Development and Validation of a New Video Extensometer System for Compressive Material Testing

Abstract

Material testing is critical to the development of new alloys, polymers, composites, and other advanced materials. In the last 50 years, the development of video extensometry has made material testing easier than ever before, by allowing researchers to measure sample deformation through image analysis. This thesis details the design, operation, and initial experiments conducted using a new video extensometer system in development under the project name SPECS (Super Portable Extensometer Camera System).

The SPECS system aims to make compressive testing of samples in the presence of barreling more accurate. Rather than reduce friction to assume a non-barreled sample, SPECS simply accounts for barreling by measuring diameter continuously across a sample's height. SPECS does this through a combination of background subtraction and edge detection algorithms. This thesis details the design of the SPECS system, along with the calibration experiments, cold upsetting tests, and finite element modeling used to verify the system.

Operation of the SPECS system is first verified through a series of calibration experiments. With an error on true stress of 4.91\%, SPECS error is about 2\% greater than stress error seen in cold upsetting tests in the literature. Much of this error is attributed to the quality of the calibration sheet, and improvements to the calibration sheet have the potential for significant reduction to this error.

The verified system is tested against aluminum 2024-T351 in a series of cold upsetting tests. The results of the aluminum alloy tests match well with similar tests in the literature. Data from the aluminum experiments is used to construct and validate a finite element model simulating a cold upsetting test. Diameter data from the finite element simulation matches closely with experimental data.

The long-term goal of SPECS is to create a coupled finite element model/video extensometer tool for fitting advanced material model parameters to experimental data. The experiments in this thesis show that SPECS is capable of collecting accurate stress/strain data which can be used to fit material models for modeling. Future work, therefore, should focus on improving the accuracy of SPECS measurements, and coupling SPECS to the finite element model to make the data fitting procedure iterative and automated.