Fast models for battery cooling system design

Abstract

Fast running battery pack electrical and thermal models were created from electrical test data and battery geometry. These fast running models were used in simulating lifetime battery temperature and cooling performance. These models include the battery, the pack structure, cold plates, and the electrical behavior. The models are capable of being integrated into either 3D or 1D models. These fast battery models allow design changes to be thoroughly evaluated at each step in a vehicle program.

Thorough and detailed evaluation is necessary because packs are very sensitive to temperature. Extended periods of elevated temperatures or short periods of high temperatures can significantly decrease battery life. Additionally, the temperature of the battery affects vehicle performance and efficiency. To ensure batteries both perform and last, a battery thermal protection system is usually required. A properly designed thermal protection system is efficient and capable of cooling the battery under a broad range of driving and storage conditions. To evaluate how well a cooling system protects a battery, the lifetime of the vehicle may be simulated. However, simulating these long time frames are a challenge due to the limited time available. Fast running simulations solve this problem.

These fast thermal models are generated from 3D models. These 3D models are used to generate the system models. The equivalent system model includes batteries, pack structure, heat sources in the pack, and cold plates. Several key assumptions allow these system models to be generated quickly and preserve accuracy. The 3D modeling step ensures these assumptions are satisfied. This fast thermal model is used to speed up the 3D models by replacing the parts of the 3D geometry. This fast running thermal system model also runs on common system tools. The fast electrical model is used to predict electrical performance and heat generation in the batteries. This electrical model is created from electrical test data. While there are many established ways to structure tests and extract electrical parameters, a unique approach was taken. This approach allows both lab data and vehicle data to be used. This fast model typically runs in dramatically less time than required by a 3D model while achieving with nearly equivalent accuracy.