The part examined is shown in the following picture.
Simulation set-up
This part of the report lists the main parameters used in the simulation.
- Die material: H13 steel.
- Metal: Al-EN-AC 47100 AlSi12Cu1(Fe).
- Heat transfer coefficient between metal and die = varying linearly from 2500 W/m^2/K at 350 °C to 7500 W/m^2/K at 700 °C.
- Average die temperature = 250 °C.
- Metal temperature on the biscuit = 650 °C.
- First phase velocity = profile varying from 0.1 m/sec to 0.8 m/sec according to the following graph, where the ordinate shows the plunger stroke in cm.
- Second phase velocity = 2.5 m/sec
- Plunger diameter = 100 mm
Evaluation of results – choosing the correct shot sleeve size
There are essentially two main problems in producing pan-like parts. First, gas porosity, which can cause the part to be rejected during machining or PTFE coating. The second problem is the formation of flaking caused by thermal differences; to avoid it, filling must be as uniform and as fast as possible. We recommend considering pre-filling before starting the second phase.
In this first part of the simulation, the filling of the cavity was compared for different shot sleeve diameters, and therefore plunger diameters. Specifically, plungers with diameters of 100 mm, 70 mm and 80 mm were tested.
In general it would be advisable to work with fill ratios above 50%; in this sense a 70 mm plunger would be advisable. With an 80 mm plunger the fill ratio is 38%, and it drops further with the 100 mm plunger. This generally causes air entrapment and therefore gas porosity in the final part. In this case, however, a vacuum system is present, so the problem becomes less important: if the extraction system is used to its best, air entrapment can be greatly reduced even while only the shot sleeve is filling. This effectively allows low fill ratios with more manageable first phase velocities, since the formation of waves as the free surface of the metal rises in the sleeve is not a problem, as it would be without vacuum.
As for metal temperature, low fill ratios cause a greater temperature drop while the sleeve fills. However, experience shows that this drop can be considered negligible: if the heat exchange surface between metal and air increases by 15-20%, the one between metal and die increases by only 3-4%. This matters because heat exchange between air and metal is negligible compared with that between die and aluminium. In short, the increase in heat loss going, for example, from a 100 mm to an 80 mm plunger is negligible, especially considering the losses due to transporting the metal and pouring it into the sleeve. A negative aspect of working with low fill ratios, however, is that the metal absorbs more hydrogen from the air.
To conclude these considerations, larger plungers allow the die to be filled at lower pressures and velocities. For pans, very high pressures are not needed to counter aluminium shrinkage: in most cases the bottom of the pan is thicker than the outer rim. Since the gate is on the rim, the thickest part of the casting is separated from the gates when the rim solidifies; it is therefore pointless to use high pressures to counter the shrinkage of the pan bottom.
On this basis, several first and second phase velocity configurations were evaluated for the three plungers mentioned. The previous picture shows filling with the 70 mm plunger: the flow front closes on itself before going past the bottom of the pan. This behaviour is certainly not optimal.
The next picture compares the cavity filling obtained with an 80 mm plunger; the problem highlighted above remains, although to a lesser extent.
The following picture shows the cavity filling obtained with the 100 mm plunger. The overall filling of the pan improves, because the metal flows across the whole bottom of the pan and then closes on the far wall.
Note that the three solutions did not use the same velocities: they were varied so as to obtain a uniform velocity at the gate. This shows that the phenomenon highlighted is not due to wrong velocity settings, but to a different distribution of the pressures in the cavity, which makes the filling velocities non-uniform. The choice of plunger therefore affects filling.
In conclusion, considering all the observations above, the 100 mm plunger is preferable to the smaller ones. Work therefore continued only on the 100 mm configuration.
Evaluation of results with the 100 mm plunger
The first consideration on the die layout concerns the vacuum system: the design of the vent channels is not optimal, regardless of where the vent inlets are placed.
The next picture shows the recommended changes, in red. The gate sections should be about 30 mm x 2 mm = 60 square millimetres. The secondary channels should have a trapezoidal section 9 mm wide and 5 mm high with 10° side draft, while the main channel to the vacuum valve should also be trapezoidal, 12 mm x 6 mm with 10° sides.
The position of the system inlets will be evaluated later, based on the simulation results.
These changes are recommended to optimise the current system: the excess "stops" (points where the metal changes direction) have been removed. They must also be made according to the following scheme.
It is important to prevent dirt from reaching the vacuum valve, otherwise the whole system may lose efficiency. In this case it is important to start extracting the air from the very first stages of sleeve filling.
Below, the filling of the pan in the original geometry is compared with the geometry modified according to the following specifications:
- Increased gate thickness, to allow higher filling velocities and therefore shorter times without affecting the gate velocity, which must be limited. Increase the existing restriction to 3.5 mm, over a length of about 280 mm, for a total section of 980 square millimetres instead of the current 715.
- Change of the runner design, with a modified plan layout and two lateral "shock absorbers" added.
- Change of the vacuum system as described above. This change is not related to the simulation results, but only to an optimisation of the current layout. Whether the vacuum inlet positions are correct will be checked below.
- First phase velocity = profile varying from 0.1 m/sec to 1 m/sec, with the same profile shown above.
- Second phase velocity between 2.6 and 3.2 m/sec, for a total filling time of 30-40 milliseconds.
Note that the cavity was pre-filled during the first phase. The second phase starts when the metal reaches the first area of the pan bottom, i.e. when the whole gate surface is already filled. This reduces possible gas porosity and also makes filling more uniform: with a substantial part of the casting already filled in the first phase, the uneven velocity distribution along the flow front during the second phase is much less evident. The drawback is a longer total filling time and therefore a lower overall quality. For this reason fairly high velocities were also used in the first phase, keeping an optimal filling time.
The two following pictures compare a first time step for the two geometries. Although there are no major differences, the filling of the pan in the picture on the right (modified geometry) is more uniform, with less prevalent side filling.
The two following pictures make the same comparison at a later step. Filling is indeed better in the picture on the right.
In the next two pictures, besides comparing cavity filling, the velocity distributions are evaluated. In the modified case the velocity pattern in the runner is much more correct and well distributed, a sign of better overall filling. In the picture on the left there are areas of stagnant metal (purple) at the sides of the runner.
To conclude the comparison, the two solutions are shown at two steps close to the end of filling. As the following pictures show, in the first case the metal flow front closes on itself, leaving air inside the bottom of the pan. The situation is clearly better with the recommended changes.
The correct position of the vacuum vent inlets remains to be evaluated. In light of the simulation results with the changes made, the evacuation system should be reviewed. Based on the last simulation, a double evacuation only at the top of the pan is likely to be sufficient. This allows the side vacuum channels to be removed, reducing the alloy poured and the projected area. The recommended evacuation section is 91 square millimetres. This can also be seen from the following picture, which shows the air inside the die in blue.
Finally, the behaviour of the metal in the shot sleeve with the recommended velocities is shown. The sleeve fills well, even though this would not be strictly necessary with a vacuum system.
