Abstract
Ti is usually considered as an important ingredient of grain refining in aluminium alloys acting through nucleation on Al3Ti (directly on as an absorption layer on TiB2) and as a growth restricting element. However, in a series of experiments we have demonstrated that in some cases adding Ti to aluminium alloys with AlTiB can be detrimental for grain refining. Two hypotheses have been tested and supported by the experimental evidence. First hypothesis is that free Ti delays dissolution and activation of an AlTiB alloy in the melt. It is shown that the grain-coarsening effect is indeed present with additional 0.1-0.15% Ti, but decreases with time after introduction of AlTiB to the melt. Second hypothesis considers the formation of primary Al3Ti followed by nucleation of aluminium through a peritectic reaction, with the related latent heat decreasing the undercooling required for nucleation of aluminium on TiB2. This is supported by the phase diagram calculation, thermal analysis and the fact that grain coarsening worsens with the larger Ti concentration. Both mechanisms could be working together. On prolonged exposure of AlTiB to the melt, more TiB2 particles become available due to deagglomeration or more complete dissolution of the master alloy, and the grain refining effect is restored as it is seen in the experiments.