour plots showing no wrinkles (BHF 200 kN).

5. Conclusions A hemispherical product has been used as a benchmark test with various BHF and drawn experimentally in a series of tests at different forming depths to spot the onset, location and mode of wrinkling. Numerically, in a comprehensive approach to wrinkling prediction analysis, discretisation errors and wrinkling prediction indicators have been used, on one hand to better approximate the curvatures and thickness distributions, and on the other hand, to better capture and describe the wrinkling phenomena. As a bonus, the use of such indicators considerably reduces the overall computational cost. In comparing the numerical results with those obtained through experimental testing, a good agreement has been found in terms of location, onset and modes of wrinkling. However, it should be noted that numerical analyses tend to little overestimate the experimentally observed amplitudes of the wrinkles as well as it predicts an onset of wrinkling (as defined here) little earlier than experimentally observed. This makes numerical analyses, little conservative compared with experimental testing. Acknowledgments This research was carried out under project number ME97033 in the framework of the Strategic Research programme of the Netherlands Institute for Metals Research in the Netherlands (www.nimr.nl). 6. References
[AMZ 91] AMEZIANE-HASSANI, H. and NEALE, K., On the analysis of sheet metal wrinkling, Int. J. Mech. Sci., 33, p. 13-30, 1991. [BON 92] BONET, J. and WOOD, R. D., Mesh enrichment procedures for the finite element analysis of thin sheet forming processes, Num. Meth. In Ind. Processes, Chenot, Wood and Zienkiewicz (eds.), Balkema, Rotterdam, p. 221 - 227, 1992. [BON 94] BONET, J., Error estimators and enrichment procedures for finite element analysis of thin sheet large deformation processes, Int. J. Num. Meth. Eng., 37, p. 1573 - 1591, 1994. [BOT 97] BOTMAN and LIEMBURG, Testmethode m.b.t. secundaire plooivorming in het product, Corus Arch. Lab. 88170, 1997. [BRU 97] BRUNET, S., BATOZ, J. L. and BOUABDALLAH, S., Sur levaluation des risques de plissement locale de pieces industrielles obtenues par emboutissage, Actes du 3eme Colloque National en Calcul des Structures, p. 753-758, Giens, France, 1997. [DIE 00] DiekA 6.1 Manual, University of Twente, 2000.

[HAA 00] ter HAAR, R. and ATZEMA, E.H., The Impact of Lubricant Buffering on the Operating Window of Deep Drawing Processes, Proceedings of the 21st Biennial Congress of the International Deep Drawing Research Group, Ann Arbor, MI, USA, 2000. [HUT 74] HUTCHINSON, J. W., Plastic buckling, Adv. Appl. Mech., 14, p. 67-144, 1974. [HUT 85] HUTCHINSON, J. W. and NEALE, K. W., Wrinkling of curved thin sheet metal, Plastic Instability, J. Salencon (Ed.), Press Ponts et Chaussees, p. 71-78, 1985. [GEL 98] GELIN, J. C., PASQUIER P. and BOUDEAU, N., Prediction of necking and wrinkling in sheet metal forming from the analysis of local equilibrium conditions, Numiform 98, J. Huetink and F.P.T. Baaijens (Eds.), Balkema, Rotterdam, 1998. [NEA 90] NEALE, K.W. and TUGCU, P., A numerical analysis of wrinkling formation tendencies in sheet metals, Int. J. Num. Meth. Eng., 30, p. 1595-1608, 1990. [NOR 97] NORDLUND, P. and HAGGBLAD, B., Prediction of wrinkle tendencies in explicit sheet metal forming simulations, Int. J. Num. Meth. Eng., 40, p. 4079-4095, 1997. [NOR 98] NORDLUND, P., Adaptivity and wrinkle indication in sheet metal forming, Comput. Methods Appl. Mech. Engrg. 161, p. 127-143, 1998. [MED 00a] MEDENBLIK and LIEMBURG, Persproeven ter controle van simulatie berekeningen secundaire plooivorming, Corus Arch. Lab. 90428, 2000. [MED 00b] MEDENBLIK, LIEMBURG and SCHOUTEN, Invloed van smeringscondities op de vorming van secundaire plooien, Corus Arch. Lab. 90433, 2000. [ROD 97] RODRIGUEZ-VILLA, A., Etude theorique et experimentale de lextrusionsouflage de corps creux en polymere, PhD thesis, Ecole Nationale Superieure des Mines de Paris, France, 1997. [SEL 90] SELMAN, A., HINTON, E. and ATAMAZ-SIBAI,W., Edge effects in MindlinReissner plates using adaptive mesh refinement, Eng. Comput., 7, p. 217 - 226, 1990. [SEL 93] SELMAN, A., and HINTON, E., Two dimensional transient dynamic analysis with adaptive mesh refinement, Comm. Num. Meth. Eng., 9, p. 825 - 836, 1993. [SEL 97] SELMAN, A., HINTON, E. and BICANIC, N., Adaptive mesh refinement for localised phenomena, Comp. & Struct., 63, p. 475 - 495, 1997. [SEL 00] SELMAN, A., MEINDERS, T., van den BOOGAARD, A.H. and HUETINK, J., Wrinkling prediction with adaptive mesh refinement, 3rd ESAFORM Conference on Material Forming, Stuttgart, 11-14 April, 2000. [SEL 01a] SELMAN, A., MEINDERS, T., van den BOOGAARD, A.H. and HUETINK, J., Adaptive Numerical Analysis of Wrinkling in Sheet Metal Forming, Submitted to Int. J. of Forming Processes, 2001. [SEL 01b] SELMAN, A., MEINDERS, T., HUETINK, J. and van den BOOGAARD, A.H., Comprehensive approach to wrinkling prediction analysis in thin sheet metal forming processes, Submitted to Int. J. Num. Meth. Eng., 2001.

[SEL 02] SELMAN, A., MEINDERS, T., van den BOOGAARD, A.H. and HUETINK, J., On Adaptive Mesh Refinement in Wrinkling Prediction Analysis, 5th ESAFORM Conference on Material Forming, Krakow, 14-17 April, 2002.

A. Selman holds a Master degree from the University of Wales Swansea. He was privileged to be in the last MSc class taught by Prof. O.C. Zienkiewicz at Swansea. He also holds a PhD degree from the same university and was honoured to be supervised by late Prof. E. Hinton. He first joined the group of Prof. J.L. Batoz at Compiegne (France) in 1993, where he worked, amongst other things, on adaptive mesh refinement for 2D and 3D problems and stability analysis of metal cans. In 1999 he joined the Netherlands Institute for Metals Forming with an office at the University of Twente where he works in the group of Prof. Hutink. His research interests have been focused on adaptive mesh refinement in static and dynamic, linear and non-linear for 2D and 3D structural problems and more recently on wrinkling prediction analysis. E. Atzema holds a Master degree in Mechanical Engineering from the University of Twente. He subsequently did his PhD in the same group supervised by Prof. Hutink. In 1995 he started working at Hoogovens R&D (currently known as Corus Research - Development & Technology in the department of Automotive Applications) where led a 4-year project on material influence on stamping defects. Currently, he is in charge of FEM analysis of stamping and general formability research mainly on High Strength Steels as well as Aluminium and Sandwiches. His research work is aimed at providing methodology for customer support in stamping problems. T. Meinders holds a Master of Science degree in Mechanical Engineering from the University of Twente, the Netherlands. He also holds a PhD degree supervised by Prof. Hutink from the same university. Recently, he has been appointed as a staff member in the Applied Mechanics group of the University of Twente under the leadership of Prof. Hutink. His research interests have been focused on sheet metal forming, and in particular on tailored blanks, drawbeads, material models, adaptive mesh refinement and improvement of the robustness of implicit codes. A.H. van den Boogaard graduated in Mechanical Engineering at the University of Twente in 1988. From 1989 to 1995 he worked at the Netherlands Organisation for Applied Scientific Research (TNO) in the Computational Mechanics section of the Institute for Building and Construction Research where he took part in the development of the Diana finite element code. In 1995 he returned to the University of Twente to become a researcher/lecturer in finite element analysis. His current research focus is on sheet metal simulations and numerical methods. J. Hutink studied Mechanical Engineering at the Universities of Delft and Twente. He received his PhD degree from the University of Twente under the supervision of Prof. A. Rijken and Prof. J.F. Besseling. He first worked at the Institute for Mechanical Engineering of the Dutch Organization for Applied Scientific Research (TNO) (1975-1980). Then, he returned to the University of Twente, where in 1992 he was appointed as a full professor in mechanics of forming processes. He is member

of the Program Council and cluster leader of the Netherlands Institute for Metals Research (NIMR). He is member of the NUMIFORM Steering Committee and organizer of NUMIFORM 98 conference and member of the Board of Directors ESAFORM (European Scientific Association for Material Forming).