A Lesson in Multiphysics

OEM engineers reveal how new software from Comsol improved their modeling and analysis of EV battery performance and audio-system sound quality.

Acoustics simulation of a sedan interior including sound sources at the typical loudspeaker locations. Results show the total acoustic pressure field inside the cabin. Image made using Comsol Multiphysics software. (Image: COMSOL)

Modeling and simulation are becoming a more critical factor in automotive design, particularly in subjective areas like passenger-cabin sound quality and complex newer technologies like battery management. Design tool suppliers are increasing the power of those tools while also making them easier to use.

Tools that handle multiple physics categories are becoming more important as engineers need to understand how complex interactions change when various parameters are adjusted. At the recent Comsol Users Conference, technologists from General Motors and Nissan detailed their efforts to understand nuances in audio systems and electrified-vehicle battery packs.

Patrick Dennis, Principal Engineer at Nissan, told attendees how the company migrated to virtual sound rooms because physical tests for analyzing audio speaker quality and configuration proved problematic. After the analysis of the sound room tests showed problems at certain frequencies, Nissan engineers ran simulations that predicted those issues, giving the team the confidence to change the roles of physical and virtual tests.

“We went into a sound room and made a binaural recording that seemed to be perfect,” Dennis said. “When we brought it back and put it in a car, it didn’t sound that good. We created a reference room using Comsol tools that gave us recordings that sound better than what we did in the expensive room. That’s our primary tool now.”

Simulation results showing the sound pressure distribution generated by a loudspeaker driver. Image made using Comsol Multiphysics software. (Image: COMSOL)

At the conference, Comsol unveiled Multiphysics Version 5.4, which the company claims offers improved performance and ease of use. The latest version provides a number of modules that make it easier and faster for users to understand how changes to one parameter ripple out to many other elements of the overall system.

“Version 5.4 is several times faster for rendering large 3D models; depending on the geometries, it can be an order-of-magnitude faster,” claimed Bjorn Sjodin, Comsol’s VP of product management. “There’s also a new comparison tool that lets you change the settings in a model and compare the new one with the model saved on file. We also added a more sophisticated module for batteries.”

That’s a helpful move for automakers and Tier 1s that are striving to meet regulatory mandates for fuel economy and emissions. Battery charging/discharging and overall lifetimes involve myriad factors that span several types of physics. The updated tools and modules can give developers deeper insight into areas that haven’t previously been examined.

“We need to supplement the information suppliers give us, so we end up with huge data sets,” reported Taylor Garrick of General Motors Global Propulsion Systems. “I deal with 40 to 50 parameters I need to model; the suppliers usually only give us 20 to 30. We need to understand all these parameters to answer questions about the size, performance, and output of battery packs, and to see when we need to cool every cell or when we should selectively cool cells.”

Multiphysics tools are becoming more important as engineers attempt to eke out more performance, since minor changes in one area can cause often-unexpected changes elsewhere.

“In battery packs, you’ve got a lot of factors to examine,” noted Sjodin. “You need to know how the ions inside the battery are reacting to heat, and you need to know how temperature changes are impacting the cooling liquids. Both factors have a big impact on performance and [battery] lifetime.”

The new multiphysics modules help GM’s engineers visualize the temperature profile in liquid-cooled Li-ion battery packs. (Image: COMSOL)

The addition of dedicated modules is part of a corporate focus on making the multiphysics software easier to use. The battery module includes 1D, 2D and 3D options, so users can work with different fidelity levels depending on the task at hand.

“When we want to put the battery pack together with inverters and other components, a 1D model gives us the quickest results,” Garrick said. “If we’re trying to work out some aspects of battery pack design, 3D models give us the best results. 3D has the highest fidelity.”

Comsol also unveiled a compiler program that makes it easier to share apps. It even lets users who don’t have the company’s software view models and simulations. Version 5.4 attempts to balance the need to provide more ways to analyze the subtleties of each design with the need to offer tools that don’t take a long time to learn. Modules dedicated to specific fields are among the company’s ease-of-use techniques.

That approach also helps specialists in these fields delve into the subtle factors that determine end performance.

“We have more modules for specific applications,” noted Valerio Marra, the company’s marketing director. “They make it easier to do the work in those fields, since many advanced factors are tailored to the application.”



Magazine cover
Automotive Engineering Magazine

This article first appeared in the January, 2019 issue of Automotive Engineering Magazine (Vol. 6 No. 1).

Read more articles from this issue here.

Read more articles from the archives here.