Battery thermal management system design, modelling and optimisation
presentationposted on 22.10.2018, 09:41 authored by Manlio valerio Morganti
3 Minute Thesis presented at the Cranfield Doctoral Network Annual Event 2018.
The growth of the electric vehicle market has led to an increase in the diffusion of batteries. Batteries are energy storage devices subject to uneven heat generation, having unisotropic thermal properties and being forced to comply with tight temperature range requirements (-10°C to 40°C), to avoid aging and permanent performance drop at low temperatures and safety concerns at high temperatures. In addition to that, the temperature gradient across a battery pack, across a module and across a single cell, needs to be monitored in order to prevent a battery life shortening. In such scenario, literature lacks of medium-fidelity coupled-physics models that can simulate both the electrical and the thermal behaviour of the battery cells at a low computational cost. In addition, no systematic methodology for battery thermal management design is currently available. The aim of this work is to provide a battery thermal management design procedure based on a coupled-physics model of a single cell that can be scaled up to a module and a pack model. The proposed simulation library allows the quick implementation of thermal models of different cell chemistries and geometries. A single cell model is then used to populate a whole module model. More modules populate a pack model. The use of a multi-physics simulation tool keeps computational costs much lower than a full-scale finite-element-method-based software. In order to ensure the reliability of results, models are validated against a single-module test rig. A liquid indirect thermal management is taken as case study. Different scenarios and options are examined and optimal solutions are proposed in each case. Objectives are minimisation of weight, cost, volume and energy consumption and operating temperature ranges and temperature gradients are taken as constraints. This methodology is lean, flexible and can be adapted to different vehicle classes reducing development time and costs.