Help with Infusion and Divinycell core

General Composites Discussions Composites Talk
#1

After using hand layup for a series of sandwich core CF parts, I’ve decided to learn the infusion process and I’m looking for some pointers.

For the first set of experiments, I kept it simple…
Part 1 from bottom to top:
[ul]
[li]flat glass plate (mold)[/li][li]PVA/wax[/li][li]3 layers, 6K 2x2 Twill 11.0oz/sq. yrd[/li][li]Nylon peel ply (1-2" larger than CF cloth)[/li][li]Perforated release film (same size as CF cloth)[/li][li]Flow media[/li][li]Vacuum bagging film[/li][/ul]

I used a degassed, <900cps epoxy (leftover from hand lay ups), MTI Hose on 3 sides of the plate and pulled <20mbar vacuum. The part came out exactly as expected with near perfect mold side cosmetics. SUCCESS!

Part 2 now includes Divinycell core and a second 3 ply stack. I also reduced the amount of MTI Hose by 1/3 and only used it on the side opposite the inlet line.
Bottom to top of Part 2:
[ul]
[li]flat glass plate (mold)[/li][li]PVA/wax[/li][li]3 layers, 6K 2x2 Twill 11.0oz/sq. yrd[/li][li]1/4" Divinycell (perforated with 1/16" drill every 2" in a grid and scored at perforations) [/li][li]3 layers, 6K 2x2 Twill 11.0oz/sq. yrd[/li][li]Nylon peel ply (1-2" larger than CF cloth)[/li][li]Perforated release film (same size as CF cloth)[/li][li]Flow media[/li][li]Vacuum bagging film[/li][/ul]

This time degassed <900cps epoxy, <20mbar vacuum, added a length of Enka channel. The top wet out nicely but the bottom was dry at each of the “islands” between the grids.

Part 3 now uses <600cps resin, a beveled core edge and I added a layer of peel ply between the glass and the lower stack. Same MTI Hose setup opposite the inlet line.
Bottom to top of Part 3:
[ul]
[li]flat glass plate (mold)[/li][li]PVA/wax[/li][li]Nylon peel ply (1-2" larger than CF cloth)[/li][li]3 layers, 6K 2x2 Twill 11.0oz/sq. yrd[/li][li]1/4" Divinycell (perforated with 1/16" drill every 2" in a grid and scored at perforations) [/li][li]3 layers, 6K 2x2 Twill 11.0oz/sq. yrd[/li][li]Nylon peel ply (1-2" larger than CF cloth)[/li][li]Perforated release film (same size as CF cloth)[/li][li]Flow media[/li][li]Vacuum bagging film[/li][/ul]

This time the epoxy penetrated further towards the MTI Hose, but was not completely wet out, similar to part #2 between the grid scores (islands).

Not knowing if it was the lower viscosity OR the addition of peelply, I tried a 4th experiment. Same stack as 3, but I removed the flow media. It infused about 20% and STOPPED.

What do you see I am doing incorrectly AND/OR what do you recommend for the next step?
Even thinner epoxy? Flow media top & bottom? Finer grid scoring (1" vice 2")?

#2

You have a few issues with your setup. First your core step is causing race tracking. This can be fixed by either beveling the core edge, or pressing the bag and consumables in to the corner of the core so that there is no bridge there. Second, your flow media is covering too much surface area of the part. You need to end it short of the laminate, and also the edges. Lastly, you have a vacuum leak in your last photo. Vacuum should be 100% tight, and should be checked by digital absolute pressure gauges at the part, for the best results

#3

You could also try without the perforated film.

Just the laminate, peel ply, flow medium and bag.

One less restriction to help get the resin through to the foam perforations.

#4

Agree, with that type of peel ply, release film is not needed, and probably not doing anything beneficial. I would also remove it

#5

Thank you!

Next, I will try without the release film and reduce the size of the flow media at all trailing edges.

Are there guidelines to how much smaller the flow media should be (75-80%) relative to laminate stack?

#6

25-50mm less than the laminate size on the sides and far end, but this is very general. It really depends on your fabrics, resin and other factors. For what you are doing I think its ok. Also, you might find better results if you start the resin infusion 50-75mm further away from the laminate. This is because the initial infusion is very fast and in some cases causes defects directly underneath the infusion line. If you start it further back it slows down enough to wet out the laminate without defects.

#7

I think it is a setup and resin problem. The perf release film will not change the result (as long as it is a infusion perf)
As Hojo said, let the flow media end about 25-50mm before the core material ends, except the side you start your infusion.
But on such a small part I would infuse from the middle and let the flow media end about 25mm before the core ends. This prevents racetracking at the edges wich is most times the biggest problem at small sandwich parts.
Also it seems your resin system has a too high viscosity, try heat to lower the viscosity or change to a infusion system.
Make sure you setup is tight and vacuum below 20mbar is fine. Make a drop test, at such a plate you should be below 5 mbar in 20min!

#8

Thanks for the feedback. I hope to get the fabric & core cut for another simple test part this weekend. For the next test I will try warming the 600cps resin a little bit as well, and will go thinner if needed.

This leads me to my next question…
My ultimate goal is to infuse a tube with dimensions:
~0.6m (24") OD. x ~1.8m (6’) length with a ~13mm (1/2") core using a mandrel.

My understanding is that for infusion using a cylindrical mandrel, the MTI hose runs along the length of the tube at the top. Underneath, the inlet line runs along the length of the tube.

Is it OK, to place the inlet line directly onto the laminate stack for infusion?

#9

X2 on center feed,flow media stops short and peelply extends over.perimeter vac.
Youll have to really heat up the resin to get it thin enough.
No advice on the tube but if thats the goal you should be doing tests on that not a flat panel as things are totally different.
The pic is a test i did, no flow media as the core is grooved and perforated.

#10

[ul]
[li]When I make my next resin purchase, what viscosity should I be using? [/li][li]Can you recommend resins available here in the US?[/li][/ul]
Thanks!

#11

That’s a big tube- what’s it for?

#12

Hi, I would try to get Hexion resin. Ther RIM 135 or 035C.

#13

Now that the weather is warming up I am ready to continue. I will start by purchasing a less viscous resin.

I see the Hexion distributors list, but does anyone know of an online retailer? Google and I are having a hard time locating them… I do see a lot of info, just no retailers.

Anyone use/recommend the Pro-Set INF 114 epoxy? The specs rate viscosity to 100-200 cPa.

#14

Carbon is always kak at infusion over long areas, because the fibers are finer and close up all the spaces etc.

If you want a Cf plate smooth both sides…

Glass BOTH sides, at least 5 mm. at full pull, with an over hang the glass will crack.

Lay up your clothetc with Spray tack onto an infusion core, such as Soric, or a designated slotted hard foam core designed for infusion.

Now lay on your tacked up part onto the glass…add on tacky tape to 2 sides TIGHT up against the CF, leave the ends clear…run the tacky tape to the ends of the glass to the CORNERS!!, it will look like a wine glass at the ends. Glass sheet not too much bigger then the core you are making etc.

Put the other glass plate on. you now have a channel…add in, at the top and bottom your mesh and spiral tube, one for epoxy and one for the vacuum. have some peel ply as a choke as normal at the vacuum end.

Absolutely no leaks please!! Bag and run it. No more issues, and no run past and perfect part.

yes it will open. BUT the epoxy must be hard and another reason to have a thickish glass.

Sy