Optimal air distribution design in enclosed spaces using an adjoint method

Abstract

This investigation studied an adjoint method to achieve the optimal design of ventilation in an enclosed environment and validated it with two, two-dimensional cases. A part of the flow field and/or temperature field was defined as the design objective, and the thermo-fluid boundary conditions were determined as the design variables. By using the adjoint method together with a steepest descent method that was implemented in OpenFOAM, this investigation found the optimal air supply parameters. With the determined air supply parameters, this study used Computational Fluid Dynamics to calculate the flow and/or temperature fields, which are in a good agreement with the design objective. However, as the adjoint method could only find the local optima, the calculations with different initial inlet air conditions may lead to multiple optimal solutions. This is common with the gradient-based method.