EV Battery Pack Simulation: Thermal Management, Structural Strength and Safety with CFD and FEA

• Lithium-ion cells work best in a fairly narrow range, often quoted as roughly 15 to 35 °C. Too hot and they age faster and the risk of thermal runaway rises; too cold and power and charging speed drop.
• Differences between cells matter as much as the peak. Cells at different temperatures age at different rates, and the weakest cell limits the whole pack, so designers often aim to keep the spread across the pack within about 5 °C.
• Fast charging is usually the hardest case: high currents produce a lot of heat in a short time, and the cooling system has to keep up.

Packs are cooled by air, by liquid cold plates (the most common choice in modern electric cars), by refrigerant, or by immersing the cells in a dielectric fluid. Whichever you choose, CFD with conjugate heat transfer solves the coolant flow and the heat conduction in the cells and plates together, and shows:

• How the coolant is shared between channels, plates and modules, so no module is starved.
• The pressure drop through the system, which sets the pump size and its energy use.
• The maximum cell temperature and the temperature spread across the pack, while driving and during fast charging.
• Hot spots at busbars and connections, where electrical resistance turns current into heat.

Cell heat can be modelled as a simple heat source or with battery models (equivalent circuit or electrochemical) that link the heat to the current and state of charge. For air-cooled packs, the ducts have to share the air evenly between modules; our CFD analysis of a cooling duct shows how the duct shape sets the flow distribution and pressure loss.

The pack structure carries the weight of the cells, holds them in place, and in many designs resists the force of cells swelling as they age. It also has to survive mounting loads, lifting and handling, and in cell-to-pack and structural pack designs it becomes part of the vehicle body itself.

In our FEA of a 48V-100Ah battery module frame, most of the frame had a comfortable safety factor, but a small region near the lower mounting area dropped to about 0.41. That is exactly the kind of weak spot simulation finds early, when local reinforcement is still cheap to add.

A battery pack sees road vibration for its whole life. Standards such as UN ECE R100, UN 38.3 and SAE J2380 define vibration and shock tests, and FEA checks the design against them before any hardware is built:

• Modal analysis finds the natural frequencies of the pack, so they can be kept away from road and motor excitation.
• Random vibration analysis uses the vibration profiles from the standards, or measured road data, to predict stresses.
• Shock and fatigue checks cover potholes and kerb strikes, and the welds, brackets and bolts that see millions of cycles.

Our guide to vibration analysis explains modal, harmonic and random vibration in more detail.

Most electric cars carry the battery under the floor, where it is exposed in side impacts, pole impacts and underbody strikes from debris or kerbs. Explicit dynamics FEA simulates these events and checks that the cells stay out of the deformation zone, while the sills, crossmembers and pack frame absorb the energy.

Our guide to automotive crash simulation and lightweighting explains how these crash models are built.

If one cell fails, it can heat its neighbours until they fail too. Regulations such as UN GTR No. 20 and China's GB 38031 require that occupants are warned and protected if a cell goes into thermal runaway, and the requirements are getting stricter. Simulation helps by:

• Modelling how heat spreads from cell to cell, and how much insulation the barriers between them need.
• Using CFD to follow the hot vent gases through the vent channels and out of the pack, away from the cabin and the other modules.
• Combining temperature and stress in one coupled thermal-structural model, as in our coupled thermal-mechanical analysis of a 1U CubeSat, to see how heat weakens and distorts the structure.

The battery shares its cooling system, and often its space, with the motor and power electronics. Electromagnetic FEA of the motor gives its torque, back-EMF and losses, and those losses feed the thermal model that keeps windings and magnets below their limits.

Our electromagnetic FEA of a wound-field synchronous machine, the type of motor some electric cars use, calculated its flux, back-EMF and torque. Inverters are usually mounted on cold plates, and CFD sizes those plates in the same way as the battery's.

• 3D CAD of the pack, modules and cooling plates (STEP, IGES, Parasolid or native files), or drawings for our 3D CAD team to model.
• The cell datasheet, and any heat generation or test data you have.
• Duty cycles: drive cycles, charging profiles and ambient temperatures.
• Coolant type, flow rate and inlet temperature, or the airflow for air-cooled packs.
• The load cases and standards you need to meet, such as vibration profiles, shock and crash.

Cover photo: CHUTTERSNAP on Unsplash