Adjoint-based Flow Control of Liquid-Gas Flows Modeled with a Sharp Interface

Spray formation plays a critical role in combustion systems where implementing control strategies can be highly beneficial. With advancements in numerical modeling of liquid-gas flows, we can now accurately predict complex two-phase flows. As a result, we are able to explore computational optimization strategies as an avenue for control. Adjoint methods are an efficient technique to calculate gradients because their cost is independent of the number of control variables. The gradients can then be input into a gradient optimization algorithm to solve the control problem. In this work, we present a computational adjoint method for incompressible, liquid-gas flows using a sharp interface. We verify our method by comparing adjoint calculated gradients against analytical gradients for canonical test cases. We then highlight the method's utility by controlling a temporally evolving liquid-gas mixing layer, a case that has relevance to spray formation.