ALGOR for Composites

[bobbywvu]

All,
I was wondering if anybody can give me a review about using ALGOR FEA analysis for composites.

Thanks in advance!

 

[GBor]

I used it for several years, but have only played with it for the last few. What is your application?

Algor has a thin and thick composite element. Thin is a layered plate to assign each ply of a laminate. It does ply orientations and individual lamina properties. Thick includes a designation for a core for one of the layers.

Global and local axis orientation isn't difficult once you understand the nature of your laminate and methods of orientation.

Generally, works similarly to that of several other FEA codes and provides similar answers in linear static. Non-linear composites are "iffy" at best in any processor in my opinion. Failure modes are poorly understood and there is no single perfect failure analysis method.

If you are doing continuous fiber reinforced composites, it is as good as any general purpose FEA. If you are building short-fiber or even non-continuous long-fiber, you probably need to stick with orthotropic plates and "smeared" properties.

 

[bobbywvu]

GBor,
Thanks for your reply. I am interested in doing some linear static analysis as of now. I would be using continuous reinforced composites.
My primary concerns are:
1. Layer wise analysis
2. Ease in ply orientations
3. Meshing capabilities
4. Various options to choose failure criteria.
Although I can get this info from ALGOR website, I would like to a review from a user.

Thanks again for your help.

Bhyrav

 

[GBor]

Layer inputs and material selections for each layer as well as layer by layer orientation is pretty easy. You can orient each layer relative to some global orientation and orient the global orientation relative to the global axes (for instance, by part may be aligned at 45 degrees to the x-axis, but then individual layers within the laminate may be based off of this). For a part not in a global plane, you can still orient the laminate with the global axes and the software will project the orientation onto the part.

Meshing is, again, pretty simple whether you are automeshing or manually meshing. Defining the part orientation even in an automeshed part is pretty simple. The biggest issue with auto-meshing (composites or not) is in the corners if you are mid-planing an imported solid model, but that has been dramatically improved over the past few years.

As I said earlier and will continue to say...failure criteria, while easy to choose from, are unreliable, but that is not a limitation of Algor, that is a limitation of the science of failure analysis in composites. How do you define failure in your particular situation...is it a certain elongation percent of the matrix or of the fiber? First ply failure theory? Are you using a brittle or ductile matrix? What about the interphase between the fiber and matrix? What are the bonding characteristics for joints in your composite? The list goes on and are EXTREMELY difficult to characterize with extensive testing. Once you have the test data, it may be difficult to determine a failure method that fits. You generally revert back to looking at directional strains either in each layer, or in the global laminate.

 

[bobbywvu]

Thanks Geoff. I would give it a try with ALGOR. Basing on your reviews, I feel more comfortable.