Investigation of tribological condition in cold forging using an optimized design of spike forging test

Abstract

The spike forging test is still commonly used although there are some new tribology test methods, but each designer would like to give their own parameters. Considering the spike height difference between low friction and high friction as the evaluation index, an optimal spike forging test developed using finite element simulations based on orthogonal optimization method and sensitivity analysis of tribological conditions of selected key design parameters was carried out. The spike height does not always monotonically decrease with the increase in friction factor in the scheme with improper parameters; therefore, the design of spike forging test should be critically evaluated before assessing tribological conditions. According to the optimized parameters, a simplified set-up of spike forging test was designed, and two retainers with a clearance fit and the billet with a chamfer were prepared to position the test billet. Finally, four different tribological conditions, including dry polytetrafluoroethylene lubricant, multipurpose grease, VG32 oil, and dry condition, in aluminum forging were chosen as a case study for the optimized design of spike forging test, and correspondingly, the lubricating effect was distinguished.

Keywords

Spike forging test spike height difference optimization tribological conditions

Introduction

With combined deformation modes of extrusion and upsetting, the spike forging process contains typical features of many forged products. Considerable interest has been given to study the spike forging process used as the finite element (FE) method.¹ Using the FE code ALPID, the effect of die chilling on metal flow during spike forging was investigated under different heat transfer conditions between the dies and the billets of Ti6242 alloy.² The generalization of a geometric linear viscoplastic model to finite strains and its numerical application was introduced and numerical computations of the spike forging process showed that the developed FE model enables accurate and reliable prediction of the studied spike forging process.³ An updated Lagrangian method was also employed in the FE simulation using an automated remeshing procedure to overcome the difficulty of severely distorted mesh during the spike forging process.⁴

An optimal die shape design was achieved by using a polynomial network and a genetic algorithm in a combined extrusion and forging process.⁵ A simple model for measuring heat transfer coefficients between the workpiece and tooling in hot spike forging was developed and implemented using a user routine in FE code DEFORM 2D.⁶ Material constitutive models were developed in spike forging of a CrMoV alloy steel using FE simulation with experimental validation.⁷

The spike forging test (SFT) was also used to investigate the formability of superplastic forming of an aluminum alloy using a mechanical press.⁸ The spike forging process can also be adopted to study material flow behavior and die filling characteristics under different hot forging conditions using different lubricants.⁹ Based on the spike height variations and forging and ejection loads, spike forging process named SFT was used to evaluate tribological conditions using different lubricants.^10,11 Since then, the SFT has been used to evaluate the lubricating properties of forging processes.

Except the SFT, there are lots of different kinds of tribology test methods which were proposed to investigate the tribological behavior during forging processes, such as ring compression test (RCT),¹² double cup extrusion test (DCET),^13,14 twist compression test,¹⁵ upsetting sliding test,¹⁶ tip test,^17,18 T-shape compression test,¹⁹ and sliding compression test (SCT)^20,21 Each tribology test method has its own advantages, and their characters have been carefully discussed in the literatures.^19,20

Although there are some new test methods, SFT is obviously more representative of different bulk deformation modes rather than simple form of material deformation under compression, and the SFT is still commonly used in industry. However, there are considerable variations in the design and use of different parameters for SFT, which often result in inconsistency in evaluation of tribological conditions. Therefore, the aim of this study is to optimize SFT parameters and to develop a simplified set-up for SFT so that the optimized SFT parameters and set-up can be commonly used to evaluate tribological conditions under different forging conditions using different lubricants.

Optimization of SFT

Design parameters

As shown in Figure 1, the spike forging die is defined by four geometric parameters, that is, top diameter d, transition radius r, cone angle α, and inclination angle β. During the test, a cylinder billet is deformed between the spike die and a flat die. The initial height H₀ and diameter D₀ of the billet are also considered to be the design parameters.

Figure 1.

Geometric parameters of spike forging die.

The ratio of the initial diameter and the height D₀/H₀ is used to define the billet geometry. The diameter ratio d/D₀ reflects the degree of material deformation. The transition radius r, cone angle α, and inclination angle β are the design parameters of SFT. Different design parameters were used in SFT by different researchers and the variations of different values are summarized in Table 1.

Table 1.

Key parameters of spike forging test.

No.	D₀/H₀	d/D₀	α (°)	r (mm)	β (°)	References
1	1.0 (40 mm/40 mm)	∼0.5 (∼20 mm/40 mm)	–	–	–	Lührs et al.³
2	1.33 (40 mm/30 mm)	0.2 (8 mm/40 mm)	2	2	–	Wu and Hsu⁵
		0.25 (10 mm/40 mm)	7	4
		0.3 (12 mm/40 mm)	12	6
3	0.89 (50.8 mm/57.15 mm)	0.3 (15.24 mm/50.8 mm)	6.7	19.05	15.7	Bennett⁷
4	–	0.6 (3 mm/5 mm)	–	2	0	Ishikawa et al.⁸
5	0.67 (20 mm/30 mm)	∼0.5 (∼10 mm/20 mm)	3	5	10	Isogawa et al.¹¹
6	1.0	0.25 (10.2 mm/40.6 mm)	2.5	0	5	Xu and Rao²²
	1.49		5	5	10
	2.0		7.5	8.5	15
	3.0		7.5	8.5	15
7	0.75 (30 mm/40 mm)	–	3	5	10	Sheljaskow²³
8	0.79 (32 mm/40.3 mm)	0.44 (14 mm/32 mm)	3	5	10	Liewald et al.²⁴

To achieve optimized design for SFT, the selection of the four parameters, including d/D₀, α, r, and β, should be chosen so as to achieve the largest possible variations of the spike height. As given in Table 1, four levels of variations of these four design parameters were used in the parametric study of SFT. According to the results that a billet of medium height is appropriate to evaluate interfacial friction,²² the value of D₀/H₀ is defined between 0.91 and 2.0, as shown in Table 2.

Table 2.

Design variables and levels of spike forging test.

Variables	Level
Variables	1	2	3	4
D₀/H₀	0.91 (20 mm/22 mm)	1.11 (20 mm/18 mm)	1.43 (20 mm/14 mm)	2.00 (20 mm/10 mm)
d/D₀	0.2 (4 mm/20 mm)	0.3 (6 mm/20 mm)	0.4 (8 mm/20 mm)	0.5 (10 mm/20 mm)
α (°)	3	5	7	9
r (mm)	2	4	6	8
β (°)	0	5	10	15

Optimization objectives

As the spike height H after deformation varies depends upon the tribological conditions at the interface between the workpiece and the spike forging die, the spike height H is an evaluation index as shown in Figure 2. Generally, the larger the spike height, the better the tribological condition.^11,22

Figure 2.

The spike height after deformation.

Considering the change of billet dimensions, the relative spike height H′ = H/H₀ is used as a measure of the material deformation under the same height reduction of the billet. To achieve the maximum sensitivity of the friction conditions by measuring the spike height from SFT, the absolute height difference between low and high frictions $δ = | {H'}_{low} - {H'}_{high} |$ is used as the optimization objective, that is, the greater the δ value, the greater the friction sensitivity of the SFT.

FE-based design of experiments

To carry out the optimization, the orthogonal experimental design is adopted and a L₁₆ (4⁵) table is determined as given in Table 3, which requires 16 simulations at low-friction (m = 0.1) and 16 simulations at high-friction (m = 0.3) conditions. A two-dimensional (2D) axisymmetric FE model in MSC (Marc) was built to run FE simulations under different friction conditions, and the velocity of spike forging die was assumed as 10 mm/s. Al6082 was used as the test material in SFT. A specific material law σ = 191.88ε^0.18 was derived from tensile testing, and the typical shear friction law was used to define the interfacial behavior. In FE simulations, the stroke is controlled to reach 70% of H₀ as shown in Figure 2.

Table 3.

Orthogonal table and analysis of results.

No.	D₀/H₀	d/D₀	α (°)	r (mm)	β (°)	${H'}_{low}$ (m = 0.1)	${H'}_{high}$ (m = 0.3)	δ
1	0.91	0.2	3	2	0	0.6132	0.5825	0.0307
2	0.91	0.3	5	4	5	0.7370	0.7198	0.0172
3	0.91	0.4	7	6	10	0.7783	0.7526	0.0257
4	0.91	0.5	9	8	15	0.8192	0.7748	0.0444
5	1.11	0.2	5	6	15	0.7773	0.7745	0.0028
6	1.11	0.3	3	8	10	0.8247	0.8051	0.0196
7	1.11	0.4	9	2	5	0.7802	0.7703	0.0099
8	1.11	0.5	7	4	0	0.7733	0.7624	0.0109
9	1.43	0.2	7	8	5	0.8834	0.8855	0.0021
10	1.43	0.3	9	6	0	0.8347	0.8368	0.0021
11	1.43	0.4	3	4	15	0.7843	0.7929	0.0086
12	1.43	0.5	5	2	10	0.7512	0.7574	0.0062
13	2.00	0.2	9	4	10	0.8539	0.8685	0.0146
14	2.00	0.3	7	2	15	0.7933	0.8093	0.0160
15	2.00	0.4	5	8	0	0.8937	0.8904	0.0033
16	2.00	0.5	3	6	5	0.8675	0.8724	0.0049
${H'}_{low}$ (m = 0.1)
k1	2.9477	3.1278	3.0897	2.9379	3.1149
k2	3.1555	3.1897	3.1592	3.1485	3.2681
k3	3.2536	3.2365	3.2283	3.2578	3.2081
k4	3.4084	3.2112	3.2880	3.4210	3.1741
R_l	0.4607	0.1087	0.1983	0.4831	0.1532
${H'}_{high}$ (m = 0.3)
k1	2.8297	3.1110	3.0529	2.9195	3.0721
k2	3.1123	3.1710	3.1421	3.1436	3.2480
k3	3.2726	3.2062	3.2098	3.2363	3.1836
k4	3.4406	3.1670	3.2504	3.3558	3.1515
R_h	0.6109	0.0952	0.1975	0.4363	0.1759
δ
k1	0.1180	0.0502	0.0638	0.0628	0.0470
k2	0.0432	0.0549	0.0295	0.0513	0.0341
k3	0.0190	0.0475	0.0547	0.0355	0.0661
k4	0.0388	0.0664	0.0710	0.0694	0.0718
R	0.0990	0.0189	0.0415	0.0339	0.0377

After each FE simulation, the spike heights of different schemes were measured, and ${H'}_{low}$ (m = 0.1), ${H'}_{high}$ (m = 0.3), and δ were calculated as given in Table 3. From Table 3, it can be found that ${H'}_{high}$ (m = 0.3) is larger than ${H'}_{low}$ (m = 0.1) in some schemes including schemes 9–14 and 16, and it means that the spike height increases with the friction factor which is different from conventional discussions.^11,22 Therefore, further discussions need to be carried out. So, all results are classified into three groups according to the level of spike height difference, that is, schemes 13, 14 in group A, schemes 11, 12 in group B, and schemes 9, 10, and 16 in group C, respectively. The simulation results of the total equivalent plastic strain of these schemes are shown in Figure 3.

Figure 3.

Simulation results with total equivalent plastic strain of some schemes (m = 0.1): (a) scheme 13, (b) scheme 14, (c) scheme 11, (d) scheme 12, (e) scheme 9, (f) scheme 10, and (g) scheme 16.

For group A spike forging testing, there is no material deformation between the flat part of dies, as shown in Figure 3(a). The friction force on the tapered surface shown in Figure 3(b) could contribute to the material filling of the spike die, and this may be the main cause for the increase in the spike height with the increase in friction factor. For group B spike forging testing, the extrusion is the main deformation mode because of the relatively large diameter ratio d/D₀. The friction force on the tapered surface results in relatively large ${H'}_{high}$ . There is a local severe deformation from scheme 12 because of a smaller transition radius used in the spike forging design, as shown in Figure 3(d). For group C spike forging testing, the large transition radius makes the material flow into the spike die much more easily and the relative height of the spike can be reached, especially in schemes 9 and 10.

According to the rules of orthogonal design method, the k value is calculated by the sum of the evaluation index of each parameter at the same level, and the R value is the difference between the maximum and minimum of the k value of each parameter. The range R listed in Table 3 indicates the effects of five design parameters on the spike height difference δ. The order of the influence of these parameters is as follows: the diameter and height ratio D₀/H₀ > the cone angle α > the inclination angle β > the transition radius r > the diameter ratio d/D₀. And the diameter and height ratio D₀/H₀ are more significant than the others, and this means that the billet geometry is most sensitive to the formed spike, as reported by Xu and Rao.²²

As shown in Figure 4, the spike height difference changes significantly with the diameter and height ratio D₀/H₀ with the maximum value obtained at the first level. The spike height difference decreases at first and then increases with the cone angle α, transition radius r, and inclination angle β with the maximum values found at the fourth level. The spike height difference changes slightly with the diameter ratio d/D₀ with the maximum value at the fourth level and the second large value at the second level.

Figure 4.

The change curve for effect of each parameter on the optimization objective: (a) diameter and height ratio D₀/H₀, (b) diameter ratio d/D₀, (c) cone angle α, (d) transition radius r, and (e) inclination angle β.

According to the above analysis, the optimal design parameters are the diameter and height ratio at level 1 and the other parameters at level 4, that is, D₀/H₀ = 0.91, d/D₀ = 0.5, α = 9 °, r = 8 mm, and β = 15 °. The optimal scheme is the fourth scheme in Table 3, and the corresponding spike height difference δ is 0.0444.

The ranges R_l and R_h indicate that the diameter ratio d/D₀ is the least influential affecting parameter. In this study, another optimal scheme with the diameter ratio d/D₀ at level 2, that is, d/D₀ = 0.3, is also chosen. FE simulation of this optimal scheme with a chamfered billet, defined as the third optimal scheme, is carried out at the same time for comparison reasons.

The simulated deformation processes of the three optimal schemes are shown in Figure 5, and the results of ${H'}_{low}$ (m = 0.1), ${H'}_{high}$ (m = 0.3), and δ are given in Table 4. In the Opt1 scheme, the deformation mode is more representative only to forward extrusion without container die (see Figure 5(b)), although the spike height difference δ is 0.0444, the combination of extrusion and upsetting feature is not obvious, so it is not a preferred scheme. In the Opt2 scheme, the combination of extrusion and upsetting deformation takes place and the spike height decreases with the friction factor. The spike height difference δ reaches 0.284 and hence considered to be a better scheme. In the Opt3 scheme, the billet has 1 × 30 ° chamfer at the top and bottom end surface, this chamfer is helpful to improve the spike height because it could make the material much easier to flow into the spike die, and the spike height difference δ increases to 0.299.

Figure 5.

Simulation results of optimal schemes: (a) height reduction = 0% (m = 0.1), (b) height reduction = 50% (m = 0.1), (c) height reduction = 70% (m = 0.1), (d) height reduction = 70% (m = 0.3), (e) height reduction = 0% (m = 0.1), (f) height reduction = 50% (m = 0.1), (g) height reduction = 70% (m = 0.1), (h) height reduction = 70% (m = 0.3), (i) height reduction = 0% (m = 0.1), (j) height reduction = 50% (m = 0.1), (k) height reduction = 70% (m = 0.1), and (l) height reduction = 70% (m = 0.3).

Table 4.

Results of optimal schemes.

	D₀/H₀	d/D₀	α (°)	r (mm)	β (°)	${H'}_{low}$ (m = 0.1)	${H'}_{high}$ (m = 0.3)	δ
Opt1	0.91 (20 mm/22 mm)	0.5 (10 mm/20 mm)	9	8	15	0.8192	0.7748	0.0444
Opt2	0.91 (20 mm/22 mm)	0.3 (6 mm/20 mm)	9	8	15	0.7890	0.7606	0.0284
Opt3	0.91 (20 mm/22 mm) with 1 × 30° chamfer	0.3 (6 mm/20 mm)	9	8	15	0.7905	0.7606	0.0299

To evaluate the sensitivity of the spike forging with respect to friction conditions, the relative spike heights under different friction factors from m = 0.1 to m = 0.5 of the optimal schemes and the worse schemes including schemes 9 and 10 were calculated as shown in Figure 6. Compared to the Opt1 scheme, the Opt3 scheme is still quite sensitive to the friction condition, and the spike height decreases with the increase in friction factor. As shown in Figure 6(b), the sensitivity of schemes 9 and 10 is too small to the whole range of friction conditions as expected. However, the spike height slightly fluctuates with the increase in friction factor. Therefore, when conducting the SFT to evaluate the tribological conditions, the corresponding key parameters should be carefully designed. Otherwise, the evaluation results would not lead to sensible conclusions.

Figure 6.

Relative spike heights under different friction factors: (a) optimal schemes and (b) worse schemes.

Design of SFT

To conduct spike forging testing, a normal forging die structure is needed. A surface with shallow concentric grooves on the platen was provided to position the billet during the SFT, and two load cells were used to collect the data of forging and ejecting loads.²³ However, it causes more complexity in the structure. A groove profile on the lower end of the punch prevents transverse material flow to the main tool motion, and a cylindrical billet with a small boss instead of cylindrical billet was mentioned²⁴ which would help the positioning of the billet in the spike die. In this work, the Opt3 scheme is adopted as the final optimal solution, and a simplified SFT set-up without shallow concentric grooves on the lower flat die is designed as shown in Figure 7. The corresponding optimized parameters of the die are as follows: top diameter d = 6 mm (D₀ = 20 mm), cone angle α = 9 °, transition radius r = 8 mm, and inclination angle β = 15 °. The fixed platen with left turn screw is used to install the set-up on the Instron 250 kN testing machine, and four bolts are used to connect the spike die with fixed platen, and the platen without any shallow concentric groove is put on the platform on the testing machine. Both the spike die and platen are made of H13 steel, hardened by heat treatment, and re-hardened by the Tufftride process. The working surfaces of dies were treated by lapping and polishing process, and the mirror surface with a surface roughness of Ra = 0.05 μm was achieved.

Figure 7.

(a) Drawing and (b) photograph of spike die test set-up.

Two retainers with clearance fit and the billet with chamfer were designed to position the billet in the center of the spike die. The fine positioning wheel on the Instron control panel was used to adjust the contact between the billet and the spike die, and the contact force was mentioned during the process. The spike die was moved down slowly until there was approximately 0.05 kN of force applied to the billet, and this allowed the billet to be automatically centered during this process. During the test, the forging load was collected by the Bluehill2 data acquisition system of the Instron machine.

Experimental testing

Three different lubricants including VG32 oil, multipurpose grease, and a dry polytetrafluoroethylene (PTFE) lubricant were used in the spike forging testing. For the VG32 oil, its density at 15 °C is 0.87 g/mL and the kinematic viscosity at 40 °C is 32 mm²/s. For the multipurpose grease, the relative density at 20 °C is 0.91 g/mL and it is insoluble in water. For the dry PTFE lubricant, the density is 0.667 g/mL and its main component is PTFE.

A group of chamfered cylinder billets were also prepared from annealed Al6082 material. Four tribological conditions were evaluated by using the optimal SFT. Under dry friction condition, the billets were forged without any lubricants. On applying VG32 oil, the billets were brushed as evenly as possible, while in the case of multipurpose grease, the billets were carefully coated manually. Using dry PTFE lubricant, both the billet and the spike die were sprayed using spray nozzle. When the tests were carried out, the target stroke and compress rate were set as 15.5 mm and 0.1 mm/s, respectively, and then the Instron testing machine worked at a constant speed and stopped when the stroke reached 15.5 mm. The low speed could help to achieve more precise final stroke of the die; therefore, the scattering of spike height could be decreased, and this could make the evaluation result more steady. Furthermore, the low speed could protect the testing machine which is daily used in the lab.

The stroke-load curves for four different lubricating conditions of all the tests are presented in Figure 8 and the corresponding maximum forging loads were extracted and recorded in Table 5. The selected samples after the SFTs are shown in Figure 9, and the spike heights of samples were measured and also listed in Table 5. Under the dry condition, the compression was stopped before the target stroke because of the initial set of 240 kN force limit of the Instron machine, and because of this, the spike heights were not measured.

Figure 8.

(a) Whole and (b) local stroke-load curves of all the tests.

Table 5.

Lubricating conditions and testing results.

No.	Lubrication condition	Spike height (mm)				Forging load (kN)
No.	Lubrication condition	Sample 1	Sample 2	Sample 3	Average	Sample 1	Sample 2	Sample 3	Average
1	Dry condition	–	–	–	–	241.98	242.05	242.11	242.05
2	VG32 oil	8.05	8.08	8.18	8.10	234.29	232.70	227.58	231.52
3	Multipurpose grease	8.21	8.14	8.16	8.17	225.40	224.87	228.55	226.27
4	Dry PTFE lubricant	8.68	8.88	8.79	8.78	219.71	217.96	220.88	219.52

Figure 9.

Selected samples of spike forged specimens.

From the data shown in Table 5, it is easy to evaluate the different tribological conditions. The order of lubricating effect is as follows: dry PTFE lubricant > multipurpose grease > VG32 oil > dry condition.

Discussion

During spike forging process, the cylindrical billet is extruded and upset at the same time, and its deformation feature is very typical in many actual forging processes. The deformation mode always changes with the design parameters. Sometimes the extrusion is the dominate form of deformation, while at other times, the upsetting can be the dominate form of deformation. In the end of spike forging process, there is a balance between extrusion and upsetting.

The design geometric parameters of the spike die and billet play an important role in affecting the spike height after deformation and the deformation force. In some cases, the spike height does not change with the friction conditions significantly and this could lead to confusing results for lubricating evaluation.

Because of these issues mentioned above, the SFT should be critically evaluated before its implementation in assessing tribological conditions in metal forming. Proper design parameters can be found by optimization based on FE simulation results.

Except for the die shape, which can directly influence the surface pressure, sliding length, and surface expansion, the temperature generated in cold forging, the deformation speed, the thickness, toughness, and other properties of lubricant film, and the micro feature of surface topography of workpiece and dies can play important effect on the friction behavior in cold forging. Therefore, more experimental works and further analysis should be performed in the future work.

Conclusion

Based on literature review, the key parameters of the SFT were selected and their variation ranges were defined. To achieve higher sensitivity from the SFT to friction condition, an optimal spike test was obtained by the orthogonal optimization method based on FE simulation results. In the optimization, the spike height difference is defined as the objective, and the optimized spike forging parameters are the diameter and height ratio D₀/H₀ = 0.91, diameter ratio d/D₀ = 0.3, cone angle α = 9 °, transition radius r = 8 mm, and inclination angle β = 15 °.

The sensitivity analysis of the spike forging to different tribological conditions of three optimal schemes and the worse schemes shows that the spike height does not always monotonically decrease with the increase in the friction factor in the worse schemes. Therefore, optimal parameters need to be used in SFT for evaluating tribological conditions.

According to the optimization results, a simplified spike die test set-up was designed. Two retainers with a clearance fit and the billet with a chamfer were used to help position the billet in the center of the spike die. And the measured spike height and the maximum forging load collected from data acquisition system were used as the evaluation indexes.

Finally, four different tribological conditions in spike forging of aluminum billets were evaluated by the optimal SFT in the lab. By comparing both the spike heights and the forging loads, the order of lubricating effect was dry PTFE lubricant > multipurpose Grease > VG32 oil > dry condition.

Footnotes

Acknowledgements

The authors wish to thank Mr Tom Buss for his support in experimental testing.

Academic Editor: Yunn-Lin Hwang

Declaration of conflicting interests

The authors declare that there is no conflict of interest.

Funding

This work was supported by the National Natural Science Foundation of China (no. 51475294) and partially supported by FP7-Marie Curie Action IRSES MatProFuture project (no. 318968).

References

. Finite element analysis of metal forming processes with arbitrarily shaped dies. Int J Mech Sci 1982; 24: 479–493.

Vardan

Shen

. Investigation of metal flow in non-isothermal forging using ring and spike tests. CIRP Ann Manuf Technol 1988; 37: 225–230.

Lührs

Hartmann

Haupt

. On the numerical treatment of finite deformations in elastoviscoplasticity. Comput Meth Appl Mech Eng 1997; 144: 1–21.

Jinn

. Recent development and applications of three-dimensional finite element modeling in bulk forming processes. J Mater Process Technol 2001; 113: 40–45.

Hsu

. Optimal shape design of an extrusion–forging die using a polynomial network and a genetic algorithm. Int J Adv Manuf Technol 2002; 20: 128–137.

Wilson

Schmid

Liu

. Advanced simulations for hot forging: heat transfer model for use with the finite element method. J Mater Process Technol 2004; 155: 1912–1917.

Bennett

. A comparison of material models for the numerical simulation of spike-forging of a CrMoV alloy steel. Comput Mater Sci 2013; 70: 114–122.

Ishikawa

Yukawa

Suzuki

. Formability of superplastic aluminum alloy processed by mechanical alloying. J Mater Process Technol 1997; 68: 236–240.

Rao

Doraivelu

Sivaram

. Physical modelling studies using spike forging to verify analytical predictions. J Mater Process Technol 1991; 28: 295–306.

10.

Lahoti

Altan

. Ring and spike test for evaluating lubricants. Battelle topical report, 1982, p.11.

11.

Isogawa

Kimura

Tozawa

. Proposal of an evaluating method on lubrication. CIRP Ann Manuf Technol 1992; 41: 263–266.

12.

Male

Cockcroft

. A method for the determination of the coefficient of friction of metals under conditions of bulk plastic deformation. J Inst Met 1964; 93: 38–46.

13.

Buschhausen

Weinmann

Lee

. Evaluation of lubrication and friction in cold forging using a double backward-extrusion process. J Mater Process Technol 1992; 33: 95–108.

14.

Schrader

Shirgaokar

Altan

. A critical evaluation of the double cup extrusion test for selection of cold forging lubricants. J Mater Process Technol 2007; 189: 36–44.

15.

Bay

Wibom

Nielsen

. A new friction and lubrication test for cold forging. CIRP Ann Manuf Technol 1995; 44: 217–221.

16.

Lazzarotto

Dubar

Dubois

. Identification of Coulomb’s coefficient of friction in real contact conditions applied to a wire drawing process. Wear 1997; 211: 54–63.

17.

Cheon

Kang

. Determination of friction condition by geometrical measurement of backward extruded aluminum alloy specimen. J Manuf Sci E Trans ASME 2002; 124: 409–415.

18.

Jung

Lee

Ajiboye

. Characterization of frictional behavior in cold forging. Tribol Lett 2010; 37: 353–359.

19.

Zhang

Felder

Bruschi

. Evaluation of friction condition in cold forging by using T-shape compression test. J Mater Process Technol 2009; 209: 5720–5729.

20.

Groche

Müller

Stahlmann

. Mechanical conditions in bulk metal forming tribometers—part one. Tribol Int 2013; 62: 223–231.

21.

Groche

Müller

Jahn

. Effects of the tool lubrication in cold forging. Tribol Lett 2014; 53: 599–605.

22.

Rao

. Analysis of the deformation characteristics of spike-forging process through FE simulations and experiments. J Mater Process Technol 1997; 70: 122–128.

23.

Sheljaskow

. Tool lubricating systems in warm forging. J Mater Process Technol 2001; 113: 16–21.

24.

Liewald

Mletzko

Felde

. Evaluation of lubricants for bulk metal forming of steel at elevated temperatures using double-cup-extrusion-test and spike-test. In: Proceedings of the 42nd ICFG plenary meeting, Shanghai, China, 20–24 September 2009.