Accelerated Fatigue Strength Prediction via Additive Manufactured Functionally Graded Materials and High-Throughput Plasticity Quantification

TL;DR

This paper introduces a rapid, high-throughput method for predicting fatigue strength of additive manufactured functionally graded materials using digital image correlation and computer vision, significantly reducing testing time.

Contribution

The study presents a novel approach combining digital image correlation and machine learning to efficiently characterize fatigue properties across multiple material compositions from a single test.

Findings

Validated on 316L additive manufactured dataset

Successfully characterized multiple compositions in a graded specimen

Achieved orders of magnitude faster testing than traditional methods

Abstract

Recent improvements in additive manufacturing and high-throughput material synthesis have enabled the discovery of novel metallic materials for extreme environments. However, high-fidelity testing of advanced mechanical properties such as fatigue strength, has often been the most time-consuming and resource-intensive step of material discovery, thereby slowing down the adoption of novel materials. This work presents a new method for rapid characterization of the fatigue properties of many compositions while only testing a single specimen. The approach utilizes high-resolution digital image correlation along with a computer vision model to extract the relationship between localized plastic deformation events and associated mechanical properties. The approach is initially validated on an additive manufactured 316L dataset, then applied to a functionally graded additive manufactured specimen with a composition gradient across the gauge length. This allows for the characterization of multiple compositions, orders of magnitude faster than traditional methods.