Fiber impregnation simulation

Elmer cases by the users for the users
Post Reply
carbon91
Posts: 2
Joined: 14 Mar 2015, 17:02
Antispam: Yes

Fiber impregnation simulation

Post by carbon91 »

Hello Community,

my name is Matthias and I am currently studying in the third year Aeronautical and Mechanical Engineering. My final year project is about a development of a suitable impregnation unit to impregnate endless carbon fibers, respectively rovings (fibrous material).
The roving runs through a sinusoidal curve (impregnation cavity). By applying a tension on this roving, three different pressure zones will be created at the radii. Below the roving will be resin injected which will be pressed through the roving by the three different pressure zones.
I attached a picture of the principle impregnation process to better understand my problem.

I am still looking for an adequate simulation of the impregnation process. I would like to know the percentage of resin (fluid) in the fibrous material after the impregnation process, if the resin is completely absorbed by the resin and the tension which is increased by the roving which is sometimes in contact with the cavity wall.
Is it possible to get the information with Elmer?
I hope you guys can help me.

Best Regards

Matthias
Attachments
Impregnation sample (Miaris, 2012)
Impregnation sample (Miaris, 2012)
impregnation cavity.PNG (94.94 KiB) Viewed 7343 times
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Fiber impregnation simulation

Post by raback »

Hi Matthias,

I appreciate the thorough introduction. This is a tough problem. Has this been modeled before in literature?

I presume you would need 3D contact model + some way to deal with resin fluid. Do I understand correctly that the resin fluid may and may not be between the rowing and resin film. This would make it a little bit similar as the EHDL lubrication models. Would you model the resin as a thin film or real 3D Navier-Stokes flow problem? Is there also air or some other fluid making it a multiphase problem?

There may be some suitable building blocks in Elmer. For fluid flow the models are more or less existing unless you want to do real multiphase flow. For the contact problem we just happen to be developing some models that could be used here. However, the problem is so difficult that without any simplification it sounds more like a doctoral thesis project.

-Peter
carbon91
Posts: 2
Joined: 14 Mar 2015, 17:02
Antispam: Yes

Re: Fiber impregnation simulation

Post by carbon91 »

Hi Peter,

first thank you very much for the quick answer and your interest.
Well, the original developer made estimations by using different formulas in Matlab. They used a self developed algorithm with approximate six different equations, including:
-capillary radius (of a single fiber, a roving consists of 1000's of fibers)
-the tension build up of the roving through the geometry
-Permeability (of the fibrous material, respectively roving)
-the flow speed (resin flow speed through the roving)
-the fiber compaction (fiber compaction takes place at the pressure zones)
-Pressure (the hydrostatic pressure of the resin build up in the pressure zones)

About your suggestion for simplification, the cavity has a cross section of ca. 1.5mm x 6mm. The roving runs in the middle of the cavity and comes only in contact with the cavity wall at the pressure, respectively impregnation zones. Ideally, the injection will create a thin resin film below the roving. Ideally, the thin resin film will be completely absorbed by the roving (Depends on different process parameters like resin viscosity, process speed etc.).

An important property of well impregnated rovings is a low void fraction. This leads to a further consideration of an additional fluid (air). Low void fractions can be reduced by a further roving tension and a further fiber compaction.

So in summary, the process has probably a multiphase problem and the resin can be seen as a formulated film below the roving.
I hope this helps you for a better understanding.


Cheers

Matthias
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Fiber impregnation simulation

Post by raback »

Hi Matthias

Still sounds very difficult. I can think of various aspects to model. I.e. the fluid pressure between the film and the roving. Or a microscale study of how the the fluid will absorb into the roving. Or the equilibrium shape of the roving under influence of the fluid pressure. All these are difficult on their own right. It would require significant amount of studying to be able to come up with a realistic plan. I would consult your supervisors on the exact scope of the problem, and also the equation that should be solved. Only then it is possible to say whether Elmer is up for the task. Certainly even then it won't be ready out-of-box but may require some tweaking to deal with the specialities of your problem.

-Peter
Post Reply