用户案例集锦

Driving EV Development with a Twin-Battery Approach

Considering energy efficiency, energy density, and environmental concerns, IAV combined complementary sodium-ion and solid-state lithium iron phosphate battery technologies in a twin-battery system optimized and validated with multiphysics simulation that opens up new possibilities for car manufacturers and battery designers.

By Joseph Carew
April 2025

Avoiding the rare raw materials required for the production of traditional batteries without sacrificing energy density is a major goal for those looking to electrify the world. Lithium-ion batteries power most of today's electric vehicles (EVs)¹ but are associated with high costs as well as sustainability and environmental concerns. Engineers and developers in the battery industry are investigating alternative chemistries and designs to find new approaches that address these concerns and reduce costs while fulfilling the demands of most lithium-ion applications.

IAV is one of the world's largest engineering companies. Within an extensive portfolio geared toward the future of mobility, battery development plays a critical role. A team of IAV engineers including Jakob Hilgert, a technical consultant at the company, felt that, with the right approach, IAV could achieve better battery designs. The team leaned on its understanding of what makes existing single-chemistry designs successful — as well as what holds each back — to develop a novel approach to solving battery energy density, sustainability, and thermal management issues: a twin-battery design.

Instead of turning solely to lithium-ion cells, IAV engineers thought a pair of alternative battery chemistries could be combined to form a less expensive and more ecofriendly system that could handle EV applications. With this approach in mind, IAV turned to multiphysics simulation to successfully design and validate its twin-battery solution.

Avoiding Lithium-Ion Battery Pain Points

While lithium-ion batteries (Figure 1) are often used for their high energy density², their creation can have environmental drawbacks. Open-pit mining for lithium removes vegetation, creates toxic soil, and releases dust that elevates the risk of illness in animals and people³. Producing these batteries is also an expensive prospect¹ and reliant on a relatively rare material. IAV engineers looked to avoid these concerns when choosing the technologies to be included in their twin-battery approach.

"We need to be prepared for batteries that have a larger focus on recycling and resources," Hilgert said. "We cannot always take the highest-energy-density cell that is theoretically possible and use that as our solution."

Instead, the team at IAV chose to pair a sodium-ion battery (SIB) and a lithium iron phosphate (LFP) solid-state battery (SSB) for its design because of the chemistries' unique ability to complement one another. SIBs are typically cheaper, more sustainable to source, and easier to recycle than conventional lithium-ion batteries⁴; however, they tend to have comparatively lower energy density and a shorter cycle life. Meanwhile, traditional LFPs are known for their stability and long cycle life but also lack in energy density when compared to conventional lithium-ion batteries. Finally, SSBs are known for having higher energy density than traditional lithium-ion battery chemistries. By combining an SIB with an LFP-SSB, the resulting design should theoretically have an improved environmental footprint (Figure 2), cost less money to create, and feature a relatively strong energy density for demanding applications such as powering EVs.

"The development of batteries for automotive use is progressing rapidly. It goes hand in hand with a rising demand for scarce raw materials," Hilgert said. "Diversification of cell chemistries is a promising approach to respond to market fluctuations and at the same time minimize system costs."

Creating Thermal Compatibility

IAV's twin-battery design was also developed, in part, to test the thermal compatibility between an SIB and LFP-SSB. The idea was that channeling the waste heat from the SIB into the LFP-SSB would rapidly activate the latter's solid-state cells and push them into the higher temperature ranges where they perform best⁵ — while simultaneously keeping the SIB from exceeding its maximum operating temperature and increasing the system's overall energy efficiency.

"If we have some cells that can operate at high temperatures and some cells that can operate at low temperatures, it is beneficial to take the exhaust heat of the higher-running cells to heat up the lower-running cells, and vice versa," Hilgert said. "That's why we came up with a cooling system that shifts the energy from cells that want to be in a cooler state to cells that want to be in a hotter state."

Cells with liquid electrolyte have limited thermal stability and require cooling (true for both sodium and lithium cells), and temperatures above ~60^°C need to be avoided. Solid-state cells can operate at higher temperatures because of their solid electrolyte, and these need an elevated temperature to reach usable ion conductivity. Therefore, the SIB cells in this concept need cooling while the SSB cells need heating, and both cells benefit from the mutual heat exchange. IAV engineers knew that this interaction in particular would be a significant optimization challenge and felt that modeling and simulation would be essential to easing the complexity. For this, the team turned to the COMSOL Multiphysics^® software.

Designing the Battery System

IAV first began using COMSOL Multiphysics^® more than a decade ago to improve its design workflow.

"We were using a large quantity of different specialized tools for different specialized topics," Hilgert said. "When we started working with batteries, it was time to say, 'We need one tool to deal with all of these topics.'"

The platform's comprehensive workspace gives IAV the opportunity to avoid building unnecessary prototypes for clients and easily optimize its designs. With the twin-battery model, IAV engineers can tweak different parameters (whether, for example, they impact the cooling of particular circuits or the maximum power that cells at a certain temperature produce) and alter the design to ensure that any real-world creation is as efficient as possible. "If you have this knowledge and you do not have to guess at all of these parameters, then the technology readiness level of the prototype will be a lot higher," Hilgert said.

Because of the multiphysics nature of battery modeling, the COMSOL^® software's capabilities were well suited for the twin-battery system (Figure 3) development project: Designing operational batteries requires proper thermal management, an understanding of how the materials of different cells are going to perform within their modules, knowledge of the varying pressures within the internal processes in the battery, as well as an electrochemical understanding of the whole. There also needs to be an understanding of how swelling or contraction during charging and discharging can impact the mechanics of these systems.

"A highly integrated model-based development process can be used to investigate the potential of different cell chemistries, designs, and cooling concepts," Hilgert said. "It reduces the need for physical prototypes and allows for performance optimization toward typical requirements of automotive applications."

Heating, Cooling, and Design Optimization

Engineers at IAV were able to verify the performance of its twin-battery concept using coupled multiscale and multidomain simulation (Figure 4). The team found that the design worked as desired during concept development, paving a path forward for better battery design. The model showed very fast on-demand activation of solid-state cells, with partial preconditioning done by the SIB's waste heat. The team has optimized the thermal management of the two cells and shortened the time and energy input needed for SSB activation in cold conditions.

"The simulations showed that it is actually possible to do what we had in mind," Hilgert said. "The waste heat actually can be transported by the cooling system, and the amount of heat is sufficient to heat up the other part of the battery."

IAV was able to run different scenarios, comparing various levels of sensitivity for different surrounding conditions or parameter selections with its model, which functions as a virtual prototype. The team successfully integrated 3D cell temperature distributions, pseudo-2D (P2D) electrochemical modeling, and 1D cooling circuit dynamics into a comprehensive electric powertrain model.

Democratizing the Twin-Battery Model with Apps

Once IAV’s simulation specialists have developed a white-box model for a customer, they often use the Application Builder in COMSOL Multiphysics to additionally package its functionality into a simulation app, a custom-configured user interface with restricted inputs and outputs that the customer can distribute internally to colleagues in different domains who use it to run simulations and evaluate results in their respective contexts. App users do not need in-depth knowledge of the underlying complex model; instead, simulation apps are designed to be easy to use and hard to break, making them ideal for IAV’s many customers who "want to distribute these simulation tasks to people that usually do not do modeling," as Hilgert put it.

“We can start with the basic functionality and hand it out to everybody, and nobody will have a problem using it. Later on, if things get more detailed, we can have the apps grow with the application and add more physics, more options, more buttons," shared Hilgert.

IAV engineers use COMSOL Compiler™ to turn their simulation apps into standalone executable files that they send to their customers alongside the white-box versions of the underlying models, who can then run them without a COMSOL license (Figure 5). This makes it easier to run simulations in distributed development environments. In the case of the twin-battery design, cooling system engineers can run parallel optimization calculations without COMSOL licenses. Streamlined access to simulation results leads to more efficient development processes and has greatly improved the acceptance of model-based development both internally and among IAV's customers.

"Having COMSOL Compiler as a distribution option is a great benefit for our work," Hilgert said. "We can use our own models for some simulation or profiling tests just by compiling some apps and then having other people do their jobs, without having to wait for the licenses."

Interfacing Java code is used to provide remote control of the apps that IAV builds thanks to the COMSOL software's API. This remote control allows users to automate repetitive modeling steps. The team also implements Functional Mock-up Unit (FMU) interfaces, which it couples to vehicle simulation environments in third-party software for cosimulation.

Users of the twin-battery app are given the voltage, state of charge (SOC), temperatures, and power dissipation as inputs to the battery management system and cooling system. Design engineers can view the internal cell states through these apps and make changes to the cooling system as they evaluate the varying battery performance.

Using Apps Internally

Apps that are used internally at IAV are often designed for cosimulation with COMSOL Multiphysics^® and external toolchains and are routed through IAV‘s virtual test bench interface. Figure 6 shows an example battery module app used for cosimulation, which provides basic user feedback about the internal states of the models like current, voltage, temperatures, etc. App results are provided as a real-time data stream to other programs in the cosimulation framework, where detailed evaluation of results is performed.

(T)Winning the Battle for a Better Battery

IAV hopes that its twin-battery design concept will function as a showcase to others in the battery industry that, even if you have demands that are contradicting, there can still be a solution.

"The twin-battery approach gives the car manufacturer or the battery designer more options to solve their problems," Hilgert said. "It also shows that there is a way of integrating future technologies with very different principles into existing frameworks."

Oracle and Java are registered trademarks of Oracle and/or its affiliates.