Simulation of mushy state solidification in stir casting

Stir casting is a promising technique used for the manufacture of Al-SiC metal matrix composites. The clustering of reinforcement particles is a serious concern in this production method. In this work, mushy-state solidification characteristics in stir casting are numerically simulated using computational fluid dynamics techniques to study the clustering of reinforcement particles.

Effects of process parameters on the distribution of particles are examined by varying stirrer speed, volume fraction of reinforcement, number of blades on stirrer and diameter ratio (ratio of crucible diameter to stirrer diameter). Further, investigation of characteristics of cooling curves during solidification process is carried out. Volume of fluid method in conjunction with a solidification model is used to simulate the multi-phase fluid flow during the mushy-state solidification. Solidification patterns thus obtained clearly indicate a strong influence of process parameters on the distribution of reinforcement particles and solidification time.

From the simulation study, it is observed that increase in stirrer speed from 50 to 150 rad/s promotes faster solidification rate. But, beyond 100 rad/s, stirrer speed limit, clustering of reinforcement particles is observed. The clustering of reinforcement particles is seen when volume fraction of reinforcement is increased beyond 10 per cent. When number of blades on stirrer are increased from three to five, an increase in solidification rate is observed, and an uneven distribution of reinforcement particles are observed for five-blade geometry. It is also seen from the simulation study that a four-blade stirrer gives a better distribution of reinforcement in the molten metal. Decrease in diameter ratio from 2.5 to 1.5 promotes faster solidification rate.

There is 90 per cent closeness in results for simulation study and the published experimental results.