Simulation App Speeds Up Electric Truck Battery Design

A consortium of universities and industry partners wrapped up the LaserBATMAN project in May 2025, after starting it in May 2022. The partners were Aurobay, Resolvent, the European Commission, the Technical University of Denmark (DTU), Vinnova (the Swedish Agency for Innovation Systems), Innovation Fund Denmark, and Volvo Group Trucks Operations. The project's full name is Multiscale Simulation of Laser Welding for Optimal Battery Pack Manufacturing. Its goal was to optimize two battery-pack metrics: robustness and lifetime, so that electric delivery trucks could cover long highway distances without the battery packs falling short of the service life those routes demand.
The simulation app itself was built by Resolvent, a Denmark-based engineering consultancy focused on green energy, and it lets Volvo's R&D department test battery lifetime performance without running physical trials for every design variant. Resolvent's team worked in COMSOL Multiphysics, initially to study how laser welding affects the electrochemistry and degradation of battery cells. Martin Refslund Nielsen, formerly chief commercial officer and partner at Resolvent, said the ambitions grew during the project: the team ended up modeling performance across the whole battery, from cell level to pack level, and then turned that model into a design tool.
The technical pipeline has three stages worth understanding. First, the team built an electrochemical model of a single battery cell, encoding the formulas for the relevant chemical reaction kinetics. That cell-layer model was then fed input data, and the resulting output data was used to train a deep neural network. The trained network acts as a reduced-order model, a common technique in engineering simulation in which a slow, physics-heavy model is approximated by a fast surrogate that can be evaluated many times. With that surrogate, the team could gain quick insight into cell behavior and then wrap it in a custom app for optimizing the busbar tabs in a pack configuration.
Laser welding sits at the center of the problem because it joins battery components, for instance connecting cells in a pack through a busbar. Done well, the weld supports battery life; done poorly, it raises the risk of degradation. That makes weld quality a manufacturing variable with electrochemical consequences, which is the kind of coupling a multiphysics model is built to capture.
The app exposes a set of inputs that a designer would otherwise have to encode in a model by hand: dimensions for the pack geometry, the number of cells, and operating conditions such as ambient temperature and charge-discharge current profiles. It can apply a simple charging current or load customized cycles, and the user sees how the battery performs for each set of inputs.
A notable control is the Busbar Weld Design Tool button, which Nielsen described as a tool within a tool. It calculates busbar weld performance after a number of cycles. Opening it presents parameters including the shape of the weld, busbar thickness, and the penetration depth of the laser. A user can vary those inputs to see what happens in production if the weld does not reach far enough into the busbar, which Nielsen framed as interesting from a robust design perspective.
A second control, Compute Battery, reports conditions such as the temperature of the weld in a given state, the state of charge, and the thickness of the solid-electrolyte interface layer. The SEI layer is a film that forms on battery electrode surfaces; its growth is widely associated with capacity fade and impedance rise over a cell's life, so exposing its thickness alongside thermal and charge data gives engineers a view of degradation rather than just instantaneous performance.
The app also allows a cooling flow to be applied so the user can see whether turbulence or standing air occurs. Nielsen gave a concrete example of the tradeoff this reveals: if results point to better cooling, the team could potentially reduce a design by two battery cells while retaining the same performance, quantifying the balance between cooling and pack setup. That kind of cell-count reduction matters for cost, mass and packaging, though the source does not state a production decision along those lines.
Nielsen's stated rationale for the app form factor is workflow, not just physics. Custom simulation apps, he said, offer more consistent analysis and reporting, and their intuitiveness makes knowledge easier to share. He summarized the outcome by saying the team is in a much better place to optimize performance and increase robustness.
The app arrived at the end of a three-year project and was presented in a keynote by Nielsen at the COMSOL Conference 2024 in Florence, in October 2024. The source material notes the piece was brought to you by COMSOL, which is the software vendor in the story, so the account of the workflow comes with that sponsorship context.
It is worth setting expectations about what the project delivered and what it did not. The output described is a modeling and design capability, not a battery product, a chemistry change, or a manufacturing line qualification. The reported benefits are faster iteration, consistent reporting, deeper insight into weld and pack behavior, and the ability to explore variants such as under-penetrated welds or altered cooling. The claim that longer highway range with fewer failures follows is framed as what the app can be used to help produce, and any link from app-driven design choices to fleet-level range and reliability figures would need validation in manufactured packs and real trucks.
Why it matters: this is a case study in squeezing costly physical testing out of electric vehicle development. For developers and product teams working on simulation tooling, battery systems or manufacturing engineering, the pattern to note is the combination of a physics model, a trained neural network surrogate, and a wrapped app aimed at non-specialist users who vary geometry, cell count, weld parameters and duty cycles. Teams facing long test cycles for cells and joints can treat this as a template; the caveat, by inference, is that surrogate models inherit the accuracy of the training data, so the app is only as trustworthy as the underlying electrochemical and thermal models it was built from.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
This article is brought to you by COMSOL . These days, no one blinks an eye at two-day, one-day, and even same-day package deliveries. Despite the plethora of localized fulfillment and distributi on centers that have cropped up to meet the demand, delivery trucks still need to make considerably long drives, often over highways. To reduce fuel and maintenance costs as well as carbon emissions, companies are investing in electric trucks for their delivery fleets. While turning to electric trucks is a promising soluti