Optimization of Vortex Promoter Design Using a Dynamic Data Driven Approach

Abstract

Thermal management of electronic equipment is one of the major technical problems in the development of electronic systems that would meet increasing future demands for speed and reliability. It is necessary to design cooling systems for removing the heat dissipated by the electronic components efficiently and with minimal cost. Vortex promoters have important implications in cooling systems for electronic devices, since these are used to enhance heat transfer from the heating elements. In this paper, an application of Dynamic Data Driven Optimization Methodology (DDDOM), which employs concurrent use of simulation and experiment, is presented for the design of the vortex promoter to maximize the heat removal rate from multiple protruding heat sources located in a channel, while keeping the pressure drop within reasonable limits. Concurrent use of computer simulation and experiment in real time is shown to be an effective tool for efficient engineering design and optimization. Numerical simulation can effectively be used for low flow rates and low heat inputs. However, with transition to oscillatory and turbulent flow at large values of these quantities, the problem becomes more involved and computational cost increases dramatically. Under these circumstances, experimental systems are used to determine the component temperatures for varying heat input and flow conditions. The design variables are taken as the Reynolds number and the shape and size of the vortex promoter. The problem is a multi-objective design optimization problem, where the objectives are maximizing the total heat transfer rate, as given by the Nusselt number, Nu, and minimizing the pressure drop, ΔP. This multi–objective problem is converted to a single-objective problem by combining the two objective functions of the form Nutota/ΔPb, where a and b are constants.