Does NX really think it has a fighting chance in the CAE world?

...Published 2018-01-29

“Don’t be silly, CAD companies can’t compete in the analyst space!” – me, only a few years ago.

I’ve been a very long time (since ±2002) FEA and MBD analyst on high-end engineering simulation and analyst software but have never been able to learn to use a proper (or any for that matter) CAD package.

For CAD generation or modifications I always turned to my design colleagues who used NX to create all their designs and if mayor cleanup was required because of their modelling practices, I recommended them to attend an FEA course to REALLY understand what I meant with NO gaps or overlaps to make my work as analyst even possible.

And then I learned that NX entered the FEA and MBD world with NX Nastran and various other solvers and I just rolled my eyes. Who did these CAD guys thought they were, entering these highly specialist worlds with their little (in my opinion then) CAD tool?

Really? I mean, come on!

So I never took it seriously and left it at that, continuing to do serious analysis work in my specialist FEA and MBD packages.

Unfortunately for me, the noise about NX CAE just never subsided and over time I couldn’t ignore it anymore and I reluctantly started looking at examples of how others used NX to prepare models for analysis. To be honest, it was always just too good to be true (and looked very difficult), so again I just left it at that. Until recently.

I am not really sure what broke this camel’s back in the end or what really flipped me over, but I am singing a totally different tune now!

Okay, now I remember what turned me: the recent (and also the not so recent) SIEMENS spending spree: Buying LMS in 2012 for ±$865m (1) and more recently in 2016, CD-Adapco for $970m (2) and Mentor Graphics for a whopping $4.5Billion (3) in the same year! And this was just in the area of serious analysis tools that I am interested in.

Now if that doesn’t grab your attention, nothing short from an electric shock will.

LMS was always well respected for their simulation AND testing tools, with strong FEA tools in the non-linear realm and a very good MBD tool, so that was an attempt in the right direction but it was the latest acquisitions that send my head spinning with StarCM++ being an exceptionally strong and easy to use CFD tool from CD Adaptco followed by the packages such as FloEFD and many others from Mentor Graphics on the thermal and electronic side.

So this meant only one thing to me, these guys (SIEMENS) are very serious about their roll to play in the simulation world, which brought me to look a bit under the covers of NX.

And then the real discovery (which I was blind to for a very long time): A very clever way in which CAD geometry stays linked to the analyst’s idealised (simplified) geometry for simulation purposes.

When I heard about this the first time (and that was indeed quite a while back), I simply said “It can’t work, maybe on a simple demo model, but not on a real complicated model.”

How wrong I was.

To be really honest, I never really believed it until I re-created a tutorial model and saw the power of this selective linking method between actual geoemtry and idealised geometry.

The video of the this eye opening moment can be downloaded from here.


Leave a Reply