Development and Implementation of a Low Cost Image Correlation System to Obtain Full-Field In-Plane Displacement and Strain Data

John Parker

M. S. Thesis Defense

Mechanical Engineering

Fiber reinforced Plastics (FRPs) are becoming increasingly popular for use as primary structures in the energy, aerospace, and other industries. With this increased use comes an increased need for an inexpensive way to test and qualify these materials and structures. Since FRPs are anisotropic materials, characterizing them generally requires multiple single-axis tests on several different test coupon geometries and fiber orientations. Multi-axial testing allows for the same coupon geometry and fiber orientation to be used for all tests, therefore eliminating the need for many different coupon geometries and fiber orientations and tests. The MSU In Plane Loader (IPL) is a multi-axial testing machine that has been in use for some time but has always been hindered by its inability to measure full field displacements on a sample during testing.

Typical methods for measuring displacements during uni-axial materials testing will not work on the IPL. The previous method used measured displacement at the grips but has been shown to be limited and inaccurate. This project developed a Digital Image Correlation (DIC) based system that can measure full-field displacements and strain directly from images taken during the testing process. Images are taken during testing by a Canon digital single-lens-reflex camera controlled by the IPL test software. These images are then processed and imported into Matlab software where a series of programs output the displacement of any number of user-selected points on the surface of the sample during the test. Strains are then calculated by numerical differentiation of the displacement values.

The system has been successfully used to measure surface and boundary displacements on composite and aluminum samples, as well as on glued aluminum joints. The software system has been shown to perform as an extensometer with a strain error of less then 0.0002 m/m or 0.5% placing it within ASTM-B2classification. It also has been shown to measure displacement to within 0.02 pixel accuracy; which for tests on the IPL translates to 0.007mm.

Full-field strain data from these tests can also be used in the development of a Dissipated Energy Density (DED) function, a subject which has been of considerable interest at MSU. Full field strain data eliminates the need to model strains on the samples using FEA as has been done in the past and increases the test data that can be used for the development of the DED function, ostensibly increasing its accuracy when used to predict the post-damage response of a composite.

Wednesday, February 25th, 2009

10:00 a.m.

Location TBA