Recent Studies on Probabilistic Modelling of Microstructural Evolution in Zr Based Bulk Metallic Glass Matrix Composites during Solidification in Additive Manufacturing

The present study aims to model nucleation of primary phase ductile crystalline particles in glassy
matrix. Bulk metallic glass and their composites (BMGMCs) are a new class of materials which
possess superior mechanical properties as compared to existing conventional materials. Owing to
this, they are potential candidates for tomorrow’s structural applications. However, they suffer from
poor ductility and little or no toughness which render them brittle and they manifest catastrophic failure
under applied force. Their behavior is dubious, unpredictable and requires extensive experimentation
to arrive at conclusive results. A cellular automata method is described for describing nucleation and
growth of primary ductile phase particles in glassy matrix in bulk metallic glass matrix composites. A
probabilistic cellular automaton (CA) model is developed and described in present study by author
which is used in conjunction with earlier developed deterministic model to predict microstructural
evolution in Zr based BMGMCs in additive manufacturing liquid melt pool. It is elaborately described
with an aim to arrive at quantitative relations which describe process and steps of operations. Results
indicate that effect of incorporating all mass transfer and diffusion coefficients under transient
conditions and precise determination of probability number play a vital role in refining the model and
bringing it closer to a level that it could be compared to actual values. It is shown that proposed
tailoring can account for microstructural evolution in metallic glasses. It is found that proper use of
transport equations and calculations of random probability number play pivotal role in describing
microscale solute diffusion and solid fraction evolution in solidifying alloy. Use of moderate size
simulation grid (cartesian) to counter mesh anisotropy along with selection of decentered square
algorithm also helps in model refinement and optimization.