When you start looking at simulation software tools, you quickly realize that there is a lot there. There are tools for preprocessing. Tools to interrogate simulation results. There are many solvers. But one of the software tools that seems to do a lot is NX from Siemens PLM. Yes, the same NX that is also Siemens PLM’s CAD software application. Let’s take a look.

Background

The history of simulation with this software is a long-standing one. However, a major leap forward in terms of capabilities came when EDS acquired SDRC and merged it into its UGS Solutions division in May 2001 (press release on Siemens PLM site). Over time, not only did they merge and transition the capabilities of the SDRC product into NX, but they also invested significantly into the ongoing development and extension of those capabilities. This effort has continued through the latest version 8 release of NX (press release on Siemens PLM site).

Capabilities Provided

The inclusion of simulation functionality in CAD modeling environments is no innovative breakthrough. Some ten years ago, a glut of CAD software providers partnered with, or outright acquired, a number of CAE software providers to embed simulation functionality in modeling environments. These were often simplified subsets of simulation capabilities meant for designers and engineers to drive design decisions during the design cycle. Siemens PLM, as well as EDS and UGS at the time, took a different tack though. Instead of just moving some subset of simulation functionality over, they moved all of the CAE capabilities from SDRC over to NX over a series of releases. The purpose was to have one single environment for both advanced geometric modeling and simulation execution. Users would range from designers and engineers to the tip-top expert analysts. We’ll dive into the implications of such a different approach during the commentary and analysis.

Supporting the Burgeoning Range of Simulation

Over the years, however, the simulation landscape has not been static. New trends have emerged and become important to organizations and users alike. First and foremost, the types of simulations have expanded significantly. New specialized simulation software providers have emerged. As of today, they provide the ability to execute the following types of analyses:

  • Linear and nonlinear structures
  • Response dynamics
  • Durability
  • Motions and controls
  • Simple and advanced thermal
  • Flow and electronics cooling
  • Laminate composites

Making Multi-Physics Real

While the types of simulations have expanded over the years, that’s not the only relevant trend here. Obviously, multi-disciplinary simulations have become increasingly important. And Siemens PLM has been investing in improving their capabilities in those areas as well. Specifically, they have invested significant resources to development capabilities to implicitly map results from one analysis to other ones. Traditionally this has been an issue because different simulations will often need different meshing models, and most likely different meshing tools, even if given the same geometry. That’s because an area of interest in one type of simulation may require a mesh of a specific type or density and an area of interest in a different simulation will have completely different modeling requirements. Additionally, results from one solver then need to become an input to the other solver, and vice-versa. Typically, this task is done by hand today, and is a major source for errors and time consumption. Siemens PLM provides functionality to implicitly automate this mapping or let you participate in explicitly mapping the overlay. However, each type of multi-physics analysis has different requirements, and the same approach cannot necessarily be used between all pairs of types of simulations. For example, the mapping between a thermal and structural analysis can be quite different between a flow and thermal analysis. Siemens PLM has been looking at and investing in a number of multi-physics analysis types over time.

Sometimes Simulation is About Geometry

In today’s simulation work, sometimes the biggest challenge is working with geometry, not setting up models, running simulations or even evaluating results. That’s because of the reality of today’s supply chains is that you are going to get CAD models in just about every format possible. Very frequently, the simulation analyst must simplify and abstract models from other CAD software applications. When these models are imported, the lack and underlying parametric features to even attempt to modify or remove geometry. Simulation analysts often end up spending a significant amount of time on these activities instead of actually preparing and running simulations.

The good news is that Siemens PLM have made other investments in NX relevant to this painful process. Specifically, they have incorporated Synchronous Technology in NX which I described in a post I published back in May 2011 titled ST in NX: It’s Not About Direct Modeling. The main point behind that post is that ST in NX can be used to manipulate features in NX models more easily than before. That’s what the folks developing NX intended. However, the ability to manipulate non-native NX models with direct modeling interactions still exist. And as such, simulation analysts can use those tools to modify and remove geometry from imported models very quickly.

Commentary and Analysis

In this post, I’ve given a lot of background as well as described the functionality that Siemens PLM has developed as part of NX. But I haven’t explained too much about why it matters. Let’s talk about that here.

One Model is Better Than Many in a Multi-Disciplinary World

Back in October 2011, I published about a very nasty problem that exists in the simulation world in a post titled The Pitfalls of Multi-Disciplinary Simulations: Divergent Model Abstractions. The main point there is that different types of simulations, or even different use cases, require different simplifications or abstractions to the geometric model. Furthermore, different simulation setup use cases lead to varying divergent simulation models as well. Ultimately you end up with lots of different simplified geometric models and simulation models. But in the world of multi-disciplinary simulation, you need those interconnected to get a complete view of product performance. An ugly outstanding question is how to reconcile all those different models with one another?

With NX, however, the picture isn’t ugly. In fact, it’s simple. There’s one model. Period. The different geometry abstractions and simplifications are just different representations that live in one model. All of the different simulation model setups also all live in the exact same model. Like the geometric representations, those simulation setups are just different representations of that one model.

Note: Upon further discussions, these multiple representations within a single NX model can also be managed separately. Teamcenter can recognize them as separable items. That means that different analysts can check out two separate representations, manipulate them and check them back in without coming into conflict. 

Now you might think this is a pretty commonplace solution to this problem. But It’s not. And think about why. If you have one application for CAD and a separate application for CAE, then you have at least two models: one for geometry and one for simulation. And more likely, you will have one CAD model for every single geometric abstraction or simplification. You’re also more likely to have a CAE model for every single simulation setup.

Where’s the downside? If there is one, it exists in this simple question: how does Siemens PLM keep up with the ever burgeoning growth of types of simulations, especially from a system’s perspective. For example, how is electromagnetics supported?

Another consideration is how this is controlled from a data management perspective. What happens if one simulation analyst checks out the model to create new abstractions and another simulation analyst wants to mesh the same model for a structural analysis?

Simulation Analysts Shouldn’t Have to Worry About Geometry

I’ve already touched on the inclusion of ST in NX and the implication for simulation analysts. Of course, it makes it will make life less painful for those simulation analysts. But it will also make it far more productive. Some analysts I talk to actually spend up to 50% of their time fiddling with geometry. No joke. For a role that is usually buried in a huge queue of work submitted by engineers, that’s a terribly waste of their skills and time. But it’s also very unfulfilling professionally. Instead with ST in NX, they can take that time and run more simulations. And who knows, maybe they wouldn’t have to work all hours of the night and weekend to do it.

Conclusions and Questions

As you can tell by the length of the post, there was a lot to talk about. Here’s the recap.

  • Simulation in NX covers a wide range of simulation types.
  • Siemens PLM has invested heavily to streamline  multi-disciplinary simulations within a single environment.
  • Different geometry abstractions and simplifications as well as simulation setups are just different representations in a single NX model, not a number of different models.
  • Siemens PLM has incorporated Synchronous Technology in NX, providing simulation analysts with direct modeling tools to modify and remove geometry from models for abstractions and simplifications.

Those are my thoughts. What are yours? What issues have you seen around managing models for multi-disciplinary simulations? I’m interested in your perspective.

Take care. Talk soon. And thanks for reading.