Session: Rising Stars of Mechanical Engineering Celebration & Showcase

Paper Number: 147875

147875 - New Experiments for Rapid Calibration of Elastoplastic Mechanical Properties 

Rapid progress in material design has been enabled by new technological capabilities such as artificial intelligence and additive manufacturing. Material characterization has become the bottleneck in the process to design and certify new materials and manufacturing methods, but this bottleneck can be alleviated by harnessing full field measurement techniques and the computational power available to engineers today. A hybrid experimental/simulation approach is expected to reduce the lengthy experimental campaigns that are currently required by using a lesser number of experiments that can reveal a larger range of material behavior in each experiment. The use of full field measurements in such experiments can then increase throughput in material characterization if the rich datasets can be reliably translated to relevant material properties. This work is focused on developing such an advanced material characterization approach to accelerate the process of calibrating constitutive models for metals with complex elastoplastic behavior. The most mature experimental and computational techniques, i.e. digital image correlation (DIC) and finite element modeling (FEM), are utilized.

The design of experimental specimens to maximize the value of measurable information is examined, and a new specimen performance indicator is developed for this purpose. The indicator considers the full distribution of strain states achieved on the surface of a specimen and favors specimens that create diverse states of strain with comparable magnitudes. The indicator also incorporates the spatial distribution of the strain field and discourages the formation of volumes with little deformation. Additionally, the indicator can be systematically tailored to specific cases, such as plane stress or constitutive models with certain symmetries assumed. Existing plane stress specimen designs in the literature are evaluated with the new performance indicator and a new specimen that incorporates 3D stress states is optimized using the indicator as the objective function. The new specimen design substantially outperforms existing designs according to the new indicator. Demonstration of the new experimental specimen fabricated from AA6061 is shown, where DIC is performed with four stereo pairs collecting images from three surfaces of the specimen. 

Attention is also given to the process of extracting constitutive properties from the new experimental specimen. Simulated data with known constitutive properties covering a broad set of hardening curves and yield surface shapes is used for this purpose. The use of this data allows for extensive pairwise comparisons directly between the constitutive models and, separately, between the resulting deformation fields. The use of both sets of comparisons is used to guide the construction of an objective function that is also a surrogate for error in the constitutive model that can be used to solve an inverse problem for material behavior.

Presenting Author: Andrew Gross University of South Carolina

Presenting Author Biography: Dr. Gross is an Assistant Professor in the Department of Mechanical Engineering at the University of South Carolina. His areas of expertise include the development of microscale testing specimens, in situ mechanical testing, ductile fracture, and the thermomechanical behavior of architected materials.

Authors:

Andrew Gross University of South Carolina