Mapping hierarchical and heterogeneous micromechanics of a transformative high entropy alloy by nanoindentation and machine learning augmented clustering

Abstract

Conventional macromechanical tests provide limited insights into complex hierarchical deformation behavior of a transformative high entropy alloy (HEA). In this work, a high throughput microstructure-micromechanical correlative study is presented by combining high-resolution nanoindentation, site-specific microscopy, and Gaussian mixture model (GMM) clustering. The investigated HEA has a heterogenous microstructure consisting of austenite and martensite phases. Comparison of elastoplastic and microstructural maps illustrate dependency of phase, crystal orientation, and interfacial constraints on the micromechanical response. The disproportionately high hardness found in martensite-rich area is attributed to its higher lattice stability to shear, creation of coherent twin boundaries, and copious dislocation activities in the twin interfaces formed in martensite phase during nanoindentation. The hierarchy in twinning behavior depends on the relative direction of loading with the c-axis of h.c.p. martensitic phase. Deformation in f.c.c. austenitic grains is slip-dominated and demonstrates orientation dependency during incipient plasticity and phase transformation. GMM based classification of hardness to modulus ratio intuitively correlates work-hardening with phase distribution due to the distinctive deformation micromechanical responses of austenite and martensite phases.

Keywords

• High entropy alloys
• Nanoindentation
• Machine learning
• Incipient plasticity
• Twinning
• Transformation-induced plasticity