Development of Computer-Vision-Based Technique for Prestress Measurement in Beams

We report the first work on demonstrating the measurement of prestresses in beams using methods of computer vision (CV). Toward this, we developed instruments, performed experiments, and implemented image processing algorithms. First, custom-made equipment were designed and developed to achieve fixed-free, fixed-fixed, fixed-pinned, and pinned-pinned boundary conditions and prestress. Beams of steel, aluminum, copper, and glass fiber epoxy composite (GFEC) were constrained to these four types of boundary conditions. The vibrations of the beams in static tensile axial loads were recorded using a smartphone camera via a lock-in mechanism. A total of 320 experiments involving all boundary conditions and ten axial loads were performed. Resonant frequencies were identified from the videos using image correlation and point-tracking algorithms. Young’s moduli were estimated from experiments with no axial load. Prestresses were determined from the resonant frequencies and were found to have excellent correlations with experimental values. This method can be easily extended to identify the prestress in beams with lateral loads and to 3-D objects.