Saturday October 25, 2014

Preparing Models for Simulation: Which 3D Modeling Paradigm is Best?

Simulation technology has its origins in the labs of universities and specialists groups within the aeronautics and automotive industries. Today, however, simulation software has moved from research labs into mainstream product design and engineers and designers are increasingly using simulation tools to guide their design choices.

How simulation tools are used today

As designs progress throughout the development cycle, a sequence of decisions and adjustments are made, affirming that the design is on the right path. When used properly, simulation tools can help guide engineers on that path by providing with a way to make better decisions in order to design better products faster and at lower cost.

To make it easier to incorporate simulation into product development, many CAE programs are tightly integrated with CAD software. The integration of CAD and FEA enables engineers to test ideas, adjust designs, explore, and verify to confirm that designs are on the right track, minimizing the risk of flawed designs moving forward when changes are most costly.

FEA1 177x300

Preparing models for simulation

When models are ready to be simulated, engineers and designers often have to simply them in order to be able to perform simulations on them. During our recent “Pros and Cons of 3D Modeling Paradigms” webinar, the question of which 3D modeling technique works best when models are going to be simulated was posed to our speakers. I think their respective responses are worth sharing.

Chad Jackson, principal analyst, Lifecycle Insights.

I think that both those technologies (direct modeling and feature-based modeling) work with simulation but in two very different scenarios. I think with direct modeling, it helps you with simplification and abstraction of model prior to analysis. Now, if you try and do that through feature-based approaches, you can run into model failures and you have got to find the guy that originally designed it to really make progress there.

Direct modeling is a really good fit there if you’re looking at one model and you want to get it ready for simulation. However, there are a lot of organizations today that deal with a lot of variance of products or parts. Feature-based modeling is a great way to automate the generation of those variances and that applies to simulation as well when you look at things like design-of-experiment studies, optimizations and things like that. I think both have a fit but very different uses.

Dan Staples, vice president, Solid Edge Product Development, Siemens PLM.

The only thing I would add is that when you talked about the abstraction, and I think when you get to that it’s typically the analyst who is getting something from the designer and he needs to abstract it for purposes of analysis. The thing about direct modeling, in general, is it’s for someone who’s not a CAD expert and hasn’t been indoctrinated into the history tree and the recipe that most of us got indoctrinated into 10 or 15 years ago.

History-based modeling techniques are intuitive, but only because we had a beat into our heads for many years. Certainly, I think if you were to ask the average guy on the street or the guys who’s an analyst and not exposed to CAD on a daily basis in terms of modeling, it’s just a lot easier to approach the kind of changes you needed to make by using direct modeling.

Brian Thompson, vice president, Creo Product Management, PTC.

It’s not merely a matter of direct modeling style changes for purposes of the abstraction or for purposes of simplification, but it’s also the analyst who does his first round analysis. He gets a result and he realizes well, ‘I’ve got an issue with the stress here or there’ and the analyst then wants to go in and modify the the model, not for simplification or abstraction purposes, but because he’s got stress problem. He could go through the pain of going back and asking the engineer to make that change or he could maybe do some exploratory work with the model himself.

With direct modeling he doesn’t need to worry about the original design intent of the model. He can make direct modeling-based changes to the model in order to explore how to reduce the stress or how to change the natural frequency or whatever type of analysis he’s doing in order to achieve the design goal without having to cycle back through the person who understood the original design intent of the model. Lots of good reasons to use direct modeling there.

If you missed the “Pros and Cons of 3D Modeling Paradigms” webinar, you can watch it in its entirety here.

Barb Schmitz