Multi Objective Adaptive PSO for Uniform and Maximum Heat Transfer over Electronic Chips in a Channel
In this paper, a numerical analysis is carried out to govern the steady laminar fluid flow and forced convection heat transfer over isothermally heated blocks affixed in lower wall of a 2-D horizontal channel. Air is considered here as a cooling fluid. A three-sided cross sectional bar is utilized as controlling element of heat flux. Parameters like different bar positions along x and y axis and bar base length are used to run the analysis with Reynolds number (25≤Re≤400). The variation of vortex creator position both horizontally as well as vertically influences much onto the thermal boundary layer over the blocks. The position of vortex creator closer to the upper wall as well as over the second block results comparatively better heat transfer over the blocks within constrained parameters. Considering local Nusselt number, it’s observed that the front and upper face of the blocks plays a significant role in heat transfer compared to the back faces. The Response Surface Methodology (RSM) prediction of Nusselt number corresponds well with the numerical outputs of continuity, momentum and energy equation solver. The solver used here is ANSYS 16.2. In accordance to maximize heat transfer and to maintain uniformity of heat flux, an Adaptive Particle Swarm Optimization (APSO) algorithm is deployed using MATLAB. A concept of the enhanced Ïµ dominance method is introduced in multi-objective PSO algorithm to overcome the premature convergence. The Pareto optimal solutions are obtained through multi-objective PSO using the proposed algorithm.
Uniform heat flux, vortex creator, Heated blocks, Adaptive PSO, RSM.