A method for calibrating a digital image correlation system for full-field strain measurements during large deformations
journal contributionposted on 21.10.2019, 10:43 authored by Robert Blenkinsopp, Jonathan RobertsJonathan Roberts, Andy HarlandAndy Harland, Paul SherrattPaul Sherratt, Paul Smith, Tim Lucas
Numerous variables can introduce errors into the measurement chain of a digital image correlation (DIC) system. These can be grouped into two categories: measurement quality and the correlation principle. Although previous studies have attempted to investigate each error source in isolation, there are still no comprehensive, standardized procedures for calibrating DIC systems for full-field strain measurement. The aim of this study, therefore, was to develop an applied experimental method that would enable a DIC practitioner to perform a traceable full-field measurement calibration to evaluate the accuracy of a particular system setup in a real-world environment related to their specific application. A sequence of Speckle Pattern Boards (SPB) that included artificial deformations of the speckle pattern were created, allowing for the calibration of in-plane deformations. Multiple deformation stages (from 10% to 50%) were created and measured using the GOM ARAMIS system; the results were analysed and statistical techniques were used to determine the accuracy. The measured strain was found to be slightly over-estimated (nominally by 0.02%), with a typical measurement error range of 0.34% strain at a 95% confidence interval. Location within the measurement volume did not have a significant effect on error distributions. It was concluded that the methods developed could be used to calibrate a DIC system in-situ for full-field measurements of large deformations. The approach could also be used to benchmark different DIC systems against each other or allow operators to better understand the influence of particular measurement variables on the measurement accuracy.
- Mechanical, Electrical and Manufacturing Engineering