Question about ANSYS of CF monocoque

[Anguspan]

Hi everyone, I have some problem with doing the FEA stuff with ANSYS for a monocoque CAD modle.

I really have no idea how to start.....even I have read some posts in this forum already. So if anyone can tell me more a bit, I am so thank you for the help.

I have done a monocoque CAD(in one part) in SolidEdge....
after I import the CAD into the ANSYS(WorkBench). What next I should do??

Hope someone would be kindy to help me out...But I really no idea with the ANSYS program

Cheers,
Gus

[Composites Guy]

Well... I hear Ansys Workbench is nice. I don't have it, though I program in Ansys 11 educational edition almost every day.

Try what I do... when I can't figure it out, I generally look in the help file, get frustrated... then swear that I'll never touch such a unituitive program again after I graduate!

What you are trying to do is not a single weekend project. To learn the program, the composites theory, and begin to optimise your chassis could take many months or even years.

For a CF monocoque I assume you have skins and a core. The first thing would be to define elements of use to you. Probably your monocoque will be thin and you can reduce your CAD model to single surfaces which well-represent the geometry. In this case you might get away away with modelling the entire thing as shell elements. I sugest the element SHELL91. If you need to model solid elements in some places (to represent core material) I suggest the element SOLID95.

You will also need material properties. You may need more than just a Young's Modulus and a poisson ratio because the fabrics have orthotropic (not isotropic) properties. Depending on how fancy you want your layup/analysis to get, these properties could be pretty difficult to understand if you don't know any composites theory. I suggest picking up a composites textbook such as the one by Robert Jones. You might be able to use a quasi-isotropic layup to simplify your analysis (equal amounts of fibers running in the 0,+45,90 and -45 degree directions.) You could do a journal search for carbon fiber and core properties to save yourself some time in testing.

The SHELL91 elements can represent skin-core- and-skin all in one infinitely thin plate. You specify the layers, material property orientations, and layer thicknesses through the use of real constants.

So... you import your geometry, specify material properties, element types, and then apply the various layup schemes with a different real constant for each unique layup area. These real constants are attributes attached to the areas of the cad model. Once you reach this point you are ready to mesh. I suggest a quadrilateral (4 sided) free mesh, with refinement where you think its needed.

Next you apply loads and constraints. There are many ways to do this, and you can search the other threads here.

Then you solve.

Then you look at the stresses and deflections.

Good luck.

[exFSAE]

I'd say Composites Guy got it pretty good..

Analysis of a complex geometry composite layup is, to put it mildly, non-trivial. I would be very wary of it, or the results, if your understanding of composite mechanics and layup isn't spot on.

Or likewise if your fabrication process isn't pretty good. Designing it is one thing. Making it happen is another!

[Gareth]

You should get a copy of HyperWorks from Altair (free for FSAE teams). HyperMesh has very good tools for building composite models (with lots of tutorials) and OptiStruct can help you do optimization on the layup (number of plies, ply areas, core thickness, ply angles etc). You definitely need to know your composite and finite element theory to use either effectively, but you're in school to learn... You may even be able to get training for free if you're near one of the offices.

Ansys is a PITA and you either need to be a very patient expert to use Prep7/Classic or be building mickey mouse models in WorkBench. Be careful using any FE package that hides the mesh from you...it's the foundation of the entire analysis. Why would you want it hidden?

OptiStruct has been used extensively on the development of the 787...if you're looking for validation. Altair is the industry leader in finite element based optimization.

[Mully]

Our chassis for the last few years has been analysed using Altair Hyperworks and now that I'm the one doing the analysis, I think it is a fanatastic bit of software.
Tip, HyperLaminate is your friend

[blister]

What do the experts think about Abaqus? I think it`s much more user friendly than Ansys.
I worked with Ansys, but i never felt in love with it.

[ad]

This is probably not the forum for asking directions on how to use ANSYS or any other FEA program, however, some of the suggestions so far have been pretty good introductions.

I have played around with dedicated pre and post solver program called LaminateTools. There are others on the market, however they are much simpler and less time demanding than ANSYS or other meshing when it comes to composites.

In the program you define both the structural properties and the manufacturing properties (eg ply draping etc, it can also tell you the effect of fibre orientation warp). You generate a mesh for each of the subsequent layers / plys.

Export this to your choice of solver (i.e. ANSYS) The tutorials even tell you how to solve it using ANSYS.

Then once the solution has converged, load the ANSYS results in and read off the results for each layer.

Simple. Plus there is an optimisation tool that can 'estimate' effects of adjusting principle fibre orientations before you resolve.

[Anguspan]

Thx a lot everyone^^
It would be very helpful for me
Thx again^^

[TG]

Hey Composites Guy (or perhaps ad),

any tips on setting up the principal axis for complex shapes in ANSYS? I've been trying to find this (help files, manuals, tutorials, etc.) but have failed to find an answer.

[Composites Guy]

I'm not sure what you are referring to by "principle axis".... please describe.

[Enghamed]

Hi to all
I saw this demo on ANSYS demo room web site which was modeling composite material in workbench.
I saw that in the material section he had a dialog box which he was defining layer orintation. Does any one know how can I do this?
Any tips on defining composite layered materials in work bench ould be help ful.
regards

[Enghamed]

where can i download the free hyperworks?
thx

[Enghamed]

Dear Andrew
Can yu please help on where can i find the laminate tools that are compatible with ansys?
thanks a lot
Hamed

Originally posted by ad:
This is probably not the forum for asking directions on how to use ANSYS or any other FEA program, however, some of the suggestions so far have been pretty good introductions.

I have played around with dedicated pre and post solver program called LaminateTools. There are others on the market, however they are much simpler and less time demanding than ANSYS or other meshing when it comes to composites.

In the program you define both the structural properties and the manufacturing properties (eg ply draping etc, it can also tell you the effect of fibre orientation warp). You generate a mesh for each of the subsequent layers / plys.

Export this to your choice of solver (i.e. ANSYS) The tutorials even tell you how to solve it using ANSYS.

Then once the solution has converged, load the ANSYS results in and read off the results for each layer.

Simple. Plus there is an optimisation tool that can 'estimate' effects of adjusting principle fibre orientations before you resolve.

[Jerry C]

Great info here, definitely helping me out with my project. Thanks composites guy!