Simscape Framework for Digital Lithium-Ion Battery Twins: Electrode & Electrolyte Domains

Abstract

To speed up the development of novel batteries, time-consuming experimental work can be supported or partially replaced with simulations of physical batteries. Since the demand for laboratory research is growing rapidly, the same can be applied to mathematical models.So far, there is a huge variety of tools used to mimic the processes occurring in different scales: from a single active material particle up to a whole battery cell.1-4 Existing battery modeling software suites such as COMSOL, Siemens Simcenter, Python, or Matlab, require the user to hold a high degree of competence in informatics, computer science, and modeling.5-7 This may limit the use of simple battery models for laboratory workers in development and research.With all the above in mind, in our work, we want to present the concept of an easy-in-use battery twin model. In our project, new domains (mass/electrode and electrolyte) have been developed in Simscape – a library provided by Simulink which is part of the Matlab suite (Mathworks TM). With an intuitive, drag-and-drop interface and clear error and warning communication, Simscape is a promising tool that can be applied to simulate battery performance and other electrochemical systems. Our first implementation follows the nomenclature from Xia et al.8 The very first digital twin proposed here is represented by Newman’s P2D model where the battery as well as the particles in each electrode are discretized (i.e., divided into layers or shells, respectively). A set of non-linear partial derivative equations8 is solved with the application of the Butler-Volmer equation used to define the electrochemical reaction rate occurring on the surface of the particle in each layer. A schematic representation of such a battery setup with 5 layers of the separator, cathode, anode electrodes, and 10 shells describing the active material particles is depicted in Figure 1. The ionic concentration gradients in the electrolyte are shown for an exemplary charge simulation with a C-rate of 1/h.The current model will be extended by modeling the thermal properties occurring inside the battery during (dis)charging after in-depth numerical validation has confirmed its validity. In future work, parameters will be determined in the laboratory, and experiments will be conducted to compare with the results obtained in the digital twin.