3.1 Filling effect of gear tooth area
In this simulation stage, the billet with a radius of 32 mm and a height of 67 mm is used. the preheating temperature of billet, preheating temperature of lower die, upper die feeding speed and friction coefficient are 800℃,300℃, 7.5 mm/s, and 0.3, respectively.
During the process of gear punching forming, it is prone to incompletely filling at the tooth corners, resulting in excessive height of the insufficient section at either end of the gear teeth.
The lower die Internal gear dendendum angle pose the greatest challenge for gear teeth forming due to their farthest distance from the billet. Meanwhile, Good filling at the crest indicates that both tooth profile are filled with equal integrity.
Figure 4 illustrates the process of punching forming of gear tooth for spur gears. The billet first forms a drum shape under the action of the pressure of the upper die, and then the gear tooth is gradually formed from the middle zone to the lower zone in turn, and finally the filling of the bottom and the upper zone of the gear tooth is completed in the final stage of punching. Five points with the same axial spacing are selected in turn on the same gear tooth in the final stage of punching forming which is shown in Fig. 4(e), and total displacement and equivalent strain are analyzed and plotted in Fig. 5 and Fig. 8, respectively. As given in Fig. 5, the billet displacement at the top zone of each gear tooth position remains consistent, which means to gain gear formation with the precision requirement. The radial displacements at the tooth face and root points in different gear areas vary due to the varying intensity of metal flow during the forming process, leading to the different displacement of the five-point tooth face and root metal across different regions.
The radial filling process of gear tooth is shown in Fig. 6. During metal flow, the billet in the α zone does not contact the tooth profile inside the tooth cavity of the lower die and directly flows to the dedendum of lower die. On the other hand, because of the filling sequence is from the dedendum to the crest, the gear tooth profile is easier to shape than the tooth crest corner. Therefore, when the filling at the crest is fuller, it means that the filling effect of the whole tooth area is good.
3.2 Axial equivalent strain of spur gear
Figure 7 is the forming process of tooth along the axial. It is found that the completion time of tooth filling varies across different regions of the spur gear during punching, the resulting outcome is as such in varying durations of metal deformation for each region of the billet. Due to the asynchronous filling process in different areas of the gear teeth and the presence of radial and axial metal flow during forming, the metal flow within gear teeth is relatively intricate. As is widely acknowledged, the intricate flow of metal can result in uneven distribution of equivalent strain, thereby impacting the mechanical properties of materials. Therefore, to investigate whether this filling method affects the forming quality of gears, it is important to analyze the equivalent strain in the gear tooth area.
Figure 8 is the equivalent strain distribution. From Fig. 8, it is found that the accumulated equivalent strain in the tooth profile is larger than that of the crest, which indicates a more intense metal flow at the tooth profile. Figure 7 (f),(g) are the metal flow direction in forming process. It can be observed from Fig. 7 (f),(g) that the direction of metal flow is along radial at the initial stage of forming and along axial at the final stage of forming. Even after the gear teeth are completely filled in this region, the metal in the tooth profile continues to flow axially upward, resulting in a greater accumulated equivalent strain of the tooth profile than that at the crest in the same region.
In theory, the shorter the deformation time of the metal, the lower its accumulated equivalent strain. However, as shown in Fig. 8, the little difference between the equivalent strain curves corresponding of the crest and tooth profile of the tooth in the different region. This indicates a high level of uniformity in plastic deformation following punching of spur gears.
The axial punching process traditional gears involves gear teeth asynchronous filling, once a tooth is filled, the billet experiences continuous extrusion pressure from the root of the tooth to either side or one side of the tooth cavity. When there is greater metal deformation resistance, metal flow becomes difficult. As illustrated in Fig. 7,this spur gear radial forming process is also a non-synchronous filling process that occurs at the different heights throughout the entire tooth profile. However, Metal flows exclusively within individual tooth cavities, which have minimum resistance to deformation because there are no corners in the tooth cavity. The graph also shows also illustrates that during the initial (f) and final stages (g) of forming, metal flow is primarily directed towards the tooth cavity and along the tooth axis, respectively. Notably, there is no occurrence of metal flowing to both sides at the dedendum, indicating excellent uniformity in plastic deformation and gear forming quality.
3.3 Optimization of process parameters
3.3.1. Optimization of billet volume and feed allowance of upper die
The final stage of the punching process is shown in Fig. 9. It can be found that there is the bottom residual of the central hole because of process, whose thickness is indicated with the feed allowance(∆h) referring to the distance between the upper die and base plate after punching completion. Figure 10 is the relationship between load, wear rate of lower die and upper die displacement. It is shown in Fig. 10 that the load and wear rate dramatically increase with the distance reduction between the upper die and base plate reach when the final stage of gear forming, which will significantly affect the quality of gear forming and the service life of die.
While gradually increasing the feed allowance of the upper die, a slight increase in billet volume is necessary to ensure quality gear formation. The maximum allowable diameter of the billet for this punching process should be at least 1mm smaller than the diameter of the addendum circle diameter of the lower die. With a 32 mm radius billet, only an increase in height can result in increased billet volume.
Figure 11 is the relationship between billet height, lower die wear and feed allowance. It is shown in Fig. 11 that the billet height increase with the increase of feed allowance in order to ensure the good forming quality of the gear. As the feed allowance increases, the wear rate of the lower die decreases first and then increases. The analysis revealed that in the later stages of punching, the rate of wear and load on the die increases rapidly due to the increased friction between the billet and the lower die caused by phenomena such as lubricant depletion and difficult metal flow. Therefore, when the feed allowance is less than 1.5 mm, the die wear decreases as the feed amount decreases in the later stages of punching. However, as the volume of the billet continues to increase, the flash formed by the residual material will also be larger, which increase contact area and time between billet and die tooth cavity leading to greater depth of die wear due to friction. Therefore, considering production cost and die wear, simulation experiments of gear forming are performed with a cylindrical billet with height of 65.7 mm, radius of 32 mm, and feed allowance of 0.15 mm, respectively.
3.3.2. Process parameters optimization with response surface analysis
Two experimental design methods that widely accepted by scholars in response surface methodology are the Box-Behnken design (BBD) and central composite experimental design[31]. The BBD method is popular of optimization because of its lack of axial points, short calculation period, and absence of extreme process parameters[32].
Frictional wear as the main form of failure of the lower die in this punching process, affects the quality of gear processing to varying degrees and has an impact on the service life of the lower die.The Archard wear model was employed to analyze die wear during the forming process and optimized process parameters. Lee and Jou[20] improved the initial Archard wear model, and the improved model expression is:
Where W is wear, K is wear coefficient, T is temperature, H is hardness value, P is contact pressure, L is relative slip length.
Eq. (1) indicates that the wear area and depth are influenced by the relative displacement between the billet and cavity region, thus necessitating determination of optimal billet volume and relative displacement to minimize lower die wear. Therefore, it is necessary to determine in 3. 3. 1. the optimal volume of the billet for this stamping process as well as the upper die feed allowance.
Meanwhile, In Eq. (1) it can also be seen that the pressure, speed, and temperature values at each node are the primary factors influencing friction. Consequently, Box-Behnken Design method was employed to develop a three-factor, three-level 17-group simulation experimental program. The process parameters and their corresponding levels are presented in Table 4, while the experimental schemes and results are shown in Table 5.
Each set of simulation results in Table 5 was imported into Design-Expert 13, and analyzed using a quadratic model containing interaction and squared terms. Significance, lack of fit term, and correlation tests were performed by analysis of variance to confirm the validity of the results. A mathematical model for predicting die wear depth during gear forming process was developed using a second-order polynomial equation:
$${y}^{’}\left(x\right)={\beta }_{0}+\sum _{i=1}^{k}{\beta }_{i}{x}_{i}+\sum _{i=1}^{k}{\beta }_{ii}{x}_{i}^{2}+\sum _{i=1}^{k}{\beta }_{ij}{x}_{i}{x}_{j}+\epsilon$$
2
where \({x}_{i}\) is the design independent variables, \({y}^{’}\) is the response prediction, \(\epsilon\) is the residual error, \({\beta }_{0}\),\({\beta }_{i}\),\({\beta }_{ii}\),\({\beta }_{ij}\) are the regression coefficients.
A response surface fit is performed according to the least squares method of Eq. (2) to obtain the response surface function between the die wear depth W and the design variables of the process parameters in the gear forming process as follows:
W = 52.36975-0.01885T-98.255µ-4.0593V + 0.026Tµ-0.00776TV + 4.3Vµ + 0.000115T²+98.3µ²+0.30812V² (3)
where T is preheating temperature of lower die; µ is friction coefficient; V is feed speed of upper die.
Meanwhile, in order to further validate the accuracy of the prediction model, three sets of test scenarios were randomly selected for simulation analysis using Design-Expert 13. The results of experimental simulation and calculation with Eq. (3) are presented in Table 6, where W is the simulation result, WC is the data calculated by Eq. (3), and error (&) is calculated by |WC-W|/W. The error in Table 6 and the correlation of die wear prediction given in Fig. 12 indicate that Eq. (3) has a good agreement with simulation results and possesses good predictive capability for relevant parameters as demonstrated by the established response surface model.
From the 3D response surface plots and contour plots in Figs. 13 and 14, it can be observed that the friction coefficient has the least significant effect on the wear of the lower die during gear punching forming when compared to the lower die preheating temperature and upper die feed speed. This phenomenon is attributed to the rapid consumption of lubricating fluid under conditions of elevated punching pressure, increased frictional forces, and higher initial temperatures for both die and billet. Therefore, the lubricant cannot act on the whole process of gear forming and its impact on mitigating die wear is considerably inferior to that of lower die preheating temperature and upper die feed rate.
Form Fig. 15, it can be observed that the impact of upper die feed speed on die wear decreases initially and then increases as the speed increases. The reason is that as the punching speed increases, the entire forming process is shortened and the metal flow time of the billet in the die cavity area is reduced, thereby decreasing frictional wear on the die. However, once the velocity surpasses a certain threshold, the billet metal flows excessively fast, leading to an increase in the relative flow rate between the billet and die. Due to the combined effects of high velocity and work-hardening, the die experiences a rapid increase in load, resulting in accelerated wear.
Meanwhile, Fig. 15 indicates that the lower die preheating temperature is a secondary factor compared to the upper die speed. The reason for this is that the lower die temperature of molding process will exceed 350℃ due to a combination of factors, including the high initial temperature of the billet, significant heat generated by friction between the billet and die cavity, and a short feed speed resulting in limited heat exchange time with the environment. The single-factor analysis indicates that the role of oxide film on metal surface decreases and die wear increases as the temperature rises above 350℃. This is not conducive to the performance of the lower die to maintain at the initial temperature, thus requiring cooling of the die below 350℃ by spraying coolant after each punching.
Therefore, the optimized parameters for achieving minimum die wear are lower die preheating temperature of 280℃, friction coefficient of 0.27, upper die feed speed of 8.3 mm/s, respectively. The lower die wear is measured to be 1.79×10− 6 mm.
Table 4
Factors and levels of simulation experiment
Factors
|
Level 1
|
Level 2
|
Level 3
|
T (℃)
|
200
|
250
|
300
|
\({\mu }\)
|
0.25
|
0.3
|
0.35
|
V (mm/s)
|
5
|
7.5
|
10
|
Table 5
Experimental schemes and results
No.
|
T (℃)
|
\({\mu }\)
|
V (mm/s)
|
W (10− 6mm)
|
1
|
250
|
0. 25
|
10
|
1. 969
|
2
|
250
|
0. 3
|
7. 5
|
1. 822
|
3
|
250
|
0. 3
|
7. 5
|
1. 815
|
4
|
300
|
0. 35
|
7. 5
|
1. 918
|
5
|
300
|
0. 3
|
5
|
2. 108
|
6
|
250
|
0. 3
|
7. 5
|
1. 817
|
7
|
250
|
0. 3
|
7. 5
|
1. 831
|
8
|
300
|
0. 25
|
7. 5
|
1. 799
|
9
|
300
|
0. 3
|
10
|
1. 871
|
10
|
250
|
0. 3
|
7. 5
|
1. 813
|
11
|
250
|
0. 35
|
5
|
1. 997
|
12
|
250
|
0. 25
|
5
|
2. 121
|
13
|
250
|
0. 35
|
10
|
2. 060
|
14
|
200
|
0. 3
|
10
|
2. 168
|
15
|
200
|
0. 25
|
7. 5
|
1. 941
|
16
|
200
|
0. 35
|
7. 5
|
1. 934
|
17
|
200
|
0. 3
|
5
|
2. 017
|
Table 6
Test schemes and results of three sets of test scenarios
No.
|
T (℃)
|
\({\mu }\)
|
V (mm/s)
|
w
|
wc
|
༆(%)
|
1
|
290
|
0. 32
|
8. 5
|
17. 94
|
17. 25
|
3. 82
|
2
|
240
|
0. 28
|
7
|
18. 49
|
18. 06
|
2. 33
|
3
|
220
|
0. 26
|
6. 5
|
19. 18
|
18. 69
|
2. 55
|
2.1. Lower die life prediction
The assessment of punching die lifespan is typically based on the extent of wear incurred during a single punching process. However, the material strength of the die and wear resistance of the cavity area are also affected by heat transfer and spray cooling during different stages of punching forming. Therefore, it is employed the identical set of dies for simulating multi-punching forming of gears and predicting die service life via wear laws generated by multi-punching.
The forming simulation is conducted using the optimal process parameters obtained from response surface analysis, and both single and cumulative wear after each punching are recorded in Table 7. From the table, it is evident that there exists a variation in the amount of wear after each punching cycle. To enhance the accuracy of predicting die's specific service life (in Fig. 16), a linear regression model was employed to fit cumulative wear data resulting in obtaining an equation:
Where, \({W}_{n}\) is the accumulated wear value after the \(n\)th forming of the gear, unit (10− 6mm); \(n\) is the number of punching forming.
According to the combined influence of gear forming accuracy and machining allowance, the maximum allowable wear value in the gear tooth area of this punching lower die is 0.25mm. Based on Eq. (4), the estimated number of gears that can be produced is approximately 110,000 pieces before the failure of die. However, in the actual production process, the actual output quantity may deviate slightly from the predicted value due to factors such as oxidation loss caused by repeated heating and cooling of the die, variability in processing environment, and errors in processing procedures.
Table 7
Single and cumulative wear of multiple punching molding dies
Wear
(10− 6\(\text{m}\text{m}\))
|
Forming times
|
1
|
2
|
3
|
4
|
5
|
6
|
7
|
8
|
9
|
10
|
Single
|
1. 79
|
2. 41
|
2. 27
|
1. 83
|
2. 33
|
2. 92
|
2. 24
|
2. 12
|
1. 77
|
2. 49
|
Cumulative
|
1. 79
|
4. 2
|
6. 47
|
8. 58
|
10. 41
|
13. 33
|
15. 57
|
17. 69
|
19. 46
|
21. 95
|