1. Tell us about the COMSOL product suite and its application.
COMSOL Multiphysics® is a modeling and simulation (M&S) software platform used in all fields of engineering, manufacturing, and scientific research. The software provides fully coupled multiphysics and single-physics modeling capabilities as well as built-in tools for managing model data and building simulation applications. The intuitive user experience of the platform product extends to a suite of add-on products that includes modules with functionality specialized for the modeling of certain applications and physics areas as well as products for integrating the COMSOL® software with CAD software.
In the field of e-mobility, our customers are mostly found in the R&D departments of large corporations and at government and university research labs. Some common applications of our software include the M&S of batteries, fuel cells, electric motors, and generators. The models involved usually couple several physics phenomena — such as those involved in electrochemistry, electromagnetic fields, heat transfer, and structural mechanics — in so-called multiphysics models.
Our software enables innovation in the development of batteries and drivetrains for electric vehicles by helping engineers and scientists to understand, predict, and optimize the behavior of the components and processes involved. Using our software in the development process is extremely time and cost efficient compared to a purely empirical development process; with COMSOL Multiphysics®, R&D teams can develop more for the money in less time.
One concrete and simple example of this efficiency is the use of M&S in the design process for the foils that conduct current into and out of a battery. (For more details, see our blog post “Improving Tabbing Design in Cylindrical Batteries“.) Before building a prototype, an R&D team can use M&S of the battery design to estimate the possible improvements in performance and answer the question, “Is it worth it or not to pursue this design idea?” It is quicker and substantially more cost effective to test new ideas using M&S than to build a prototype.
2. How do you see EV technology trends evolving in the industry this year?
While our focus is on the possibilities of using simulation software, in looking at the literature regarding EV technology, there are some trends reflected in M&S.
One of the biggest problems in automotive applications is that we want batteries to recharge very fast, preferably in a matter of minutes, like fueling a gasoline car. We also want to be able to drive for hours on this charge. However, the recharge current is orders of magnitude greater than the typical discharge current, which is problematic. Many of the electrical systems in the drivetrain have to be designed for coping with currents that are orders of magnitude different between discharge and recharge. These systems involve the battery, inverters, DC links (capacitors), current collectors, feeders, cooling and heating systems, etc.
This current disproportion is still an unsolved problem, and is therefore the focus of many R&D projects. We think that the trend for systems adapted for shorter recharge will continue. New battery types and new designs are evolving and will likely continue to evolve in the near future.
3. How will COMSOL be contributing to the higher technology revolution in the coming future?
The development in the field is very rapid, with huge investments being made. We have to match this pace.
Many of the engineers that are thrown into the field are not yet experts in modeling batteries, generators, motors, electrical components, etc. We need to constantly improve our software’s user interfaces to make it easy for these new engineers to benefit from multiphysics M&S. We must implement the latest models, incorporate the latest material properties of new materials being developed, and make sure that they are available in easy-to-use interfaces.
In addition, we have to continue to make it possible for M&S experts to build simulation applications based on multiphysics models and then share them with project stakeholders who are not M&S experts. The use of dedicated simulation applications built in house in our customers’ R&D departments allows them to gain even greater benefits from multiphysics M&S.
4. What makes COMSOL’s offerings different from other players in the market?
Based on our customers’ feedback, and also based on our patents, we believe that we offer unique M&S capabilities. For instance:
- The Model Builder gives users the ability to create mathematical models and generate numerical models “on the fly”. This makes it very easy for our developers and our customers to add new models for new materials and new processes accounting for any physics phenomenon.
- The Application Builder — for creating simulation apps based on multiphysics models — is unique; we have not heard of any other tool that is so specifically adapted for M&S. The Application Builder enables users to drag and drop widgets and other components into a simulation app’s user interface and link them to the settings of the underlying multiphysics model. Then, using the COMSOL Compiler™, the app can be compiled into a completely standalone app that can be run without a COMSOL license or installation, which means that the benefits of M&S can be shared with a larger community.
- The Model Manager provides a centralized repository to help users organize, reuse, and maintain version control of models, apps, simulation data, and associated files, such as experimental data, CAD files, presentations, and reports. While there are other tools for simulation process data management (SPDM), we are not aware of any that is as specialized and adapted for multiphysics M&S as the Model Manager.
5. Anything you would like our readers to know?Artificial intelligence (AI) and machine learning (ML) are on everybody’s agenda these days. We see the combination of M&S and AI (and ML) as a way of creating very fast and compact yet accurate models. One application for this combination of technologies is using it to predict the status of battery systems, such as for accurately determining their state of charge and state of health. In this case, we would use high-fidelity models to train an AI. The process would require a large number of computations, but the resulting AI-based model would be very compact and extremely fast. It could be used for real-time decision making and could be incorporated into a digital twin (DT), where the high-fidelity model would be used to continuously update the AI and the DT. The DT could then be used for diagnostics, characterization, selection of materials, and other processes involved in the development and design of battery systems. We think this integration with AI and ML is something to look forward to regarding multiphysics M&S.
