The transition to hybrid and electric vehicles is helping push another change as product developers move to model-based design tools. Engineers in Detroit and China are both making the move, saying it’s a necessity as engineers strive to stay in the forefront of the shift to electrification.
Automakers have been moving to model-based design for years, but that shift has been slow in part because it’s difficult to combine model-based design with the techniques used to develop legacy systems. As systems become more complex, as is the case with hybrids, many engineering managers say that it’s time to take the next step in the evolution of product development.
“Thirty years ago, we programmed using assembly language,” explained Xiaokang Liu, Principal Engineer at Dongfeng Electric Vehicle. Dongfeng is one of China’s largest automakers. “Twenty years ago we programmed using C. For today’s more advanced tools we believe what’s needed is a model-based design methodology.”
Designing electric powertrains poses myriad challenges. There are few precedents to borrow from, and there are many different topologies to explore. Model-based design lets engineers try out many concepts and strategies with virtual simulations in far less time with lower costs than needed to build prototypes and run them through tests.
This design approach also makes it easier for companies to deploy successful designs in a range of systems. “Model-based design is a key enabler for reuse,” said Greg Hubbard, Senior Manager, Hybrid and Electric Drive Controls at General Motors.
One of the central challenges of hybrids is to integrate the electric and internal-combustion engine systems so they provide the utmost efficiency while also making the vehicle fun to drive. The two power sources must interact in ways that save fuel without drivers noticing the changes between electric and ICE.
“There’s a strong trend towards seamless interfaces in the powertrain,” Hubbard said. Though drivers don’t want to feel the changes, many enjoy watching display screens that illustrate the current mileage and show which source is providing power, he added.
GM and Dongfeng use The MathWorks modeling software to help them understand the intricacies of their designs. As engineers dive deeper into the minutia of fuel burning and battery control, they have hundreds or thousands of variables than can’t be managed with conventional tools.
During simulations, they can change a few parameters to see how performance changes. On the hardware side, modeling and simulation help engineers compare one technology to another. That is helpful when powertrains are used in different vehicles.
“When we base our control strategies on physics, it’s easier to move from one application to another,” Hubbard said. “When parameters change, whether it’s as simple as a tire radius or something subtle like friction losses in the transmission, the equations help us interpret what’s driving the factors we’re looking for.”
Engineers can also examine equations in real time, working their way backward from parameters such as axle torque. Once that torque is determined, they can determine the torque level that is necessary for both the engine and the electric motors. “We can break it down to 120 times per second and determine the optimal place to operate the engine and the electric motors to get optimal efficiency,” Hubbard said.
Another major challenge for powertrain engineers comes with the different types of batteries. Now-popular NiMH batteries have far different characteristics than the lithium-ion batteries of next-generation systems. There are a number of different lithium-ion variants, each with its own operating parameters. Modeling the batteries and related systems is not a simple task.
“The internal chemical reactions of batteries are very complex; it’s very difficult to use accurate mathematical modeling to describe their condition,” Liu said. “The calculation of state of charge is even more of a world-class dilemma. We needed to focus more on the algorithms, but using traditional methods of programming would no doubt have impacted our progress.”
Maintaining battery packs is tricky, since the cells must be charged and discharged at roughly the same rate. They must also be kept at the correct temperature or lifetimes will be impacted. Engineers must also adjust operation to account for temperature extremes.
“We have to hide the weaknesses of the batteries and make sure they operate under hot and cold conditions,” Hubbard said. We also have to make sure we don’t blow the fans so loudly that we annoy the drivers.”
The task of managing batteries extends beyond temperature and cell load leveling. “We designed a battery management system which not only plays an important role in protecting the battery, but at the same time it’s also directly related to the vehicle performance and economics,” Liu said.
The economics are one aspect of a total strategy. At GM, there has been a major effort to come up with one architecture that can be used throughout the company’s product lines so that engineers in all departments are working within the same environment on the same challenges.
“We spent a lot of time focusing on the best topology for hybrids and electric vehicles,” Hubbard said. “That’s critically important. Once you select the topology, it’s three, four, or five years before production, and there are still thousands of issues that will determine fuel economy, quality, and whether the vehicle’s fun to drive.”
That approach helped the company utilize regeneration during breaking on all its vehicles. That gives GM an edge in mild hybrids that use mainly start-stop techniques to conserve fuel. Some automakers’ mild hybrid systems don’t recharge batteries during braking, so they’re not as efficient.
“Regenerating energy gives us a 15% fuel-economy improvement,” Hubbard said.
GM’s start-stop system always engages belts, letting the electric motor provide a boost at high speeds while also enabling energy recovery. Constantly maintaining belt engagement also makes start-up smoother, Hubbard explained.