Matlab Finds New Playground in Simulation Tools
It’s no secret that going virtual to engineer and test new cars, planes or even consumer products saves money, particularly if it means going through multiple versions of a single design. But at October’s Matlab Expo, software engineers and scientists came together to paint a new image of software as something beyond a mere sandbox: a playground.
But imagery aside, Kevin Daffey, head of electrical power and control systems for Rolls Royce, explained software’s role for anyone in doubt, saying, “You couldn’t design an aircraft engine without it.”
Although many of the roughly 900 attendees at the Silverstone, Northamptonshire event were already familiar with popular applications like finite element analysis or computational fluid dynamics, what had audiences surprised was the sheer scope of what digital modeling and simulation tools have to offer.
But beyond these more common tools, Daffey focused on Design of Experiments (DOE), and its growing role in managing multi-variable optimization problems. For Rolls Royce, he explained that the need arises because so many parameters were being manipulated that many would interact with one another, which can be too much for an engineer, but just the sort of problem that the right computer can handle with ease.
Speaking for Ford, Bob Lygoe, a powertrain calibration computer-aided engineering and optimization specialist, noted that DOE was high on his company’s priorities as well. In Ford’s case, boosting fuel efficiency and reducing engine emissions are at odds, which means that in recent years, the company’s engine efficiency has actually been dropping in order to meet emissions standards.
Lygoe explained that by solving these problems “one factor at a time, we can end up with a non-optimal solution.” With DOE, on the other hand, engineers can more easily find the Pareto front, or the maximum point to which you can improve one parameter without compromising one or more others.
Solving for the Pareto frontier allows engineers to identify a finite list of configurations that in this case improve upon either emissions or efficiency without coming at a cost to either. For the infinite configurations that would come at a cost to either variable, engineers can easily discard them. Instead, a designer can look examine the full range of Pareto-efficient choices to determine which the best fit without dealing with information overload.
To improve engines while curbing CO2 emissions by 15 percent, Ford is working with Land Rover, Johnson Matthey, ITM Power, Revolve Technologies, Cambustion, the Universities of Bradford, Liverpool, and Birmingham. The project is called CREO (CO2 Reductions through Emission Optimisation), and it is looking at on-board hydrogen generation to boost combustion and after-treatment efficiency, multi-objective optimization of the entire powertrain, as well as catalyst formulations.
Going against previous trends, Lygoe expects that this optimization-based approach could actually improve fuel economy by three to five percent.
But Sanjiv Sharma, who works on modeling and simulation for Airbus, says that modeling and simulation-based engineering takes time to display its full results. He reminded audiences that although this promises cost reductions when it means cutting down on the manufacture and testing of countless designs building reliable models is a time-consuming challenge all of its own.
Because of this, Sharma emphasized that if time is going to be taken to implement digital modeling and simulation tools, making mistakes or faulty designs should come at no cost.
But whether they were speaking on behalf of Ford or Airbus, both presenters clearly underscored education as part of the investment. In Sharma’s case, he noted the importance of having designers work alongside modellers to help ingrain knowledge of the tool. Meanwhile, Daffey highlighted the fact that compared to talent, software licenses are a small part of the budget: for Rolls Royce, he says that many students who learned Matlab informally have a number of bad habits that require additional training in order to break.