An efficient technique for surface strain recovery from photogrammetric data using mesh less interpolation

Abstract

In recent years, techniques for monitoring displacements and surface strains in laboratorial tests have been introduced. Some of those methods are based on photogrammetry and image post-processing, allowing both displacement and strain fields to be monitored at a significant number of target points at any stage. This manuscript aims at contributing by proposing a simple and comprehensive approach for monitoring experimental tests using photogrammetry. An innovative strategy is introduced to efficiently recover the strain fields at the surface of a loaded specimen. This technique is based on a meshless approach and does not require any triangulation of data points. In this approach, shape functions are constructed from radial basis function interpolation in local subdomains, for which different numbers of data points may be included. To allow accounting for the presence of generic geometries, which may include concavities or convexities, the selection of interpolating points is performed considering a line-of-sight algorithm, avoiding inappropriate selection of neighbouring points. A number of practical examples are here presented for different concrete specimens tested in the laboratory. The presented results indicate that the proposed method can provide smooth strain fields, which allow an easy identification of microcracking development at early stages of loading.