In order to examine the quality of the hydro-mechanically drawn cups, the maximum percentage reduction in thickness is calculated for each case, based on the following relationship [18]:
(1)
where t0 is the initial thickness of the blank and tf is the minimum thickness of the product at the end of the process. Having a desirable quality is a unanimous objective in various sheet metal forming processes. Different criteria can be employed in order to inspect the quality of a hydro-mechanically drawn cup. As the previous investigations, the maximum thinning percentage introduced by Eq. (1) is one of the criteria which also utilized in this research work. However, another quality index is additionally proposed and employed in the present study. This index is the uniformity of the final product. In other words, when the difference between the thicknesses of the final drawn cup at its various regions in comparison with the average value is the minimum, one can conclude that the product possesses an almost uniform wall thickness. This difference can mathematically be calculated using the below equation [19]:
(2)
where n is the number of points selected on the profile of the deformed part for measuring the thickness, tave is the average thickness of the whole product and ti is the sheet thickness at each specific point on the profile of the drawn cup. It is worth mentioning that both the variables introduced in Eqs. (1) and (2) are dimensionless and this is a favorable advantage. In current research work, the maximum reduction in both parameters of Rth and Vth in the produced products is desirable. A piece of this article is assigned to studying the influences of different process parameters on these variables. These effects are not the same for various regions of the workpiece. For this reason, the product profile is divided into three main areas, namely the bottom or the base of the cup (A), the area which is in contact with the punch edge (B); and the main wall of the cup (C).
4.1. Design of experiments
In this section, in order to evaluate the degree of effect of each process variable and to examine the importance of the interactions between different factors, the design of experiments (DOE) employed in this study is described. The parameters investigated for HMDD of Al 2024 sheets consist of the maximum fluid pressure (in three levels of 5 MPa, 10 MPa and 15 MPa), process temperature (in two levels of 200 0C and 250 0C) and the initial diameter of the workpiece (in two levels of 70 mm and 73 mm). Therefore, altogether 12 practical HMDD experiments based on the full factorial approach were performed. The percentage effects together with the design matrix of the tests are demonstrated in Fig. 4. As shown in Fig. 4, any change in each process variable, definitely affects the response obtained from the practical experiments. This figure also implies that the most important factors affecting the quality of the drawn cup are the blank diameter, the process temperature and the maximum fluid pressure, respectively. On the other hand, the maximum interaction in production of the Al cups has occurred between the temperature and the initial blank diameter. However, the levels of interactions are negligible in comparison with the main effects of the process variables. For this reason, in specifying the optimum conditions for HMDD of 2024 Al sheets, the effect of interactions can reasonably be ignored.
4.2. Validation of FE simulations results
Fig. 5 shows two cups drawn under the same test conditions but with different heights. Both the experimental sample (Fig. 5(a)) and the relevant numerical model (Fig. 5(b)) are illustrated together for a better comparison between the real and FE components. It should be noted that the samples in Fig. 5 are produced by applying 15 MPa fluid pressure and at 250°C. In order to validate the findings obtained from the finite-element simulations, the experimental and numerical thickness distributions for these specimens are illustrated in Fig. 5(c). The distance from the center of the workpiece is obtained along the diametrical section of the final product. The thickness of the deformed sheet was measured at different regions of the workpiece using a QLR point micrometer. The correlation between the FE model and experimental results is encouragingly very good and the maximum difference between these two sets of findings is about 6%. Moreover, it is obvious from Fig. 5 that, the minimum thickness of the drawn cup is around the punch edge-blank contact area.
4.3. Repeatability of experimental results
In order to verify the repeatability of the practical HMDD experiments, a typical test considered for studying each process variable was carried out twice (Table 1). The results of the repeated experiments together with those of the original ones are given in the Table 1. The outcomes presented in this table confirm a good consistency of the experimental findings and, consequently, one can deduce that the HMDD tests are repeatable.
Table 1
Results for inspecting the repeatability of HMDD experiments for Al 2024 alloy.
Specimen No.
|
P (MPa)
|
Pb (MPa)
|
T (0C)
|
H (mm)
|
Rth
|
Vth
|
1
|
15
|
0
|
250
|
22.6
|
10.7
|
0.842
|
2 (R)
|
15
|
0
|
250
|
22.6
|
10.1
|
0.825
|
3
|
15
|
0
|
200
|
22.6
|
15.6
|
0.737
|
4 (R)
|
15
|
0
|
200
|
22.6
|
15.9
|
0.761
|
5
|
15
|
3
|
250
|
22.6
|
torn
|
torn
|
6 (R)
|
15
|
3
|
250
|
22.6
|
torn
|
torn
|
4.4 The effect of fluid pressure
The fluid pressure is one of the important operation variables which could affect the uniformity of the product in sheet hydroforming process. The pressure paths employed for practical HMDD tests in this research are shown in Fig. 6. The mechanism of applying the pressure is such that by moving the punch downwards, the oil in the container underneath the blank is compressed up to a specific level. Afterwards, the safety valve is opened and the excess pressure is relieved. For the three pressure paths shown in Fig. 6, the relevant HMDD experiments were conducted and the thickness distributions of the products are plotted in Fig. 7. The corresponding numerical distributions of the thickness are also included in the same figure. The process temperature and the height of the deformed sheet in these experiments were considered 250°C and 22.6 mm, respectively.
Figure 7 implies that the greatest effect of the fluid pressure can be observed in regions C and, then, B, where the ultimate variations (either thickening or thinning) with respect to the initial thickness of the blank occur. This is due to the presence of fluid pressure on the underneath of the workpiece, which can prevent the cups from sudden rupture. Furthermore, the fluid pressure causes a uniform distribution of the punch force during the forming process. According to Fig. 7, there is the least thickness change at the bottom of the product (region A). The reason for this phenomenon is the reduction of tensile strains in region A due to the frictional force occurred at the punch-workpiece contact surface and the slight slipping of the sheet on the punch face. The largest influence of change in fluid pressure inside the die chamber on the variation of thickness is related to the corner (region B) and wall (region C) of the final drawn cup. In the region B, the workpiece is bent and stretched simultaneously, and this can cause relatively large strains at this area. The maximum compressive stress, instead, is created in the flange and wall regions of the product. Drawing the sheet into the die cavity and, consequently, reducing its radius and perimeter can increase the thickness at the flange area. The increase in thickness is due to hoop compressive stresses.
Table 2 summarizes Rth and Vth of cups with initial blank diameters of 70 mm and 73 mm, drawn with various maximum fluid pressures and at two different process temperatures. By increasing the maximum fluid pressure from 5 MPa to 15 MPa, Rth of the final product with the initial blank diameter of 70 mm and process temperature of 200 0C and 250 0C decreased by 17.1% and 28.3%, respectively. The maximum improvements in thickness variation of the product for the same pressure range and at process temperatures of 200°C and 250°C were obtained equal to 27.9% and 31.7%, respectively. As the initial diameter of the blank increases, the amount of contact surface and the related friction at the workpiece-matrix interface in the flange area increases and more areas of the sheet are subjected to large bending strains due to the die curvature. Therefore, with the downward movement of the punch, the blank enters the die cavity with more difficulties and the thinning and consequently the thickness variation in the deformed cup with a larger initial diameter increases. The results demonstrated in Table 2are in accordance with the real conditions of the HMDD process.
Table 2
Rth and Vth values due to changes in maximum fluid pressure, process temperature and initial blank diameter.
Fluid pressure
(MPa)
|
70 mm
|
73 mm
|
200 0C
|
250 0C
|
200 0C
|
250 0C
|
Rth
|
Vth
|
Rth
|
Vth
|
Rth
|
Vth
|
Rth
|
Vth
|
5
|
18.7
|
1.04
|
14.5
|
1.23
|
Torn
|
22.8
|
1.72
|
10
|
16.8
|
0.86
|
11.3
|
0.95
|
22.6
|
1.46
|
16.9
|
1.59
|
15
|
15.5
|
0.75
|
10.4
|
0.84
|
18.2
|
1.21
|
14.3
|
1.36
|
The maximum thickness reduction versus changes in the fluid pressure at 200 0C and 250°C are shown in Fig. 8. The trend of variation in the maximum thickness reduction is similar for both the temperatures. Moreover, it can be seen that with increasing the fluid pressure in the process at 200 0C, the slope of the curve is less than that of 250°C. This is mainly due to the increased formability of the material at higher temperatures. Again, a good correlation exists between the FE and experimental findings shown in Fig. 8. It is obvious that for pressures greater than 15 MPa, no significant enhancement in the maximum thickness reduction can be achieved. For this reason and since excessive fluid pressure in the container considerably increases the required forming load, a pressure of 15 MPa has been selected for the remaining HMDD tests in the present investigation.
4.5 Influence of the pre-bulge pressure
Pre-bulge pressure (Pb) is another process parameter influencing the quality of the HMDD products. This pressure is different from the fluid pressure in the container and is applied underneath the blank just before starting the punch movement. In fact, the pre-bulge pressure causes the upwards swelling of the central region of the work sheet before downwards displacement of the punch. Fig. 9 shows that how the amount of pre-bulge pressure affected the soundness of the drawn cup. The percentage maximum reductions (Rth) together with the values of the uniformity index (Vth) for these three cases are summarized in this figure. For these experiments, the maximum fluid pressure and process temperature were 15 MPa and 250°C, respectively. The variations in the product thickness are also plotted in Fig. 9 for a couple of defect-free parts deformed with various pre-bulge pressures. Applying pre-bulge pressure at the beginning of the HMDD process can reduce the negative effects of the die curvature on both the Rth and Vth. This initial pressure has the role of a holder at the underneath of the sheet and it is expected that this initial pressure prevent application of extra load from the punch to the sheet before movement of punch and the contact with the workpiece. This parameter could also have a positive influence on the uniformity of the final product. On the other hand, with an excessive increase in the pre-bulge pressure, the punch needs greater force at the beginning of the movement, and the interaction between the mechanical load and the fluid pressure on both sides of the worksheet has a negative effect on it. As shown in Fig. 9, the application of pre-bulge pressure had initially positive impacts on the maximum percentage of thinning (about 17%) and the thickness variation of the final product (about 30%). Nevertheless, an excessive level for the pre-bulge pressure has caused tearing of the product. All these figures imply that for 2024 aluminum alloy under consideration, the most uniform and sound cup can be hydro-mechanically drawn with a pre-bulge pressure of 1 MPa.
4.6 Influence of the process temperature
By increasing the drawing temperature, one can improve the ductility of the blank to attain deeper cups. Nevertheless, the energy consumption increases and the strain hardening behavior decreases when the HMDD operation is carried out at elevated temperatures. For this reason, four levels, namely 100 0C, 200 0C, 250 0C and 300 0C, around the transition temperature of the alloy have been selected for conducting the HMDD experiments. Fig. 10 demonstrates the products drawn at different temperatures and a fluid pressure of 15 MPa together with the relevant values of Rth and Vth for regions A and B of the cups. It should be noted that no pre-bulge pressure has been considered for drawing the specimens shown in Fig. 10. It is clear that there is no significant reduction in thickness in region A of the workpiece. This is mainly due to the increased adhesion between the sheet and the punch face, which consequently decreased the plastic deformation of the sheet. By increasing the temperature, the value of Rth at region A is slightly and monotonically decreased. The main effect of operation temperature occurs at region B, where at 100 0C the blank is torn, at 200 0C value of \Rth is 15.6%, at 250 0C it decreases to 10.7% and finally at 300 0C thickness reduction is again increased to 17.9%. In current study, with increasing the process temperature from 200 0C to 250 0C and from 250 0C to 300 0C, the thickness variation respectively increased 14.2% and 41.2%. Considering that the increase in the temperature of the forming process (due to the decrease in the flow stress and increased formability at higher temperatures) reduces the quality of the final product, the reduction in the uniformity of the product is quite expected. For this reason, and considering both the quality indices (i.e. Rth and Vth), the appropriate temperature for the HMDD process of Al 2024 alloy is suggested to be 250 0C.
4.7 Microstructural studies
One of the important effects of a forming operation is the change in the microstructure of the component which in turn affects the mechanical properties of the product. After grinding and polishing the test samples with different sandpapers having various grades, a Buehler Ltd micro-hardness tester was employed to evaluate the hardness distributions, shown in Fig. 11. With this regard, a force of 100 grf was applied for 20 sec [20]. For each point, the hardness was evaluated three times and their average was considered as the final value. The average hardness in regions A, B, and C was calculated to be 142 HV, 152 HV, and 175 HV, respectively. It should be mentioned that the Al cups studied in this section were drawn at a temperature of 250°C.
A barker’s solution was used to obtain an appropriate microstructural image of 2024 aluminum alloy sheet [21]. The proper time for etching process was considered to be about 60 second. A micrograph of region A (with maximum average grain size) and another for region C (with minimum average grain size) are shown in Fig. 12. Using the linear intercept method, the average grain size for these two areas was calculated to be 15 µm and 12 µm, respectively. In order to establish a relationship between the grain size and the strength of the material in the present research, the Hall-Petch relation can be employed [22]:
σy = σ0 + kd-0.5 (3)
where, d and σy represent the grain diameter and the yield strength, respectively; and k and σ0 are constant material parameters. Based on Eq. (3), one can be realized that the grain size has an inverse relation with the strength and, consequently, with the hardness of the material. Comparing the grain sizes obtained from microstructural studies (Fig. 12) with the hardness values (Fig. 11) for different regions of the component, one can confirm the accuracy and validity of the experimental results for the present investigation.
Another important analysis in microstructural studies is related to the texture of the material, which is one of the important factors in changing the mechanical properties of polycrystalline structures. Macrotexture measurements were carried out on the plane of RD-TD surface using X-ray diffraction (XRD). The texture was determined employing a PANalytical X-ray diffractometer with Cr Kα radiation up to a tilt angle of 75˚. Two incomplete pole figures (PFs) including those of {1 1 1} and { 2 0 0} planes were utilized to calculate the inverse pole figures (IPFs) using the TexTools software. Since the maximum and minimum hardness and grain sizes were related to regions A and C, the texture study was carried out for these two areas. In FCC crystal structure, the {1 1 1} plane is the main slip plane and <1 1 0> is the main slip direction. The best condition for plastic deformation is when the {1 1 1} plane of majority of grains are oriented parallel to the sheet surface. The IPFs of rolling direction (RD), transverse direction (TD) and normal direction (ND) for regions A and C are shown in Fig. 13. The IPF of RD for region A demonstrates contour lines with the maximum intensity of 1.2 MRD which is concentrated in [1 1 1] direction, implying that this direction of many grains is parallel to the RD. The IPF of TD shows that the main texture is [2 0 3]║TD with the maximum intensity of 1.4 MRD. Finally, the IPF of ND expresses that there is a very strong [0 0 1]║ND texture in the region A with a maximum intensity of 1.9 MRD. On the other hand, for region C, the IPF of RD does not show any significant change in maximum texture intensity. However, the main texture direction was changed from [1 1 1] to [1 1 0]. The maximum texture intensities of TD and ND IPFs for the cup wall region are decreased to 1.1 MRD and 1.7 MRD, respectively. The main texture direction of ND for this region was changed into [3 2 3] plane.
The preferred crystallographic orientation is one of the most important and influential factors on the micro hardness of the specimens. For aluminum alloy with FCC structure, [1 1 1] direction is introduced as the high-density crystallography direction, because of the closed-pack atomic arrangement in this direction. It can be claimed that when the texture is such that the [1 1 1] direction is aligned to the axis of the indenter in a hardness test, a higher HV should be perceived in the experiment. The texture intensities of [1 1 1]║RD for regions A and C were calculated to be 1.15 MRD and 0.83 MRD, respectively. Based on this observation, region A should possess a higher hardness in comparison with region C. This outcome is in contrasting with the experimental results of the hardness test and Hall-Petch estimation for these two areas, i.e. the finer the grains (region C) the harder the material.
After facing the difference between evaluations of the Hall-Petch equation and the crystallographic texture in interpreting the hardness test results, it was decided to examine the precipitation hardening which is another factor influencing the hardness of the material. Indeed, the precipitation hardening is one of the main mechanisms for strengthening 2024 Al alloy. For this reason, the study of development and morphology of precipitations, secondary phases and intermetallic compounds in this Al alloy can be helpful to investigate the factors changing its mechanical properties such as hardness. Therefore, the distribution of precipitations in different regions of a typical 2024 Al alloy drawn cup has been examined. The weight percentages of the elements constituting 2024 Al alloy are given in Table 3. Considering that after aluminum, the copper element has the maximum weight percentage in the chemical composition of 2024 Al alloy, the presence of copper precipitations in the material is likely. In order to investigate the precipitations produced in the workpiece during the forming process, the energy dispersive spectroscopy (EDS) test was carried out. Fig. 14 illustrates the results of EDS tests for different regions, namely A and C, of the final product. As can be seen in this figure, Al2Cu precipitations could be considered as the main intermetallic compounds distributed discontinuously at the grain boundaries in region A. However, in region C, Al2Cu precipitations are observed more continuously at the grain boundaries, compared with the dispersion in region A. The presence of continuous precipitations at the grain peripheries is a sign of increase in the hardness on the alloy. The grain boundaries are usually considered as one of the most sensitive areas for the formation of defects in a material. Thus, the presence of precipitations at the grain boundaries prevents creation of defects at these areas and, consequently, increases the hardness of the alloy. On the other hand, accumulative or colonial precipitations are observed in both the regions. The image obtained from the EDS test demonstrates that the colony precipitations formed in region C are smaller than those of region A. Moreover, they are scattered inside the grains with comparatively smaller dimensions. This type of precipitations distribution is another source of hardness enhancement in region C. Therefore, it is reasonable that an intensified increase in the hardness of region C, in comparison with region A, should be detected. Now it can be claimed that the presence and sort of distribution of Al2Cu particles as high-hard intermetallic compounds has a more significant and meaningful influence on the micro hardness of 2024 Al cups than the preferred grain orientation.
Table 3
The weight percentages of the elements making up 2024 Al alloy under consideration.
Element
|
Al
|
Cu
|
Mg
|
Mn
|
Weight Percentage
|
93.50
|
4.40
|
1.50
|
0.60
|
In continuation of this section, the amount of induced plastic strain as well as the interfacial contact pressure of the deformed specimen and their relationships with the hardness and grain size of each region of the Al cup drawn via HMDD process is investigated. Regarding this concern, the plastic strain distribution contour and the diagram of the contact pressure of the deformed part are shown in Fig. 15. The region of the sheet, which is between the die wall and the punch during the HMDD process (region C), has experienced the most elongation and, consequently, the plastic strain created in this area is greater than other regions. These values are in agreement with distribution of the effective strain in the product (Fig. 15(a)), where more severe deformation occurred in region C, in comparison with region A. The number of grains per unit area was found to be respectively 5202 and 5951 for regions A and C. It is reasonable that the greater the number of grains per unit area, the harder is the material. The contact pressure variation at the tool-cup interface is shown in Fig. 15(b) for a drawn depth of 11 mm. When the punch pulls the sheet into the die cavity, high bending strains are induced into the sheet by the die edge. The interfacial friction between the flange area and the die edge (region C in Fig. 15(b)) intensify this situation and increases the amount of induced plastic strain. On the other hand, the regions of the flange farther from the center of the blank experience greater effective strains when arriving on the die edge. That is why, more or less, the trend of the contour shown in Fig. 15(a) governs almost the whole HMDD operation.