spring back composite artifacts

[Nicolas]

Good morning to all,
I have to face the definition of a forming equipment for carbon laminates and I come to put the following problem.
being the thickness of the high artifact (about 15mm) occurs, after the care, a contraction of the fibers such as to cause a form error known as "spring-back" (positive) variable according to the size characteristics of the piece.
The practice of equipment compensation (negative spring-back) is of common use on plastics, but there is nothing about the behaviour of carbon artifacts.
I ask, please, if you are aware of the phenomenon and where to find bibliography?

thanks early.

 

[Onda]

generally the phenomenon is due to a lack of laminate symmetry. non-symmetrical fibers compared to the average geometric axis of thickness have a different expansion (thermal or resin withdrawal) and the artifact deforms.
Is your 15mm laminate all carbon or is it a sandwich?
Is it impregnated or wet laminate?

in any case are not aware of specific literature on how to bypass the problem, that is to change the shape of the mold to get the finished piece with the desired shape.
what I have always read is on the way to minimize the deformations of the finished piece, using symmetric laminates, low shrink resins, and special shortcomings during production.

wave

 

[Nicolas]

generally the phenomenon is due to a lack of laminate symmetry. non-symmetrical fibers compared to the average geometric axis of thickness have a different expansion (thermal or resin withdrawal) and the artifact deforms.
Is your 15mm laminate all carbon or is it a sandwich?
Is it impregnated or wet laminate?

in any case are not aware of specific literature on how to bypass the problem, that is to change the shape of the mold to get the finished piece with the desired shape.
what I have always read is on the way to minimize the deformations of the finished piece, using symmetric laminates, low shrink resins, and special shortcomings during production.

wave

Good morning wave,
from further questions to the customer, I confirm that there is a symmetric lamination on a carbon solid (not sandwich). the material is a certified epoxy prepreg and therefore not derogable.
from previous experiences I have noticed that the phenomenon is always present but its size (visibility) is function of the net thickness of the skins.
thanks:wink:

 

[numero1]

http://www.acciaiterni.com/db/eng/docpubblicazioni/301 l.pdfhere is quoted something, if it can serve......

 

[MacGG]

Hello nicolas,

My experience to solve these problems is to adopt simulation techniques to predict the return of the material according to the manufacturing technique.

you need fe technologies with non-linear solutor, both for the modeling of the training process, and for the characterization of the used materials.
therefore take into account the costs of the software and any training necessary to adopt such techniques.

Otherwise, you go to attempts, trying to make pieces and seeing what happens. In this way, with some experience and if the costs allow it, it is possible to reach the desired dimensions after a few iterations (3-4). ovviamante this is possible if you then use modest series that shock the initial waste (and if the forms are not particularly complex for which experience helps).

Hi.

♪

 

[Nicolas]

Hello nicolas,

My experience to solve these problems is to adopt simulation techniques to predict the return of the material according to the manufacturing technique.

you need fe technologies with non-linear solutor, both for the modeling of the training process, and for the characterization of the used materials.
therefore take into account the costs of the software and any training necessary to adopt such techniques.

Otherwise, you go to attempts, trying to make pieces and seeing what happens. In this way, with some experience and if the costs allow it, it is possible to reach the desired dimensions after a few iterations (3-4). ovviamante this is possible if you then use modest series that shock the initial waste (and if the forms are not particularly complex for which experience helps).

Hi.

♪

hi macgg/number1,
Thank you for your suggestions. The only solution is that of macgg, but I think the customer is more willing to lose "some" money rather than wait for training. :biggrin:
bye:

 

[Onda]

15 mm of laminate non sandwich in pre-impregnated are many, when I happened to make similar pieces I also had spring back problems that I never completely solved.
one of the techniques used to eliminate them is to laminate in different steps and do more care in order to consolidate the material. in this way you minimize the withdrawal but you do not eliminate. laminate must be packaged to present not only global symmetries but also for each single package.
Moreover with this system at every lamination recovery you get a secondary bonding between the skins that lowers the interlaminating cutting resistance, so it is not always a system that can be adopted.

I have noticed that sometimes the problem is due to the use of too reduced bent rays, especially on female molds where the radius is reduced to increasing thickness.

I don't know how to tell you, I should see the piece to think about it.

If you find literature about it, I'd like you to be a part of it because I'm interested in the subject. Thank you.
wave

 

[Nicolas]

hello wave,
It can be seen that there are two phenomena that could cause such defects:
- contractions induced by thermal gradients on fibres;
- gradient of the matrix in thickness.
I saw a fem analysis with the only thermal part on a "c" section. simulated values were 70% of the real contraction, but this is valid for a simple section. for more complex pieces we have to try.
I also noticed the same problem on curved rays, but also on male mold (though reduced).
the problem (not totally resolute) I think there are. It is necessary to have time to carry out the necessary studies/considerations, (testing samples, validations, etc.) inherent in the composite.
to every good account I will keep you updated.

good work:finger: