Dynamic Design Optimization for Thermal Management: A Case Study on Shell and Tube Heat Exchangers

Abstract

We present a new methodology for designing a heat exchanger that explicitly considers both static and transient performance characteristics. The proposed approach leverages 1) a highly detailed, albeit static model that captures the complex nonlinear relationship between heat exchanger geometry and heat transfer coefficients, and 2) a reduced-order dynamic model of the heat exchanger that approximates the geometry detailed in the static model. In order to optimize the component design for both static and transient performance metrics, pole locations of the corresponding linearized model are penalized in the cost function of the proposed optimization algorithm in order to move dominant poles further into the left half complex plane. Through a simulated case study for a shell and tube heat exchanger, we demonstrate how the proposed algorithm exploits the trade off between static design metrics, including mass and footprint, and the rate at which heat is removed from the primary fluid.