What is Classical FEA and How Does It Differ?

  • Thread starter Thread starter Observables
  • Start date Start date
  • Tags Tags
    Classical Fea
Click For Summary

Discussion Overview

The discussion revolves around the concept of Classical Finite Element Analysis (FEA) and its differences from modern FEA techniques used in software like Ansys and Abaqus. Participants explore the definitions, applications, and evolution of FEA methods, including their historical context and technological advancements.

Discussion Character

  • Debate/contested
  • Technical explanation
  • Conceptual clarification

Main Points Raised

  • Some participants suggest that "Classical FEA" may refer to older methods or software, possibly indicating a lack of clarity in the term's definition.
  • One participant mentions that Classical FEA typically uses geometry that is not identical to CAD models and employs h-version finite elements, which simulate detailed features by decreasing mesh size.
  • Another participant counters that decreasing mesh size at details does not yield higher resolution and questions the accuracy of the claims regarding Classical FEA's practices.
  • There is a discussion about modern FEA utilizing actual CAD geometry and potentially employing mesh-free methods or p-version finite elements, with some participants expressing skepticism about the fundamental differences between Classical and modern approaches.
  • Some participants express curiosity about the practical applications of mesh-free methods and their advantages in typical scenarios.
  • One participant clarifies that the term "Classical" in software contexts may arise from the need to differentiate between older and newer interfaces for practical purposes, such as bidding and hiring.

Areas of Agreement / Disagreement

Participants do not reach a consensus on the definition of Classical FEA, with multiple competing views on its characteristics and distinctions from modern FEA. The discussion remains unresolved regarding the implications of these differences.

Contextual Notes

Some limitations include the ambiguity of the term "Classical FEA," varying interpretations of mesh refinement techniques, and the evolving nature of FEA software capabilities. The discussion reflects differing experiences and practices among participants.

Observables
Messages
14
Reaction score
0
Hi all,

I was speaking to a design engineer and he mentioned that he uses Classical Finite Element Analysis in his work designing valves and cylinders.

Can someone explain to me what is Classical FEA, and how does it differ from FEA that you use in a computer program (Ansys, abaqsus,...), or am I being dumb, and they are the same thing.

thanks
 
Engineering news on Phys.org
"Classical FEA" probably means he either uses old software, does it by hand, or uses old methods. In any case, it's a nonsensical term.
 
The only think I can think of is Ansys Classic. Now known as Mechanical APDL in Workbench.
 
Finite element methods have been developing/evolving ever since they were invented, not least because of the steady increase in computer power over the past 50 years or so.

I have never heard the term "Classic FE" used to describe anything definite, beyond a general nostalgia for the "good old days".
 
(1) Classical FEA uses geometry that generally is not exactly the same as the CAD (computer-aided design) solid model part geometry, and generally uses h-version finite elements, which means detailed features are simulated by decreasing the mesh size at details.

(2) "Modern" FEA uses the actual CAD part geometry (isogeometric analysis), and either uses mesh-free (meshless) analysis methods, or perhaps uses a p-version finite element mesh, which means detailed features are simulated by automatically increasing the mesh polynomial order at details, instead of decreasing the mesh size.

Both of the above categories of FEA are used in computer programs.
 
nvn, I don't think any of your post is really true, sorry...

nvn said:
(1) Classical FEA uses geometry that generally is not exactly the same as the CAD (computer-aided design) solid model part geometry,

Standard practice for any analysis engineer. Removing features that do not affect the solution redudces mesh size and solve time. Standard examples would be removing holes from low stress-gradient areas, and removing fillets/chamfers from external corners.

nvn said:
and generally uses h-version finite elements, which means detailed features are simulated by decreasing the mesh size at details.

This doesn't make sense to me. Deacreasing mesh size at details would just give you lower resolution at a detail feature (unless you correspondingly increase element order), this has been true ever since FEA was first invented.

nvn said:
(2) "Modern" FEA uses the actual CAD part geometry (isogeometric analysis),

It's true that you can use geometry straight out of a CAD model. Additionally less simplification is required these days due to more powerful computers. Still, I don't think this is really a fundamental difference between your two examples...

nvn said:
and either uses mesh-free (meshless) analysis methods, or perhaps uses a p-version finite element mesh, which means detailed features are simulated by automatically increasing the mesh polynomial order at details, instead of decreasing the mesh size.

I would consider ANSYS to be a state-of-the-art FEA package, and it utilizes standard meshing techniques at least for structural models (differences in the FLUENT vs. CFX world are more subtle). What "modern" FEA packages are you referring to, considering "Finite Element Analysis" basically requires the use of finite elements in a mesh?

Additionally, increasing mesh density in detailed features has been practice in FEA since day 1.
 
nvn said:
(1) Classical FEA... uses h-version
Isn't it the other way round? Very old software used high order (above quadratic) elements. That's nothing new. While very new software uses local mesh refinement.

(2) "Modern" FEA ... mesh-free (meshless)

I'm curious what software applications in practical use use mesh-free methods. It sounded exciting when it became popular a few years ago, but I wonder if it actually offered any advantages in common situations.

I tend to agree with other posters that "classic" doesn't have any clear definition. I've heard it used (and use it myself) to refer to older applications where you build the mesh directly, or generate it from CAD geometry and work on it from there - meaning if you make changes to the CAD geometry you have to reapply a lot of the boundary conditions etc. That's in contrast to something like ANSYS Workbench or Solidworks Simulation where you hardly have to think about the mesh at all.
 
ANSYS has two different interfaces. One is the "Classical" Ansys, that has been around for many years. WorkBench is the newer interface that is easier to use but is limited. The reason the word "classical" came about was that companies wanted to be able to differentiate between the two for bidding purposes and hiring purposes.

Thanks
Matt
 

Similar threads

  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 9 ·
Replies
9
Views
4K
Replies
2
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K