Ultrahigh-speed imaging for high-impact concrete deformation analysis in pre- and post-cracking stages.

The evaluation of high-speed camera image sequence analysis results in concrete material testing under high-impact loading necessitates the consideration of the effect of the image quality on the measurement accuracy and thus on the potential of the geometric measurements derived from the image sequences. In this contribution, we evaluate the application potential of three ultrahigh-speed cameras with frame rates up to 10 Mfps to analyze the deformation of concrete specimens before and after main crack formation in bending and compression tests. Specifically, we evaluate the Kirana 7M and Shimadzu HPV-X2 cameras with ISIS sensor architecture, and the Phantom TMX 7510 camera with BSI CMOS sensor technology. Three-point bending tests and split-Hopkinson pressure bar tests are performed on 160×40×40m m 3 cuboids and on 80 mm long, 50 mm diameter cylinders. Prior to main crack formation, the displacement vector field represents the specimen deformation, with higher values indicating the position where main cracks will initiate and propagate. Deformations of 80 µm in 54 µs for a bending test and of 154 µm in 36.67 µs for a compression test could be measured. The main cracks are then detected using displacement vector field discontinuity analysis techniques, and their evolution is followed to estimate the crack propagation velocity. Average velocities in bending tests between 603 and 854 m/s have been determined over a time interval up to 40 µs. An investigation of the camera sensor operation of the three optical devices is presented to assess their suitability for deformation analysis. Laboratory tests and real experimental results show that the quality of the propagation vector field, the crack detection, and the crack tip tracking are obviously affected by the image quality, but more significantly by the spatial and temporal resolution due to the small relative step deformations.

Analysis of the Accuracy Potential of a Stereo High-Speed Camera System in 3D Measurements in Highly Dynamic Experiments.

In the context of setting up a stereo high-speed camera system for accurate 3D measurements in highly dynamic experiments, the potential of a "Fastcam SA-X2" stereo system is evaluated by testing different camera configurations and motion scenarios. A thorough accuracy analysis is performed using spatial rigid-body transformations and relative measurement analyses of photogrammetrically reconstructed surfaces of nondeformable objects. The effects of camera calibration, exposure time, object velocity, and object surface pattern quality on the quality of adjusted 3D coordinates are taken into consideration. While the exposure time does not significantly influence the quality of the static measurements, the results of dynamic experiments demonstrate that not only an insufficient frame rate but also an increased noise level resulting from short exposure times affects 3D coordinate accuracy. Using appropriate configurations to capture dynamic events, the errors in dynamic experiments do not differ significantly from the errors obtained in static measurements. A spatial mapping error of less than 1 µm is obtained through the experiments, with proper testing configurations for an object surface area of 5×20 mm. These findings are relevant for users of high-speed stereo imaging techniques to perform geometric 3D measurements, deformation, and crack analyses.