carbon fibre analysis 2

[T3ster]

Hello,
I have to do a finite element analysis with ANSYS of a commercial carbon fibre tube. If have no information about the different layers of the tube.
All material properties I have are the following:

I assume that the material data I have to define in ansys is:
EX=2.5*10^6
EY=8.7*10^6
GXY=0.8*10^6
PRXY=0.43

Am I right?

But with these propertys I will only be able to do a 2-D analysis with plane elements. Will it provide me satisfactory results?
What data may I use, if I want to do an analysis with shell elements or volume elements?

I would be deeply grateful if you could give me any hints.

 

[rb1957]

"If have no information about the different layers of the tube" ... then there is really very little you can do. The laminate properties depend Very strongly on ply directions.
of course, you can make a whole mess of assumptions, but then you know what "assume" does ...

Quando Omni Flunkus Moritati

 

[T3ster]

I just need to know if the tube withstands the loads without failing. Is this possible with these material properties?

 

[rb1957]

i guess you could assume (see that's the word i was talking about) that all the fibers are oriented so that the minimum in-plane modulus (what i think your table calls "flexural modulus, long" but flexural modulus has it's own technical meaning (and not the same as E) so maybe it's applicable, maybe not. this sort of looks like your suggestion (assuming that X is axial)

i think you can use this material for planar elements, which includes shells and plates, and also 3D (solid) elements ... why not?

Quando Omni Flunkus Moritati

 

[T3ster]

Thanks for your help!

And what parameters can I define for EZ, GYZ, GZX, PRYZ, PRXY if I use 3D elements?

 

[rb1957]

good question, hadn't thought about that ...

if this is "real work" i'd suggest calibrating your model by testing. deflections are the easiest thing to measure.

Quando Omni Flunkus Moritati

 

[T3ster]

It must be possible to use the material properties from the manufacturer. Why does he provide these data if nobody can use it?

 

[rothers]

Use a cylindrical co-ordinate system

 

[rb1957]

in most cases the through-thickness properties are dependent on the resin/matrix (what's the out-of-plane property of a planar laminate? ideally, zero)

Quando Omni Flunkus Moritati

 

[T3ster]

@rothers
Could you explain that more precisely?

 

[rickfischer51]

I think rb1957's comment about the through thickness properties is due to the fact that there are usually no fibers in the radial direction.

Why use solid elements to model a tube? Unless its a thick wall, use a multi layer shell like SHELL281 and set up your ply data with the SECDATA cards. Also, I think you're need some data for the matrix. Also also, you might want to read up abit on composite materials. I took a class on plastics and we used Mechanical Properties of Solid Polymers by Ward and Sweeney. See chapter 8.

Rick Fischer
Principal Engineer
Argonne National Laboratory