The connected car is an electromagnetic mire. Miles of cable are hidden under the hood and behind doors, while multiple antennas are mounted on the body and inside bumpers. Protecting these electronics from impeding each other is hard work.
The changes have made electromagnetic simulation software indispensable to the auto industry, said Ulrich Jakobis, Altair Engineering's vice president of electromagnetic solutions. It is vital for prototyping antennas as well as finding the best place for them on connected cars, where they are safe from interference.
“In the beginning, it was all about antennas built by suppliers and not by automotive companies,” Jakobis said in a recent interview with Microwaves & RF. Automakers are more familiar with software that can virtually render a car's body and test aerodynamics – less so with tools that simulate antenna patterns and other invisible effects.
That has been changing over the last two decades, said Jakobis, an original author of the simulation software Feko, which went on sale commercially in 1994. Now, engineers are using tools like Feko to analyze electromagnetic fields and navigate new levels of complexity in connected cars.
“The increased usage of electromagnetics simulation is driven by the electrification inside the car,” Jakobis said. “So much is being added, like tire pressure monitoring systems that communicate wirelessly, keyless entry systems, and navigation systems.” Push button ignitions and Bluetooth connectivity are two other additions he singled out.
“For years, it was more about A.M. and F.M. broadcasting antennas, which were directional and had only a few dB tolerance,” Jakobis said. “These days, if you want to use high resolution radar antennas installed on bumpers, you need to be in a completely different range of precision at predicting antenna performance.”
The shape of the bumper affects the performance of the radar antenna, which sends and receives signals to help alert drivers of impending collisions. Other performance issues arise from coupling, in which antennas absorb electromagnetic energy leaking from other antennas and secreted by cables, which causes interference.
The wide range of frequencies used by connected cars is another potential concern. Already, cars connect to satellites for navigation and allow passengers to surf the internet over Wi-Fi. Cars could also be linked to cellular networks to chat with each other about traffic conditions and update digital maps in the cloud.
All this requires lots of electronics. BMW said that it was using around 2.35 miles of cable and 200 electronic control units in its cars in 2015. That same year, the cars used 12 communications technologies and handled 12,500 onboard signals. Separately, experts have estimated that connected cars will require around 18 antennas.
To handle this, Altair has been improving Feko with new mathematical solvers that can be swapped out based on specific simulation problems. Like rivals Ansys and Computer Simulation Technology, Altair has added automatic meshing and the ability to speed simulations using racks of graphics chips in data centers.
“With all these changes, one needs a wide range of solvers,” Jakobis said.
With Feko, engineers can start with solvers based on method of moments – also known as MoM equations – to design a circuit board. Users can then switch over to other algorithms based on the finite element method (FEM) to simulate antenna patterns. Users can also feed these patterns into another program called WinProp, which Altair acquired last year to visualize how radio waves ripple through an environment.
Feko also comes with specialized tool for modeling cables, which are taking increasingly complex paths under the hood and in the flanks of connected cars. The software also lends a helping hand with integrated antennas, offering a tool for simulating antennas installed on windshields.
“There is no method perfectly suited for the whole wide range of high frequency electromagnetic problems,” Jakobis admitted. “These days, popular topics are radar sensors for autonomous driving and mobile communications, and 20 years ago people were not thinking of these, at least not to the extent they are now.”