Hi, i think i have an all-singing, all dancing excel file somewhere which seems to do just about anythin you want with composites, will try to dig it out for you, was a pretty big file if i remember.
Anyway, have a look at NASA ref pub 1351 (Basic mechanic of laminated plates). Go to the NASA tech records online and type in the 1351.
It outlines the method to calculate the stress-strain behaviour of a laminate as a whole, but it also then goes onto explaining how you can obtain the stresses and strains for each ply. Basically you calculate the strain in each ply and transform them into strains in the principle material directions, which you in turn relate back to stresses.
I sent materials-sciences an email saying that the calculator is broken. If I browse to it and see it is up and running I will submit another post. If anybody else notices please do the same.
Robert Stupplebeen
Just been bounced back to me. I will try to see if theres anything i can do to it to reduce its size tomorrow at work. (uk time so 12 hours).
I will send you a small file that you might also find useful.
40818, the file is very useful but i need something about the interlaminar stresses. can you try to send the other files to this address rveroll@alice.it?
Will do when i get to work tomorrow and see what i can do to get your email to accept the file size. Have you looked at the NASA report? It does tell you how to do what you want?
Classical laminate theory doesn't directly address through-thickness stresses/strains, and that includes interlaminar shears. The NASA document appears to address CLT only, and is no immediate use for through-thickness stresses.
{Note on RP1351:
RP1351 also has a couple of errors. One is unimportant and is in the bit where the author solves the ABD matrix using determinants, which nobody now uses since fully inverting a 6x6 is trivial these days. However, the other appears relatively serious. RP 1351 P. 9 says that
[Qbar] = [T'][Q][R][T] ([T'] is the inverse of [T])
whereas I think it should be
[Qbar] = [T'][Q][R][T][R']
or, alternatively,
[Qbar] = [T'][Q][T'T] ([T'T] is the transpose of [T'])
This has to do with keeping consistency between tensor and engineering shear strains, and it's possible that the author (Nettles) has compensated elsewhere, but as far as I'm aware coding something up using the RP1351 equations will give the wrong answers.
}
The through-thickness stresses and strains due to *in-plane and hygrothermal* loads peak at the laminate edges and drop off fairly sharply with distance in from the edge, typically trending to almost zero at about 1.0*thickness in for shear and less than that for direct 3-direction stress. (See Datoo's book for a reasonable description and analysis of edge-effects.) NB: codes such as NASTRAN that work with layered elements typically cannot calculate these stresses. It's necessary to model the layers explicitly to recover them.
The through-thickness shears due to *applied shear loads* are distributed through the thickness a bit differently from in a non-layered plate or beam.
NASTRAN and other codes with interlaminar shear recovery in layered elements will calculate interlaminar shear stresses due to applied mechanical shear loads correctly, and FE of a unit plate is often the simplest way to go. I don't currently have a spreadsheet-style solution available, though I hope to recover an old McDonnell Douglas one next week.
Essentally you take the plate-axis (X-direction, not 1-direction) endload in each ply due to the bending from the moment caused by the shear load and add up the endloads between the plate/beam extreme fibre and the point of interest through the thickness. This is exactly what the the classical S.A.y_bar/I aka V.Q/(I.b) formulas used for isotropic materials do. To put it another way, you have to do a ply-by-ply analysis of a plate with a unit bending load on it to find the plate-axis direction endloads in each ply then add them up. So, classical laminate theory and NASA RP1351 are appropriate as a means to find these endloads (if TP1351 isn't as wrong as I think it may be).
I'll update with the appropriate formulae when I can.
Alyakir, yes, your Matlab code would definitely be of interest. Perhaps you could e-mail it to
evpuneq.creel@hxrf.nrebfcnpr.txacyp.pbz
You have to perform a Rot13 on this to "de-munge" it. If you don't have Rot13 routine than you can paste it into the box at
and click the cypher button to get my actual e-address. (The people who run Eng-Tips are a bit sensitive about having actual e-mail addresses displayed, as spammers and other miscreants tend to misuse them.)
If you're wanting actual 3-direction stress then it depends on how it's being created. If it is due to free edge effects from in-plane loads or temperature, then M. Datoo's book is a reasonable reference but does not quantify it accurately. (I can e-mail you a brief extract if you like.) While there are a lot of papers analysing the problem on the web, the only more accurate practical method that I'm aware of is to build a detailed finite element model with brick elements with at least one element through the thickness of each ply.
If you want the Z stress for some applied through-thickness tension or compression, then how this loading is applied determines how the Z stress is distributed. This is not so much of a composites-specific problem as a general stressing problem. You will need to supply a few details of your loading and structure to get some help.
hello,
i just got a simillar problem to the ones presented. i need a module to calculate interlaminar stress in composites.
i am using a beam element in a FEM model that i develloped in MATLAB.
can you send me any matlab file that you might have for this?
thanks!
