Experimental Demonstration of a Topology Optimization Method to Generate a Capacity-Specific Device With Cavitation Resistance for Control Valve Applications

Topology optimization (TO) is a powerful method of generating structures that have desirable functional performance, to date most commonly used to improve structural behavior or to optimize pressure drops in laminar flow environments. In this study, we use TO to generate free-form pressure-staging geometries for the purposes of cavitation suppression in a turbulent flow device, an industrial flow control application which has not heretofore been addressed. Using variable permeability gradient-based adjoint TO in conjunction with both an out-of-plane resistance modified two-dimensional (2D) flow model and a penalty-term extended k–ε turbulence model, we generated flow channels of predetermined capacity that gradually reduce static pressure to suppress the initiation of cavitation. Three-dimensional (3D) extrusions of the 2D geometries were then printed using a masked stereolithography apparatus and evaluated using a water flow test in conjunction with acoustic cavitation detection. After testing, the results were compared to single and dual orifice baseline devices of equivalent capacity. The results of the experimental validations showed capacity deviations from target of up to 7% with performance improvements, as characterized by the delay of incipient cavitation, of up to 13% over the capacity-equivalent two-stage baseline device. This study demonstrates a new ability to rapidly generate fit-to-purpose devices at significantly reduced engineering effort using topology optimization methods.