Below are the questions asked during the presentation, along with their respective answers.

Q: Which solvers are you using for the simulation of electric machines?
A: SIMULIA CST provides 2D and 3D modeling and simulation techniques for the simulation of electric machines. With the acquisition of Opera we were able to extend our capabilities to magnetic hysteresis analysis, iron loss modeling, template driven motor design, and to perform 2D and 3D rotational motion analyses.

Q: How are you able to reduce the time for mesh generation of vehicles from weeks to hours?
A: The input data for electromagnetic simulations in the automotive industry is based on a mid-surface mesh created from mechanical CAD models. The problem is that originally thick (metallic) parts are reduced to infinitesimally thin sheets, which results in tiny gaps between them, thus interrupting DC current paths. In the past engineers had to modify the mesh manually and to connect adjacent parts. It took about 3-4 weeks in average to create an improved mesh that could be used for electromagnetic simulations. SIMULIA CST implemented new meshing algorithms that automatically consider connections, couplings, ties, bolts or fasteners and makes it possible to reduce the time for mesh generation from weeks to hours.

Q: Are s-parameters used to model the 3D EM fields for circuit co-simulation?
A: All of our 3D (time or frequency domain) field solvers create s-parameters. Either co-simulations are based on s-parameters or reduced order models depending on if the simulation is run in frequency or time domain.

Q: What would a distributed TLM look like? Which one is more effective – lumped TLM or distributed TLM? Where and when to use them?
A: TLM is a time domain simulation method that makes it a perfect candidate for GPU computing. Cell warping or mesh lumping is a great tool to vastly decrease the mesh count of a model by reducing so-called mesh bleeding. In general, mesh lumping is highly recommended because the number of unknowns can be reduced tremendously to 10% -5% or even less. Distributed computing can be used in addition, e.g. for parallel port excitations, thus reducing the total calculation time even more.

Q: Do you measure OR simulate eddy currents?
A: We are a simulation software vendor, so we simulate eddy currents. In general there is the possibility to import measurement data into our software, e.g. from near field scanners or network analyzers, and to include them in electromagnetic simulations.

Q: The Wall Street Journal reported recently on EV manufacturers are eliminating AM radio (medium wave) reception due to inverter and motor EMI. Have you had a chance to study mitigation measures?
A: Yes. In 2016 my colleague Patrick DeRoy gave a presentation at EMC LIVE about Emissions Simulation for Power Electronics Printed Circuit Boards: One of the examples demonstrates electromagnetic interference with the radio (courtesy of Bosch US).