Carbon fiber in FEA 3

[nemo78]

Hi all,
I'm wondering how to model a carbon fiber laminate material with a finite element software (it doesn't matter which one, it's just a theoretical question).
It's an un-isothropic material, and I've always used isothropic ones.. let's say that the laminate is made by 6 plies, 3 are unidirectional (so they have one modulus in one direction), and 3 have fibers laying at +/- 45°, so there are 2 different modulus in 2 different directions.
How should I set the material properties?

 

[johnhors]

The answer to your question lies with your chosen FEA software, as this is really a pre-processor task. My company after looking at all the leading FEA solutions, chose to use LUSAS (

http://www.lusas.com

)as it has the most capability and a simple windows based interface in which to describe any particular lay up.

 

[nemo78]

Ok now I found a way to calculate the A,B,D matrixes, should I use those values to build the stiffness matrix?

 

[candychan]

I think ANSYS or ABAQUS would play this role.If it's ANSYS,solid46 element will allow you set the ply angle.the orthotropic material setting is not a problem in ANSYS.If you use ABAQUS,I suggest you read the example documentation,there is a very detailed example on the composite material as you concerned.

 

[jxc]

I use FEMAP & though I do not have a lot of experience with un-isothropic material, FEMAP does allow for your material input as candychan describe. I would try my manual & online help also a resources for help.

jxc

 

[annebai]

Have you looked at Strand7 ? I was using a different code until I was introduced to Strand7 for composite capabilities.
Ann

 

[mloew]

Note: The original poster, newmo78 has not logged in since 8 May 2004. It is not clear that additional answers are being sought. Likely a one thread wonder.


Best regards,

Matthew Ian Loew
"I don't grow up. In me is the small child of my early days" -- M.C. Escher

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[annebai]

Thank you for the note. I am new to this forum and still learning the tricks of the trade ...
Ann

 

[gregkress]

nemo78,

Hope that you got the answer that you were looking for. I teach composite laminate analysis for Abaris Training, so I will give you my response. Modeled as a thin plate, a composite material needs four material constants to analyze stiffness, E1, E2, v12, and G12. This is what separates composites from isotropic materials. Although a symmetric woven fabric will have an equal (almost equal) E1 and E2, but it will still not be considered isotropic because G12 will not be a function of E and v (Poisson's Ratio). As long as your FEA program allows you to enter a longitudinal and transverse modulus value, shear modulus, and poisson's ratio, you will be OK. Of course once you start talking strength, then you need to look at Failure Criteria, and that is a whole other discussion, but basically you should look at failure occuring when the fiber strain value is achieved in any fiber direction of the laminate.

Greg

 

[sgdon]

Hi Greg

Can you explain abit further on the following statement:

"Although a symmetric woven fabric will have an equal (almost equal) E1 and E2, but it will still not be considered isotropic because G12 will not be a function of E and v (Poisson's Ratio)"

We know that G12=E1/(1+2v+E1/E2). Therefore when E1=E2=E, it reduces to the well known isotropic equation: G12=E/2(1+v).

Thanks
sgdon

 

[Eap2n]

Hi sgdon,

I believe the relationship

G12=E1/(1+2v+E1/E2)

should be

Gxy=E1/(1+2v12+E1/E2)

where 1,2 are principal material coordinates and x,y are of an off-axis rotated coord system. You get this equation by substituting an off-axis angle of 45 deg into the full relationship between Gxy, E1, E2, and v12 which also includes the direction cosines m and n. Its kind of confusing, but what this is saying is that for a woven composite with E1=E2 the isotropic E,G,v relationship is recovered between the shear modulus at 45 degrees and the E and v at 0 degrees. Here is an example:

Woven graphite/epoxy props:

E1=E2=57000 MPa, v12=0.05, G12=4000 MPa

It is clear that E1/E2, v12, G12 do NOT satisfy the isotropic relationship, as they should not. However, if you calculate the shear modulus Gxy at 45 degrees you get a value of approx 27143 MPa. If you use this value for G in the isotropic relationship (along with E1/E2 and v12) you will see it is satisfied. The usefulness of this property is that it gives you a quick way to calculate the maximum shear modulus at 45 degrees for a woven composite with E1=E2.

Are you looking at Herakovich's text?

Hope this helps.
Erik

 

[sgdon]

Hi Eap2n

Thanks you so much for the explanation and i apologise for the error in the given equation.

The difficult part to is to understand why G12 has no relationship with E1/E2 and v12. Thanks for reminding me about the Herakovich's bk. I tried to search for additional explanation and it is noted that for a transversely isotropic material (isotropic in 2-3 palne), the relationship G23=E2/2(1+v23) exists. This reaffirms that for an isotropic plane, we still can make use of the isotropic relation.

I am wondering if we extend this to a WR, assuming isotropic in the 1-2 plane (for the sake of argument), doesnt the relationship hold? I guess i still have some missing links on this..

Regards
sgdon

 

[Eap2n]

Hi sgdon,

I think what is confusing you is the fact the E1 and E2 are the same for a woven composites is leading you to believe that the 1-2 plane should be isotropic. Dont let this fool you, the 1-2 plane is definitely not istropic, youve got fiber in the 1 direction and youve got fiber in the 2 direction but at any orientation between 0 and 90 degrees its a different situation. If you have some woven composite properties I would suggest you plot Ex as a function of theta, the equation can be found in any composites text (4.59 or 4.63 in Herakovich). You will see that Ex is equal to E1 at 0 degrees and equal to E2 at 90 degrees but in between it varies significantly.

Erik

 

[sgdon]

Thanks Erik,

I totally agree with you. The point we trying to understand now is why G12 has no relationship with E1 and E2 and v12, all in principle material direction with no regards to XY coordinates.

For the sake of discussion again, if we have a lamina with all fibres running in every direction, then this can be regarded as isotropic and i.e G12=G12(E) holds. The E here obviously is equivalent to E1, E2 as well as E's in other directions. If we start slowly reducing the fibres in directions other than the 0 and 90 degrees, when will that relationship becomes invalid? There must be some ways of explaining this.

sgdon

 

[Eap2n]

Hi sgdon,

"The point we trying to understand now is why G12 has no relationship with E1 and E2 and v12, all in principle material direction with no regards to XY coordinates."

The shear modulus for a plane X of any material is a function of the poisson ratio and the two direct stiffnesses at the two planes oriented at 45 and -45 degrees from plane X. Why is this? Because pure shear stress is equivalent to pure tension and compression along planes oriented 45/-45 degrees to the application of the shear stress (see fig 4.13 if you have the Herakovich text, the signs are incorrect but the idea is the same). This is why for an orthotropic material E1,E2, and v12 have nothing to do with G12, it is in fact E and v at 45 degrees which govern G12 which is why the full equation for Gxy reduces to a function of E1,E2, and v12 when you look at Gxy at 45 degrees.

Erik

 

[sgdon]

Thanks Erik. The explanation is very helpful!

sgdon