Abstract
Errors in strain measurements in stereo-digital image correlation (stereo-DIC) caused by camera self-heating have been experimentally observed in previous research, and have been shown to widely vary from one system configuration to another. Such “thermal errors” are sometimes so large that they strongly compromise the accuracy of the measurements. Despite correcting such errors is crucial when aiming at high-accuracy measurements, the mechanism of the thermal error generation and how it relates to the camera parameters in stereo-DIC are still not clear. In this paper, we first explain in detail how self-heating can introduce large artifacts in the strains measured by stereo-DIC. Using a simplified stereovision model, we provide the main equations that describe the theoretical errors in 3D coordinate reconstruction and 3D deformation measurement. Through several sets of simulations, the effect of camera self-heating on the 3D coordinate, displacement and strain measurements, and the effect of camera parameters on the thermal errors in stereo-DIC were explicitly presented based on the derived theoretical formulas. Finally, several real self-heating tests using a smartphone-based single-camera stereo-DIC system confirm the correctness of theoretical analyses and simulation results.
Highlights
Errors in strain measurements in stereo-digital image correlation caused by camera selfheating have been experimentally observed in previous research, and have been shown to widely vary from one system configuration to another
Our recent research further indicated that the thermal errors would significantly increase to tens of thousands microstrains for a single-camera stereo-DIC system with small baseline distance and short focal length[26]
The mechanism of the thermal error generation in stereo-DIC is first explained, and the theoretical errors www.nature.com/scientificreports in 3D coordinate reconstruction and displacement measurement are derived based on a simplified stereovision model
Summary
Note that, in these simulations, the translations of the camera lenses and sensors are constant and taken from the data in Fig. 5(a) at the time of 60th min. In these simulations, the translations of the camera lenses and sensors are constant and taken from the data in Fig. 5(a) at the time of 60th min From these figures, it can be concluded that the measured W displacement errors are more sensitive to camera parameters than the U and V displacement errors, especially for the system configuration with small baseline distance and included angle. As no external loading was applied to the plate, the actual displacements and strains on the plate should be zero and the measured displacements and strains can only be attributed to the errors due to camera self-heating
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call