A panel of experts gathered at an engineering conference called for software that allows everyday engineers to quickly analyze how a range of phenomena would affect design performance.
Designers and engineers have traditionally used successive simulations and analyses, first simulating one physical effect, such as structure mechanics then entering the results from that package as inputs for a subsequent analysis, such as for fluid flow.
But the progressive method doesn’t offer engineers an easy way or give them full range to explore their design space, according to a group of experts who attended the Council for the Future of Engineering Software, held earlier this month in Scottsdale, Ariz.
Also, results must often be entered into the successive package manually, which of course slows the process and introduce errors, they added.
Simulations and analyses give engineers a way to understand the complex physical phenomena that act upon or are acted upon by a design. These include structural mechanics, heat transfer, and fluid flow, to name but a few. Often, engineers use the simulations to help arrive at an acceptable or optimized design. So they’ll design, simulate and analyze results and then—depending on those results—change the design and begin the simulation and analysis process again.
The capability to quickly simulate and analyze how multiple physical phenomena would affect the design allows designers to fully explore their design space, said Don Tolle, director of systems engineering, simulation and analysis at PLM consulting firm CimData.
Tolle guided conversation on the topic during a COFES analyst and user briefing.
He and others attendees argued that too often simulation and analysis software speaks the language of mathematicians, not engineers. The tools require users have expert knowledge in quantitative and algorithmic methods, design of experiments, and Pareto optimization, wrote Tolle in an introduction to his session.
“But we’re beginning to see intelligence built into these tools to enable designers to describe their problems in simple engineering terms,” he added.
For instance, tools available within a web environment can upend the manual process and make multiple simulations available to a greater number of designers and engineers, Tolle added.
“The web space is something going to be a game changer,” he said. “Extracting simulation results and metrics and understanding design parameters within a web environment is helpful without needing a lot of CPU.”
Those who attended the session agreed the paired Isight and Simulia Executive Engine (the latter was formerly known as Fiper), from Simulia, and the Dakota system, from Sandia National Laboratories, allow engineers to explore the design space.
Isight is an open system for integrating design and simulation models created with various CAD, analysis and other software applications, to automate the execution of hundreds or thousands of simulations. It does this by allowing the user to logically connect various blocks, to simulate a desired workflow. The software may be executed either by a single user or—through the Executive Engine–by multiple users working as a team.
Simulia calls this process automation and design automation. Isight simulates multiple processes while the Executive Engine distributes and parallelizes execution of these simulations, according to the company.
Dakota allows engineers use computational methods developed in heat transfer, fluid flow, structural mechanics, and other engineering fields as design tools. So rather than predicting how a single physical phenomenon will affect the design, engineers can determination systems performance is affected by range of physical phenomena that act on the design.
A systematic, rapid method of determining the optimal design solution based on simulation and analysis results will lead to better designs and improved system performance and will reduce dependence on prototypes and testing. That, in turn, will shorten the design cycle and reduce development costs, according to members on the COFES panel.
