CAD-based and CAD-integrated simulation continues to be attractive to engineers, even though there is a limit to the capabilities it provides. “Certainly, more design engineers are taking up the task that used to be left to the most dedicated analyst. This is due mainly to the advancement in the intuitive use of modern software,” said Bruce Klimpke, Technical Director at Integrated Engineering Software.
“The appetite for affordable, easy-touse engineering simulation tools has been present for a considerable timeframe, and the recent economic downturn itself was not the catalyst for such interest,” explained Josh Fredberg, Vice President of Marketing at ANSYS.
“Although more design engineers are performing tasks typically assigned to specialized analysts, many organizations realize this comes at a significant risk, since a single individual is not likely to be sufficiently experienced in all the disciplines involved,” said Snyder. “An approach successfully taken by many organizations is to tear down the ‘over-the-wall’ relationship between the designer and analyst, and instead develop close collaboration from the beginning of the development process.” Specialized analysis tools, he explained, need to seamlessly integrate with the CAD environment used by the designer.
Collaboration and Data Management
As analysis and simulation tools become more user friendly and moreclosely integrated with CAD tools, collaboration becomes more important, and along with collaboration comes the need to manage and control the data more closely. “As analysis and simulation are incorporated into the early stages of the development cycle, seamless collaboration between the designer and analyst is critical,” stated Snyder. “In addition, the number and complexity of simulations are increasing rapidly, necessitating efficient organization and accessibility to models and results.”
This means that designers want their CAD and other data sources to link closely with their simulation and analysis data. “It is important to have simulation reports with images and numbers stored with the product data,” explained Marovic. “Therefore, reports in Word, PDF, or HTML are important to prove that a certain variant could fail or was the better choice for a product.”
Littmarck agrees that the connection between tools remains necessary. “Our focus is on making the best multiphysics simulation product on the planet. Having said that, connecting our solution with other engineering tools is critical. This can range from spreadsheets, to mechanical CAD, to technical computing environments, to presentation and reporting tools.”
This need to manage simulation data more carefully increases when there are both designers and analysts (experts) working with the same information. “The value of simulation is the insight provided to designers about what works and what doesn’t work,” said Berry. “This intellectual property should be, and is being, managed today by customers using PLM-like functionality within their data management software.” Berry explained that this is especially necessary as simulation moves from the “one-of-a-kind analysis provided by the expert, to the simulation assembly-line methods used by designers.”
Simulation requires collaboration and data sharing tools that manage analysis from a project perspective, not a bill of materials view, according to Fredberg. In general, he believes, strict PLM offerings from CAD vendors are not aligned with the needs of simulation engineers because they tend to be very document-centric. “Since managing simulation processes and data is a specialized subset of the larger PLM vision, it is often overlooked or poorly add ressed by the PLM solutions being offered today,” Fredberg added.
HPC and Cloud Computing
As larger, more complex simulations are achievable using advanced software, the hardware must be able to keep pace. The ability for high-performance computing (HPC) platforms to run complex simulations also depends upon basic software capabilities being available to take advantage of the hardware. “Simulation software can only utilize high-performance computing platforms after more basic software, such as compilers, has been developed to make use of the hardware,” said Klimpke. “Having machines with many CPUs or GPUs is changing, and will continue to radically change, the reality of simulations being performed.”