Automotive engineers have been combining electronics and mechanical systems for years, but they are still figuring out how to efficiently model and simulate mechatronic systems. Sharing files between disciplines remains an issue along with the challenge of creating models for use in different applications.
Solutions for these issues were examined during a crowded SAE 2013 World Congress panel during the "Simulation and Modeling Mechatronics" technical session. One that is particularly vexing is the difficulty of developing models that require input from a number of development programs.
William Milam of Ford Motor Co. noted that the vast number of tools used to design vehicle systems makes it very difficult to move data seamlessly across disciplines. In response to an attendee question, panelists downplayed the likelihood that a standard will make it simpler to share data. Each field has its own requirements, which often aren’t acknowledged in tools used by other design engineers.
“Thermal modeling tools have their own language for thermal issues,” Milam said. “When you try to meld things together, the real issue is do they work in a different environment? When you’re doing things like mixing analog and digital, it’s a very tough problem with no off-the-shelf solution. I don’t know if there’s a standard on the horizon.”
He noted that the issue can consume a fair amount of time.
“I’ve seen people spend afternoons doing battle over tools instead of focusing on the problem at hand,” Milam said.
Companies have developed a range of techniques to resolve this matter. Some have created their own converters.
“We’ve developed tools to communicate,” said Nathan Crosty of Magna Electronics. “It’s too difficult to do it by hand. We’ve been pushed to automate this task.”
Determining when existing models are effective and when models must be changed was another topic raised by audience members. Panelists said there isn’t any “one size fits all” solution. Engineers must create models for different applications. Sometimes that’s an issue when systems are combined, since different parameters are needed when other elements within the larger system interact. Engineers must also change models when they move to real-time simulations.
“Reducing models down is an art more than it’s a science,” said Dan Robinson of John Deere Intelligent Vehicle Systems. “You need to know what the model will be used for, then you need to pick out the dynamics. If you have models developed for other applications and you try to run them in real time, it’s not going to work.”