Вы находитесь на странице: 1из 7

Wear 271 (2011) 1651– 1657

Wear 271 (2011) 1651– 1657 Contents lists available at ScienceDirect Wear journal homepage:

Contents lists available at ScienceDirect

Wear

journal homepage: www.elsevier.com/locate/wear

Wear journal homepage: www.elsevier.com/locate/wear Short communication Experimental study of friction in sheet

Short communication

Experimental study of friction in sheet metal forming

L. Figueiredo, A. Ramalho , M.C. Oliveira, L.F. Menezes

CEMUC – Mechanical Engineering Department, University of Coimbra, Portugal

a r t i c

l e

i n f o

Article history:

Received 3 September 2010 Received in revised form 4 February 2011 Accepted 4 February 2011

Keywords:

Friction Tribology Sheet metal forming

a b s t r a c t

During deformation of the sheet metal over a tool, contact occurs only at the peak asperities of both surfaces. In the contact areas the processed material flows over the tool’s surface, therefore all the mod- els used to study forming processes must include a way to take into account the contact with friction phenomena. More widely used friction models are based in the Amontons-Coulomb theories. Unfortunately experience shows that for most applications the available models cannot accurately describe the friction phenomena. The determination of the friction coefficient in a sheet metal forming process is a complex procedure, because many variables influence the friction mechanisms. The aim of this research work is to apply an experimental approach in order to bridge simple benchmark friction experiments with real sheet forming applications. Two different techniques were used to assess friction, namely unidirectional crossed cylinders slid- ing with linear increase of the load and an equipment which allows measuring the friction coefficient under stretch-forming conditions in a sheet metal forming process. The tested materials are a cold-rolled advanced high-strength steel, DP600, and an aluminium 1100 alloy against heat-treated AISI D3 steel. The test protocols were established to allow the study of several effects: sliding speed, the surface roughness, the lubricant effect, the load and the running-in effect. The differences between the two techniques are widely discussed and laser profilometry and scanning electron microscopy are used to help understand the prevalent friction mechanisms. The present study allows concluding that: the friction results obtained by a load-scanning test are always higher than values assessed by a draw-bead test; roughness of the die material plays an important role on the friction coefficient; a significant reduction of friction was attained in multi-pass load-scanning tests due to running-in effect.

© 2011 Elsevier B.V. All rights reserved.

1. Introduction

Nowadays, numerical simulation has been widely accepted in the optimisation of forming processes owing to the advantage of the notable progress of computer capabilities. Significant benefits can be obtained especially on time-to-market and start-up costs by utilizing simulations. However, the advantage of applying numer- ical simulation of sheet metal forming operations results depends on the correct modelling of several topics [1,2]. Among these, con- tact conditions definition in conjunction with friction modelling assumes a decisive role. In fact, tribological properties, and fric- tional processes, are important factors determining the result of forming [3]. However, tribology itself comprises the interaction of different factors connected to the sheet metal surface. Thus, exper- imental research in sheet metal forming follows two directions:

Corresponding author. Tel.: +351 239790735. E-mail address: amilcar.ramalho@dem.uc.pt (A. Ramalho).

0043-1648/$ – see front matter © 2011 Elsevier B.V. All rights reserved.

doi:10.1016/j.wear.2011.02.020

- to understand contact conditions during sheet metal forming;

- to assess the influence of specific variables in sheet metal forming operations.

Therefore, experimental research should include classic tests, where each influencing parameter could be isolated and controlled, and technological modelling tests to allow transferability of the results from the laboratory to the different industrial metal forming operations. Concerning classic tests, Podgornik et al. [4] performed a study to evaluate galling properties of tool materials for metal forming operations comparing several tribological test methods. Among several possibilities, it was proved that load scanning is a very simple and suitable method to evaluate galling properties of tool materials. This method was successfully applied to study the effect of the surface treatment and the roughness of tool material as well the behaviour of different lubricants [4–8]. The assessment of friction during sheet metal forming oper- ations is a very complex task, and the laboratory test selected presents a fundamental importance on the friction results. Tisza and Fülöp [3] classified friction tests for sheet metal forming as a

1652

L. Figueiredo et al. / Wear 271 (2011) 1651–1657

function of the main performed operations: stretch forming; deep drawing; stretch drawing. Stretch forming was used by several authors [9–11] to study the influence of several parameters on friction during sheet metal forming. New die materials and surface treatments have been investigated [10,11] as well as the effect of blank material, surface roughness and lubricant viscosity. Instrumented deep-drawing tests were used especially to evaluate the effect of lubricants [12–14]. The stretch-drawing type tests are widely used to investi- gate the influence of several material and technological parameters

[15–23].

Among stretch-drawing tests, the draw-bead type test intro- duced by Nine [24] assumes an outstanding role. Sanchez [25] published results from an interlaboratory exercise with this tech- nique which reveals very accurate experimental results. In spite of the abundant number of laboratory tests that have been reported, effective numerical simulation is still hindered by accuracy of the contact and friction modelling. This difficulty can be understood due to the nature of the friction phenomena. The main objectives of this study are:

-

Understand friction during sheet metal forming using comple- mentary experimental methods: load-scanning and draw-bead friction tests.

-

Compare the results obtained by two different experimental tech- niques: one, the load scanning, with good control and which allows the study of the effect of each variable; the second, draw bead, replicates the sheet metal forming conditions, so the trans- ferability of the results is assured.

-

Testing the abilities of the recently developed draw-bead tester in different contact conditions.

2.

Experimental details

2.1. Load-scanning tests

The load-scanning test was done with two opposite cylindrical surfaces, with cross relative position, i.e. with the axes in perpen- dicular directions. Relative sliding motion during testing forms 45 in relation to each specimen axis, therefore the contact spot moves along a contact path on each specimen. This test procedure derives from the research work of Hogmark [4–7,26]. This type of test, with point-contact geometry, can be done with

a constant normal load or varying the load, using different loading waves, during the test. The sliding velocity is another test param- eter that can be adjusted. The equipment also allows changing the diameter of specimens, their roughness and the lubrication. More- over tests can be done applying single or multi-pass conditions. The equipment developed at the University of Coimbra has a high precision of motion and positioning and is numerically con- trolled [8]. The sliding motion corresponds to the movement of the horizontal base, where the specimen is fixed on a three-axis piezoelectric load cell. Normal load is applied by a spring, with well-defined constant rigidity, controlling the vertical motion of the upper specimen. Therefore, both the specimen path and the loading wave are numerically controlled. Both normal and tangen- tial forces, measured by the load cell, were acquired in real-time during the test. Default test conditions in this research were: load linear increase from 0 to 75 N with a sliding speed of 1 mm/s. Contact occurs between a cylinder with 5 mm diameter of the die steel, against

a sheet specimen with cylindrical shape. In this study sheets with

a thickness of 1 mm were conformed to obtain a cylindrical shell with a contact radius of 5 mm Fig. 1.

a cylindrical shell with a contact radius of 5 mm Fig. 1 . Fig. 1. Detail

Fig. 1. Detail of load-scanning test assembly.

The friction coefficient, as established by the Amontons- Coulomb model, corresponds to the linear proportionality ratio between the friction force and the normal load [27], Eq. (1). There- fore, each pair (friction force; normal force) allows the estimation of the friction coefficient value, as plotted in Fig. 2a.

F = N

(1)

coefficient value, as plotted in Fig. 2 a. F = N (1) Fig. 2. Effect of

Fig. 2. Effect of the normal load in the friction coefficient.

L. Figueiredo et al. / Wear 271 (2011) 1651–1657

1653

L. Figueiredo et al. / Wear 271 (2011) 1651–1657 1653 Fig. 3. Draw-bead friction test. (a)

Fig. 3. Draw-bead friction test. (a) Configuration of the deformation geometry. (b) Outline of the test assembly.

For these tests performed with increasing load, the friction coef- ficient is better determined as a slope, because in that case the friction coefficient is the relationship between the friction force and normal load increase and thus the offset errors are nullified [27]. This approach, Fig. 2b, allows the verification of the applicability of the linear Amontons-Coulomb model and permits the calculation of a safe friction coefficient value that is applicable to the entire loading range.

2.2. Draw-bead friction test

The draw-bead test allows the simulation of the bending and unbending in a sheet metal forming process and to measure the

friction coefficient during the sliding of the sheet against a die dur- ing the forming process [25]. To do this type of test, test equipment was especially designed in order to be used in conjunction with

a classical electromechanical tensile test machine. With this test equipment it is possible to measure the forces associated with the forming process. The required configuration is shown in Fig. 3. These tests are performed with a constant velocity and the pulling and normal force data is acquired by two load cells with

a rate of 100 Hz.

data is acquired by two load cells with a rate of 100 Hz. Fig. 4. Evolution

Fig. 4. Evolution of friction coefficient with increasing roughness (DP600 against

D3).

The pulling force of the sheet is the sum of the force of bending plus the friction force. Thus, in order to find the friction coefficient in the process it is very important know the contribution of these two effects to the pulling force. To resolve this problem two types of test are made. The first one, the “A type test” is made with the five rotary rolls built with bearings. In this assembly the pulling force measurement just takes into account the deformation force, because with these rollers the tests are done with negligible fric- tion. In the second one, the “B type test”, the rollers 1, 2 and 3, Fig. 4, are changed to fixed rollers. In this type of test the pulling measurement takes into account the deformation and the friction force. The friction coefficient is evaluated in these tests according to Eq. (2) [24], where F fs is the total pulling force, acquired in “B type test”, F fr is the pulling force without friction, acquired in “A type test” and F n is the normal force, measured during “B type test” by the corresponding load cell, Fig. 3b.

= F fs F fr F n

(2)

Default test conditions in this research were: fixed relative posi- tions of the rolls, constant speed of 1 mm/s and a sliding distance of 100 mm. The values of the force components used in Eq. (2) to calculate the friction coefficient are the average of the measure- ments in the steady-state part of the test. For each test condition a minimum of three repetitions were considered, further the aver- age value of friction coefficient and the standard deviation were also considered.

2.3. Materials

The tool steel AISI D3, tempered and quenched, was used as die material. Two cold-rolled sheet materials, with a thickness of 1 mm, were used in the friction test against AISI D3, namely: DP600 dual- phase steel and aluminium alloy AA 1100. Table 1 summarises the material properties. Table 2 summarises the different material combinations used in the different series of tests. All tests were done with lubrica- tion. Before testing, both sliding surfaces were cleaned with ethylic alcohol and lubricated with stamping oil, Fuchs Renoform MZAN 54. The average quantity of lubricant applied was around 25 g/m 2 .

Table 1

Mechanical properties.

Property

Material

AA 1100

DP600

AISI D3

Hardness (GPa) Yield stress (MPa) Ult. tensile stress (MPa)

0.8

5.87

7.75

105

362

110

651

1654

Table 2 Summary of test conditions and material combination.

L. Figueiredo et al. / Wear 271 (2011) 1651–1657

Study

Materials

Experimental technique

 

Load-scanning

Draw-bead

Comparison load-

DP600 against D3 AA1100 against D3 Lubricant Fuchs MZAN54

Normal load: linear 0–75 N Contact radius: 5 mm Sliding speed: 1 mm/s R z of D3: 1.66 and 2.48 m

Roll diameter: 21 mm Sliding speed: 1 mm/s R z of D3: 1.66 and 2.48 m

scanning/draw-bead

Influence of roughness

DP600 against D3 Lubricant Fuchs MZAN54

Roll diameter: 21 mm Sliding speed: 1 mm/s R z of D3: 1.66, 2.48, 3.66 and 4.47 m

Influence of running-in

DP600 against D3 AA1100 against D3 Lubricant Fuchs MZAN54

Normal load: linear 0–75 N Contact radius: 5 mm Multi-pass test: 5 passes R z of D3: 1.66 m

Lubrication regime

DP600 against D3 Lubricant viscosity:

Roll diameter: 21 mm Sliding speed: 1–8.47 mm/s R z of D3: 1.66 m

28.03–60.18 mPa s

In the lubrication tests, in addition to the stamping oil, three paraffinic industrial mineral oils, with viscosities ISO 32, 46 and 68, were used. Mahr-Rodenstock RM600 3D laser topography equipment was used to assess the roughness. Philips XL30 scanning electron micro- scope was used to investigate the wear surfaces.

3. Results and discussion

3.1.

Comparison between draw-bead and load-scanning friction

tests

In order to compare the results obtained by load-scanning and draw-bead testing systems, a set of tests was performed for both pairs of materials under study. Table 3 summarises the obtained results. Experimental results obtained for the aluminium alloy demon- strated the effect of the die material roughness on the friction. In this case, the effect of abrasion by steel asperities is determinant in the result. Therefore, an increase of the roughness induced a rise in the friction coefficient. Comparing friction coefficients obtained for the two tested pairs, in the same roughness conditions, the highest value corresponds to the steel sheet material, which should be due to the higher yield stress value. Comparing the results obtained by both techniques one can see that the same evolution has been obtained in both cases. How- ever, load scanning always produced lower friction values. This difference should be due to the highest contact pressure on the load-scanning test. In fact, the increase in contact pressure could induce a reduction in the friction by changing the transition from a mixed to a boundary lubrication regime [28,29]. The contact pres- sure in the draw-bead test can be calculated applying the formula developed for flat belt drives and band type brakes [30]. The contact pressure, p, is a function of the maximum pulling force in the metal sheet band, F fs , the band width, w, and the roll radius, r, Eq. (3). The values obtained for the maximum pressure are around 8 MPa. In the load-scanning friction test the contact occurs between crossed cylinders with equal radius; therefore, is a Hertzian type contact.

Table 3 Die roughness and friction coefficients (average values and standard deviation) measured by draw-bead and by load-scanning tests.

Materials

D3 Roughness R z /R a ( m)

Draw bead COF

Load scanning COF

DP600/D3

1.66/0.22

0.14 ± 0.019 0.11 ± 0.0014 0.13 ± 0.0007

0.083 ± 0.005 0.075 ± 0.018 0.11 ± 0.015

AA1100/D3

1.66/0.22

AA1100/D3

2.48/0.27

Considering the contact geometry and the maximum normal load, the maximum pressure value was 3130 MPa.

p =

F fs

wr

(3)

In spite of the differences obtained by the two techniques, the values agree with the majority of the published results. Using a hybrid numerical–experimental approach, Subramonian [31] achieved values in the range of 0.08–0.09, therefore similar to those obtained from load scanning. However, other authors [25,32] mea- sured values around 0.15, which agree with the values obtained in the current work applying the draw-bead technique.

3.2. Influence of roughness

In order to further evaluate the effect of surface roughness, draw-bead tests were done with the DP600 sheet against AISI D3 sliding cylinders, under different conditions. The D3 specimens were polished with different polishing routines, from emery paper grit P320 up to P2500, which corresponds to the test conditions of the results presented in the previous section. The different emery papers were applied with the rolls in rotation in a lathe, using a similar procedure to that used to prepare metallographic speci- mens. Before the tested rolls were cleaned in an ultrasonic bath of ethylic alcohol. The roughness of the cylinder sliding surfaces was measured by laser roughmeter equipment; Table 4 summarises the obtained results. Fig. 4 shows the evolution of the friction coefficient with the die surface roughness. One can see a clear tendency of an increase of the friction coefficient with the increase of the R z peak-to-peak roughness parameter. However this growth tends to stabilise for the highest roughness values. Considering the difference of hardness between the tested materials, and the low sliding speed, abrasion was the main contact mechanism. In fact, as displayed in Fig. 5, the wear occurs mainly by grooving. Therefore single asperity models [33] can be used to explain the increase of the friction coefficient with the raise of sur- face roughness. Furthermore, bearing in mind the low sliding speed used in these tests (1 mm/s) and the lubricant viscosity (40 mm 2 /s),

Table 4 Roughness parameters of different tested die surfaces ( m).

 

R a

R z

R k

R pk

R vk

P2500

0.22

1.66

0.57

0.48

0.14

P1000

0.27

2.48

0.84

0.55

0.44

P600

0.48

3.66

1.22

0.67

0.38

P320

0.52

4.47

1.97

1.45

0.38

L. Figueiredo et al. / Wear 271 (2011) 1651–1657

1655

L. Figueiredo et al. / Wear 271 (2011) 1651–1657 1655 Fig. 5. Micrograph of the contact

Fig. 5. Micrograph of the contact surface. The scratches parallel to the sliding direc- tion allow identification of abrasion by grooving as the main contact mechanism.

of abrasion by grooving as the main contact mechanism. Fig. 6. Running-in effect in (a) dual-phase

Fig. 6. Running-in effect in (a) dual-phase steel DP600 and (b) aluminium alloy

AA1100.

(a) dual-phase steel DP600 and (b) aluminium alloy AA1100. Fig. 7. Evolution of the friction coefficient

Fig. 7. Evolution of the friction coefficient for the pairs D3/DP600 D3/AA1100 with the number of passes.

for the pairs D3/DP600 D3/AA1100 with the number of passes. Fig. 8. Effect of the number

Fig. 8. Effect of the number of sliding passes in the aluminium wear track surface morphology for the D3/AA1100 pair: (a) 1 pass; (b) 3 passes and (c) 5 passes.

boundary lubrication was the expected regime. In this lubrication regime, friction is significantly dependent on the roughness.

3.3. Effect of running-in

Load scanning was used to investigate the evolution of the fric- tion in multi-pass tests. Two pairs of material were investigated in this study: AISI D3 cylinders against DP600 sheet steel and AISI D3 cylinders against aluminium alloy 1100. To verify the running- in effect multi-pass tests are made from one to five passes in the same track. The roughness of the AISI D3 cylinder was always R a = 0.27 m and R z = 2.48 m. Both materials tested reveal a reduction of the friction coeffi- cient with the increasing of the number of passes, Fig. 6. Fig. 6a and

1656

L. Figueiredo et al. / Wear 271 (2011) 1651–1657

1656 L. Figueiredo et al. / Wear 271 (2011) 1651–1657 Fig. 9. Effect of the number

Fig. 9. Effect of the number of sliding passes in the spectra of the aluminium rough- ness profiles for the D3/AA1100 pair (AR - as-received AA1100 sheet).

b displays the evolution of the tests corresponding to the 1st and 5th passes respectively for the pair D3/DP600 and D3/Al1100. A significant decrease of the friction coefficient has been ver- ified in both cases; however the pair D3/AA1100 demonstrated a stronger reduction. In fact the friction coefficient tends to be very small after the first four passes, Figs. 6a and b and 7. Blau [34] identified several parameters contributing to the running-in phe- nomenon. Among them, roughness is one of the most important, especially on boundary-lubricated contacts, as is the present case. In fact observing the wear tracks obtained in the AA1100 surface after the 1st, 3rd and 5th passes, Fig. 8, it is clear that abrasion is the main wear mechanism involved and that the surface becomes smoother with the increase in the number of passes. To understand this effect, the spectra of the roughness profiles were analysed for the as-received AA1100 sheet and on the wear track after the 1st and 3rd passes, Fig. 9. By these results one can conclude that the running-in effect is strongly effective in the reduction of the rough- ness, especially the higher wavelength components. Furthermore, the running-in effect occurs predominantly in the first pass. This strong and quick effect of running-in could be influenced by the higher value of pressure resulting from the point-contact geome- try. The present study was restricted to the running-in effect of the metal sheet; therefore, the results can only be used for understand- ing the cases where the sheet contacts several times with the tool, as is the case of progressive metal forming.

3.4. Effect of lubrication regime

Sheet metal forming operations involve a wide range of contact pressures and relative sliding speeds. Therefore, considering lubri- cated contacts, which is the current practice, the friction depends strongly on both surface and oil properties. The Stribeck curve is the usual way to analyse the evolution of the friction coefficient accord- ing to both operation parameters and lubricant characteristics. The Stribeck curve plots the coefficient of friction as a function of the Hersey parameter H, defined as H = v /p, where is the dynamic viscosity of the lubricant, v the speed and p the apparent contact pressure.

Table 5 Draw-bead test conditions to study lubrication effect.

Sliding speed [mm/s]

Dynamic viscosity [mPa s]

H [m]

1.00

28.03

1.44E-11

1.00

39.20

1.96E-11

2.00

39.20

4.00E-11

3.00

28.03

4.39E-11

3.00

39.20

6.02E-11

3.00

40.71

6.51E-11

4.00

39.20

8.01E-11

5.00

39.20

1.01E-10

6.00

39.20

1.24E-10

6.00

60.18

1.93E-10

7.50

39.20

1.53E-10

7.50

60.18

2.41E-10

8.47

39.20

1.75E-10

8.47

40.71

1.82E-10

8.47 39.20 1.75E-10 8.47 40.71 1.82E-10 Fig. 10. Variation of the friction coefficient as a

Fig. 10. Variation of the friction coefficient as a function of the contact conditions using a Stribeck type curve.

Especially important for metal forming is to know the transition conditions boundary/mixed/hydrodynamic lubrication regimes. To obtain a wide range of the Hersey parameter, as well as the stamp- ing oil, three paraffinic industrial mineral oils were used and the test speed was ranged from 1 to 8.3 mm/s. Table 5 summarises the test conditions and the obtained Stribeck curve is plotted in Fig. 10.

4. Conclusions

Tribological characteristics involved in sheet metal forming have been investigated using two experimental approaches: a load- scanning type tester and a recently developed draw-bead type device. The following conclusions can be drawn from this study:

1. Comparing the results obtained by both experimental tech- niques load scanning always produced lower friction values. This difference could be due to the highest contact pressure on the load-scanning test.

2. Roughness of the die material has a significant effect on the fric- tion coefficient.

3. A significant effect of reduction of friction by the running-in effect has been achieved by multi-pass load-scanning tests. The reduction of the friction occurs especially by the attenuation of the roughness components with high wavelength.

Acknowledgement

The authors are grateful to the Portuguese Foundation for Sci- ence and Technology (FCT) for the financial support for this work (project PTDC/EMETME/74152/2006).

L. Figueiredo et al. / Wear 271 (2011) 1651–1657

1657

References

[1] K. Roll, Simulation of sheet metal forming – necessary developments in the future, in: P. Hora (Ed.), Proceedings NUMISHEET’08, 7th International Con- ference and Workshop on Numerical Simulation of 3D Sheet Metal Forming Processes – Part A, 2008, pp. 859–863, ISBN: 978-3-909386-80-2. [2] Miklós Tisza, Numerical modeling and simulation in sheet metal forming aca- demic and industrial perspectives, Mater. Sci. Forum 473–474 (2005) 135–140. [3] Miklós Tisza, Tibor Fülöp, A general overview of tribology of sheet metal form- ing, J. Technol. Plast. 2 (2001) 11–25. [4] B. Podgornik, S. Hogmark, J. Pezdirnik, Comparison between different test methods for evaluation of galling properties of surface engineered tool surfaces, Wear 257 (2004) 843–851. [5] B. Podgornik, S. Hogmark, O. Sandberg, V. Leskovsek, Wear resistance and anti-sticking properties of duplex treated forming tool steel, Wear 254 (2003)

1113–1121.

[6] B. Podgornik, S. Hogmark, Surface modification to improve friction and galling properties of forming tools, J. Mater. Process. Technol. 174 (2006) 334–341. [7] B. Podgornik, S. Hogmark, O. Sandberg, Influence of surface roughness and coating type on the galling properties of coated forming tool steel, Surf. Coat. Technol. 184 (2004) 338–348. [8] A. Ramalho, Study of the relationships friction – contact stresses using

load scanner tests, in: Proceedings 11th Nordic Symposium on Tribology – NORDTRIB2004, 2004, pp. 699–707. [9] B.H. Lee, Y.T. Keum, R.H. Wagoner, Modeling of the friction caused by lubrica- tion and surface roughness in sheet metal forming, J. Mater. Process. Technol. 130–131 (2002) 60–63. [10] J.H.C. de Souza, M. Liewald, Analysis of the tribological behaviour of poly- mer composite tool materials for sheet metal forming, Wear 268 (2010)

241–248.

[11] V. Severo, L. Vilhena, P.N. Silva, J.P. Dias, D. Becker, S. Wagner, A. Cavaleiro, Tribological behaviour of W–Ti–N coatings in semi-industrial strip-drawing tests, J. Mater. Process. Technol. 209–10 (2009) 4662–4667. [12] D.D. Olsson, N. Bay, J.L. Andreasen, Int. J. Surf. Sci. Eng. 4 (1) (2010) 2–12. [13] Hyunok Kim, Ji Hyun Sung, Rajesh Sivakumar, Taylan Altan, Evaluation of stamping lubricants using the deep drawing test, Int. J. Mach. Tool Manuf. 47 (2007) 2120–2132. [14] Hongying Gong, Zhonling Lou, Zhiling Zhang, Studies on the friction and lubri- cation characteristics in the sheet steel drawing process, J. Mater. Process. Technol. 151 (2004) 328–333.

[15]

S. Hao, B.E. Klamecki, S. Ramalingam, Friction measurement apparatus for sheet metal forming, Wear 224 (1999) 1–7.

[16] K.J. Weinmann, S.K. Kernosky, Ann. CIRP 45 (1) (1996) 269–272. [17] H.Y. Kim, B.C. Whang, W.B. Bae, An experimental study on forming char- acteristics of pre-coated sheet steels, J. Mater. Process. Technol. 120 (2002)

290–292.

[18] S.S. Han, The influence of tool geometry on friction behavior in sheet metal forming, J. Mater. Process. Technol. 63 (1997) 129–133.

Z. Deng, M.R. Lovell, Effects of lubrication and die radius on the friction behavior of Pb-coated sheet steels, Wear 244 (2000) 41–51.

[20] M.R. Lovell, Z. Deng, Characterization of interfacial friction in coated sheet steels: influence of stamping process parameters and wear mechanisms, Tribol. Int. 35 (2002) 85–95.

[21] D. Wiklund, B.-G. Roseˇın, A. Wihlborg, A friction model evaluated with results from a bending-under-tension test, Tribol. Int. 42 (2009) 1448–1452. [22] A. Azushima, M. Sakuramoto, Effects of plastic strain on surface roughness and coefficient of friction in tension-bending test, Ann. CIRP 55 (1) (2006). [23] M. Samuel, Influence of drawbead geometry on sheet metal forming, J. Mater. Process. Technol. 122 (2002) 94–103. [24] H.D. Nine, Draw bead forces in sheet metal forming, in: Proceedings of a Sym- posium on Mechanics of Sheet Metal Forming: Behaviour and Deformation Analysis, Plenum Press, Warren, Michigan, 1978, pp. 179–211. [25] L.R. Sanchez, Characterisation of a measurement system for reproducible fric- tion testing on sheet metal under plane strain, Tribol. Int. 32 (1999) 575–586. [26] S. Hogmark, S. Jacobson, O. Wänstrand, A new universal test for tribological evaluation, in: D.J. Schipper (Ed.), Proceedings 21st Meeting of International Research Group on Wear of Engineering Materials, IRG-OECD, Amsterdam, March, 1999, pp. 25–26. [27] A. Ramalho, A reliability model for friction and wear experimental data, Wear 269 (2010) 213–223. [28] W.C. Emmens, J. Bottema, Friction of aluminium in deep drawing, in: Proceed- ings IDDRG Congress, 1998. [29] A. Azushima, J. Miyamoto, H. Kudo, Effect of surface topography of workpiece on pressure dependence of coefficient of friction in sheet metal forming, Ann. CIRP 47 (1) (1998) 479–482.

[30]

Shigley, Joseph Edward, et al., Mechanical Engineering Design, 7th ed., McGraw- Hill, 2004.

[19]

[31] B.E. Soumya Subramonian, Evaluation of lubricants for stamping deep draw quality sheet metal in industrial environment; M.Sc. Thesis, Ohio State Univer- sity, 2009. [32] ZhuWei, Z.L. Zhang, X.H. Dong, Deep drawing of rectangle parts using variable blank holder force, Int. J. Adv. Manuf. Technol. 29 (2006) 885–889.

[33]

[34] Peter J. Blau, Running-in: art or engineering? J. Mater. Eng. 13 (1991) 47–53.

E. Rabinowicz, Friction and Wear of Materials, 2nd ed., John Wiley & Sons, 1995.