type of model 1

[Delogeo]

Can anybody tell me what type of model do I need for a 3D FEA analysis: a 3D surface or a 3D solid model?

 

[GBor]

It depends on the FEA software and the CAD Modeler. Generally, a 3D solid model exported in a STEP file works pretty well. I've had success with Trimmed Surface IGES files. A great deal of FEA software these days has some method of embedding the analysis within a CAD system so that you don't have to translate...it automatically does it for you.

If you would name the CAD package and the FEA software, someone may be able to give you a better idea. You may also want to post this to the FEA forum (forum727).

Garland

Garland E. Borowski, PE

 

[Drej]

Er, it depends on what you're trying to model. Is it solid or is it a surface is what I'm getting at? Remember also that 3D means a lot of things, not just the geometry (for example, 3D using 1D line elements).

etc., etc...


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[ctopher]

Delogeo,
Do you have 3D CAD software? If yes, which one?
Or are you looking into purchase?

Chris
Sr. Mechanical Designer, CAD
SolidWorks 05 SP2.0 / PDMWorks 05
ctopher's home site

FAQ559-1100
FAQ559-716

 

[delgeo]

I obtained objects made by 3D surfaces in AutoCAD 2000 and in MDT 6. Can I use these 3D objects to do FEA analysis (entering supports and forces) in order to obtain deformations and Von Misses stresses ?
I use ALGOR 13.

 

[GBor]

Algor 13 is a little aged at this point, but the answer, I'm pretty certain is, yes, if the surfaces are Trimmed Surface IGES. This is an export option from AutoCAD 2000/MDT 6. If you have MDT 6, type igesout and it should bring up the IGES export dialog box. Make sure it is set to Trimmed Surface IGES (144), not Bounded Surface (143). If you only have AutoCAD 2000, I think you had to purchase the IGES translator separately.

I believe Algor v.13 was the first version to have "FEMPRO". You should be able to open FEMPRO and open an IGES file. If FEMPRO didn't come out until later, then from SuperDraw III go to File: Import: CAD Solid Model. Set the import file type to .igs and try and pull it in.

Garland E. Borowski, PE

 

[ctopher]

I agree with GBor. We use CosmosWorks and prefer it over Algor.
(I did not intend to rhyme)

Chris
Sr. Mechanical Designer, CAD
SolidWorks 05 SP2.0 / PDMWorks 05
ctopher's home site

FAQ559-1100
FAQ559-716

 

[Drej]

Is it me or is the question very ambiguous? The poster is asking what type of model is needed for 3D FEA analysis [sic]? That's like asking "What type of connection do I need for 'this' joint?" - you can't answer the poster until you know what "the question" really is.

In reply to: "I obtained objects made by 3D surfaces in AutoCAD 2000 and in MDT 6. Can I use these 3D objects to do FEA analysis (entering supports and forces) in order to obtain deformations and Von Misses stresses ?" which is a slightly clearer question, the answer is yes, of course you can.

I believe a clearer definition of the question is needed here before any more advice is given out.

Cheers,

-- drej --


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[ctopher]

I agree, that is what I was trying to find out.

Chris
Sr. Mechanical Designer, CAD
SolidWorks 05 SP2.0 / PDMWorks 05
ctopher's home site

FAQ559-1100
FAQ559-716

 

[GBor]

I was trying to be a little gentle by giving a broad, generic response, but I'm certainly interested in a little more information...

Garland E. Borowski, PE

 

[SWComposites]

Delogeo -
You can use the 3D surfaces to build a FE model and get deformations and stresses, BUT, the real question is: will the results mean anything relative to the actual structural response? To answer this question we need a lot more information regarding what you are trying to model, how the model represents the actual structure, and for what purpose are you going to use the results.

Even when a CAD model is meshed and the FE analysis is run sucessfully, the results can still be rubbish (something not always clearly pointed out by software vendors promoting intergrated software packages). Remember, ALL FE analyses are APPROXIMATIONS of actual structural behavior.

 

[delgeo]

Delgeo=Delogeo
I am a biginner and I try to understand how does Algor works.
Thanks for your explanations.

 

[Drej]

> I am a biginner and I try to understand how does Algor works.

How does what work? You need to be a lot more specific and give a lot more information.


------------
See faq569-1083 for details on how to make best use of Eng-Tips.com

 

[GBor]

delgeo,

CAD surfaces are sometimes the start of an FE model. Each of us are trying to understand whether those surfaces represent a structure of thin plates or a mass block. The CAD represetation just gives geometry. Now you need to understand the element theories so you know whether to use that geometry in building a brick model, plates, beams...what are you trying to model. AS SWComposites pointed out, if you use beams where you need bricks, your results could be "rubbish".

Algor is a great tool, but you need to understand degrees of freedom, element types, even convergence theories to start to realize the limitations and power of FEA. This is true of any tool, not just Algor.

It sounds like you need a good course in FEA, not just some training using Algor. Even back to your original question of a 3D surface model or 3D solid model...the answer is you can use either one, but do you understand what to do with it when you get it into your FEA package?