Soldering Behavior of JIS ADC12 Alloy Die Castings and Its Mechanism
2018 Volume 59 Issue 9 Pages 1471-1476
Details
2018 Volume 59 Issue 9 Pages 1471-1476
Die casting, a highly efficient production method, is widely applied to the manufacturing of automotive components. On the other hand, extending the die life is required to meet the needs of reducing the production cost because die costs are included in the manufacturing costs.
As one of the primary factors leading to die failures, soldering was investigated in this study by observing the adhesion of ADC12 alloy to the die and the reaction between ADC12 alloy and die in die casting process, analyzing the diffusion of concerned elements in ADC12 alloy, and numerically simulating the flow and the solidification of the molten metal. It was revealed that ADC12 alloy adhered to the surface of the die and caused the surface layer of the die castings to remain on the surface of the die by breaking the die castings with shearing stress at the time of ejection. In addition, Al–Si–Fe compounds were found at the interface between the die and the remaining ADC12 alloy layer with the repetition of die casting. These compounds grew gradually with the concentration of Si in the remaining ADC12 alloy layer towards the interface by diffusion with the further increase of shot numbers.
This Paper was Originally Published in Japanese in J. JFS 89 (2017) 757–763.
Die casting, a process in which molten metal fills a die cavity at high speed and solidifies under high pressure, is widely applied in the manufacture of various automobile components because of its high productivity. In order to reduce the cost of die castings, much research has been conducted1–3) on the behavior of molten metal during cavity-filling and solidification and the associated effects on the internal and external defects of die castings. Another important consideration is the lifetime of dies, whose cost accounts for a large percentage of the overall cost. The main failure modes of dies, including soldering and heat cracking, have thus received increasing research attention.4)
Soldering is caused by a reaction between a die and an aluminum alloy adhered to the die during die casting.5) This reaction forms a compound on the surface of the die, the amount of which increases with increasing shot number, eventually leading to die failure. Although mold release agents are applied to prevent soldering, uniform coating of the die is difficult because of high temperature and complicated die shape.6,7) In addition, the oils in mold release agents become carbonized and affect soldering.8) Therefore, it is important to understand the reaction behavior between a die and aluminum alloys.
Various investigations have been conducted on the reaction between a die and an aluminum alloy when the former is immersed in the latter.9) During die casting, aluminum alloys solidify within several seconds after mold filling. The die is thus exposed to the molten aluminum alloy for a very short duration; in the remaining time, the die is in contact with the solidified aluminum alloy. Therefore, studies have also been conducted on the relation between the start time of the reaction and the temperature at the point when the die makes contact with the solidified aluminum alloy.10) However, the reaction behavior between the die and the aluminum alloy in the practical die casting process, in which the die is simultaneously exposed to both liquid and solid states of an aluminum alloy, requires further clarification. The release force required for die castings, as an index of soldering, has been examined for aluminum alloys (JIS ADC10 alloy)11) with various addition levels of Fe, Mn, etc. and types of die material.12) Nevertheless, various aspects of soldering in practical die casting remain unclear. The present report examines the adhesion of a practical aluminum alloy (JIS ADC12) to the die and the reaction behavior between them in the repetitive die casting process without the use of a mold release agent to clarify soldering behavior.
The die castings were made by injecting molten aluminum alloy (JIS ADC12) at 924 K into a rectangular cavity [100 mm (l) × 120 (w) × 10 mm (h)] with a mold pin (ϕ10 × 10 mm) (hereafter referred to as “pin”) made of JIS SKD61 tool steel at the center of the cavity, as shown in Fig. 1, with a 135-ton horizontal die casting machine. The draft angle of the pin was set to 0°. The die casting conditions are shown in Table 1. The die castings were made without the use of a mold release agent using two-step injection (at low and high speeds, respectively) after die casting twice at a low speed for die heating. Pins used for different numbers of shots (1–9, 100 and 300) were prepared using the above die casting method. A K-type thermocouple with a diameter of 0.1 mm was installed in the center of the pin to monitor the temperature during die casting. After die casting, the appearance of the pins was observed before and after the adhered aluminum alloy was removed via dissolution with 10% NaOH aqueous solution. Furthermore, the occurrence and development of soldering were investigated by observing the cross-sections of pins using scanning electron microscopy (SEM). Since Fe in aluminum alloys is effective in preventing soldering,11) in this study, JIS AD12.2 alloy containing a relatively low Fe content was used to facilitate the investigation of soldering behavior. Table 2 shows the composition of the used alloy (in terms of mass fractions).
Die castings for soldering test.
The filling and solidification behaviors of the molten aluminum alloy during die casting were simulated using the casting simulation software TopCast (Toyota Communication System Co., Ltd.). The parameter values used in the simulation are shown in Table 3.
The diffusion distance of each element was estimated as follows. The diffusion of an element in a solid can be expressed according to Fick’s second law as:
| \begin{equation} \frac{\partial C}{\partial t} = D_{k}\frac{\partial^{2}C}{\partial x^{2}} \end{equation} | (1) |
| \begin{equation} D_{K} = D_{0}\exp \left(\frac{-Q}{RT}\right) \end{equation} | (2) |
The variation of the concentration with time t, C(x, t), can be derived from eq. (1) as:
| \begin{equation} C(x,t) = A\exp \left(- \frac{x^{2}}{4D_{k}t} \right) \end{equation} | (3) |
The change in concentration on the left side of eq. (1) reflects the macroscopic change in the element amount and thus the diffusion that results from the jumping motion of the atoms of the element. The diffusion distance, x, of an atom can be expressed according to random walk theory14) [parabolic rule of diffusion, see eq. (4)], which states that the diffusion coefficient and the jumping motion are proportional to each other. The diffusion distance of each element was estimated as:
| \begin{equation} x = \sqrt{2D_{k}t} \end{equation} | (4) |
Table 4 shows the physical property values15) used in the calculations.
Figure 2 shows the flow behavior of the molten metal in the die cavity for the tested castings. The molten metal flowing out of the gate hits the pin at the center of the cavity and flows into the left and right sides of the cavity. After reaching the upper side of the cavity, the molten metal begins to fill the left and right sides of the cavity near the gate and then near the back side of the pin. Figure 3 shows the temperature distribution on the movable die surface at the time of ejection (6 s from the start of injection) from the 1st to the 4th shot. The temperature of the pin after the 1st shot is 693 to 733 K (yellow region) and its distribution is almost uniform. With increasing number of die casting cycles, the temperature of the pin increases, starting from the tip of the fixed die side; the temperature throughout the pin exceeds 733 K (orange region). The temperature of the root portion of the pin (at the movable die side) is lower than that of the tip of the pin. Figure 4 shows the changes in the maximum and minimum values of the measured pin temperature with increasing number of die casting cycles and a comparison of the measured and calculated temperatures T of the pin after the 5th shot. As shown, the temperature of the pin rises with increasing number of die casting cycles. The temperature obtained in the simulation agrees with the measured value.
Molten metal flow in cavity by numerical simulation.
Temperature distributions of movable die at ejection time by numerical simulation.
Comparison of calculated and measured temperatures of core pin during die-casting.
Figure 5 shows the results of the soldering of the aluminum alloy on the pin after the 1st to 5th, 8th and 9th shots, respectively. The aluminum alloy adhered to the entire surface of the tip of the pin after 3 shots, when the maximum temperature of the pin exceeded 730 K, as shown in Fig. 4. The portion of the pin to which the aluminum alloy adhered corresponds to the fixed die side, where the molten metal hit the pin at a higher velocity; the pin thus had the highest temperature, as described above. Then, the adhesion area spread gradually after the 4th shot, covering about 60% of the entire surface from the pin tip, until the 9th shot.
Change of soldering behavior with the increase of shot number.
Figure 6 shows the variation in the thickness L of the aluminum alloy layer adhered to the pin with the number of shots up to the 300th shot. The thickness of the aluminum layer adhered to the pin was 10 µm from the 3rd shot to the 9th shot. After that, with increasing shot number, the aluminum layer became thicker, peeled away, and a new aluminum layer adhered; this process then repeated. The thickness of the aluminum layer after the 300th shot was about 18 µm. The surfaces of the pins to which aluminum layers adhered were examined.
Change of Al layer thickness with the increase of shot number.
Figure 7 shows the surfaces of the pins after the aluminum adhesion layer was removed. Only tool marks (machining striations) were observed on the surface of the pin before die casting. After the 4th shot (two shots after the aluminum alloy began to adhere), still only tool marks were observed (i.e., there were no erosion traces such as pits). Then, after the 9th shot, a number of pits of submicron size were observed on the surface of the pin. After the 100th shot, large continuous pits were clearly observed, and after the 300th shot, many large erosion marks were also observed beside the pits.
Change of pin appearance with the increase of shot number.
Figure 8 shows the simulated temperature distributions of the 1/4 cross-sections of the die castings near the gate side at the time of the completion of filling and at the time of die opening for the 5th shot in Fig. 4. Immediately after the completion of filling, the temperature of the die castings around the pin was 840 K or lower (as indicated by the blue color), which was lower than that of the region away from the pin (860 K, as indicated by the light blue color). Similarly, at the time of die opening, the temperature (750 K or lower, as indicated by the light blue color) of the die castings around the pin was lower than that (790 K, as indicated by the green color) of the region away from the pin.
Temperature distribution of castings at 5 shots by numerical simulation.
These results suggest that the soldering was due to the breakage of the die castings, as reported by Aoyama et al.16) The strength of the region away from the pin was lower because of its higher temperature, compared to that of the region just around the pin; the region away from the pin was thus broken by the shear force generated when the die casting was pulled out from the pin.
3.2 Reaction between adhered aluminum alloy layer and pinIn order to investigate the state of the adhered aluminum alloy layers after the 9th shot when fine pits (erosion traces) began to appear on the surface of the pin, the 100th shot, and the 300th shot, the cross-section (circle in Fig. 5) was observed using SEM. The SEM images of the adhered aluminum alloy layers for each shot are shown in Fig. 9. A number of fine crystals with a size of 1 µm or less were observed at the interface between the adhered aluminum alloy layer and the pin after the 9th shot (i.e., 6 shots after soldering). After the 100th shot, besides the fine crystals, large crystals with a size of several micrometers were also observed at the interface. After the 300th shot, many large crystals with sizes of several micrometers were observed.
SEM micrographs of cross sections near the interface of Al layer.
In order to clarify the causes of the precipitation of these crystals, the distributions of elements on the cross-sections were measured using energy-dispersive X-ray spectroscopy (EDS). Figure 10 shows the distributions of Al, Si, Fe, and Mn along the cross-section of the adhered aluminum alloy layer (point A in Fig. 9) near the interface after the 9th shot. Although the adhered aluminum alloy layer was exposed to the high-temperature molten metal alloy for 6 shots after soldering took place, all of the elements were almost uniformly distributed throughout the layer without remarkable segregation.
Energy dispersive X-ray spectroscopy of Al layer (Shot number: 9).
Figure 11 shows the EDS analysis results of the fine crystals at the interface between the adhered aluminum alloy layer and the pin after the 9th shot. The concentrations of Fe and Si in the fine crystals (Point 1) were slightly higher than those in an area without crystals (Point 2) located at the same distance from the interface. These results suggest that the fine crystals are Al–Si–Fe-based compounds. The Fe in these compounds likely comes from impurities in the aluminum alloy and the pin (via elution). In order to confirm this supposition, the diffusion behavior of Fe in a die casting cycle was examined.
Energy dispersive X-ray spectroscopy of each point after 9 shots.
As shown in Fig. 4, the temperature of the pin increased sharply after the injection of molten metal and varied in the range of 470 to 760 K during die casting. Figure 12 shows the calculated diffusion concentrations of Fe in the aluminum alloy layer when the pin was kept in contact with the aluminum alloy layer in the temperature range shown in Fig. 4 for a die casting cycle time of 39 s. The diffusion rate of Fe depends on the temperature of the adhered aluminum alloy layer; the diffusion concentration increases with increasing temperature. For example, if the temperature of the adhered aluminum alloy layer is 542 K, the Fe concentration at a distance of 0.1 µm from the surface of the pin is 0% and thus Fe hardly diffuses. When the temperature of the adhered aluminum alloy layer increases to 752 K, the concentration of Fe rises to about 30% at a distance of 1 µm from the surface of the pin.
Relationship between iron concentration in aluminum layer and distance from pin surface.
Figure 13 shows the calculated diffusion distances d of the considered elements when the pin was kept in contact with the aluminum alloy layer at 650 K (the average temperature in Fig. 4) for 39 s. Although aluminum hardly moves in iron, iron moves about 0.2 µm in aluminum. The moving distance of silicon in aluminum is as long as 0.82 µm. Therefore, the erosion (pits) of the pin can be considered to be caused by the diffusion of iron from the pin into the adhered aluminum alloy layer. In addition, for the 1st shot after the aluminum alloy adhered to the pin, the diffusion distance of iron is as small as about 0.2 µm, and thus no Al–Si–Fe-based compound is likely to have formed. However, 6 cycles after the aluminum alloy adhered to the pin, the diffusion distance of iron increased to 1.2 µm, which is six times as large as that after the 1st shot.
Calculated values of diffusion distance of each element (Al, Fe, Si).
According to a report on the reaction initiation temperature of Fe/Al by Aoyama et al.,10) the reaction proceeds in a very short time at a temperature high enough. For example, the reaction of Fe/Al occurs in 6 s (corresponding to the curing time) at 822 K and in 36 s (total curing time for 6 cycles) at 799 K. From Fig. 8, it can be seen that the lower limit of the temperature range of the die castings around the pin changes from 740 to 820 K from the completion of filling to die opening. The reaction initiation temperature during the 6th shot is in this temperature range. During the 6th cycle, iron can diffuse into the adhered aluminum alloy layer to a distance of 1.2 µm from the surface of the pin and the temperature of the adhered aluminum alloy layer exceeds the reaction initiation temperature of Fe/Al due to the heat supplied from the side of the die casting. As a result, compounds with a size equal to or less than the diffusion distance of iron (1.2 µm) form on the surface of the pin. This is confirmed by Fig. 9, where the thickness of the compound observed after the 9th shot (6 cycles after the adhesion of the aluminum alloy) is 1 µm or less.
Figure 14 shows the EDS line spectra of aluminum and silicon in the adhered aluminum alloy layer (point A in Fig. 9) after the 100th shot. A backscattered electron image is also shown in the figure. Figure 15 shows the EDS analysis results of the compound. The silicon concentration in the adhered aluminum alloy layer near the surface of the pin after the 100th shot is higher than that after the 9th shot. Although the detailed mechanism requires further examination, considering that the diffusion distance of silicon in aluminum is relatively long (see Fig. 13), silicon in the adhered aluminum alloy layer can diffuse to and concentrate at the interface under high temperature during 100 cycles of die casting. Therefore, it can be considered that the Al–Si–Fe-based compound forms and grows on the surface of the pin through a reaction between the concentrated silicon and the iron diffused into the adhered aluminum alloy layer from the pin. Therefore, the concentration of aluminum near the surface of the adhered aluminum alloy layer (Point 3 in Fig. 15) is high and the concentration of silicon is higher than that of aluminum in the area without crystals near the interface with the pin (Point 2). Furthermore, the compound observed on the pin after the 100th shot is an Al–Si–Fe-based compound, the same as the fine compounds precipitated after the 9th shot.
Energy dispersive X-ray spectroscopy of Al layer (Shot number: 100).
Energy dispersive X-ray spectroscopy of each point after 100 shots.
Once the aluminum alloy adheres to the pin, fine compounds comprising aluminum, silicon, and iron form at the interface between the pin and the aluminum alloy layer over several shots of die casting. With an increasing number of die casting cycles, the compound becomes coarse due to the accumulation of silicon on the surface of the pin in the adhered aluminum alloy layer and the elution of iron from the pin.
To prevent soldering in the die casting process, we investigated the adhesion and reaction behavior between an aluminum alloy and a mold pin without the use of a mold release agent. The following conclusions were obtained.
