Composite plate (different layer thicknesses)

[Docq]

Hi everyone,

I come here for some helps as iam a beginner with Nastran. I have not followed any trainings and youtube videos from MSC are not helping for what iam trying to do. My issue is mainly in the modeling hence i don't really use Nastran for the analysis.

I would like to model a composite plate (with 2D elements).
I attached a picture of the plate main geometry. There are 3 regions (surfaces 4,5,6). The surface 7 corresponds to the overall plate.

http://files.engineering.com/getfil...4835-8f37-81401d56628e&file=baseplate_geo.png

Defining a composites material with different layers (thickness and orientation) that i can do. However the problem is that here, the plate i need to model has not the same number of layers in the different sections. Furthermore the core material is different in each section.

I guess a picture could explain it better.

http://files.engineering.com/getfil...9bc7cbb089&file=baseplate_compositedesign.png

As i said iam only starting in Nastran but i however had some thinkings about the problem. Hence, here is my questions:
- is it possible to model the plate by using only surface 7, with composite properties (layer stiffness, layer type,...) different in each section. (Iam not really enthousiastic about the answer but who knows).
- If the plate is modelled in 3 parts, how can I 'connect' these 3 parts, knowing that some layers are placed over the overall plate when some are only defined in the side sections...

Iam not asking you to do it for me but as iam starting i might need some instructions. So please don't hesitate if you have any ideas or comments !

Regards,
Docq

 

[Docq]

Hi,

is there anybody who could help me out with my situation.
I don't have knowledge in composite material that much. I would be glad to learn from you.

BR
Docq

 

[APav]

Hi,

Yes what you describe is possible. If you are a beginner with NASTRAN let's take it back a bit: there are different levels/steps when defining laminate composites in NASTRAN:

1) individual ply/core material: this uses MAT cards, up to you to choose between isotropic (MAT1), orthotropic (MAT8), anisotropic (MAT9) behavior

2) layup (ply stacking): this is done at property level (PCOMP card). In the PCOMP you will describe the ply layup, i.e number of plies, thickness, angle, material, failure criteria...

3) actual finite elements: the component level (CQUAD4 and CTRIA3). Here you will define the element orientation (stack direction) and the reference "0° angle", i.e. the reference direction from wich ply angles are defined.


So in what you describe, you simply need to mesh your complete part continuously, then apply different properties (PCOMP) to you regions. From you drawing you middle region will have 2 plies less than the outer regions.

The difference will then be in post-treatment: NASTRAN will give you results by "ply indices". There will be a shift in indices between your regions.

If you want to avoid this shift you can use "Global plies", for wich you can specify a precise ply ID instead of an index corresponding to the stack order. But if you're beginning in NASTRAN I would recommend sticking to the easy PCOMP card.

APav

 

[stressebookllc]

Apav pretty much covered the basics. You need to learn to model plies.

Start with a simple plate and experiment. You need to define all the things Apav mentioned.

http://www.stressebook.com

Stressing Stresslessly!

 

[Docq]

Guys !

Thank you for your answer it helped me a lot ! I was missing a small thing to understand properly how to model the different sections of the baseplate. your responses surely helped me ! Again thank you !


best regards,
Docq