Simulation of the Tension Strength Test Composites matierial

[madmonkey]

HI ALL
i am trying to simulate in ANSYS the Tension Strength Test (TST).The TST has been already done. i got the values(grapf stress-strain).
Now im traing to create the simulation of the TST in ANSYS. I got all the material values:

fiber E=73GPa, V(volume)fiber=0.32 Poisson=0.25
matrix e=3Gpa Poisson=0.316

number of layers 14 (i use 29 layers first and last is half height of matrix, another 13 of matrix in full heihgt and 14 of fibers)
i got the dimension of the sample (whole and the middle secion).
i'm using the element SOLID46
but the solution is different as a TST.

test value for the composite is around 16GPa
but in Ansys it is only 3GPa

any tips where i make the mistake?

thanks for all answers

 

[tstanley]

I was hoping there would be some discussion around this thread. It seems to me that being able to simulate the test is fundamental to simulating the structure as a whole that is planned to be made.

I have a number of questions. Is your material unidirectional or woven? Are you using micromechanics theory to come up with the properties of the combined material, fiber and matrix? Have you got your material oriented properly with the elements? Your fiber volume looks kind of low at 0.32 for a good quality laminate.

Why did you use solid elements as opposed to 2d laminate elements?

How did your displacement values compare with the real world test?

Tom Stanley

 

[madmonkey]

Hi tstanley

i have already done the "research". but lets start with answer of your questions.

material was woven, but there is "no" different. cause sample look like:
|`````\_____/``````|
|
|_____/`````\______|
and the middle fibers what are perpendicular to tensile force has almost none signification for tensile strength.
(book mechanics of composite material - Robert M. Jones, the best book what I have read yet)

I use macromechanical behavior of lamina, rule mixure theory
Efiber + Ematrix = ? E of lamina? or laminate

elements:
yes i have right orientated element

solid elements, cause my aim was to apply it on 3d part after the tensile test simulation.

and the solution?
I have compare real solution with ansys, and I find out that the volume fraction, from creator of the composite was wrong(he gave me wrong numbers). And after I correct the volume fraction, the test and simulation were the same.
the stress-strain curves were the same.
If you want I can put here the link, to look for it in my thesis.

Vladimir Zachar
just finish with my school, what now? )

 

[tstanley]

I like Jones' text as well. I've been reading a number of papers on the effects that the various types of weave have on the properties of the composite. It does seem to have some effect in reducing the strength in the direction of the fiber because the fiber undulates. There are a number of papers on the topic in the "Journal of Composite Materials" available from Sage Publications on the web.

I would be interested in your thesis if you could put a link to it here.

It is good to here that your problem was solved and that the analysis now replicates the physical test.

Tom Stanley

 

[madmonkey]

hi Stanley

the only problem with the thesis is the language, it is in Slovak language,now im preparing for final state examination(B.A.)-17th of june 2005
i was not having time to translate it into English. may be during the next few weeks i will have the time for it.
but maybe you will understand it. (at least i will try to translate the TST, comparation and solution)

http://www.madmonkey.ahoj.sk/dpsvk.pdf

(my home page is

http://www.madmonkey.ahoj.sk)

but basically the first few pages(capters) are from books about composites(pictures from mechanics of composites - jones)
then there is a verification example(ansys),
and then the TST. comparation with ansys and exact solution.
and last capter is the worst, cause I was in hurry, i spend too much time with comparing the results between real test and ansys. (cause of the wrong input data from the company and how to fix the sample in ansys)
im little bit angry with my self and little bit
disappointed, cause i was thinking about the bending test of composites. but at the end we havent time to look on it, we have measure it i got the data.......
and last thing is, the fatigue of composites, but at my university, we havent measure it yet.

ok stanley, any question, feel free to ask me.

what do you do with composites? what for software do you use?
Vladimir Zachar
(im studing mechanical engineering-almost done-only the final state examination) wish me luck )

 

[tstanley]

Hi:

I took a look at your thesis but will have to study it more closely. I'm interested in what you have to say about anisotropic materials. I can understand the graphs and material properties but will have to study the rest of it. I live in a multicultural area of Canada so may be able to get some help with the language. Other than Jones, I also use "Mechanics of Composite Materials" by Kaw. It is similar to Jones, but more modern and explains some things better (not necessarily anisotropic materials though).

I am interested in the simulation of composites because of powerboat racing. See

http://www.f1boat.com/index.shtml

.
These boats use composites to construct a safety cell around the driver to protect him during a crash. They are sandwich
panels made from Carbon fiber, Kevlar and glass fiber with an Airex or balsa core. A variety of woven and unidirectional
fabrics are used. Each boat builder has to have his layup certified and he submits a sample that is tested with a four bar bend test. I have been working with the associations and have obtained the results of these tests and have simulated the bend test. Once the test can be accurately simulated then a model of the safety cell can be made so that the actual failure loads can be predicted and that can be correlated with the loads expected in a crash. This way the standards can be set to better protect the drivers. This is a labour of love and not a way to earn money. On the other hand the process can be used in other applications.

I use MSC Nastran for Windows for my analysis. My web site has some of my work on it, see

http://members.axion.net/~tstanley/index.htm

.

Good luck with your final exam.

Tom Stanley

 

[Geraud]

Hi,

Actually I am trying to solve the same problem as yours. So I had a very interested look at your research report and more especially at the Ansys simulation part.

To begin with I tryied your example.

My problem is that I always obtain 75 GPa (??) for the Young's Modulus, that is close to the fibre modulus value.

Would you tell me what Ansys postproc. options did you use (element result at the centre of the specimen for instance ?) to obtain the good value ?

Thank you in advance,

Géraud

 

[madmonkey]

75 is exact fiber Young's Modul
but for result you have to use the Mixure teory.
that means

Elaminate = Efiber*Vfiber+Ematrix*Vmatrix
73 * 0.171 + 3*0.829 = 73 GPA

i havent find any other output.
i hope this has help you....

 

[clocklight]

Mixture theory is not good for the estimation of properties of composite materials.

Try Mori-Tanaka or Self consistent method to calculation the Young's Modulus

It should be lower than the value from Mixture theory