Implementation of the direct evaluation of strains using a phase analysis code for random patterns

TL;DR

This paper introduces a novel phase analysis method for directly measuring strains from surfaces with random patterns, demonstrating promising resolution and application to composite materials.

Contribution

It adapts grid method concepts to random patterns for direct strain evaluation, validated through numerical and experimental results.

Findings

Resolution is proportional to strain level, at 9% of the nominal value.

Random noise must be carefully controlled for accurate measurements.

Successful application to carbon fiber composite reveals clear over-strain detection.

Abstract

A new approach for decoding directly strains from surfaces encoded with random patterns has been developed and validated. It is based on phase analysis of small region of interest. Here we adapt to random patterns new concepts proposed by Badulescu (2009) on the grid method. First metrological results are encouraging: resolution is proportional to strain level, being 9% of the nominal value, for a spatial resolution of 9 pixels (ZOI 64 \times 64 pixels2). Random noise has to be carefully controlled. A numerical example shows the relevance of the approach. Then, first application on a carbon fiber reinforced composite is developed. Fabric intertwining is studied using a tensile test. Over-strains are clearly visible, and results connect well with the previous studies