Numerical modeling of castings in the production process.

Two case histories show how numerical simulation of the filling sequence and solidification provides the map for casting quality.

Dramatic improvements in quality and productivity in metalcasting are possible using modern computer technology. Numerical simulation has been used at Ford Motor Co.'s Casting Div. for process optimization since the beginning of 1991.

Flow Field

While working on a ductile iron casting manifold, foundrymen faced several problems with the casting's filling sequence. On the side of the casting where the downsprue was located, the melt filled the cavity with a higher flow rate than on the opposite side. This resulted in a nonuniform temperature distribution after the filling sequence, leading to cold shuts.

Figure 1 shows the ductile iron exhaust manifold, which the foundry produced for a 2.01 engine. Using a numerical modeling software program, Fig. 2 shows the casting with attached risers and a gating system. The enmeshment is completed automatically. This fact is important for a short processing time.

Figure 3 depicts the filling sequence for the old gating design. The colors refer to metal temperatures, according to the attached color scale. Gray elements are the areas that have not been filled.

Figures 4a and 4b provide cross sections of the two risers, which have differing flow rates after four seconds. Flow distribution problems occur because as the first riser already begins filling, the last riser has not even been reached by the melt. The arrows show the direction and the velocity profile of the fluid.

Figures 5 and 6 show the casting during the filling sequence at five and seven seconds, when 70% of the cavity is filled. The unbalanced flow pattern results in an unbalanced temperature distribution, causing problems during the following solidification.

Based on the results of this simulation, engineers changed the pattern to optimize the metal flow into the fourth riser. As seen in the optimized design (Fig. 7), the flow profile in the risers is now balanced (Fig. 8a and 8b). Figure 9 shows the flow pattern when 60% of the total cavity has been filled. This figure is comparable to Fig. 6. As a result, the temperature profile after the filling sequence is optimized (Fig. 10).

The riser design was also optimized in the exhaust flange area. The riser size was found to be sufficient. Enlarging the riser neck, however, enabled the riser to better feed the casting. Compare Fig. 11 and 12, which show the old and new design. Red represents liquid metal, while blue shows areas where the metal is completely solidified.

Riser Design

The following optimization was performed for an aluminum piston cast in permanent mold for a 7.51 truck engine. Figure 13 shows the enmeshment of the piston with risers and gating system. Figures 14, 15 and 16 depict the flow sequence of the piston. Cooling channels in the mold are used to optimize the productivity.

Figures 17 and 18 show the casting after 38 and 82 seconds. The simulation indicates that the thermal center moves into the piston instead of staying in the risers. This results in defects typically referred to as hot spots.

These defects can be detected further with higher accuracy using a newly developed feeding criteria. During solidification, the actual flow of liquid feed metal is calculated. If a critical section in the casting--such as the feeder neck--reaches a certain stage of solidification higher than its critical feeding value, the dendritic network of the primary phase prevents proper feeding flow.

According to this metallurgical phenomena, the feeding simulation detects areas where feeding flow hasn't been possible and automatically maps these regions for easy detection. The feeding value is highly alloy-dependent and, for aluminum alloys, ranges between 25-30%.

The colors in Fig. 19 show areas of a calculated density between 80% (blue) and 100% (white). Gray color presents areas where no metal is left due to shrinkage compensation. Typical gray areas should be the tops of the risers. The piston is cut in the same cross section for direct comparison. The feeding criteria detects even the smallest defects.

The optimized riser design using a single riser is shown in Fig. 20. The solidification simulation predicts the hot spots to be in the riser (Fig. 21 and 22). The feeding criteria resulted in excellent feeding conditions (Fig. 23). The newly designed piston is again depicted for direct comparison.

The success of applying numerical simulation also is presented on a piston for a 1.91 engine. Figure 24 shows the feeding criteria of the old design with defects occurring in the piston area. The modified design led to the production of defect-free castings (Fig. 25).

Head Start Over Competitors

Simulation is not an expert system. It assists foundry workers who have gating and risering expertise or tooling and process background to solve real life casting problems. Putting both capabilities together, however, generates a highly productive package for developing a robust process. All this can be done up front before any tooling or pattern has been built.

This results in cost and time savings, giving foundries using simulation a definite edge over the competition.