Comparative study of multiphysics modelling and simulation software for lifetime performance evaluation of battery systems.

Abstract

Battery systems play a vital role in numerous applications, ranging from electric vehicles to renewable energy storage. Accurate modelling and simulation of battery systems are crucial for assessing their lifetime performance. Currently, there are various simulation software on the market to assist industry and academia to implement multiphysics models and simulate battery systems. Every tool has unique capabilities, which can suit different needs of users.

This thesis investigates and compares four widely used multiphysics modelling and simulation software tools— Simcenter Amesim, MATLAB/Simulink, MATLAB/Simulink/Simscape and Dymola—with respect to four main aspects, namely computational performance and accuracy, ease of use, features, and licensing. The scope of the study is to highlight peculiarities of every software, to provide insights to simulation engineers, researchers and students that need to select the right simulation software for their investigations.

The very same multiphysics model of a battery pack, composed by an Equivalent Circuit Model (ECM) with two RC branches and liquid cooling will be described in detail and implemented on the four tools. The system includes electric, thermal, and ageing model. The main output of the simulation is battery lifetime degradation in different operating conditions. Accuracy of the four software will be benchmarked against test data, while computational speed will be estimated based on the lifetime simulation.

The main findings of this comparative study is that Dymola proved to be the tool with the best speed performance, requiring only 0.25 seconds to perform a 24h simulation and 288 seconds to perform a 10 years simulation. In terms of ease of use, Amesim resulted as the most user-friendly, with a simple user interface, smooth workflow, and clear documentation. In terms of features and interoperability, Simulink has the great advantage of being closely interconnected with MATLAB environment. The same is true for Simscape, whose speed and accuracy performance were the poorest, but is more user-friendly since it provides multiphysics component.

The findings of this comparative study will assist researchers, engineers, and industry professionals in selecting the most suitable software tool for lifetime performance evaluation of battery systems, based on their specific case.