Simcenter Amsterdam: making simulation a routine part of design
Last December I went to Amsterdam to attend Siemens’ Simcenter Symposium. It’s one of my favorite events because it brings together so many different … everythings: auto talks to aero, who talk to marine; STAR-CCM+ people learn about Amesim; and we all learned that pigeons can detect cancer (yep, a scientist was paid to figure that out).
In total, over 600 people attended the more than 150 sessions, on everything from CFD to EMAG to process changes needed to move simulation into the hands of more designers. That’s a huge increase in all ways that matter from just a year ago: more presentations, more attendees, more diversity of application … it’s hard to know why with any certainty but it would appear that Siemens’ intentions to create a portfolio multiplier is working, as more of its users take advantage of more of the tools. This used to be a CFD-weighted event (since it’s based on the old CD-adapco event) that’s now truly multi-physics, if I may be allowed to use the term in this way.
Many take-aways. Here are my top items:
- The Simcenter portfolio has its roots in NX Simulation, which came about largely because of the acquisition of a Nastran variant over 15 years ago. Nastran is still one core of the offering, but is rarely mentioned. I asked Jan Leuridan, Siemens SVP of Simulation and Test solutions about this and he pointed out that most people use it but don’t see it as worthy of mention. Interesting: where others highlight their solvers as points of differentiation, to Siemens and its users, they’re tools, important but not where actual action happens.
- I attended last year’s event in Prague (you can read about that here) and was impressed by how many of the users were exploring the breadth of the Siemens offering. Much of it was new to them at the time. Just one year later, the focus wasn’t on point solutions but on combining physics, more advanced usage of Siemens’ optimization tools, and generally up-skilling rather than exploration. Maybe it’s just because different individuals attended the two events, but the tone in Amsterdam was more strategic than that in Prague.
- The event kicked off with customer keynotes that gave a high-level look at how simulation changes outcomes. Henrik Alfredsson of Aker Solution talked about how CAE and IoT created an entire new line of business that enables Aker to both design and install subsea gas production systems, and then (the new offering) to monitor and predict performance.
- Gugliemo Caviasso of Maserati took us in a different direction. Many people buy a Maserati because of its distinctive engine growl; what would that be in a world of electric engines? Should the manufacturer mask mechanical noises that we can’t hear over the sound of today’s combustion engine? Yes, Maserati does a significant amount validation and verification, but is also using CAE to redefine its brand values. (I checked with a car-enthusiast who once owned a Maserati. She fell in love with the car’s look but bought it because of its craftsmanship and the engine’s rumble. Creating that rumble without combustion is clearly a tough problem for Maserati and other car companies to crack.)
- The keynote that most people were waiting for was, of course, Siemens’. Dr. Leuridan gave a quick tour through the offering, from 0D/1D to 3D, from generative design in concept stages to test in just-before-production, and from the automotive through shipbuilding industries. He also noted that Simcenter will soon be available via a token model (STAR-CCM+ and HEEDS already are), which got a lot of interest from attendees I spoke with; said that the machine learning component of artificial intelligence will play an increasingly important role in the world of CAE (more on both, below). Finally, Dr. Leuridan hammered home Siemens’ unique value proposition: the integration of simulation and test in one platform. The ability to simulate and validate with test; to simulate in advance of test to target test; and the ability to use test to define areas for detailed simulation at the system and component level takes Siemens’ offering from the theoretical to the practical.
- One thing that was missing, for me, was the next step: using real-time data to drive simulation. Say you have an IoT system in place to gather data (in Siemens’ world, that’s Mindsphere). Wouldn’t it be incredibly useful to direct some of that flow to a simulation, if a potential problem is indicated? Mr. Alfredsson of Aker is taking the first steps; I can’t wait to see how this develops.
- The Simcenter team was generous with its roadmaps, with each major product set offering an update. Nearly all were standing-room only, which means that attendees want to know where the tools they’ve invested in are going — and also, as a couple told me, to figure out where they need focus their training and up-skilling efforts.
- I can’t cover every product, but the general themes of the roadmaps were this: improved and consistent UI across the platform, the ability to launch more simulation types from Simcenter 3D environment, more physics across the applications, workflow improvements to speed FEA model creation,
- One thing worth mentioning from the Simcenter 3D roadmap session ties back to test. With the coming release, Siemens added new test/analysis correlation tools for test planning. This should allow analysis to help test engineers position sensors during physical test. There was also something about transfer path analysis between test and simulation that I didn’t quite get that had the audience take notice: apparently, testers can capture these forces during test and pass them back to CAE which can compute loads for NVH? Dunno. If this matters to you, ask Siemens!
- HEEDS, the design space exploration solution, is still one of my favorite CAE apps. As per usual, I didn’t understand every part of the presentation on the latest release, but support for Python 3.6 and the ability to record and play back macros seemed to please people looking to automate design exploration. With version 2019.2, users can use AI (well, adaptive machine learning) to sample the design space and improves response surface accuracy. As across the platform, there are also usability enhancements, template tools, and streamlined connections to Autodesk Inventor and Aspen HYSYS, as well as Simcenter 3D.
- And, while we’re on HEEDS, that was another tool that wasn’t explicitly mentioned as often as it used to be. In years past, customer presenters would spent a good chunk of their allotted time talking about how they set up HEEDS, how hard it was and how confident they were of their results. As with NX Nastran, it seems to have become an understood, unmentioned part of the tool set that’s, nevertheless, an integral part of what many people do. Perhaps it’s the UI improvements over the last few years, maybe it’s the more affordable licensing — it’s definitely much more the norm than it used to be.
The customer presentations were, as always, fascinating and highlight how much there is to learn about the world around us and the products we rely on every day. They all had a couple of things in common: a problem, some exploration and discovery and then a solution — that’s engineering, after all. But in many cases, the discovery led to new questions which led to new discoveries, which led to … and so on. Somewhat new for me was how many of the speakers mentioned using Teamcenter (Siemens’ data management and PLM solution) to drive these processes and manage the massive amounts of data generated.
This conference used to be a simulation guru experience (the early STAR-CCM+ events I attended were expert-fests). In 2019, it broadened to include new roles (designers, material specialists, test engineers and others) because they need to understand simulation –how it works, where it does and doesn’t apply, what the limitations might be for their use case– to make a bigger business impact. That was true again this year, as several presenters talked about using the automation tools across Simcenter to create advanced simulation tools for designers.
One last thing: One of the most interesting presentations had absolutely nothing to do with CAE. Dr. Hannah Fry, co-author of The Indisputable Existence of Santa Claus: The Mathematics of Christmas (actually a lot of fun; I read it on the flight home), talked about AI and its impact on humans. She covered whether pigeons can diagnose cancer (yes), that AI can go hideously wrong when applying algorithms to prison sentencing guidelines, and pointed out that humans write the algorithms and need to both plan for the worst and take responsibility for the outcomes. It was an interesting choice for Siemens, since many people wonder if they will lose their jobs to an algorithm. Dr. Fry’s examples make clear that some tasks can be done by pigeons or humans or algorithms — but not all. My take: humans can’t outsource the hard choices, at least for now.
The title shot is a view of one of Amsterdam’s many canals in an early morning fog. Gorgeous, no?
This is a video, produced by Statoil, of the Aker undersea compressor: https://www.youtube.com/watch?v=Ew1h9aU4odo . They don’t mention Aker, so here’s a video Aker has created about the subsea compressor system, hinting at the analytics at 1:30 or so: https://www.youtube.com/watch?v=ZzQhfIv3mgc . If I can find a public video of the analytics, I’ll post that — it’s brilliant.
Note: Siemens graciously covered some of the expenses associated with my participation in the event but did not in any way influence the content of this post.