- Email: [email protected]
Numerical and experimental analysis of a cross wedge rolling process for producing ball studs
archives of civil and mechanical engineering 17 (2017) 729–737
Available online at www.sciencedirect.com
ScienceDirect journal homepage: http://www.elsevier.com/locate/acme
Original Research Article
Numerical and experimental analysis of a cross wedge rolling process for producing ball studs Tomasz Bulzak *, Zbigniew Pater, Janusz Tomczak Department of Computer Modelling and Metal Forming Technologies, Lublin University of Technology, Poland
article info
abstract
Article history:
The paper reports the results of theoretical and experimental tests of a forming process for
Received 21 June 2016
producing ball studs which are widely used in the automotive industry. It is proposed that
Accepted 20 February 2017
semi-finished ball studs are produced by cross wedge rolling in a double configuration. The
Available online
theoretical analysis was performed by numerical techniques based on the finite element method. Numerical computations were made using the simulation software DEFORM v 11.0.
Keywords:
During the simulations, the accuracy of the adopted tool design was verified, and optimal
Ball stud
parameters of the process along with the effect of selected parameters of the process and the
Ball-and-socket joint
quality of produced parts were determined. The proposed rolling process was verified under
Cross wedge rolling
laboratory conditions using a flat-wedge forging machine available at the Lublin University
FEM
of Technology. The experimental findings show a high agreement with the numerical results, in terms of both quality and quantity. The results confirm that ball studs can be produced by the proposed cross wedge rolling technique. © 2017 Politechnika Wrocławska. Published by Elsevier Sp. z o.o. All rights reserved.
1.
Introduction
Ball studs make part of ball-and-socket joints which are used, among others, to join rigid suspension elements or elements of the steering system [1,2]. Fig. 1 shows the application of a ball stud in the ball-and-socket joint as well as the dimensions of the ball stud investigated in this paper. Ball studs can be produced by flashless cold forging. This process is, however, little effective due to high consumption of energy caused by low plasticity of material and the necessity of performing the process in several stages (Fig. 2) [3,4]. An additional problem in cold forging is heavy wear of forging dies [5,6]. Regarding the forging process for the ball stud, the design of which is shown
in Fig. 1, it is necessary to use tools with a more complex design than is the case with the forging process for producing the ball stud shown in Fig. 2. A schematic design of a forging process for producing a ball stud with necking is shown in Fig. 3. Compared to rolling processes, the forging technique for ball studs has the following advantages: a simple design of the process, a higher availability of forging machines and less complicated tool geometry [7–9]. An alternative method for producing ball studs is cross wedge rolling (CWR). In this process parts are formed by wedge-shaped tools [12]. The tools are mounted on the wedges or on flat or concave rolling mill plates. In contrast to forging, CWR is characterized by reduced energy consumption [13,14]. Its rolling efficiency can be additionally increased by running
* Corresponding author. E-mail addresses: [email protected] (T. Bulzak), [email protected] (Z. Pater), [email protected] (J. Tomczak). http://dx.doi.org/10.1016/j.acme.2017.02.002 1644-9665/© 2017 Politechnika Wrocławska. Published by Elsevier Sp. z o.o. All rights reserved.
730
archives of civil and mechanical engineering 17 (2017) 729–737
the forming process for ball studs in a double configuration. A particularly interesting variant of CWR is multi-wedge cross rolling (MCWR) which is currently used for forming elongated parts such as car axles [15]. The use of multi-wedge tools for forming short parts such as ball studs makes it possible to produce more parts simultaneously. The production of parts in a double configuration is also vital given the design of both the wedge tool and the rolling process itself. Asymmetrical parts are formed in a double configuration [12,16]. During the forming of asymmetrical parts, there occur unbalanced axial loads and strains are located in a non-uniform manner, which can lead to a series of undesired phenomena [17,18]. Failure modes which may occur during the forming of asymmetrical parts include the loss of position stability of the workpiece relative to the tool, formation of helical grooves on the workpiece surface, and overlap.
2.
Fig. 1 – Schematic design of a ball stud and its dimensions [10].
Methods and procedures
This paper reports the results of a study on a cross wedge rolling process for producing a ball stud in a double configuration. The process was performed using tools described by the forming angle a = 308 and the spreading angle b = 108, as shown in Fig. 4. The design of the wedges for cross rolling is based on the assumption that the values of a and b should satisfy the condition that 0.04 ≤ tga tgb ≤ 0.08 [19]. In the analyzed case, the condition is not met because tgatgb = 0.1, which means that uncontrolled slipping of the workpiece may occur during rolling. The value of b was set to 108 in order to reduce the length of the tool segment. In addition, this enabled reducing the duration of a work cycle of the flat-wedge rolling mill. The above condition can be satisfied by decreasing the value of a. However, at smaller values of a may cause such problems as underfilling of the
Fig. 2 – Schematic design of a forging process for producing a ball stud in 5 operations [11].
archives of civil and mechanical engineering 17 (2017) 729–737
731
Fig. 3 – Schematic design of a forging process for producing a ball stud with necking [11].
wedge profile in the spherical part of the forging and the cracking of forgings in the axial zone. Taking into account the above tool design, a numerical analysis was performed to verify if ball studs could be produced the rolling process in a double configuration, assuming that the ball studs are connected by a cylindrical bridge with a diameter of 16 mm. Three cases of the rolling process were investigated, each for a different workpiece diameter: do = 28 mm; 29 mm; and 30 mm.
2.1. (FEM)
Numerical modelling by the finite element method
The design of the tools for producing ball studs by rolling was verified in preliminary tests by numerical modelling. The
modelling was performed using the DEFORM v 11.0 simulation software. Fig. 5 shows the geometrical model of the analyzed process. The geometrical model consists of two wedges (upper and lower (Fig. 4b)) and a workpiece. The workpiece was modelled as a rigid-plastic object and discretized by tetragonal elements. The workpiece was assigned the properties of C45 steel, the model of which was obtained from the material database library of the applied software. The numerical simulations were performed in hot forming conditions and thermal conditions were taken into account. The thermal conditions were described by the workpiece – tools heat transfer coefficient (10 kW/m2 K) and the workpiece – air heat transfer coefficient (0.2 kW/m2 K). It was also assumed that 90% of the plastic deformation work is converted into heat. The initial temperature of the workpiece was set to 1150 8C, while the temperature of the tool was maintained constant at 20 8C. The contact between the tools and the workpiece was described by a constant friction model with the friction factor m set to 1. Such a high value of the friction factor results from the fact that in real conditions the wedges have no technological serrations on their surface to increase roughness of the tools. Given the difficulty of performing material separation in FEM, the numerical modelling did not include the operation of cutting off the scrap material on the forging's ends or separating the produced parts.
2.2.
Fig. 4 – Wedges used in the study: (a) experimental, (b) CAD model used in numerical analysis.
Experimental tests
The rolling process was performed on cylindrical workpieces, their dimensions being: ; 28 mm 110 mm, ; 29 mm 100 mm and ; 30 mm 100 mm. Prior to rolling, the workpiece was heated in an electric chamber furnace. The rolling process for forming ball studs was performed under laboratory conditions using the flat-wedge rolling mill as shown in Fig. 6. The tools were fixed in the work space of the machine as
732
archives of civil and mechanical engineering 17 (2017) 729–737
configuration using the assumed values of a and b angles. Numerical methods were employed to determine effective strain, damage criterion, variations in the mean stress and force parameters. The experimental findings regarding the characteristics of forces in the rolling process were compared with the FEM numerical results.
3.1.
Fig. 5 – Geometrical model of a cross wedge rolling process for producing a ball stud.
Fig. 6 – Flat-wedge rolling mill with a hydraulic drive available at the Lublin University of Technology.
shown in Fig. 7. The machine's slide with the fixed upper tool moved with a velocity of 0.3 m/s. Force parameters in the rolling process were measured by indirect method via measuring variations in the pressure in the hydraulic cylinder which drives the slide.
3.
Results
The numerical and experimental results confirm that ball studs can be formed by cross wedge rolling in a double
Distributions of strains and damage criterion
The distributions of strain effective is illustrated in Fig. 8a. In all investigated cases of the rolling process, the strains are distributed in a similar fashion. The distribution of strains along the centerline of the produced parts is non-uniform. The lowest strains occur in the spherical region of the ball stud while the highest strains are located on the part's end. This is naturally caused by the degree of cross-sectional reduction in the workpiece. The degree of cross-sectional reduction on the part's end is the highest (the highest strains), whereas in the spherical region of the part it is the lowest, which is reflected in the lowest strains. When ball studs are formed from the workpiece with the diameter do = 28 mm, the cross-sectional reduction in the spherical region of the forging is negative, which means that the material in this region should be subjected to upsetting during rolling. The distribution of strains along the radial line of the produced parts is nonuniform, too. In all investigated cases, the highest nonuniformity of the strains is located on the part's end. One can observe that the distribution of strains along the radial line is the most non-uniform in the region of necking which is in direct contact with the ball. This non-uniformity increases with increasing the diameter of the workpiece. The finite element method can be used to predict cracking in deformed material based on empirical criteria. The criteria are based on macroscopic parameters and do not take account of changes in the material's structure during deformation. The criteria have the form of an integral from the function describing the effect of selected parameters on the investigated process. The function representing the deformation process predominantly depends on values of stresses and strains. One of the criteria widely used for predicting cracking of material is a value of the Cockroft–Latham integral expressed as: Z C¼
e 0
s1 de; si
(1)
where s1 is the maximum principal stress, si is the effective stress, e is the effective strain, C is a value of the Cockroft– Latham integral.
Fig. 7 – Wedge tool mounted in the work space of the flatwedge rolling machine.
The distributions of normalized Cockroft–Latham damage criterion is illustrated in Fig. 8b. Examining the distributions of the Cockroft–Latham integral it can be observed that to a certain degree they resemble the distributions of effective strain. In both cases, the highest values are located on the forging's ends while the lowest – in the spherical region of the forging. With increasing the workpiece's diameter, the value of C increases too, hence leading to a higher risk of material cracking. The value of C increases in the region of the connector, too; here, however, it is desired due to the fact that the forgings are separated from each other in this particular region.
archives of civil and mechanical engineering 17 (2017) 729–737
733
Fig. 8 – Distributions of strain effective (a) and normalized damage Cockroft–Latham damage criterion (b).
Fig. 9 – Arrangement of measuring points in the forging.
cut into the workpiece to form a cylindrical connector between the balls of both forgings. After that, the mean stresses at the selected measuring points vary periodically, from positive to negative values. The variations in the mean stresses at points P1 and P2 reflect the pattern of the rolling process. At point P3, the variations in the mean stresses are the lowest. The variable character of the stresses at points P1 and P2 can lead to low-cycle fatigue of the material and thus to a loss of material cohesion in the central region of the produced part. The final stage of the rolling process consists in forming the end step of the part, which leads to an increase in the stresses at Point P3. On separating the forgings, the compressive stresses at Point P1 significantly increase while the tensile stresses remain relatively insignificant. Regardless of the applied diameter of the workpiece, the distribution of stresses in the central region of the workpiece is similar.
3.3.
Fig. 10 – Variations in the mean stress at measuring points for do = 28 mm.
3.2.
Variations in mean stress
The history of evolution of the mean stress during rolling is plotted in Figs. 10–12. The variations in the mean stress were measured at three points initially located in the centerline of the workpiece. During rolling, the points moved in accordance with the kinematics of material flow and their final position is illustrated in Fig. 9. At the initial stage of rolling, the compressive stresses at Point P1 significantly increase to a value of about 150 MPa. This results from the fact that the tools
Force parameters
The knowledge about the forces in metal forming processes is vital for tool design, machinery selection and process specification. Figs. 13–17 illustrate characteristics of the forces in the rolling process for producing ball studs. Fig. 13 compares the tangential forces measured for three workpiece diameters in the experimental tests. As expected, with increasing the diameter of the workpiece, the rolling force increases, too. The increase in the force is more visible between the diameters do = 28 mm and do = 29 mm than between the diameters do = 29 mm and do = 30 mm. At the beginning of the rolling process, the rolling force increases (the wedge cuts into the material); then, it decreases (the sizing of balls takes place). On covering the distance s = 200 mm by the wedge, one can observe another increase in the force, which corresponds to the beginning of forming of a subsequent step of the forging. On a distance ranging from s = 400 mm to s = 1000 mm, the force remains relatively stable. Next, during the sizing of forgings, the rolling force decreases. The increase in the force during the final stage of the rolling process is caused by
734
archives of civil and mechanical engineering 17 (2017) 729–737
Fig. 11 – Variations in the mean stress at measuring points for do = 29 mm.
Fig. 14 – Comparison of the tangential force for do = 28 mm.
Fig. 12 – Variations in the mean stress at measuring points do = 30 mm.
Fig. 15 – Comparison of the tangential force for do = 29 mm.
Fig. 13 – Variations in the tangential force for different workpiece diameters (experimental).
Fig. 16 – Comparison of the tangential force for do = 30 mm.
resistances which are generated when the waste material on the part's end is cut off and the parts are separated by removing the connector between them. Figs. 14–16 offer a comparison of the variations in the forces obtained in the experiments and numerical simulations. The
plots demonstrate that the forces obtained in the numerical simulations show a high agreement with the experimental results. The highest differences can be observed at the final stage of the process, i.e. when the forgings are separated. The variations in the forces obtained by FEM are similar at the
archives of civil and mechanical engineering 17 (2017) 729–737
735
Fig. 18 – Ball studs formed in experimental tests from workpieces with different diameters. Fig. 17 – FEM distributions of radial forces for different workpiece diameters.
end of the process for all investigated workpiece diameters. In the final stage of the forming process, one can observe that the variations in the force corresponding to the increase in the force on cutting off the scrap material on the forging's ends (the first peak) and separating the forgings (the second peak). In the numerical simulation, the variations in the forces at the first and second peak are similar. In the experiments, the forces are higher when the scrap material is cut off than when the forgings are separated. The forces measured in the experiments on cutting off the scrap material are much higher than those observed in the numerical simulations. The differences may be due to the fact that the scrap material on the forging's ends is cut off with two sets of cutters while the separation of forgings is done using only one set of cutters. The discrepancies between the experimental findings and numerical results for this stage of the rolling process (separation of forgings) can result from the difficulties in numerical modelling of material separation. During rolling the tangential forces remain relatively low at about 30 kN. In contrast, the radial forces are nearly twice as high as the tangential forces. The FEM-determined distribution of the radial forces is illustrated in Fig. 17. Similarly to the tangential forces (Fig. 13), the values of the radial forces are similar for the workpieces with do = 29 mm and do = 30 mm.
3.4.
Experimental results
The experimental tests involved forming a test batch of ball studs from workpieces with different diameters. These ball studs are shown in Fig. 18. As for the diameters do = 29 mm and do = 30 mm, the tool impression is filled correctly. Regarding the diameter do = 28 mm, one can observe that the tool profile is underfilled in the spherical region of the part. The forming process included separating the forgings in the spherical region. The application of a cutter for separating the forgings resulted in distorting the profile of the ball. Fig. 19 shows the ball studs produced according to two variants of the forming process: with and without separation of the forgings. It can be observed that the quality of the spherical region of the forging significantly improved due to the fact that we did
Fig. 19 – Ball studs formed from workpiece with diameter do = 30 mm. Two variants of the forming process: with and without separation of forgings.
not use the cutter for separating the forgings. The profile of the ball in the ball studs produced without the use of the cutter is correct, and the spherical part of the forgings is free from any defects. When the cutter is used to separate the forgings in the rolling process, the workpiece gets tilted and thus the spherical region of the forging clamps on the cutter, which results in part defects. In order to enable the separation of forgings, it is necessary to modify the geometry of the active region of the cutter. Given the limited length of the tool, and thus the necessity of performing the forming process for balls studs at a higher b, there is a risk of internal cracking. The forgings produced in the experiments were inspected for internal cracks. The inspection was conducted by non-destructive testing using 3D X-ray tomography. The numerical results demonstrate that the highest risk of internal cracking during rolling may occur when the workpiece has the diameter do = 30 mm. In this case of the investigated rolling process, the normalized Cockroft–Latham damage criterion is the highest. Fig. 20 shows the X-ray image of a ball stud formed from the workpiece with the diameter do = 30 mm. Despite the highest values of the Cockroft–Latham
Fig. 20 – X-ray image of a ball stud in longitudinal section.
736
archives of civil and mechanical engineering 17 (2017) 729–737
Table 1 – Comparison of the main dimensions of ball studs.
Workpiece
; 28 mm
; 29 mm
; 30 mm
; A [mm]
CAD FEM Experiment
29.00 28.44 28.56
29.00 29.14 29.16
29.00 29.10 29.10
; B [mm]
CAD FEM Experiment
19.36 19.38 19.50
19.36 19.38 19.39
19.36 19.42 19.50
; C [mm]
CAD FEM Experiment
22.97 22.84 23.39
22.97 23.04 23.20
22.97 22.94 23.03
; D [mm]
CAD FEM Experiment
16.00 16.00 16.28
16.00 16.11 16.21
16.00 16.10 16.24
L [mm]
CAD FEM Experiment
82.50 82.50 83.13
82.50 82.79 82.85
82.50 82.67 82.70
integral, the forging does not reveal any signs of material cohesion loss. Table 1 offers a comparison of the main dimensions of ball studs produced by numerical modelling and in the experimental tests. Machining allowance for the ball stud forging illustrated in Fig. 1 was set to 2 mm. The ball stud forging formed from the 28 mm diameter billet does not meet the requirements for diameter A, as shown by the numerical findings and experimental results. Nonetheless, the diameter of the ball formed from the 28 mm diameter billet is within the assumed machining allowance. In other cases, the dimensions of the forgings comply with the assumed dimensions of the CAD-designed forging. The ball studs produced in the experimental tests are bigger than the numerically modelled ball studs. This phenomenon can be explained by elastic strain of the rolling mill's body or incorrect distance setting between the tools.
4.
Conclusions
The paper reported the results of studies on producing ball studs for automotive applications. The use of numerical modelling in the design of forming methods for new parts is very useful, particularly when designing innovative processes. The proposed rolling process was verified under laboratory conditions. The experimental findings show a high agreement with the numerical results, in terms of both quality and quantity. The results confirm that ball studs can be formed from workpieces with different diameters by cross wedge rolling. In addition, it has been found that the process can be performed using wedge tools with higher values of the spreading angle b. During the numerical simulations and
experimental tests we did not observe uncontrolled slipping due to the application of tools with high values of b. Regarding the investigated range of workpiece diameters, the numerical results and experimental findings demonstrate that the variations in the maximum rolling force are insignificant. The diameter of the workpiece has impact on the damage criterion, as the risk of internal cracking increases with increasing the diameter of the workpiece. The results of Xray tomography do not reveal the presence of internal cracks in the parts produced from the workpiece with the largest diameter. Given that ball studs can be formed from workpieces with different diameters in the proposed cross wedge rolling process, this process can be performed using workpieces with high dimensional tolerance.
references
[1] C. Weiss, M.M. Morlock, N.P. Hoffmann, Friction induced dynamics of ball joints: instability and post bifurcation behavior, European Journal of Mechanics A/Solids 45 (2014) 161–173. [2] P. Alexandru, I. Visa, C. Alexandru, Modeling the angular capability of the ball joints in a complex mechanism with two degrees of mobility, Applied Mathematical Modelling 38 (2014) 5456–5470. [3] J.G. Eom, Y.H. Son, S.W. Jeong, et al., Effect of strain hardening capability on plastic deformation behaviors of material during metal forming, Materials and Design 54 (2014) 1010– 1018. [4] J.H. Song, Y.T. Im, Expert system for the process sequence design of a ball stud, Journal of Material Processing Technology 89–90 (1999) 72–78. [5] Z. Gronostajski, M. Hawryluk, M. Kaszuba, et al., Die profile optimization for forging constant velocity joint casings, Archives of Metallurgy and Materials 56 (2011) 551–558. [6] Z. Gronostajski, M. Hawryluk, The main aspects of precision forging, Archives of Civil and Mechanical Engineering 8 (2008) 39–55. [7] C.S. Im, S.R. Suh, M.C. Lee, et al., Computer aided process design in cold-former forging using a forging simulator and a commercial CAD software, Journal of Materials Processing Technology 95 (1999) 155–163. [8] W.S. Weroński, A. Gontarz, Z. Pater, The research of forging process of eccentric part on three slide forging press, Journal of Materials Processing Technology 177 (2006) 214–217. [9] G. Winiarski, A. Gontarz, Z. Pater, A new process for the forming of a triangular flange in hollow shafts from Ti6Al4V alloy, Archives of Civil and Mechanical Engineering 15 (2015) 911–916. [10] B. Heibing, M. Ersoy, Chassis Handbook. Fundamentals, Driving Dynamics, Components, Mechatronics, Perspectives, Deutsche Nationalbibliothek, Berlin, 2011. [11] http://www.afdex.com [accessed 15.03.16]. [12] J. Zhou, Y. Yu, Q. Zeng, Analysis and experimental studies of internal voids in multi-wedge cross wedge rolling stepped shaft, International Journal of Advanced Manufacturing Technology 72 (2014) 1559–1566. [13] H. Ji, J. Liu, B. Wang, et al., Cross-wedge rolling of a 4Cr9Si2 hollow valve: explorative experiment and finite element simulation, International Journal of Advanced Manufacturing Technology 77 (2015) 15–26. [14] J. Tomczak, Z. Pater, T. Bulzak, The influence of hollow billet thickness in rotary compression, International Journal of Advanced Manufacturing Technology 82 (2016) 1281–1291.
archives of civil and mechanical engineering 17 (2017) 729–737
[15] J. Zhao, L.Q. Lu, Z.H. Hu, Study on varying rule of mechanical parameters in forming automobile semi-axes with multiwedge cross wedge rolling, Proceedings of the International Conference on Mechanical Engineering and Mechanics 1–2 (2007) 1723–1727. [16] W. Yuan, L. Wang, T. Yuan, Experimental study on cross wedge rolling process of aluminum alloy material, in: 5th International Conference on Advanced Engineering Materials and Technology 5 (AEMT 2015), 2015, 415–422.
737
[17] Z. Pater, A. Tofil, J. Tomczak, T. Bulzak, Numerical analysis of the cross wedge rolling process (CWR) for a stepped shaft, Metalurgija 54 (2015) 177–180. [18] J. Tomczak, Z. Pater, T. Bulzak, Thermo-mechanical analysis of a lever preform forming from magnesium alloy AZ31, Archives of Metallurgy and Materials 57 (2012) 1211–1218. [19] Z. Pater, Analysis of the helical-wedge rolling process for producing a long stepped shaft, Key Engineering Materials 622–623 (2014) 893–898.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: http://www.elsevier.com/locate/acme
Original Research Article
Numerical and experimental analysis of a cross wedge rolling process for producing ball studs Tomasz Bulzak *, Zbigniew Pater, Janusz Tomczak Department of Computer Modelling and Metal Forming Technologies, Lublin University of Technology, Poland
article info
abstract
Article history:
The paper reports the results of theoretical and experimental tests of a forming process for
Received 21 June 2016
producing ball studs which are widely used in the automotive industry. It is proposed that
Accepted 20 February 2017
semi-finished ball studs are produced by cross wedge rolling in a double configuration. The
Available online
theoretical analysis was performed by numerical techniques based on the finite element method. Numerical computations were made using the simulation software DEFORM v 11.0.
Keywords:
During the simulations, the accuracy of the adopted tool design was verified, and optimal
Ball stud
parameters of the process along with the effect of selected parameters of the process and the
Ball-and-socket joint
quality of produced parts were determined. The proposed rolling process was verified under
Cross wedge rolling
laboratory conditions using a flat-wedge forging machine available at the Lublin University
FEM
of Technology. The experimental findings show a high agreement with the numerical results, in terms of both quality and quantity. The results confirm that ball studs can be produced by the proposed cross wedge rolling technique. © 2017 Politechnika Wrocławska. Published by Elsevier Sp. z o.o. All rights reserved.
1.
Introduction
Ball studs make part of ball-and-socket joints which are used, among others, to join rigid suspension elements or elements of the steering system [1,2]. Fig. 1 shows the application of a ball stud in the ball-and-socket joint as well as the dimensions of the ball stud investigated in this paper. Ball studs can be produced by flashless cold forging. This process is, however, little effective due to high consumption of energy caused by low plasticity of material and the necessity of performing the process in several stages (Fig. 2) [3,4]. An additional problem in cold forging is heavy wear of forging dies [5,6]. Regarding the forging process for the ball stud, the design of which is shown
in Fig. 1, it is necessary to use tools with a more complex design than is the case with the forging process for producing the ball stud shown in Fig. 2. A schematic design of a forging process for producing a ball stud with necking is shown in Fig. 3. Compared to rolling processes, the forging technique for ball studs has the following advantages: a simple design of the process, a higher availability of forging machines and less complicated tool geometry [7–9]. An alternative method for producing ball studs is cross wedge rolling (CWR). In this process parts are formed by wedge-shaped tools [12]. The tools are mounted on the wedges or on flat or concave rolling mill plates. In contrast to forging, CWR is characterized by reduced energy consumption [13,14]. Its rolling efficiency can be additionally increased by running
* Corresponding author. E-mail addresses: [email protected] (T. Bulzak), [email protected] (Z. Pater), [email protected] (J. Tomczak). http://dx.doi.org/10.1016/j.acme.2017.02.002 1644-9665/© 2017 Politechnika Wrocławska. Published by Elsevier Sp. z o.o. All rights reserved.
730
archives of civil and mechanical engineering 17 (2017) 729–737
the forming process for ball studs in a double configuration. A particularly interesting variant of CWR is multi-wedge cross rolling (MCWR) which is currently used for forming elongated parts such as car axles [15]. The use of multi-wedge tools for forming short parts such as ball studs makes it possible to produce more parts simultaneously. The production of parts in a double configuration is also vital given the design of both the wedge tool and the rolling process itself. Asymmetrical parts are formed in a double configuration [12,16]. During the forming of asymmetrical parts, there occur unbalanced axial loads and strains are located in a non-uniform manner, which can lead to a series of undesired phenomena [17,18]. Failure modes which may occur during the forming of asymmetrical parts include the loss of position stability of the workpiece relative to the tool, formation of helical grooves on the workpiece surface, and overlap.
2.
Fig. 1 – Schematic design of a ball stud and its dimensions [10].
Methods and procedures
This paper reports the results of a study on a cross wedge rolling process for producing a ball stud in a double configuration. The process was performed using tools described by the forming angle a = 308 and the spreading angle b = 108, as shown in Fig. 4. The design of the wedges for cross rolling is based on the assumption that the values of a and b should satisfy the condition that 0.04 ≤ tga tgb ≤ 0.08 [19]. In the analyzed case, the condition is not met because tgatgb = 0.1, which means that uncontrolled slipping of the workpiece may occur during rolling. The value of b was set to 108 in order to reduce the length of the tool segment. In addition, this enabled reducing the duration of a work cycle of the flat-wedge rolling mill. The above condition can be satisfied by decreasing the value of a. However, at smaller values of a may cause such problems as underfilling of the
Fig. 2 – Schematic design of a forging process for producing a ball stud in 5 operations [11].
archives of civil and mechanical engineering 17 (2017) 729–737
731
Fig. 3 – Schematic design of a forging process for producing a ball stud with necking [11].
wedge profile in the spherical part of the forging and the cracking of forgings in the axial zone. Taking into account the above tool design, a numerical analysis was performed to verify if ball studs could be produced the rolling process in a double configuration, assuming that the ball studs are connected by a cylindrical bridge with a diameter of 16 mm. Three cases of the rolling process were investigated, each for a different workpiece diameter: do = 28 mm; 29 mm; and 30 mm.
2.1. (FEM)
Numerical modelling by the finite element method
The design of the tools for producing ball studs by rolling was verified in preliminary tests by numerical modelling. The
modelling was performed using the DEFORM v 11.0 simulation software. Fig. 5 shows the geometrical model of the analyzed process. The geometrical model consists of two wedges (upper and lower (Fig. 4b)) and a workpiece. The workpiece was modelled as a rigid-plastic object and discretized by tetragonal elements. The workpiece was assigned the properties of C45 steel, the model of which was obtained from the material database library of the applied software. The numerical simulations were performed in hot forming conditions and thermal conditions were taken into account. The thermal conditions were described by the workpiece – tools heat transfer coefficient (10 kW/m2 K) and the workpiece – air heat transfer coefficient (0.2 kW/m2 K). It was also assumed that 90% of the plastic deformation work is converted into heat. The initial temperature of the workpiece was set to 1150 8C, while the temperature of the tool was maintained constant at 20 8C. The contact between the tools and the workpiece was described by a constant friction model with the friction factor m set to 1. Such a high value of the friction factor results from the fact that in real conditions the wedges have no technological serrations on their surface to increase roughness of the tools. Given the difficulty of performing material separation in FEM, the numerical modelling did not include the operation of cutting off the scrap material on the forging's ends or separating the produced parts.
2.2.
Fig. 4 – Wedges used in the study: (a) experimental, (b) CAD model used in numerical analysis.
Experimental tests
The rolling process was performed on cylindrical workpieces, their dimensions being: ; 28 mm 110 mm, ; 29 mm 100 mm and ; 30 mm 100 mm. Prior to rolling, the workpiece was heated in an electric chamber furnace. The rolling process for forming ball studs was performed under laboratory conditions using the flat-wedge rolling mill as shown in Fig. 6. The tools were fixed in the work space of the machine as
732
archives of civil and mechanical engineering 17 (2017) 729–737
configuration using the assumed values of a and b angles. Numerical methods were employed to determine effective strain, damage criterion, variations in the mean stress and force parameters. The experimental findings regarding the characteristics of forces in the rolling process were compared with the FEM numerical results.
3.1.
Fig. 5 – Geometrical model of a cross wedge rolling process for producing a ball stud.
Fig. 6 – Flat-wedge rolling mill with a hydraulic drive available at the Lublin University of Technology.
shown in Fig. 7. The machine's slide with the fixed upper tool moved with a velocity of 0.3 m/s. Force parameters in the rolling process were measured by indirect method via measuring variations in the pressure in the hydraulic cylinder which drives the slide.
3.
Results
The numerical and experimental results confirm that ball studs can be formed by cross wedge rolling in a double
Distributions of strains and damage criterion
The distributions of strain effective is illustrated in Fig. 8a. In all investigated cases of the rolling process, the strains are distributed in a similar fashion. The distribution of strains along the centerline of the produced parts is non-uniform. The lowest strains occur in the spherical region of the ball stud while the highest strains are located on the part's end. This is naturally caused by the degree of cross-sectional reduction in the workpiece. The degree of cross-sectional reduction on the part's end is the highest (the highest strains), whereas in the spherical region of the part it is the lowest, which is reflected in the lowest strains. When ball studs are formed from the workpiece with the diameter do = 28 mm, the cross-sectional reduction in the spherical region of the forging is negative, which means that the material in this region should be subjected to upsetting during rolling. The distribution of strains along the radial line of the produced parts is nonuniform, too. In all investigated cases, the highest nonuniformity of the strains is located on the part's end. One can observe that the distribution of strains along the radial line is the most non-uniform in the region of necking which is in direct contact with the ball. This non-uniformity increases with increasing the diameter of the workpiece. The finite element method can be used to predict cracking in deformed material based on empirical criteria. The criteria are based on macroscopic parameters and do not take account of changes in the material's structure during deformation. The criteria have the form of an integral from the function describing the effect of selected parameters on the investigated process. The function representing the deformation process predominantly depends on values of stresses and strains. One of the criteria widely used for predicting cracking of material is a value of the Cockroft–Latham integral expressed as: Z C¼
e 0
s1 de; si
(1)
where s1 is the maximum principal stress, si is the effective stress, e is the effective strain, C is a value of the Cockroft– Latham integral.
Fig. 7 – Wedge tool mounted in the work space of the flatwedge rolling machine.
The distributions of normalized Cockroft–Latham damage criterion is illustrated in Fig. 8b. Examining the distributions of the Cockroft–Latham integral it can be observed that to a certain degree they resemble the distributions of effective strain. In both cases, the highest values are located on the forging's ends while the lowest – in the spherical region of the forging. With increasing the workpiece's diameter, the value of C increases too, hence leading to a higher risk of material cracking. The value of C increases in the region of the connector, too; here, however, it is desired due to the fact that the forgings are separated from each other in this particular region.
archives of civil and mechanical engineering 17 (2017) 729–737
733
Fig. 8 – Distributions of strain effective (a) and normalized damage Cockroft–Latham damage criterion (b).
Fig. 9 – Arrangement of measuring points in the forging.
cut into the workpiece to form a cylindrical connector between the balls of both forgings. After that, the mean stresses at the selected measuring points vary periodically, from positive to negative values. The variations in the mean stresses at points P1 and P2 reflect the pattern of the rolling process. At point P3, the variations in the mean stresses are the lowest. The variable character of the stresses at points P1 and P2 can lead to low-cycle fatigue of the material and thus to a loss of material cohesion in the central region of the produced part. The final stage of the rolling process consists in forming the end step of the part, which leads to an increase in the stresses at Point P3. On separating the forgings, the compressive stresses at Point P1 significantly increase while the tensile stresses remain relatively insignificant. Regardless of the applied diameter of the workpiece, the distribution of stresses in the central region of the workpiece is similar.
3.3.
Fig. 10 – Variations in the mean stress at measuring points for do = 28 mm.
3.2.
Variations in mean stress
The history of evolution of the mean stress during rolling is plotted in Figs. 10–12. The variations in the mean stress were measured at three points initially located in the centerline of the workpiece. During rolling, the points moved in accordance with the kinematics of material flow and their final position is illustrated in Fig. 9. At the initial stage of rolling, the compressive stresses at Point P1 significantly increase to a value of about 150 MPa. This results from the fact that the tools
Force parameters
The knowledge about the forces in metal forming processes is vital for tool design, machinery selection and process specification. Figs. 13–17 illustrate characteristics of the forces in the rolling process for producing ball studs. Fig. 13 compares the tangential forces measured for three workpiece diameters in the experimental tests. As expected, with increasing the diameter of the workpiece, the rolling force increases, too. The increase in the force is more visible between the diameters do = 28 mm and do = 29 mm than between the diameters do = 29 mm and do = 30 mm. At the beginning of the rolling process, the rolling force increases (the wedge cuts into the material); then, it decreases (the sizing of balls takes place). On covering the distance s = 200 mm by the wedge, one can observe another increase in the force, which corresponds to the beginning of forming of a subsequent step of the forging. On a distance ranging from s = 400 mm to s = 1000 mm, the force remains relatively stable. Next, during the sizing of forgings, the rolling force decreases. The increase in the force during the final stage of the rolling process is caused by
734
archives of civil and mechanical engineering 17 (2017) 729–737
Fig. 11 – Variations in the mean stress at measuring points for do = 29 mm.
Fig. 14 – Comparison of the tangential force for do = 28 mm.
Fig. 12 – Variations in the mean stress at measuring points do = 30 mm.
Fig. 15 – Comparison of the tangential force for do = 29 mm.
Fig. 13 – Variations in the tangential force for different workpiece diameters (experimental).
Fig. 16 – Comparison of the tangential force for do = 30 mm.
resistances which are generated when the waste material on the part's end is cut off and the parts are separated by removing the connector between them. Figs. 14–16 offer a comparison of the variations in the forces obtained in the experiments and numerical simulations. The
plots demonstrate that the forces obtained in the numerical simulations show a high agreement with the experimental results. The highest differences can be observed at the final stage of the process, i.e. when the forgings are separated. The variations in the forces obtained by FEM are similar at the
archives of civil and mechanical engineering 17 (2017) 729–737
735
Fig. 18 – Ball studs formed in experimental tests from workpieces with different diameters. Fig. 17 – FEM distributions of radial forces for different workpiece diameters.
end of the process for all investigated workpiece diameters. In the final stage of the forming process, one can observe that the variations in the force corresponding to the increase in the force on cutting off the scrap material on the forging's ends (the first peak) and separating the forgings (the second peak). In the numerical simulation, the variations in the forces at the first and second peak are similar. In the experiments, the forces are higher when the scrap material is cut off than when the forgings are separated. The forces measured in the experiments on cutting off the scrap material are much higher than those observed in the numerical simulations. The differences may be due to the fact that the scrap material on the forging's ends is cut off with two sets of cutters while the separation of forgings is done using only one set of cutters. The discrepancies between the experimental findings and numerical results for this stage of the rolling process (separation of forgings) can result from the difficulties in numerical modelling of material separation. During rolling the tangential forces remain relatively low at about 30 kN. In contrast, the radial forces are nearly twice as high as the tangential forces. The FEM-determined distribution of the radial forces is illustrated in Fig. 17. Similarly to the tangential forces (Fig. 13), the values of the radial forces are similar for the workpieces with do = 29 mm and do = 30 mm.
3.4.
Experimental results
The experimental tests involved forming a test batch of ball studs from workpieces with different diameters. These ball studs are shown in Fig. 18. As for the diameters do = 29 mm and do = 30 mm, the tool impression is filled correctly. Regarding the diameter do = 28 mm, one can observe that the tool profile is underfilled in the spherical region of the part. The forming process included separating the forgings in the spherical region. The application of a cutter for separating the forgings resulted in distorting the profile of the ball. Fig. 19 shows the ball studs produced according to two variants of the forming process: with and without separation of the forgings. It can be observed that the quality of the spherical region of the forging significantly improved due to the fact that we did
Fig. 19 – Ball studs formed from workpiece with diameter do = 30 mm. Two variants of the forming process: with and without separation of forgings.
not use the cutter for separating the forgings. The profile of the ball in the ball studs produced without the use of the cutter is correct, and the spherical part of the forgings is free from any defects. When the cutter is used to separate the forgings in the rolling process, the workpiece gets tilted and thus the spherical region of the forging clamps on the cutter, which results in part defects. In order to enable the separation of forgings, it is necessary to modify the geometry of the active region of the cutter. Given the limited length of the tool, and thus the necessity of performing the forming process for balls studs at a higher b, there is a risk of internal cracking. The forgings produced in the experiments were inspected for internal cracks. The inspection was conducted by non-destructive testing using 3D X-ray tomography. The numerical results demonstrate that the highest risk of internal cracking during rolling may occur when the workpiece has the diameter do = 30 mm. In this case of the investigated rolling process, the normalized Cockroft–Latham damage criterion is the highest. Fig. 20 shows the X-ray image of a ball stud formed from the workpiece with the diameter do = 30 mm. Despite the highest values of the Cockroft–Latham
Fig. 20 – X-ray image of a ball stud in longitudinal section.
736
archives of civil and mechanical engineering 17 (2017) 729–737
Table 1 – Comparison of the main dimensions of ball studs.
Workpiece
; 28 mm
; 29 mm
; 30 mm
; A [mm]
CAD FEM Experiment
29.00 28.44 28.56
29.00 29.14 29.16
29.00 29.10 29.10
; B [mm]
CAD FEM Experiment
19.36 19.38 19.50
19.36 19.38 19.39
19.36 19.42 19.50
; C [mm]
CAD FEM Experiment
22.97 22.84 23.39
22.97 23.04 23.20
22.97 22.94 23.03
; D [mm]
CAD FEM Experiment
16.00 16.00 16.28
16.00 16.11 16.21
16.00 16.10 16.24
L [mm]
CAD FEM Experiment
82.50 82.50 83.13
82.50 82.79 82.85
82.50 82.67 82.70
integral, the forging does not reveal any signs of material cohesion loss. Table 1 offers a comparison of the main dimensions of ball studs produced by numerical modelling and in the experimental tests. Machining allowance for the ball stud forging illustrated in Fig. 1 was set to 2 mm. The ball stud forging formed from the 28 mm diameter billet does not meet the requirements for diameter A, as shown by the numerical findings and experimental results. Nonetheless, the diameter of the ball formed from the 28 mm diameter billet is within the assumed machining allowance. In other cases, the dimensions of the forgings comply with the assumed dimensions of the CAD-designed forging. The ball studs produced in the experimental tests are bigger than the numerically modelled ball studs. This phenomenon can be explained by elastic strain of the rolling mill's body or incorrect distance setting between the tools.
4.
Conclusions
The paper reported the results of studies on producing ball studs for automotive applications. The use of numerical modelling in the design of forming methods for new parts is very useful, particularly when designing innovative processes. The proposed rolling process was verified under laboratory conditions. The experimental findings show a high agreement with the numerical results, in terms of both quality and quantity. The results confirm that ball studs can be formed from workpieces with different diameters by cross wedge rolling. In addition, it has been found that the process can be performed using wedge tools with higher values of the spreading angle b. During the numerical simulations and
experimental tests we did not observe uncontrolled slipping due to the application of tools with high values of b. Regarding the investigated range of workpiece diameters, the numerical results and experimental findings demonstrate that the variations in the maximum rolling force are insignificant. The diameter of the workpiece has impact on the damage criterion, as the risk of internal cracking increases with increasing the diameter of the workpiece. The results of Xray tomography do not reveal the presence of internal cracks in the parts produced from the workpiece with the largest diameter. Given that ball studs can be formed from workpieces with different diameters in the proposed cross wedge rolling process, this process can be performed using workpieces with high dimensional tolerance.
references
[1] C. Weiss, M.M. Morlock, N.P. Hoffmann, Friction induced dynamics of ball joints: instability and post bifurcation behavior, European Journal of Mechanics A/Solids 45 (2014) 161–173. [2] P. Alexandru, I. Visa, C. Alexandru, Modeling the angular capability of the ball joints in a complex mechanism with two degrees of mobility, Applied Mathematical Modelling 38 (2014) 5456–5470. [3] J.G. Eom, Y.H. Son, S.W. Jeong, et al., Effect of strain hardening capability on plastic deformation behaviors of material during metal forming, Materials and Design 54 (2014) 1010– 1018. [4] J.H. Song, Y.T. Im, Expert system for the process sequence design of a ball stud, Journal of Material Processing Technology 89–90 (1999) 72–78. [5] Z. Gronostajski, M. Hawryluk, M. Kaszuba, et al., Die profile optimization for forging constant velocity joint casings, Archives of Metallurgy and Materials 56 (2011) 551–558. [6] Z. Gronostajski, M. Hawryluk, The main aspects of precision forging, Archives of Civil and Mechanical Engineering 8 (2008) 39–55. [7] C.S. Im, S.R. Suh, M.C. Lee, et al., Computer aided process design in cold-former forging using a forging simulator and a commercial CAD software, Journal of Materials Processing Technology 95 (1999) 155–163. [8] W.S. Weroński, A. Gontarz, Z. Pater, The research of forging process of eccentric part on three slide forging press, Journal of Materials Processing Technology 177 (2006) 214–217. [9] G. Winiarski, A. Gontarz, Z. Pater, A new process for the forming of a triangular flange in hollow shafts from Ti6Al4V alloy, Archives of Civil and Mechanical Engineering 15 (2015) 911–916. [10] B. Heibing, M. Ersoy, Chassis Handbook. Fundamentals, Driving Dynamics, Components, Mechatronics, Perspectives, Deutsche Nationalbibliothek, Berlin, 2011. [11] http://www.afdex.com [accessed 15.03.16]. [12] J. Zhou, Y. Yu, Q. Zeng, Analysis and experimental studies of internal voids in multi-wedge cross wedge rolling stepped shaft, International Journal of Advanced Manufacturing Technology 72 (2014) 1559–1566. [13] H. Ji, J. Liu, B. Wang, et al., Cross-wedge rolling of a 4Cr9Si2 hollow valve: explorative experiment and finite element simulation, International Journal of Advanced Manufacturing Technology 77 (2015) 15–26. [14] J. Tomczak, Z. Pater, T. Bulzak, The influence of hollow billet thickness in rotary compression, International Journal of Advanced Manufacturing Technology 82 (2016) 1281–1291.
archives of civil and mechanical engineering 17 (2017) 729–737
[15] J. Zhao, L.Q. Lu, Z.H. Hu, Study on varying rule of mechanical parameters in forming automobile semi-axes with multiwedge cross wedge rolling, Proceedings of the International Conference on Mechanical Engineering and Mechanics 1–2 (2007) 1723–1727. [16] W. Yuan, L. Wang, T. Yuan, Experimental study on cross wedge rolling process of aluminum alloy material, in: 5th International Conference on Advanced Engineering Materials and Technology 5 (AEMT 2015), 2015, 415–422.
737
[17] Z. Pater, A. Tofil, J. Tomczak, T. Bulzak, Numerical analysis of the cross wedge rolling process (CWR) for a stepped shaft, Metalurgija 54 (2015) 177–180. [18] J. Tomczak, Z. Pater, T. Bulzak, Thermo-mechanical analysis of a lever preform forming from magnesium alloy AZ31, Archives of Metallurgy and Materials 57 (2012) 1211–1218. [19] Z. Pater, Analysis of the helical-wedge rolling process for producing a long stepped shaft, Key Engineering Materials 622–623 (2014) 893–898.