Critical raw material-free multi-principal alloy design for a net-zero future

Area of Science:

• Materials Science and Engineering
• Computational Materials Design
• Alloy Development

Background:

• Refractory High-Entropy Alloys (RHEAs) offer superior high-temperature performance for engine components compared to traditional superalloys.
• The development of RHEAs is often reliant on critical raw materials (CRMs), posing supply chain risks and hindering Net-zero goals due to emissions from recycling.
• Existing RHEA research heavily depends on materials like Niobium (Nb) and Tantalum (Ta).

Purpose of the Study:

• To develop an inverse prediction approach for novel multicomponent alloy compositions that eliminate the need for CRMs.
• To achieve hardness levels comparable to CRM-containing multi-principal element alloys (MPEAs) without using CRMs.
• To accelerate the discovery of reduced-CRM MPEAs (R-CRM-MPEAs) to mitigate supply chain vulnerabilities and environmental concerns.

Main Methods:

• A machine learning (ML) model was trained on a computational database of 3,608 unary and binary materials using Thermo-Calc 2024a.
• The Extra Trees Regressor (ETR) model demonstrated superior performance and was integrated with metaheuristic optimization techniques, specifically Cuckoo Search Optimization (CSO).
• The CSO method identified novel MPEA compositions with reduced CRM content, achieving predictions within a ±20% error margin compared to Thermo-Calc.

Main Results:

• A novel CRM-free alloy composition, Ti_0.01111NiFe_0.4Cu_0.4, was computationally designed, exhibiting a Vickers hardness of 488 HV, comparable to CRM-containing alloys like CoCrFeNb_0.309Ni (480 HV).
• The study focused on facilitating the development of R-CRM-MPEAs rather than solely high-hardness alloys.
• Experimental synthesis and validation of an FCC-phase alloy (Al_6.25Cu_18.75Fe₂₅Co₂₅Ni₂₅) showed excellent agreement between Thermo-Calc, ML predictions, and experimental Vickers hardness values.

Conclusions:

• This pioneering work presents a robust framework for accelerating the discovery of novel R-CRM-MPEAs.
• The developed ML-driven approach effectively addresses challenges associated with CRM supply chain vulnerabilities, import dependence, and associated environmental impacts.
• The study validates the efficacy of computational inverse design in creating high-performance alloys with reduced reliance on critical raw materials.