Stereo Image Based Motion Measurements in Fluids: Experimental Validation and Application in Friction Extrusion

Optical refraction at interfaces is a common issue when viewing a submerged specimen through a transparent window. The distortions that are introduced during such imaging must be minimized when employing stereo-vision systems to make accurate quantitative (a) position, displacement, velocity and strain measurements on structures immersed in fluids via marker tracking and (b) velocity measurements along particle paths in a fluid via particle tracking. In this study, for the first time an optical model with refraction at multiple media interfaces that was developed previously for stereo-vision based measurements in fluids is experimentally validated via (a) motion and strain measurements on an immersed, rotating surface and (b) particle tracking in a fluid. Consistent with the model, the calibration and reconstruction processes are developed and demonstrated experimentally to be effective in removing distortions. To demonstrate the utility of the method for accurate measurements on submerged objects, the authors measured positions and strains on a rotating, rigid surface through stereo-imaging of specific features. Results confirmed that the vision-based method is accurate and repeatable for measuring 3D positions and strains on submerged bodies. The methodology is then applied to the study of extrusion processes. Using a transparent small scale lab model extruder with a highly viscous fluid designed for use with stereo vision measurement systems, the 3D motions of neutrally buoyant spherical particles are measured during the extrusion process and compared to CFD simulation predictions. Results confirm that the dual calibration-stereo imaging approach is more accurate and effective for particle tracking than typical approaches used in such studies.