PLM for Simulation

Consider Toyota’s sudden acceleration problem. Before that, the Ford-Firestone tire blowout problem. And in an entirely different industry, the BP-Deepwater Horizon oil spill in the Gulf of Mexico. Each of these cases involved physical phenomena in complex systems. Each involved multiple physical properties and multiphysics domains. Each could have undoubtedly been better understood with more computer-aided engineering (CAE) and simulation. But there’s a catch. More simulation increases the complexity of simulation management, organizing and making sense of the simulation results, and making simulation data available to people (corporate and regulatory) and to other information systems (applications and data repositories, on-site and remote, within a company and with partners and vendors).

Much is at stake. When a product goes out into the field and breaks down, blows up, or does something that smacks of personal liability, bad publicity, or a loss of profits, who takes the blame, handles the warranty issues, pays the price?

At a more mundane level are the basic problems in storing historical CAE and simulation data and making those data accessible. CAE analysts know they need to save their product simulation trials, but what data do they save? One consideration, says Ken Versprille, PLM research director for Collaborative Product Development Associates, LLC (cpd-associates.com), is to look back after product development, after the fact, and say, “This is what was done” and “This is why changes were made here and here. Companies want to feel very comfortable in reusing those [parts and assemblies] in other products. Unless you store that information, that history is just word-of-mouth. Companies can’t rely on word-of-mouth anymore.”

Another consideration is the process management in simulation (versus data management, though those two areas do overlap). Simulation in product development is as much an iterative effort as product development itself. The analyst takes the first-cut version, meshes the product design, applies materials and loads, sets up a simulation run, then watches what happens. The results typically drive design changes, which require reapplying materials and loads, redoing meshes, and rerunning the simulation. Templates can capture these design properties, analysis parameters, and simulation settings for reuse. Ideally, if the same type of simulation needs to be run in the future, an analyst need only log into a repository of prepared templates of best practices, mouse click around to the appropriate template for the specific type of product design model, attach it to the product model under investigation, then run it. “Storing and reapplying what they’ve done in the past not only speeds the whole analysis/simulation process, but it also adds consistency to the whole process,” says Versprille.

Scott Del Porte, lead product manager for Ansys EKM at Ansys, Inc. (ansys.com), ticks off some other effective CAE data management benefits: “Accessing and reusing historical design information and expertise to aid in the progress of new designs. Capturing and leveraging existing engineering knowledge. Addressing the loss of engineering expertise, while protecting intellectual property. Reducing future development costs and risks by simulating a wider range of operating conditions.” Sound familiar? “While these goals are shared with many PLM-based initiatives, the unique nature of CAE simulation processes and data demands a special treatment,” continues Del Porte. In the world of CAE and simulation, these PLM-like initiatives are satisfied by simulation process and data management (SPDM) systems.

Data management of its own
Appliance, automotive, aerospace, medical device, and other companies have PLM strategies that include linking data from multiple systems, such as product data management (PDM) for CAD, enterprise resource planning (ERP) for inventory and financials, and customer relationship management (CRM) for customers, explains Del Porte. “Simulation warrants its own data management system.” He gives this folksy analogy: If you’re a farmer, you want to store diesel fuel, grain, and hay in different kinds of silos because hay can’t be stored in the same silo as grain or diesel fuel, diesel fuel can’t be stored in the same silo as grain and hay, and so on.  The same is true with simulation, CAD, and customer data. Separate systems. “You wouldn’t want your CAD data stored in your CRM database,” says Del Porte. The primary reason is because CAE and simulation databases are huge compared to what’s generated in the CAD and manufacturing worlds. There’s data about pressures, temperatures, material properties, boundary conditions, solver algorithms used, and more. “If you try to put simulation data in a PDM system, you’ll crush it,” posits Del Porte.

One wouldn’t necessarily hear about SPDM when talking to PLM vendors who provide CAD, CAM, and CAE software. They also provide PLM, such as Teamcenter from Siemens PLM Software (plm.automation.siemens.com/en_us), V6 from Dassault Systèmes (3ds.com), and Windchill from PTC (ptc.com). SAP (sap.com) doesn’t provide CAD, but it does provide the data management foundation for a network of applications and utilities that provide PLM. Then there are the CAE vendors. Notable SPDM systems include Engineering Knowledge Manager (EKM) from Ansys and SimManager from MSC.Software (mscsoftware.com).

“SPDM is really a microcosm of what PLM is doing. It ‘snaps’ in underneath this umbrella called ‘PLM’ by providing specialized data management and process management for simulation,” says Del Porte. Or as Versprille puts it, “These systems play nicely with the PLM systems from the other vendors. This is by necessity. You don’t use CAE tools in a vacuum.”