Surface quality (voids) in vacuum infused products

Currently I am researching vacuum infusion for my final thesis in Aerospace Engineering at the Delft University of Technology in the Netherlands. One of the problems with this production technique is the voids that can occur inside- and on the surface of the final laminate. The void content has to be minimised as voids are weaknesses in the material and if they are located on the surface they form pinholes which will require extra processing steps to get rid off.

Possible causes for these void formations are variations in permeability on a filament and filament bundle scale, poor outgassing of dissolved gas in the resin, evaporation of volatile components in the resin, shrinkage of the resin and leakage in connections and mold.

I am looking at possible solutions to minimise the void content in a relatively thick (5-6mm) carbon/epoxy laminate. I am thinking of doing tests with increased temperatures (injection temps, curing temps, mold temps), different outgassing techniques, different supplemental layers (breathers/bleeders), various vacuum pressures and a number of different release agents.

Does anybody know any other possible causes, have any insights which may be able to help my research, or have any practical solutions for the reduction of voids? I am grateful for any information or tips.

Regards,
Jan Koeleman

My basic test set-up has a U-type shape: |__| and is approximately 80 cm long. The resin is an epoxy resin (Rutapox 4908) and the fibres are Carbon.
The bottom is approx. 25 cm. across and slightly curved; the centreline is about two centimeters above the edges of the bottom.
The sides are approx. 7 cm in height.
The corners where the side and bottom meet are not at right angles but have a small radius.

The fibre lay-up is as follows:
1(0/90)
3(UD)
1(+/- 45)
4(UD)
1(+/- 45)
3(UD)
1(0/90)

The orientation of the fibre laminate is such that the UD fibres lay parallel to the centreline of the mold bottom.

To aid infusion, a runner (a metal spiral), on top of a 4 cm. wide piece of Diatex runnermaterial, is used. It is placed on the laminate, parallel to- and at the location of- the centre line of the mold bottom and across the length of the mold. A perforated release film is placed between the runnermaterial and the laminate.

The resin input is located at one end of the mold in connection with the runner and runnermaterial. Two vacuum ports are located on the other side of the mold: on both sides of the mold at the edges of the laminate.

I have pictures of small voids (pinholes) and large voids however I can’t upload them and I don’t know how to place them on a public server. I can of course mail them directly.

Thanks again for any tips.

jkoeleman,
Do you leak test the laminate before you shoot it? Also you might want use some heat on it, I take my laminates up to 150 (f) this helps pull some of the air out and makes for a tighter laminate.

As for the leak test; I make sure that there is absolutely no leak. I have noticed that even the smallest leak can lead to large voids (I’ve already blown a test by thinking that the leak was ‘small enough’). As my test set-up is reasonably simple I can easily check for leaks.

Where I learned vacuum infusion; the area I am specialising in for my studies (Aerospace Engineering) is the design and production of composite parts. Vacuum infusion is a promising production method but cannot be used at present for commericial aviation because the quality of the components that are produced fluctuates too much. This necessitates a lot of testing, which is needless to say an expensive and time consuming process. Basically I am investigating for my final thesis whether it is possible to better control the process thereby making it possible to ‘predict’ (and thus ‘guarantee’) a certain quality level of the part.

My longest resin flow from resin inlet to the vacuum outlet (I presume this is what you’re asking) is about 80 cm. But because I use a runner across the entire mold, the longest flow of the resin through the laminate (perpendicular to the runner) is about 20 cm. You presumed correctly that I don’t use a flow medium throughout the entire laminate (neither on top nor in the laminate). I have put it on my list of experiments.

The vacuum I infuse with is 100 mbar (90% vacuum) and my laminate is only carbon (I only use a strip of surface medium below the metal spiral runner). I’m going to check out the continuous strand fabric, sounds like a possibility. I use a leak detector to check for leaks (doesn’t always work but if I size the vacuum foil right, not too many folds, then it usually does). I also use a digital vacuum pressure meter to measure what the vacuum pressure does (i connect the mold directly to the pump so that the resin trap doesn’t act as a buffer).

I haven’t figure out the cps of my epoxy resin. As you say it also depends on the temp, I don’t wait to long before injection because the exothermic reaction of the resin may cause overheating. It’s a bit of a balance I guess. I don’t have an infrared temp gun to measure temps and a problem I have is controlling the temp. I don’t really have a method of keeping the temp constant (i’m using electric blankets to heat my mold up to 37 deg C but this is a quite unprofesional).

As for the resin, it is a common epoxy resin (i believe it is also used for the vacuum infusion of yachts) but I don’t know whether it’s Russian. I’ll find out later.

So to sum it all up, I think 'm going to mess about with the temp, the mould release agent, the flow media, and the vacuum pressure.

Any other ideas? In any case, thanks for all the tips.

Void is the result of not pulling a full vacuum. Without an ultrasonic leak detector, the best thing you can do is clamp off the vacuum line and leave it for about 10-20 mins. If you haven’t lost more than 1" of mercury in that time then you’re ok. Everyone I have talked to could not tell the difference in strength between a vacuum infused part vs. a prepreg part.

Evan