A European consortium is using multiscale simulation and AI surrogates to optimize laser-welded battery packs for heavy-duty trucks, boosting durability and lifetime.
Fast shipping has reset expectations: two-day, one-day, even same-day delivery is now routine. Meeting that demand requires fleets of trucks that can still cover long highway stretches, even as logistics networks add more local distribution hubs. For many operators, battery-electric trucks offer a path to lower fuel and maintenance costs while cutting carbon emissions—but the battery packs that make these vehicles possible do not always deliver the longevity needed for sustained, long-distance use.
A European research-and-industry consortium has been tackling that bottleneck by bringing advanced simulation directly into battery-pack design and manufacturing. The group—spanning companies, universities, and public funders—combined multiscale computational modeling with a purpose-built software application to speed up decisions about how battery packs should be assembled for durability. Participants included Volvo Group Trucks Operations, the Technical University of Denmark (DTU), engineering consultancy Resolvent, Aurobay, and Vinnova (Sweden’s innovation agency), alongside support from the European Commission and Innovation Fund Denmark.
The effort, formally called Multiscale Simulation of Laser Welding for Optimal Battery Pack Manufacturing and nicknamed “LaserBATMAN” (for battery manufacturing, not the comic-book character), ran from May 2022 to May 2025. Its central objective was to improve two intertwined performance targets for heavy-duty battery packs: robustness and lifetime.
One manufacturing step sits at the heart of both goals: laser welding, which is widely used to join components during pack assembly—for example, when connecting cells via a busbar. A high-quality weld can help enable consistent electrical performance over time. A poor weld, by contrast, can introduce weaknesses that accelerate cell degradation, undermining range and reliability.
To understand and anticipate these effects, the project team relied on multiphysics, multiscale simulation using COMSOL Multiphysics. Resolvent’s Martin Refslund Nielsen, formerly the firm’s chief commercial officer and a partner, described COMSOL as well suited to the coupled physics involved in welding, electrical behavior, and battery aging. The initial focus was on how welding influences electrochemistry and degradation at the cell level, but the scope expanded. Over time, the team developed models that extended from individual cell layers up to pack-level performance, ultimately translating that modeling capability into a practical design workflow.
A key step was building an electrochemical model of a battery cell by encoding the relevant reaction-kinetics equations. The team then generated input–output data from this detailed model and used it to train a deep neural network (DNN) as a reduced-order surrogate. With the DNN standing in for the full model when speed was critical, engineers could explore how design and manufacturing choices affect cell behavior far more quickly than would be possible with high-fidelity simulation alone.
That reduced-order approach enabled the consortium to create a custom simulation app, developed by Resolvent, intended for Volvo’s R&D teams. The app is designed to streamline testing and optimization—particularly around busbar tab configuration in laser-welded pack architectures—so engineers can evaluate battery lifetime implications earlier and iterate faster.
Why it matters
Heavy-duty electric trucks are only as practical as the battery packs powering them, and premature degradation can make long-haul routes uneconomical. By linking welding decisions to electrochemical aging through simulation and an engineer-friendly app, the LaserBATMAN partners aim to shorten development cycles while improving durability. What to watch next is how widely these model-based workflows spread across battery manufacturing—and whether they translate into measurable lifetime gains in real-world fleets.