Executive Summary : | The present project proposal is on development of a coupled Computational Fluid Dynamics (CFD) and Machine Learning (ML) model, and its application for analysis of energy harvesting from flow-induced vibrations (FIVs) of cylinder and sphere with a detached splitter-plate. The Fluid multi-Structure Interaction (FmSI) study is proposed for the effect of the splitter plate-based enhancement in energy harvesting capabilities of cylinder and sphere of various shapes. Further, the study is proposed for six different problems—three each in two-dimension (2D) and three-dimension (3D) for the cylinder and sphere, respectively. The three 2D FIV problems corresponds to free stream flow across an elastically-mounted (a) D-section, inverted C-section, and triangular section cylinders, (b) various circular-triangular cylinders, and (c) triangular cylinders at various angle of attack; with a detached splitter-plate. Whereas, the three 3D FIV problems corresponds to free stream flow across an elastically-mounted sphere, D-shaped sphere, and inverted C-shaped sphere with a detached splitter-plate. The motivation for the novel CFD-ML model is to reduce the computational time for the unsteady simulations considered here, which should lead to a substantial speed-up; specially for the 3D problems. The series of 2D and 3D studies are proposed for a mass ratio of 10, damping ratio of 0.005, reduced velocity varying from 3 to 20, and non-dimensional gap varying from 0.3 to 2.5; and Reynolds number Re=100 for the 2D studies while Re=300 for the 3D studies. The results are envisaged for analysis of the near-wake as well as far-wake flow patterns from the temporal variation of vorticity contours and velocity vectors. For the vibrational characteristic of the cylinder/sphere, the variation of amplitude and frequency of the transverse-vibration, and phase-difference between the temporal variation of lift-coefficient and transverse-vibration with increasing reduced velocity will be investigated. Cause and effect-based analysis will be presented for strongly coupled multi-physics characteristics of the present FmSI system. The proposed study should lead to an efficient FmSI-based energy harvesting system. |