Infusion with core pics

General Composites Discussions Composites Talk
#1

Showing you how infusion is done when using core materials.

In this case, the part is approximately 48"x48" and consists of carbon and glass plies. Layup schedule was basically:
Carbon plies, 1 stitched glass ply, 4 small aluminum plates for mounting points, core, 1 glass ply, carbon plies. Then peel ply, and flow media.

It is critical to work the material properly onto the lay-up tool. Do not depend on the vacuum bag to consolidate the ply stack.
The resin feed line was placed along 1 side, and a vacuum line centered on the opposite side. Note: feed line is lower than vacuum outlet. Resin source always needs to be lower than pump, and our pumps are located on the second level with drop lines throughout the clean room.

The whole things is bagged using a good sealant tape and bagging film. It is critical to ensure there is enough slack in the bag to accomodate the shape of the tooling and any raised/recessed portions. Note: “ears” in bagging film create excess bag to follow the shape of the core material and in the angles of the tooling to prevent bridging.

Once bagged, the vacuum line is open slightly, and as the bag draws down slowly, the dry lay-up and vaccum bag are worked into position and around all the details of the core material. Polyethylene paddles are used to assit in working the upper plies into the chamfer and radii of the core, through the bag.
We check vacuum, and measure over 29". Then a leak test is performed by closing the vacuum line, and the gauge must not drop more than 1" in 5 minutes. This ensures there is no ingress of air, which will DESTROY an infusion.
Vacuum is held for 4 hours to evacuate any moisture and remaining air trapped in the lay-up. We often leave parts bagged for over 24 hours while other lay-ups are done so we can infuse numerous tools simultaneously. Another leak test is performed, and ready to infuse!

The epoxy resin is measured with accurate scales, and mixed very carefully in a clean container. (It is critical to get the ratio exact, not “close enough” like cheap polyester resin can tolerate.) Then transferred into another container and mixed some more to ensure there are no unmixed part A or part B ingredients anywhere on the container, which will not cure and leave weak/soft spots in the component.

The end of the resin feed line has a slight cut made on the side to prevent the line from sucking onto the bottom of the container. A simple mixing stick is zip tied to the end of the line to keep it straight and always reaching the bottom. The resin feed line is unclamped and away we go with the infusion!

Starting time for the infusion was 6:58, by 7:08 the infusion was complete and the feed line clamped shut. Yup, only 10 minutes to infuse something of this size. Notice how little resin is left, but it’s imperative to never let the resin run out or allow air into the line. This will destroy the infusion and produce many voids in the components. After making over one hundred of these, the resin consumption is kept within +/- 1% and we can precisely mix enough resin without wasting more than a few ounces.

#2

Rest of infusion pics

#3

Only a few ounces of resin left, enough to keep the line from going dry.

Some finished pics.

#4

I am suprised that you would only do a 5 minute leak check and allow up to a 1" drop. I find that any amount of drop in vacuum will lead to air in the part. We do a minimum of 15 minute leak check and do not allow any drop. Am I understanding you correctly?

#5

Tet another top thread.

can i ask where the vacuum guage gets plumbed in? Between part and pressure pot or pump and pressure pot?

#6

Is there any reason you use the green flow media instead of blue Enkafusion? Or have any preference with the two?

#7

That is our maximum allowable. But if it’s leaking that much (which is really a tiny leak anyways) we double check all connections and taped areas. Normally we drop less than .5" in 4 hours.
Still, our numbers are within spec of Boeing, Raytheon, and Bell helicopter requirements. :wink:

First of all, we don’t use pressure pots. Our resin traps are something smaller, simpler, and more effective. :slight_smile:
Each lay-up tool (mold) has a gauge that stays on the line between resin trap and pump. This way we know how much vacuum is reaching the lay-up, and check for leaks on the whole system including the resin trap, connections, and vacuum bag. There is also a gauge placed on the resin feed line to make sure it is not leaking and that we are pulling as much vacuum as on the pump line. Obviously this gauge is removed prior to infusion, so the gauge on the pump line allows us to monitor through the whole infusion and curing process.

We use at least 7 different types of flow media. All depends on the resin being used, permeability of the ply stack, and part size or spacing between feed lines (if multiple).
In this case, green mesh is used because the epoxy is fairly low viscosity, has a long (70 minute) gel time, and the part is not too big. There’s only about 48" distance where the resin has to travel. A problem with some of the higher volume flow medias is that they leave a lot of resin on top of the peel ply and can be very difficult to remove, aside from wasting a bunch of resin. It is possible that the resin flows too fast through the flow media, as it may not go down into the ply stack at the same rate.

#8

TET, thanks for showing this example of infusion with a core.
Quick questions,
Is your core closed cell , and non perforated? ie is the resin just making contact with the surfaces and not going through the core while infusing the bottom side?
How do you make sure the vacuum does not constrict flow on the underside (mold-side) of the 2 sided infusion?

Say you were creating an object like the T shaped core you have .
After infusing both sides, if you were to trim excess fiberglass to the shape of the core , are the sides of the core then coated somehow…I am trying to figure out how one would approach something like that.

I am new to the world of composites (currently mainly in the research stage), about to soon do a little test panel, with core, just to get a feel of it.

Finally, is there a reason why the core does not extend to the edges of the whole part?

thanks

myxo

#9

Good questions, I will try to break down the answers to make it clear.

The foam core is closed cell so that it doesn’t fill up with resin. We use many variations depending on the product. I.E. Different thickness, composition, perforated or non, grooved one side or both, grooved one direction or both, etc… It’s available in various densities to suit specific applications, the higher densities absorb less resin on the surface. But they are heavier and more expensive, so weight of the resin in lighter foams is a trade-off.
This particular core material is perforated with about 6" spacing, simply to aid in resin flow. There is little structural benefit from the small “pillars” of resin created by the perforations. The core is also slit to aid conformability, these slits are very thin (facing tool side in pic), and although they do help with resin flow they close even more under vacuum and don’t take up much resin. These slits are very different from grooves cut in other core materials we use.

As for making sure flow is not constricted on the tool side, we use the appropriate infusion specific fabric that aids in flow. The fiberglass in the pics above is not woven, but quadaxial stiched. There are no crimps to hinder flow, as it’s made with 4 layers of unidirectional fibers stitched together, which creates small flow channels in all directions through the thickness of the material.
The perforations in the core are mainly what transfer resin from the flow media above the core to the plies below. The resin then only has to travel through about 3" of reinforcements before meeting another flow front. The slits do help a little (although they almost close under vacuum) with resin flow, so then the resin only has to travel about 1/2" under each square of core.

If core material is exposed after trimming or drilling, the edges must be cleaned back and sealed. That is, remove or push back some of the core at the edges of the skin to the desired thickness of the edge filler. This can be done with a thickened epoxy, or lightweight filler putty. This will help prevent delamination and the ingress of moisture which affects most cores.
For the most part, having core material extend to the EOP is not desireable. However, there are instances where it is necessary such as making several parts from a large sandwich panel.

Yes, there are several reasons the core does not extend to the edge of the parts. There will be mounting points attached on the sides, and the core could be crushed if a fastener is put through it. Also, the increased skin thickness provides greater fatigue and impact resistance in the mounting areas.

Hope that helps, think I got everything!

Good luck with your composite adventures. Work on some solid laminates first. Then bond skins to a core. Then finally try to make a sandwhich panel in one shot.

#10

One small addition: If you use a normal laminate, not infusion optimised, as the quadrax TET is using, you can quite easily make grooves in the foam core by pressing it onto chicken wire. One of my customers does this with Core-Cell, and it works perfectly. The chicken wire print allows enough resin flow to succesfully impregnate the bottom laminate.

#11

Thanks TET, for the in depth explanation on the different points. It clears up a lot of my questions.
OK, I will start simple as you suggested (solid laminate panel >>single sided skin/core >> sandwich). :slight_smile:
Herman : that’s a nice tip about the chicken wire , thanks for sharing it.

cheers

#12

I haven’t read what TET suggested but the usual progression usually is, solid laminate, then add core by post bonding (Lay up two solid laminates, then glue to core) and then finally dry layup everything and create the panel in a single process. You never really want your core exposed so single sided core is good for practice but bad for application!

Cheers!

#13

The mistake is mine. I misunderstood what TET had said.

Good luck with your composite adventures. Work on some solid laminates first. Then bond skins to a core.Then finally try to make a sandwhich panel in one shot.

I see he was saying what you are saying ie do solid laminate>>bond core to skins>>then progress to straight shot infusion.
Thanks for pointing it out.

#14

No problem at all. With composites everything is more complex, bidirectional vs uniaxial + core + resin + fillers etc etc etc. All that to describe one panel! My welding buddies only need a metal and a number to fully describe their materials they commonly work with eg. 1040 steel.

I’m curious about your research though. Mind telling me more about it?

#15

I’m curious about your research though. Mind telling me more about it?

Oh, when I said research , I didnt mean formal research (ie institutional) , Its independent …however I have still approached it from the perspective of trying to understand physical /chemical properties and so on. I have been fascinated with composites and natural fibers and its uses.And I have been doing a lot of reading about that.
I want to be able to try different mixes of materials (fibers/resins) and test the effects on its physical properties. So initially I will be starting out with flat panels and will see where it goes from there.