CFD Simulation Strategies for Battery Modules in a Rack Cabinet

Challenges

Thermal management is critical in battery modules to ensure optimal performance, longevity, and safety. As these modules operate, they generate heat, which, if not properly managed, can lead to overheating and potential failure.

Effective thermal management helps maintain the battery cells within a safe temperature range, preventing thermal runaway, enhancing efficiency, and extending the battery's operational life.

Computational Fluid Dynamics (CFD) modeling for battery modules presents several significant challenges due to the complex nature of battery systems. These challenges arise from the need to accurately simulate various physical phenomena that occur during battery operation. Some of the primary challenges include accurate modeling of heat transfer and the identification of thermal hotspots crucial for designing safe systems.

Challenges are further compounded when battery modules are stacked together in racks/cabinets. The simulation of a module can contain many mesh elements. When modules are stacked together in a rack, the rack simulation would contain multiples of module mesh count plus the mesh count for the rack domain surrounding the modules. The high mesh counts can cause challenges for hardware memory as well as for simulation solution time.

Engineering Solutions

CFD (Computational Fluid Dynamics) simulation offers significant benefits for the thermal management of battery modules. It allows engineers to visualize and analyze the heat distribution and fluid flow within the battery module and rack/cabinet without the need for physical prototypes.

The engineering solution shown here is to use a two-stage approach for CFD simulation of a rack/cabinet. First, a Fluent CFD simulation is performed on one module/enclosure. Heat flux from the external surfaces is stored in a profile file. Second, a Fluent CFD simulation is performed on the rack with empty enclosures along with heat flux from stage 1 that is read and applied to enclosure surfaces.
The models shown here are for steady-state thermal/airflow conditions. They include an example module in an enclosure as well as a rack with 7 enclosures.

Step-by-Step Guide to Setting Up a Two-Stage CFD Simulation for a Battery Rack/Cabinet with Multiple Modules/Enclosures

Setting up a CFD simulation for a battery module/enclosure involves several key steps:

Create a detailed 3D mesh model of the battery module/enclosure, including all relevant components, physics models, materials, and boundary conditions.
Run module/enclosure simulation and then write out profiles of the heat flux on the external wall surfaces.
Generate translated copies of the heat flux profile for each rack enclosure.
Create a 3D mesh model of the rack/cabinet, including multiple empty enclosures, and all relevant rack components, physics models, materials, and boundary conditions.
Update the rack enclosure wall boundary conditions with the profiles.

Stage 1: Setting Up the Module/Enclosure Simulation

Set up module/enclosure geometry. It is best to have watertight geometry. Below is an example module of 10 cells sitting on a cold plate with liquid cooling. Generate a mesh using Fluent Meshing with appropriate mesh sizing.
Set up the Fluent case.
Solver = Steady-State; Gravity = off; SST k-omega turbulence
Battery model = on; Solution Method = CHT-Coupling; Set Active and Passive conductive zones
Set Battery Model Parameters, including Energy Source per Battery.
Set up materials for air, coolant, solid components.
Set up Cell Zone conditions for fluid and solid zones.
Set up Boundary Conditions for coolant inlet and outlet. Set up thermal boundary conditions for external walls using convection. It is recommended to rename the prefix of this zones to something like wx-*
Generate a custom field function for the negative of total surface heat flux.
Run module/enclosure simulation to convergence. Confirm results are appropriate.
Define New Profiles. Select external walls with convection thermal boundary condition (wx-* prefix).
Activate “Write Merge Profiles” under Merge Profile Options. Select "neg_heat_flux" as the value.
Open .csv profile file in a spreadsheet
Create new profiles with a translation offset (for example deltaY = 0.2meter).
Copy individual profiles to a separate worksheet and save sheet as a .csv file. Repeat for all profiles.

Stage 2: Setting Up the Rack/Cabinet Simulation

Set up geometry of rack cabinet with patterned battery module/enclosures.

Pattern module/enclosure subassemblies.

Exclude internal components of battery such as cells, tabs, busbars, and cold plate.

Retain the enclosure shell components but fill with a solid body labeled as "void." (see below).

Include rack/cabinet components like panels, covers, support, etc.

Cap openings to obtain a watertight geometry.

A subassembly structure tree as shown below is recommended for component management and for profile assignment.
Mesh geometry with appropriate settings. If the recommended subassembly structure is used, the regions should look as follows:
After volume meshing is complete, delete all enclosure regions.
Set up case in Fluent:

Solver = Steady-State; Gravity = on; SST k-omega turbulence; Energy = On; Battery Model = Off.
Set up materials for air and solid rack components.
Set up Cell Zone conditions for fluid and solid zones.
Set up Boundary Conditions for rack openings. Set up convective thermal boundary conditions for external rack walls.
Read each profile file (sequentially) and select neg_heat_flux for the FIelds
Set up Boundary Conditions for the enclosure walls using the profiles just read. The profile needs to be applied to each face zone of a given enclosure. Copy or multi-select should be used to expedite the process. The copy functionality is shown below.
Run the simulation a brief number of iterations and verify the application of the heat flux boundary conditions.
Run to convergence and evaluate results. For example, the contour plot of the enclosure surfaces shows an increase in enclosure temperature with elevation due to natural convection inside the rack cabinet.

Ansys Solution Benefits

One common challenge in CFD simulation of battery modules and rack closets is accurately modeling the complex geometry while simulating within the constraints of compute and time resources.

In the above example the mesh count for one module/enclosure was 3 million elements. The mesh count for the rack model was 1.2 million. If 7 module/enclosures were modeled in detail, the rack model would have had 7*3 + 1.2 = 22.2 million elements.

Model	Mesh Element Count
Module/Enclosure	3.0 million
Rack with 7 module/enclosures	21.0 million
Rack with 7 black boxes	1.2 million

The above procedure could be extended to transient analysis, including thermal abuse runaway, by writing profiles from the module/enclosure at regular intervals, processing these profiles in spreadsheet, and reading these profiles during rack/cabinet simulation at the same regular intervals. In both cases the writing and reading should be done via a journal file and should specify .csv file type.

Fluid flow simulation with Ansys Fluent can be further leveraged for passive and/or active venting design. Structural simulation with Ansys Mechanical can be leveraged for further analysis of the rack/cabinet structural integrity. Sensitivity simulation with Ansys OptiSLang can be leveraged for further analysis of design, operation, and noise parameters. Reduced Order Model simulations with Ansys TwinBuilder can be leveraged for rapid parameter evaluation.

Ozen Engineering Expertise

Ozen Engineering Inc. leverages its extensive consulting expertise in CFD, FEA, optics, photonics, and electromagnetic simulations to achieve exceptional results across various engineering projects, addressing complex challenges like battery thermal behavior under normal and abnormal operating conditions.

We offer support, mentoring, and consulting services to enhance the performance and reliability of your battery system. Trust our proven track record to accelerate projects, optimize performance, and deliver high-quality, cost-effective results for both new and existing systems. For more information, please visit https://ozeninc.com.

Suggested Blogs

Tags:

CFD, Fluent, ANSYS fluent, battery, battery modeling, Transient Profiles

Post by Brian Peschke
January 8, 2025