Composite FEA 2

[Spirit]

Hi pals.
In my job I have often to cope with composite configuration for which the usual modelling with 2-D CLT models are pushing the limit... as far as you know, is there some commercial software or modelling technique allowing to model composite structures with their full layup, for example with solid elements? Your help would be appreciated, thanks in advance.


Cheers


Spirit


'Ability is 10% inspiration and 90% perspiration.'

 

[prashant77]

U can use ABAQUS for that this is very compatible with non linear analysis like composite analysis

 

[orangepeel]

Spirit,

Ansys, MSC.Nastran, NENastran, MARC, and IDEAS also have composite plate elements that one can use. Most FEA packages will model laminated materials..... although if you use MATLAB, I don't think that FEMLab does.

A number of the FEA companies have 300 node demo packages that you can obtain and try out.

take care,

LDP

 

[RPstress]

Interestingly, I'm not aware of any really good laminate modelling software. The usual laminated plate elements in general FE codes work ok, but accurately modelling a curved flange, for instance, needs a LOT of elements. Also, getting good through-thickness forces is a real problem. Laminated plate elements will generally recover interlaminar shears, but more sophistication is usually beyond them, by their very nature.

We could do with some good support for creating and post-processing laminates made from solid elements. You can, of course, model this explicitly, but it's a LOT of work, both creating the model and post-processing it.

Using a general FE code is arguably over-kill for many problems, which are more complicated than a CLT plate analysis allows for, but do not necessitate the analysis of an entire structure. Also, repeatedly modding an FE model when sizing is not an efficient way to do things. At the moment we tend to analyse such problems with an ad-hoc set of Mathcad sheets and spreadsheets (with VB macros) created individually for each problem. This is hardly ideal, but does give us the needed flexibility. There is a reasonable suite of such halfway house software in ESDU, but it doesn't tie together very well.

Perhaps there's someone out there who's working on a nice little program which will do edge effects, through-thickness tension and shear around curved laminates, stress concentrations at holes, ply-drops and transitions to honeycomb from monolithic laminates, with some stability calcs...maybe it already exists. Any takers?

In the mean time, small FE models with lots of detail, running several ESDU programs, or writing in-house spreadsheets, etc., seem to be the main options. Of these, ESDU and most FE require serious money (though increasingly reliable FE on the cheap is becoming available), and writing your own stuff takes time (=money) to set up.

PS: Spirit, in what way were you finding the 2D CLT inadequate? Just curious.

 

[CSrac]

Maybe you have heard of Fibersim, sold by the Vistagy Co. (Vistagy.com). I have used it for the past 5 years at Raytheon Aircraft. I use as a Manufacturing tool , but it has great analysis and documentation capability. It must run with a 3d software such as Pro E , Catia or Unigraphics.

 

[HeinBau]

ANSYS finite element package has solid composite elements over and above the normal laminated shell elements.

Regards