Quantifying alignment deviations for uniaxial material mechanical testing via automated machine learning

Abstract

The loading coaxiality of uniaxial material mechanical testing (UMMT) introduces an additional bending stress, significantly affecting the admeasurement of material properties. Owing to the lack of a calculation method, only qualitative analysis of misalignment can be performed using the current technologies. To quantify the alignment deviations of the UMMT, the relationship between the strain measurement points on the specimen and alignment deviations were mapped using automated machine learning. A finite element model was used to obtain the dataset. The numerical results indicated that the established deep learning model can be utilized to quantify the alignment deviations for UMMT.