Numerical optimization of the thermal insulation of an elbow pipe type connection of an electric bus cooling system using computational open source code

Abstract

When considering heat exchange inside an internal fluid flow, factors such as the geometry of the pipe and the material and thickness of insulation layers becomes of great importance. As the pipe changes direction, the flow inside alters its dynamics, causing the heat exchange with the ambient to vary as well as the internal gradient of temperature of the fluid. These type of pipe circuit can be found in the connections between cooling systems and battery packs of electric urban transport vehicles (BEV – Battery Electrict Vehicle), in which the rise of cooling the fluid’s temperature is an undesirable phenomena. Therefore, the present work proposes to study and minimize the temperature difference between the inlet and outlet of an elbow pipe used for this purpose, subject to a constant heat flux from external environment, allowing to maintain a maximum difference threshold. The study were conducted by changing parameters of the pipe such as insulation layer thickness and material, maintaining the external wall temperature and the inlet fluid temperature as constants. To perform such study was used the free, open source computational Fluid Dynamics (CFD) software OpenFOAM to construct and parametrize the computational mesh as well as to obtain the flow profile and temperature distribution inside the fluid in steady-state. Coupled with the CFD software, was used the free, open source software Dakota to manage the simulations realized by the CFD software and vary the design parameters, with the minimization of the objective function as the goal. The principal method of optimization used was the Genetic Algorithm (GA) approach in which each combination of parameters represent a solution and by evaluating and comparing those the algorithm is capable to find a global optima in parameter space without the need to compute gradients, allowing for discrete variables optimization, although in a lengthy process. Utilizing the optimization method described, was possible to obtain an optimal set of parameters that minimizes the temperature difference from the inlet to the outlet in the elbow shaped pipe, therefore the study suggest that the set of design parameters chosen exert great influence in the final temperature obtained at the outlet.