Laminate Pre-processing and Post-processing

[MarvinGopal]

Hi All,

I've got some questions regarding laminate modelling and post-processing.

Pre-processing

1. Does it matter in what order the layups are? I mean does it have to be according to the actual order as the laminate is physically? any pros and cons for either of them?

2. What is a Global Ply? Whats the significance of it? Does it come in handy somewhere?

Post-processing

1. The results are always shown in separate plies. The allowable that i have for the laminate is a tested data of the whole laminate and not the individual layers of the core or skin. How do i sum and compare the results (normal and shear stresses especially)? Am i suppose to just add up all the similar stresses?

2. Since the results are shown in individual plies, which ply order does it follow? Does it follow the ply Id in the defined layups? If so, I could have several different layups with different configs right?

I appreciate any help at all on this matter. BTW pardon the questions if they are too basic. I'm a new FE user and have loads to learn.

 

[jotunn]

I'll take a stab. Always interested to hear others' opinions too...especially if differing.

Pre-processing
1. Not sure I follow the question exactly, but I think what you're asking is: "If I have a [0/45/45/0] fabric layup, should I model as such, or can I model differently (e.g. [0/0/45/45] or [45/0/0/45])? The answer to that would be no...model as laid-up. In that hypothetical example, the flexural stiffness of [45/0/0/45] would be less than actual and the asymmetrical/unbalanced coupling effects of [0/0/45/45] would be reflected whereas actual would be minimal (if fabric is relatively balanced).

2. Global ply essentially assists in easing the post-processing work load. For example, if you have a model with a ply step-up (thickness increase) where plies A are full coverage and plies X are localized buildups (e.g. [A/A] turns to [A/X/X/A] turns to [X/A/X/X/A/X]) you could model plies A as global and keep better tabs on them during post-processing (by visualizing them across the whole model).

Post-Processing
1. Generally Nastran likes lamina allowables (for computing failure index or strength ratio (FI/SR)). Option 1: if you're wanting to know FI/SR directly from output, you have to give Nastran lamina allowables. There are methods to compute lamina allowbles from laminate testing (c.f.

http://jcm.sagepub.com/content/39/14/1295.abstract).

Option 2: This is just a thought...I've never done it...I don't know how acceptable it would be...but: if you have laminate allowables for your layup, you might be able to manually calculate margins by comparing the allowable to something like membrane force divided by thickness. Typing before thinking...I'd stick with option 1. I don't know how you'd account for biaxial/shear loading, etc. without a huge amount of testing. I wouldn't feel comfortable with that approach. I am curious, however, if others ever do anything different than option 1 in the same situation.

2. Yes, ply ID defined in layups is what will show up during post-processing (although, as mentioned, global plies operate differently). And yes, depending on the complexity of the structure you're modeling and how detailed the model is, you could end up with quite a few layups/properties defined.

 

[MarvinGopal]

Hey Jotunn,

Thanks for the explanation. I understood more or less point number 1 for both section (pre and post). I do have a follow up question on points number 2 though. I'm out stationed at the moment and will reply with the follow up question as soon as im back if you dont mind.

Thanks for the help.