Bowing condition on large carbon fiber panel

[HPedraza]

Hello: I have a problem with a large carbon fiber panel 61" x 30". The issue is the panel has a bowing condition (from 0.150" to 0.085"). However my tolerance is 0.030" maximum.
We tried some different ways in order to reach at least the 0.030" in bowing condition.
We use a compresion tool, material of tool is aluminum.
So, We left cool down the part inside the tool at room temperatureand then demold the part. We demold with hot part. We demold at 200ºF.
About the cooling process, we tried left cooling down at room temperature, we use a chiller, even we use a big fan to cool down the tool.
No matter what, the part is bowing.
Somebody can share information about how can avoid or at least reduce this condition?
All the help will be very appreciated!
Thank you.

 

[Compositepro]

While warping can be caused by tooling or cure cycle issues, an unbalanced laminate lay-up will cause what you describe. How thick is your panel and what weave style is being used? Do you understand what a balanced lay-up is?

 

[HPedraza]

The panel thickness is 0.510". About the weave style, the information that I found is the following:
Prepreg 200 GSM Carbon Fiber CF0302 2/2 Twill.
The part has a foam core (Polyethermide).
Could you clarify the unbalanced lay up? I'm not familiar with this process, but I want to learn more (Knowledge is always welcome).
By the way, about cure, we use a recipe and the steps are as follows:
1.- 1 Minute @ 285ºF , Pressure: 100 PSI.
2.- 10 Minutes @ 285ºF, Pressure: 100 PSI.
3.- 1 Minute @ 285ºF, Pressure: 0 PSI.
4.- 1 Minute @ 285ºF, Pressure: 0 PSI.
5.- 1 Minute @ 285ºF, Pressure: 1250 PSI.
6.- 80 Minutes @ 285ºF, Pressure: 2100 PSI.
7.- 1 Minute @ 285ºF, Pressure: 100 PSI.

Then, for cooling down was at room temperature.
Thank you Compositepro.

 

[JeniaL]

0.51 thick it's a lot of plies. as compositepro said balanced lay-up is important but more important tool compensation. do you take into account spring in/back on tooling? cure cycle looks a little bit strange. you don't have any dwells. can you please attach data sheet for matrix?

p.s 2100 psi=144.78 bar our autoclave is going to blow to hell at that pressure.

 

[HPedraza]

We lay up 3 plys of carbon fiber (45º-90º-90º), then put foam as core, next step is another 3 plys of carbon fiber (90º-90º-45º). The 45º plys are facing the suface of tool.
We use a shrink coefficient, this is to compensate the spring in / spring back tool behaviour.
About the pressure, I'm not in charge of determine the recipes. So, that's why I'm not familiar with the amount of pressure required. We don't use autoclave; we are using hydraulic presses.

 

[Compositepro]

It is difficult to know where to start. Foam cores cannot take that kind of pressure so obviously you are closing to stops, at which point press pressure is meaningless.

A balanced lay-up is one where there is a mirror image of the amount and orientation of carbon fiber on each side of the part centerline. A twill-weave is supposed to be a a balanced weave but in practice it will never be perfectly balanced. A satin-weave is an example of an unbalanced weave. Carbon fabrics are woven on looms and the spacing of the warp yarns is controlled by by a non-adjustable reed (a comb like device). Yield (grams/meter) of carbon fiber varies a little from spool to spool, and lot to lot. To get the correct areal fiber weight (AFW), the pick-count (spacing between fill yarns) will be adjusted. This results in the AFW of the warp being different from the AFW of the fill. This can cause an unbalanced condition, but may not be a problem in your case.

Is your part flat? Is the bow always the same? Bowing is caused by internal stresses caused by resin cure shrinkage, and CTE (coefficient of thermal expansion) differences between fiber, resin and tooling. The ease of release between tool and part can bean important factor.

 

[HPedraza]

the part is flat, the bow is not always the same. Tomorrow we'll try a new one part with a different profile and press.
I'll let you know the results.