Lightweight, low cost compositionally complex multiphase alloys with optimized strength, ductility and corrosion resistance: Discovery, design and mechanistic understandings

Abstract

A strategy for designing compositionally complex alloys (CCAs) achieving multiple objectives is articulated. In this specific case, the objectives are low density and cost, along with good strength, ductility, and aqueous corrosion resistance. The present strategy has produced a two-phase material, having a ductile FCC matrix and a L21 intermetallic reinforcing phase. A machine learning algorithm has been trained using attributes of binary phase diagrams as inputs and the existing experimentally informed CCA database as output values, which efficiently predicts the phases which will result from combinations of selected elements. Corrosion resistance is achieved through tailoring the composition such that solute partitioning promotes passivity of both phases. Surprising results of composition optimization, such as preferred levels of Mn and Mo in an Al-Cr-Fe-Mo-Mn-Ni-Ti alloy are presented. While raw material cost, density, corrosion resistance, and stiffness are strong functions of alloy chemistry, the strength and ductility are highly sensitive to microstructure as well. A genetic algorithm is used to optimize a crystal plasticity-based prediction of toughness, defined as the product of ultimate tensile strength and uniform elongation, by varying microstructural attributes. Surprising dependencies of this toughness on the microstructure inputs are highlighted. Experimental results demonstrate the efficacy of the approach.

Keywords

• Machine learning
• Genetic algorithm
• Corrosion
• Heusler phase
• Ni2TiAl
• Crystal plasticity