Date of Award
Master of Science in Engineering
Mechanical and Aerospace Engineering
Dr. Jinseok Kim
Dr. Daniel Kujawski
Dr. Richard Meyer
Instrumented indentation, finite element method, gradient boosting, material characterization, strain hardening exponents
Masters Thesis-Open Access
In this study a new technique is proposed for the determination of elastic-plastic stress-strain relations for isotropic materials using the force-displacement output from an instrumented indentation test, finite element simulation, and machine learning methods. This non-destructive testing methodology has promising potential for industry implementation. Applications to benefit from this method include the characterization of localized material properties on surface engineered components, and the post-manufacturing assessment of material properties for items such as additively manufactured metallic components and load-bearing welded joints.
Currently, the capability of the instrumented indentation test for determining material properties is limited to the elastic modulus and surface hardness. Numerous experimental and numerical approaches have been suggested for determination of full-range isotropic relations, although past efforts to characterize the stress-strain behavior using a single instrumented indentation test were not successful because there exists no direct relationship between force-displacement output and the elastic-plastic relation. The contents of this study reveal that the development of an accurate algorithm for the determination of full-range stress strain curves from the output of a single sharp instrumented indentation procedure is possible through the leverage of finite-element numerical simulations, gradient boosted regression tree analysis, and hyperparameter optimization methods.
Promer, Darren R., "Prediction of Isotropic Strain Hardening Material Properties Using Gradient Boosted Regression Tree Method and Hyperparameter Optimization" (2020). Master's Theses. 5164.