A technique for fluid-structure interaction diagnostics using a single plenoptic camera.

This work presents a new methodology for experimentally studying fluid-structure interaction (FSI) using a plenoptic camera. Current approaches to 3D measurements in FSI problems use 4-camera tomographic particle image velocimetry to capture flow motion and 2-camera digital image correlation to capture surface motion, but face limitations due to the optical access required for a 6-camera arrangement. Thus, we propose using a plenoptic camera that encodes the 4D light field information, to capture both 3D flow field and surface motion using a single camera. We utilize proper orthogonal decomposition to separate the particle and surface information in a plenoptic image. Furthermore, we introduce a novel correlation-based plenoptic depth estimation technique for robust surface reconstruction. 3D vector fields are generated using algebraic reconstruction followed by cross-correlation. Validation of the methodology is performed with synthetic analysis of a moving flat plate immersed in a uniformly moving particle field. Results show an uncertainty of less than 1% of full depth for the reconstructed surface and an uncertainty of less than 3.5% of maximum velocity for the flow field. Finally, we demonstrate this method experimentally by capturing the flapping motion of a flexible flag in a water tunnel.