The KOBO method [1, 2] is an unconventional bulk material forming method based on cyclic change of the deformation path and localized plastic flow of material [3]. In most cases, the KOBO method is applied in the form of a direct extrusion, where the change of deformation path is realized by cyclic oscillation of the die, with given frequency and at a given angle (Fig. 1).
The most significant advantage of the KOBO extrusion is that it enables the cold extrusion (without preheating the billet and press/tooling parts) with a high extrusion ratio (on the order of several hundred [4]) of nonferrous metals and alloys, even those considered hard to extrude (e.g. EN-AW 7075 aluminum alloy). These properties of the KOBO method are associated with the concentration of lattice defects (Frenkel defects) in the billet zone closest to the die face, which is subjected to cyclic, high deformation in complex strain conditions. This leads to fluid-like flow of extruded material [5, 6] with radial plastic flow scheme [7].
The KOBO extrusion method, besides being an effective method of grain refinement in extruded material [8], gives an opportunity to model the effect of process parameters, such as die oscillation frequency and extrusion speed, on the extrudate’s microstructure and mechanical properties [9]. These parameters may also be changed during the process (constant force extrusion). By changing those parameters, it is possible to significantly lower extrusion force as well [10].
Constitutive models of the extrusion of commercially pure lead with cyclic torsion (not the KOBO method
The attempts to model the KOBO method with FEM methods that have been presented in a few publications [14, 15] are based on specific cases, simplified models and a very narrow range of process parameters. No numerical model of KOBO extrusion has been validated through experimentation, especially considering both mechanical and thermal aspects of the process as well as tooling geometry.
The tooling for KOBO extrusion differs from that used in conventional extrusion. The front portion of the oscillating die has a direct effect on the introduction of complex strain conditions in the extruded billet, yet the geometry of the dies’ front surface is mostly neglected in experimental work. Therefore, this work is focused on the impact of die geometry on the KOBO extrusion process. Based on preliminary research [7], the most favorable geometry of the front portion of the die was chosen. The experimental research presented in this paper shows the effect of die face geometry modification on extrusion force and the obtained extrudate’s structure and mechanical properties.
Extrusion tests were carried out on a direct extrusion KOBO press (Fig. 2) with characteristic parameters: maximum extrusion force of 2500 kN and maximum torque on the die of 2606 Nm.
The material chosen for experimental research was commercially bought EN AW-7075 T6 aluminum alloy in the form of an extruded round bar 60 mm in diameter. The material was cut into pieces 60 mm in length and annealed in the following cycle: heating and stabilizing the temperature to 415°C, holding for 180 minutes, cooling at 10°C/h to below 200°C and finally air cooling to room temperature. Obtained billets have tensile strength of R
The effect of the annealing process on the internal structure of the material was grain growth in the axial direction (Fig. 3).
The microstructure of alloy 7075 in the initial state (reference sample) is typical of the alloy after the annealing process (Fig. 4) [16, 17]. It consists of a matrix, a solid solution of alloying elements in aluminum with fine separations of irregularly shaped MgZn2 phase particles, large separations of FeAl3 intermetallic phase particles and a dispersive strengthening phase of Mg2Si (Fig. 4).
All tests were carried out with the same process parameters: extrusion speed 0.1 mm/s, die oscillation angle ±8°, die oscillation frequency 7Hz, cold extrusion (without preheating billets and tooling), extrusion ratio
For comparison, a conventional extrusion test with heating of the press container and ingot to 450°C was also performed.
Based on preliminary research [7], a die face geometry with radial grooves and a small prechamber was chosen (Fig. 5). In all cases, the bearing length was 3 mm. The only modification to the die face geometry was varying the number and depth of the grooves (Fig. 6, Table 1). The active diameter of the die was 44 mm (for a 60 mm billet). All dies were made of 1.2343 tool steel, hardened and tempered to ~56 HRC. A total of seven dies were used for experimental research, six with grooves and one with a flat face part for comparison. For easier identification of die geometry and corresponding results, each die geometry was given a code in the form GxDx: Gx represents the number of grooves, Dx represents the depth of the grooves, and NG indicates no grooves (Fig. 6). The tests were carried out in three series. Each series tested three different dies (one die per test). Adequate time elapsed between each test for press parts to cool down to room temperature.
Dimensions of the characteristic features of the dies used for experimental research
Code | Number of grooves | Grooves depth [mm] |
---|---|---|
G8D1 | 8 | 1 |
G8D2 | 8 | 2 |
G8D3 | 8 | 3 |
G6D1 | 6 | 1 |
G10D1 | 10 | 1 |
G12D1 | 12 | 1 |
NG | – | – |
Samples for tensile test and microscopic examination were prepared from front, middle and end sections of obtained extrudate (shortened to F, M, E later in text). Tensile tests were made according to the ISO 6892-1:2016 standard. Metallographic samples were prepared by mechanical grinding and polishing and etched with Kellers reagent according to the ASTM E407 standard. Macro and microstructure examination were made with a metallographic microscope and SEM.
The graph in Figure 7 records the extrusion forces as a function of stem displacement.
The first G8D1 test was disturbed by starting the extrusion process without first cleaning the container sleeve, which resulted in an increase in the resistance to pushing the butt through the container sleeve during the execution of the test. The force of pushing the butt during the cleaning of the press container sleeve is given in Table 2. After cleaning the sleeve, the force of pushing the butt was negligible.
Maximum recorded extrusion forces in the KOBO process using different die types
Test | Maximum extrusion force [kN] | Comment |
---|---|---|
G8D1 | 2086 | No container sleeve cleaning before test |
G8D2 | 1962 | |
G8D3 | 1962 | |
G6D1 | 1924 | |
G10D1 | 1906 | |
G12D1 | 1890 | |
NG KOBO | 1923 | Flat face die |
NG | 1458 | Conventional extrusion at 450° C |
Butt | 79 | Butt pushing force while cleaning container sleeve |
The maximum value of the extrusion force varies depending on the die variant used (Fig. 7, Table 2). For the first three tests, excluding the first trial, the maximum extrusion force reached the same values regardless of the die variant used. For the next three tests (series 2), the maximum extrusion force decreases with the change in tool geometry: the minimum force, 1890 kN, was obtained for the test made with the die with the largest number (12) of grooves with the smallest depth (1 mm). For the conventional hot extrusion test, an extrusion force lower than in all cases of cold extrusion in the KOBO process was obtained. There is a noticeable increase in force at the final stage of extrusion, which is characteristic of the conventional process but does not occur in the case of KOBO extrusion. The maximum extrusion force of the press (2500 kN) was not reached for any of the cases.
A static tensile test was carried out in each of the three series, adequate for the course of the experimental study. The results of the first series of tests showed that the material underwent deformational strengthening. Regardless of the tool used, the tensile plots (Fig. 8) and the determined strength properties (Table 3) are comparable. In two cases, the test showed a slightly higher value of tensile strength; both cases were samples from the final part of the extrudate and were probably cooled slightly faster than the rest of the extrudate during the opening of the press container after the extrusion process. The difference is ~20 MPa higher than the series average.
Mechanical properties of extrudate determined by static tensile test – series 1
R |
R |
A50 | |
---|---|---|---|
MPa | MPa | % | |
G8D1F | 194.3767 | 319.763 | 16.82835 |
G8D1M | 182.0377 | 325.0069 | 15.53416 |
G8D1E | 191.4832 | 351.6311 | 17.21147 |
G8D2F | 191.682 | 329.3694 | 16.09939 |
G8D2M | 184.0032 | 325.3811 | 17.21048 |
G8D2E | 174.6589 | 327.0506 | 14.48359 |
G8D3F | 191.8396 | 326.371 | 18.55923 |
G8D3M | 179.7999 | 322.9885 | 17.81401 |
G8D3E | 187.1591 | 344.8309 | 14.46475 |
The results of the second series showed that the material underwent deformational strengthening adequately with the first series. Regardless of the tool used, the tensile plots (Fig. 9) and the determined strength properties (Table 4) are comparable. The yield strength value marked in red in Table 4 is the result of software error, the value is not correct.
Mechanical properties of extrudate determined by static tensile test – series 2
R |
R |
A50 | |
---|---|---|---|
MPa | MPa | % | |
G6D1F | 139.9892 | 375.7772 | 18.3135 |
G6D1M | 203.2703 | 351.6911 | 15.75259 |
G6D1E | 203.6512 | 357.1238 | 15.73096 |
G10D1F | 219.1892 | 359.8165 | 17.40554 |
G10D1M | 213.132 | 362.4292 | 17.7133 |
G10D1E | 211.7307 | 372.7265 | 15.55199 |
G12D1F | 228.0526 | 372.5014 | 16.29631 |
G12D1M | 204.0853 | 361.9522 | 17.48726 |
G12D1E | 217.7237 | 377.1663 | 13.81927 |
The results of the third series of samples (Fig. 10, Table 5) showed that a die without grooves (NG KOBO) produced extrudate with strength properties comparable to those of the extrudate produced by dies with grooves. However, there is a noticeable decreasing trend with the sample location along the length of the extrudate. The value of the yield strength decreases noticeably with the length of the extrudate.
Mechanical properties of extrudate determined by static tensile test – series 3
R |
R |
A50 | |
---|---|---|---|
NG_KOBO F | 227.5532 | 363.6127 | 11.61306 |
NG_KOBO M | 210.6277 | 368.3837 | 37.3797 |
NG_KOBO E | 194.2238 | 354.8399 | 16.83634 |
NG F | 178.6093 | 314.8302 | 12.87594 |
NG M | 178.4273 | 318.8969 | 27.68791 |
NG E | 179.8115 | 303.1239 | 9.513819 |
The hot extrusion test using the conventional process produced extrudate with noticeably lower tensile strength and relative elongation.
In both cases, the elongation values marked in red in Table 5 are the result of specimens breaking outside the measurement area during the test.
The macrostructure of the alloys was evaluated on the samples made in cross sections. Observations of the samples in the undigested state did not reveal the presence of any defects formed during the extrusion process. The etching process revealed visible flow lines forming bands arranged circumferentially in the near-surface areas of all tested samples, regardless of the extrusion stage. Their density varies depending on the examined variant, reaching a maximum depth of about 0.5 mm (with a predominant depth of 0.2 to 0.4 mm) (Fig. 11).
For test NG, conventional hot extrusion, the presence of distinct peripheral flow lines was not observed after any of the deformation stages. Only after the initial stage (F), at a distance of about 0.8÷1.0 mm from the surface, a slight disruption of the microstructure is observed, arranged in the form of a ring about 0.2 mm wide (Fig. 12).
Observations of the microstructure of the specimens representing each sample and extrusion stage (F, M, E) included the near-surface area, where flow lines (high strain) were found, and the area in the rod axis. The microstructure of the 7075 alloy for all test specimens observed in the crosssection is characterized by fine, equiaxial grain sizes not exceeding 5
The microstructure of the alloy in the nearsurface area for all examined samples (Fig. 13) is closely correlated with the macrostructure. The morphology of the particle precipitates of the intermetallic phases corresponds to the course and arrangement of the flow lines. In the area of lower density of the lines, we observed an almost complete absence of particles of the strengthening phase MgZn2, appearing in the form of large, dark precipitates, and a much smaller amount of the other intermetallic phases visible in the form of very fine, bright precipitates. In the areas of higher concentration of the flow lines, there is a greater number of intermetallic phases. Outside the area where the bands/flux lines are present, the particle distribution of the separated intermetallic phases is uniform.
However, between the initial state and after the extrusion process, the number and size of the MgZn2 phase precipitates in the microstructure of the alloy decreased (fig. 14). A tendency for these parameters to decrease depending on the stage of extrusion was also observed, according to the F>M>E relationship.
The microstructure of the NG test (conventional hot extrusion) samples after deformation (Fig. 15) is the same for all stages of deformation (F, M, E). However, discrepancies were observed in its structure compared to the “typical” microstructure characteristic of alloy 7075 after the KOBO deformation process, which was observed for the variants examined earlier. Although the phase composition of the alloy did not change, differences were found in the morphology of the phase components. Observations using light optical microscopy showed variations in the relative volume and particle size of the Al2CuMg strengthening phase, which could be observed on both the cross-sectional and longitudinal sections of the rods. Separations of the strengthening phase are finer and there are significantly fewer of them compared to other deformation variants. The use of a scanning electron microscope (SEM) magnifying up to 3000x (Fig. 16) revealed the presence of dispersed superfine particles of the strengthening intermetallic phases MgZn2 and Mg(Zn, Al, Cu), which have been partially or completely spheroidized. Their relative volume also increased compared to the alloy after the KOBO deformation process. Such changes in microstructure are characteristic of the alloy’s aging process, which occurs under the influence of temperatures higher than the aging temperature.
In the case of the microstructure of samples from test NG, significant inhomogeneity of the morphology of the phase components is apparent, related to the banding typical of conventionally extruded bars.
The mechanical properties of the samples from the obtained extrudate show similar properties for all samples within a given series, which shows that the use of grooves with different geometries on the die faces does not significantly affect the tensile properties of the obtained extrudate. Extruded samples obtained using a die with a flat face in the KOBO process show a decrease in mechanical properties, especially the yield strength with the location of the sample. These results indicate that in this case the implementation of the process generates higher heats that decrease the properties of the pressboard, higher than in the initial phase of the process with a downward trend. The strength properties of the extruded samples obtained in the conventional hot extrusion test are lower than for the rest of the samples obtained from the KOBO process, which was predictable due to the realization of the process at a temperature above the recrystallization temperature of the extruded material.
All of the samples of extrudate produced by dies with face grooves show analogous internal macro and microstructure and morphology. The internal structure of the specimens correlates with the results obtained from testing the strength properties of the extruded product. The use of annealed material for testing and the analysis of the test results allow us to conclude that the input material, aluminum alloy 7075 subjected to precipitation strengthening, was strengthened only by plastic deformation. This fact indicates that the temperature emitted during the process is lower (<~470°C) and its influence is too short to activate the supersaturation process, and the material after extrusion did not undergo natural aging until the realization of strength tests. On all specimens obtained from the pressboard after the KOBO process, there is a clearly visible encirclement of the outer area of the expression bars that differs in structure from the structure closer to the axis of the extruded rod. This zone reaches a thickness of ~0.2–0.5 mm and, in relation to the structure of the material in the rest of the product volume, does not show clear features of deformation in the axial direction band structure. In the near-surface areas of the extrudate, a higher concentration of fragmented intermetallic phases was found. This gives rise to the conclusion that the material in the near-surface zone comes from an area of the billet that has been subjected to more intense plastic deformation. Regardless of the location of the sample from the pressboard (F, M, E) obtained by the KOBO method, the internal structure of the product is the same, showing no significant change in microstructure along the length of the product.
The KOBO method is an efficient method for extruding light metal alloys (in this case, aluminum alloy) in a process carried out without heating the billets and press components. The extrudate obtained during the experimental research is characterized by the same mechanical and structural properties regardless of the location of the test samples. The products obtained by the cold KOBO process do not bear the features of the influence of elevated temperature. The use of dies with different geometries of the front part reduced the extrusion force. The lowest extrusion force registered was 1890 kN, compared to >1900 kN in other performed tests. This indicates that the modification of the die face has a direct effect on the formation of a zone of intense plastic deformation near the die face, and conventional cold extrusion in the examined case is impossible to perform. Attention should be paid to the effects of microstructure transformation shown in this paper, specific for “cold” extrusion of KOBO, in comparison to the data presented in numerous publications on alloy 7075 extruded by traditional “hot” processes [18–20]. Extrusion by the KOBO method using a flat die was feasible; however, the extrudate bears signs of increased heat generation during the process, most likely due to friction. Circumferential shear bands in the near-surface zone of the extrudate, occurring in all analyzed samples obtained by the KOBO method, with an internal structure different from that observed near the axis of the rods, indicate that this is a volume of material that comes from the zone of intense plastic deformation. The volume of material closer to the bar axis shows features corresponding to axial plastic flow, which indicates that the billet material is deformed by moving along the zone of intense plastic deformation without being subjected to deformation in complex strain conditions.