Industry Update: Analysis & Simulation Software
- Thursday, 30 October 2008
In our annual poll of executives at leading analysis and simulation software companies, we posed questions dealing with virtual prototyping, multiphysics, the roles of the designer and analyst, and how software vendors are helping their customers remain competitive. Here’s what they had to say about major changes in their market, and how new technology will benefit users.
While the term “virtual prototyping” has been used for a number of years, it is defined differently by various industries. For some, virtual prototyping can mean simply simulating a design on a computer instead of building a physical prototype. But for others, it is more complex, and the tools required to achieve a true virtual prototype are not always user friendly.
“Virtual prototyping is being able to do everything you do in the lab through simulation,” said Bruce Klimpke, Technical Director of Integrated Engineering Software. “Not only would it mean how to simulate the physical parameters such as electrical, thermal, and mechanical, but it would also include aesthetics and how people may react to the product.”
Svante Littmarck, President and CEO of COMSOL, Inc., agrees. With virtual prototyping, he said, “Instead of building a physical prototype, you simulate it using a computerized model representing all interesting aspects of the realworld prototype, such as physical behavior following the laws of science, interaction interfaces with other objects and media, and other relevant constraints. It streamlines the path between ideas for design modifications and their consequences, which contributes insight into the physical physics underlying performance, and in turn, drastically improves designs.”
This complete product performance analysis also helps designers and engineers get their products to market faster, often at a lower cost, by eliminating expensive prototyping and testing phases. “Virtual prototyping expresses what engineering analysis software can accomplish as a product development tool, which basically is testing a 3D CAD part in computer space,” said David Weinberg, President and CEO of NEi Software. “The word ‘virtual’ conveys the most powerful benefit of engineering simulation — no fabricating of parts, no test fixtures, no sensors, and no test runs required,” Weinberg added. “Product designers can catch flaws and problems early, intelligently optimize, explore innovative alternatives, improve quality, and deliver finished designs faster.”
“The CAE industry is evolving toward its ultimate concept — ‘black box’ analysis, where a product designer won’t need to know about FEA details such as analysis type and finite element mesh,” explained Michael Bussler, President and CEO of ALGOR, Inc. “Instead, the software will function as a black box virtual prototyping tool, requiring the user to simply describe the physical characteristics of the product’s environment. The black box tool will then automatically handle all of the analysis set-up and processing details behind the scenes and allow the user to see how the product performs.”
Virtual prototyping not only helps designers and engineers learn about a product’s performance in a controlled, but real-world, environment. It also can simulate entire systems or subsystems. “Since real-world behavior is affected by multiple physics concurrently, the only way to truly simulate that is with a comprehensive multiphysics system,” said Chris Reid, Vice President of Marketing for ANSYS, Inc. “Virtual prototyping can identify how and why a prototype test failed or succeeded in cases where physical testing was unable to explain the results.”
Multiphysics analysis capabilities play an important role in true virtual prototyping, and have been an important component in simulation for a number of years. “Over the past decade, multiphysics analysis capabilities have become more tightly integrated into a single process so that engineers can simulate an entire scenario that incorporates the whole product and the environment in which it will be used,” explained Bussler. “No longer do users analyze one instant in time as with linear static stress analysis. Instead, simulation includes large-scale motion and impact and stress analysis, while also considering other multiphysics effects.”
While virtual prototyping has its obvious benefits, it also requires more detailed data, extensive computing power, and of course, the right software. “Unless your software supports multiphysics modeling, you’re not going to be able to tackle problems in a realistic fashion,” according to Littmarck. “The more physical phenomena your software can take into account, the fewer physical prototypes you’ll need and the closer you’ll get to ‘true’ virtual prototyping.”
Reid agrees that the choice of software is critical. “The technology enables the engineer to go beyond physical constraints and perform simulated tests that otherwise would not be possible. It is critical to exploring and expanding operational boundaries in developing products and processes,” he added. “In this way, modeling and simulation can be used to drive new solutions, rather than to merely verify existing ones.”
Extinction of the Analyst
As CAE software enables more complex analysis, including virtual prototyping, vendors also are working to keep it easy to use. And as more analysis and simulation capabilities become available within CAD tools, is the dedicated analyst becoming a dinosaur?
“Dedicated analysts are by no means a dinosaur in today’s product development process,” stated Dale Berry, Director of Technical Marketing for Dassault Systemes SIMULIA. “There are many engineering applications that require in-depth analysis and review to understand the behavior of various physical interactions. Often, some of the most complex simulations are not even directly linked to a CAD model.” Berry added that “Analysis experts are being tasked to define, develop, deploy, and monitor the analysis tools and methods being used to ensure that performance-based decisions made with CAD integrated analysis tools are valid and accurate.”
“The dedicated analyst is not in danger of becoming extinct,” Bussler agreed. “Still, people whose primary job function is to perform computer simulation of product tests have seen their role evolve and change. Twenty years ago, the analyst’s work was commonly done near the end of the product development cycle, often as a final verification of physical prototyping data,” he explained. “Today, CAE software is commonly used much earlier in the product development cycle.”
Jon Friedman, Aerospace, Defense, and Automotive Marketing Manager for The MathWorks, agrees that the analyst’s role continues to change. “In the past, the dedicated analyst only got involved once the physical testing began. Now, design, testing, and verification have become a continuous function, rather than something performed at the end of a design,” he said. “The domain expertise and experience of analysts are more critical to the design flow than ever before. The dedicated analyst must now get involved early and often to analyze the virtual results, through the initial prototype testing, and all the way through the final testing and sign-off.”
The ability of design engineers to perform the required analysis and simulation tasks within their CAD program depends upon the software’s capabilities. “In most cases, design engineers are able to perform analysis and simulation tasks,” said Bussler. “Whether the analysis tool is within CAD or works with CAD doesn’t really matter because FEA software has a CAD-like interface. Anyone who uses CAD software can quickly and easily learn how to use today’s CAE software. If they know how to properly set up and perform a valid product test, then they can successfully use FEA software,” he added.
Weinberg agrees that a user friendly CAD environment can help designers accomplish their analysis goals. “The introduction of CAD-based FE tools has not at all diminished the role of the dedicated analyst. What FEA within CAD has accomplished is to extend the benefit of simulation to a different and earlier point in the product development process,” he explained. “The ‘dinosaur’ is any FEA system that does not have a common solver platform that can be shared by both designers and analysts. Designers and analysts have different skills, capabilities, objectives, and responsibilities. Their FEA software needs to reflect that, while not isolating one from the other.”
Many vendors have built scalability into their software, enabling users to access only what capabilities they need, or are able to use. “There will always be a need for experts who understand deep, comprehensive physics concepts,” said Reid. “Not every user will require the same level of technology because he or she may not have the need for an answer of the highest accuracy.”
Berry also believes in scalability of capabilities within CAD and simulation tools. “It’s a false notion that nonexperts do not require powerful underlying simulation technology. Tools for non-experts need to be just as powerful — or more powerful — than those used by experts since they have less tolerance for working around issues where the technology is insufficiently robust.”
“As designs become more complicated, more dedicated analysis and niche software will become more important,” added Klimpke. “The real question will be how well they all can be integrated into the total design.”
The analysis and simulation software market is looking at trends in hardware advancement and in technology innovation. Multicore processing and 64-bit computing are both helping users increase the size and complexity of problems, and get answers faster. “Though the increase in speed of serial computing has not increased significantly, the next major push will be in parallel and distributed computing,” predicted Klimpke. “Both these avenues will completely revolutionize the size and complexity of problems being solved.”
Added Littmarck, “Problems are becoming more complex because software developers can implement more sophisticated algorithms that need more computational power. These problems are being solved faster because the latest hardware runs faster and provides more memory at a lower cost.”
But with complexity come other problems. “Very few software applications have been modified to truly leverage multiple cores. This is a much more complex problem,” said Friedman. “While peripheral tasks can be offloaded easily, the core computations of simulation often are very interconnected and hard to break into distributable tasks.”
As for the software itself, companies are becoming more savvy about the types of tools they need. “Companies recognize that the earlier FEA is implemented in the design process, the more pronounced the benefits,” said Weinberg. “Companies with a tradition of FEA use are looking to leverage their investment by getting tools that are more industry-specific. They are looking to their software provider to add features and modules that enable them to do simulations very specific to their unique problems.”
Reid agrees. “Healthcare and biomedical, energy, and the process industries are pioneering new techniques using simulation,” he said. “There are companies developing breakthrough products in advanced-technology healthcare that would not exist were it not for the ability to design and develop their products using simulation.”
“Perhaps the most significant trend is the industry’s acknowledgement of the need for simulation lifecycle management solutions,” stated Berry. “Com - panies are now starting to understand that managing the data, methods, and design decisions that are an integral part of the simulation process will become a key competitive advantage.”
Added Reid, “As the technology is used more, organizations will need to find efficient ways to capture, manage, and re-use the resulting data, engineering insight, and best practices. The value of these types of knowledge assets is inestimable, and, likewise, the cost of losing it is huge.”