In our annual poll of executives at leading analysis and simulation software vendors, we found that both 3D printing and the cloud will have increasing impact on CAE in 2015, and users will demand additional multi-physics capabilities to simulate larger, more complex models.
Last year, our roundtable participants predicted that simulation would become a necessity for survival in many industries, given the economic climate and the need to reduce the costs of physical prototypes. As the proliferation of 3D printing continues to decrease the number of prototypes needed, more industries are turning to additive manufacturing technology, which in turn impacts how — and when — simulation is performed.
“The maker mindset that takes advantage of 3D printing is certainly having a big effect on analysis and simulation usage and scope,” said Dale Berry, Senior Director of User Experience for Dassault Systemes SIMULIA. “When you can print a 3D object that could not be manufactured using traditional means, then simulation is no longer a luxury — it’s a necessity to ensure that the final product design is functional, long-lived, cost-effective, and robust.”
Optimized designs that are virtually tested by simulation can be printed rapidly in 3D models, which can have a dramatic effect on the way design cycles are perceived, according to David Vaughn, Vice President of Marketing for CD-adapco. “Any technology that reduces the time to prototype a design is further supporting the growing trend toward using more engineering simulation earlier in the design process.”
Dominic Gallello, President and CEO of MSC Software, agreed that being able to go directly from the computer to part also means that simulation must be pushed farther up the design chain, and in many cases, even before CAD. “3D printing opens up the possibilities for the math (simulation) to come before CAD, not just as a checking tool after CAD. You can deliver all kinds of unusual shapes for load-bearing parts without having to change expensive hard tooling — you just print.”
Svante Littmarck, President and CEO of COMSOL, Inc., agreed. “The ability to print 3D drawings is a very positive thing for software companies in the modeling and simulation market. It takes engineers one step closer to production by allowing them to print prototypes directly from their analysis or design software.” This type of manufacturing, added Littmarck, eliminates at least one level of communication, since a 3D prototype can be produced by the analysis expert right away.
“From the simulation perspective, whether the model is created by 3D printing, or by using nuts and bolts to hold components together, the simulation data is similar,” said Bruce Klimpke, Technical Director at Integrated Engineering Software. “Of course, the CAD data for 3D printing has to be precise. These models are usually more amenable to simulation.”
The leap from physical testing to 3D printing means that old design rules no longer apply. Added Berry, “It is ironic that despite the fact that you can print such objects faster than ever before, you have to rely on simulation rather than testing more so than ever before to get it right.”
The increasing dependency on simulation in the manufacturing process has caused software vendors to think differently about integrating 3D printing into the current workflow, according to Derrek Cooper, Director of Simulation Products for Autodesk. “In the CAD space, for example, you’re seeing more support for ‘output’ to 3D printers, similar to the emergence of integration with CAM many years ago. The main difference is that CAM is an established and understood market, where the additive processes are being developed and tweaked in real time.”
Cloud-Based Engineering
Engineering and simulation in the cloud was an important trend our executives were watching for in 2014, and as 2015 approaches, cloud-based software is poised to change the paradigm of how — and where — simulation can be performed. Software vendors are finding the cloud to be “an important, emerging way for our customers to access the power of CAE,” said Jim Rusk, Senior Vice President for Product Engineering Software at Siemens PLM Software. Added Mark Sherman, Director of FEMAP Product Development at Siemens, “Cloud technology has reduced the latency in running GUI-based software, but there’s nothing better for interacting with a large FEA model than local high-performance hardware.”
Before trying to help their customers navigate cloud-based simulation, vendors must understand its benefits and pitfalls. “Cloud-based engineering simulation is a complex undertaking for end-to-end workflows, involving a focus on access to data, interactive HPC job management, and high-performance 3D graphics,” said Barbara Hutchings, Director of Strategic Partnerships at ANSYS. “We’re working with customers who want to implement simpler cloud-based workflows — template workflows that can be exposed in a browser, without complex application access.”
For users in high-tech, high-security industries like aerospace, cloud-based engineering has not hit in a big way yet, said Gallello. “CIOs are certainly thinking about private clouds. Security for these firms is of huge concern. ‘Anytime, anywhere’ does not exactly fall in line with this, especially when it comes to product IP. I believe the walls of the fortresses are going to be built higher going forward.”
And offering cloud-based simulation and analysis tools just for the novelty of it is not useful, stated Darrell Teegarden, Mechatronics Product Manager at Mentor Graphics. “If the cloud products that emerge over the next couple of years don’t address fundamental issues, then they won’t be successful.
“Simulation and analysis using cloud services will most certainly become an everyday use paradigm,” Teegarden added, but “it will be most valuable as it becomes integrated into solutions that bridge the innovation-to-implementation bottlenecks.”
Managing Data — and Humans
Managing simulation data continues to be a bottleneck. As high-performance computers enable more simulation data to be aggregated, analysts and engineers must be able to interpret the results. This bottleneck requires management of both the data and the humans. “It’s up to the analysis software engineer to create the tools necessary for efficiently generating and analyzing big data sets,” said Littmarck. “Of course, it’s also a human management issue to use such tools efficiently.”
Gallello agreed that the key is to use data management tools efficiently and effectively. “High-performance computing delivers more results faster, but there are not enough trained engineers to interpret the results. The key is to deliver these results side-by-side with past decisions that were made,” he said. Klimpke added that “the bottom line is to determine when enough time is spent on analysis in order to make a decision.”
Management of engineering data in context with design data is essential for companies seeking to improve collaboration and speed up development, said Rusk. But “adopting a simulation data management solution also requires a cultural shift. Engineers need to embrace a new way of working and demonstrate a willingness to share data more openly.”
Controlling both the data and those accessing it remains an issue, given the volume of data and the growing interest in most organizations to protect, archive, and leverage simulation results, said Hutchings. “But there is certainly a human aspect to the data proliferation question, both in terms of evolving away from individuals independently managing their own data, and in terms of the time involved to sort through it and extract understanding.”
Big data also creates big access issues, which turn data management into “information management.” Data management tends to imply managing files. Simulation creates a large number of files, and file sizes tend to be large. “I believe the industry will shift to managing the information in these files, rather than the files themselves,” stated Cooper. “Today, most simulation users store their files on file servers or external devices. When they need information or a result, they usually go back to these files and reopen them to search for an answer, which requires managing proper versions. The cloud opens the door to a variety of technologies like search, where it really doesn’t matter where the files are stored. What matters is how you can access the information.”
Results analytics is an emerging area in simulation that Berry believes will transform the traditional post-processing exercise into a modern decision-support process. “The process will enable users to identify the needle in the haystack across multiple, very large results datasets that will be located remotely rather than on the user’s desktop.” These methods, he added, “will open the door to a future where simulation is at the heart of the corporate go-forward product development strategy.”
In addition to data management and human management, Vaughn sees a third option. “There needs to be an adaptation of both the technology and the users. It’s possible that a truly innovative technology breakthrough will be the result of a human adaptation to using or viewing data differently. No doubt, this generation of engineers sees the world differently, having been raised while connected to electronic devices with access to exponentially more data than probably the past several generations combined.”
Trends for 2015
As described earlier, the cloud has opened the door to more and easier access to data and results, and its use will continue to become more commonplace among simulation software users. As computing capabilities evolve, so does the level of detail in an FEA model. “With high-performance desktop hardware, cluster-based computing, and now the cloud, the resolution of FEA meshes can become finer and finer,” said Sherman.
“Only a few years ago, many simulation vendors were quiet and skeptical around the cloud,” Cooper said. “But with the need for heavier computation on the rise, companies are looking at alternative computing power. I believe you will see more and more cloud offerings on the market.”
This trend implies challenges for end-user productivity, but also contributes to global collaboration and IP reuse and protection, according to Hutchings. “One very clear trend is in the IT environment for simulation, where we see continued consolidation of enterprise infrastructure — essentially moving simulation from the desktop or department server to the local datacenter or the cloud, with users remote and globally dispersed.”
Berry sees a paradigm shift occurring in which simulation technology is put into an enterprise-level platform built for all. This shift is to “full-blown complex simulation of the most difficult problems facing a company, all performed within a collaborative environment. Such practices require re-imagined user interfaces, as well as an entire truckload of technology.”
The shift of the simulation paradigm is great news for end users, said Gallello. “2015 is the year that users should poke their heads up from their work to see what is out there. I think they will be amazed with the leaps forward.”