Engineering simulation is unique compared to other data used in product development, such as CAD or live test data. “Context is a key element in the value of simulation data,” explains Keith Meintjes, Research Director, CAE for Collaborative Product Development Associates (CPDA). Context means knowing what inputs produced the simulation result, such as the level of maturity of the design of the part or vehicle and other key, critical pieces of data. It also means knowing what simulation tools, simplifying assumptions, and versions of software were used. “CAE results data are temporary—these jobs produce many GB of data in each results file,” explains Meintjes. Engineers analyze these large data sets and develop reports and recommendations, which are long-lived. The raw results are not. Users delete them because of their large size. Add to this the fact that simulation data tends to be scattered throughout an organization—from corporate databases to shared drives to an engineer’s personal hard drive—means that a systematic means of tracking and using that data is vital.
This will become even more important as organizations rely on simulation to replace development hardware and physical testing. “All companies are basing their product development decisions on simulation,” explains Meintjes. He goes on to note that automotive OEMs are in some ways leading the development of SDM because of the sheer volume of simulations they use to develop designs and to assess critical vehicle performance requirements, from fuel economy to crash.
He explains that engineering software vendors are offering technologies from three different perspectives, based on their expertise and company history. Product Data Management (PDM) vendors such as Siemens and SIMULIA (from Dassault Systèmes) extend their PDM tools—typically used for CAD management—to the more context-oriented needs of SDM. Analysis vendors, such as Ansys and MSC Software, grew their offerings from the need to manage simulation projects. “More recently, there has been a third group of vendors, such as Altair and Phoenix Integration, who are very much approaching this as a desktop first application. Their emphasis is on first helping the individual engineers manage their data,” he explains. In all cases, these companies are striving to solve the same problem—organizing and maintaining the context of a simulation so engineers can examine and reuse simulations in the future.
He emphasizes that managing simulation data is most definitely not the same as managing CAD data. “CAE data is routinely deleted, and this is a completely different concept of operation from working with CAD data. [Many automotive OEMs] never delete anything out of their CAD data vault,” he explains. “Also, context is much less critical for CAD files. Think of it this way: CAD is a snapshot, CAE is a frame in a video.”
Reusability reduces cost and increases confidence. Knowing the context of a past simulation means those same results can be the starting point in a new, similar prototype design subjected to the same stress and boundary conditions. An SDM also speeds things up and reduces cost by automating inputs.This means more simulation runs and more variations in inputs contributing to robust design, understanding how a design will react to slight variations in use and loads. “We have seen companies that have taken [simulation] processes that used to take weeks and now are doing it in days or hours,” he says.
He goes on to note that many companies remain unsure of the benefits of SDM, especially when comparing those benefits to cost. “For some of these solutions, the cost is very high,” he says. However, the realization that this is not a PDM solution as used for CAD is growing. “There is a better realization of why the PDM business process does not apply.” He notes that managers of successful SDM projects that his company has spoken to are clear in their objectives and expectations. They typically first address individual and workgroup productivity rather than taking on broader corporate requirements such as global multisite synchronization and collaboration.
“Approach the elephant in small steps,” he cautions.