(New?) Composite Panel Construction Technique

[StrongFeather]

If you'd like to steal some of my experiences, feel free - ultralight teardrop build in my sig, seven years built and still getting miles put on!

I finally got around to reading your build journal. Wow! Very cool, Whitney. Had I read that before starting my build, I just may have gone with a composite deck. Maybe on build number two .

If my plan works, it won’t be as light as yours, but it will have two advantages; no filleting and no fairing. The panels will (hopefully) be a class A finish out of the mold, so no painting either. The panels will be bonded with six10 or something similar.

I’m still not sure what my layup will be, but I’m starting to think the 18 oz. plain weave might be too heavy.

Steve


Sent from my iPhone using Tapatalk

[saywhatthat]

Think two layers of 7 oz

[StrongFeather]

Just wanted to apologize for the lack of updates on here. Over the last two weeks I've been working on the trailer floor, frame, axle, etc... and doing some final decision-making research on the composite panels. I finally ordered the bulk of the materials and tools to start working on the panels and I spent most of the day yesterday cleaning the garage and getting it set up for panel prototyping and (hopefully) production. I still have to get a few things locally, but I hope to have some progress and pictures to post next weekend. Maybe even panel #1 will be born

A note on the research side of things, most of it was focused on the surface coatings. Epoxy degrades and turns chalky-white when exposed to sunlight, so any carbon fiber (CF) panels will have to be clear coated with a UV resistant coating of some kind. Option 2 is an automotive type, post-layup clear coat that would be sprayed on each panel after they come out of the mold, or they could all be sprayed together after the shell is assembled. Either way, it would likely involve surface prep, taping and wet-sanding, etc... I'm hoping to avoid post processing of any kind, so Option 1 is an in-mold epoxy-compatible polyester clear coat made by Duratec called SunShield. Unfortunately, they recommend that it be sprayed into the mold, which I also hope to avoid, so I'm going to see how it works being brushed in and go from there.

The surface coating for the fiberglass panels will be a colored polyester gelcoat - black in my case - but there are a lot of conflicting opinions out there as to whether epoxy and polyester are compatible. In the end, the majority seem to agree that epoxy will bond to polyester, or anything else for that matter, but polyester will not bond (well) to epoxy. With that said, my plan is to apply the polyester gelcoat, mixed with a Duratec High Gloss Additive, in-mold. The fiberglass layup will be placed on the tacky gelcoat and infused.

Once the panels are made, I'll do some unscientific destructive testing to see how much torture they can take.

Stay tuned.


Steve

[StrongFeather]

Quick update... if you want to call it that since very little progress has been made.

I finally received the supply order that I was waiting on and was able to lay-up the first test panel to the point of getting ready to pull a vacuum. That's when I noticed that they sent me spiral tubing instead of vacuum tubing. And since you can't pull a vacuum through spiral tubing - I was stuck and had to wait for them to send the correct item. Then, after a trip to the 'Depot and ACE, I realized that my new vacuum pump has 37° flare fittings instead of 45° as I assumed. Unfortunately, a 37° flare to NPT adapter is impossible to source locally so I had to order one. If only I would have verified that sooner.

I'm chomping at the bit to start infusing, so stand by. Hopefully it won't be much longer.


Steve

[Atomic77]

I have laid up and infused a lot of parts. Some were primed, some were gelcoated and some were clearcoat. Most all my parts were layered in the mold so I am familiar with spraying the mold. Which is my strong suggestion because if you brush it you will run into a rash of problems you can't even imagine at this point. It's almost impossible to get it even and remember, whatever goes into the mold first is what you see when it's done so you want it to look nice. Just a couple cents worth... Good luck!

[OP827]

Strongfeather, as a reference did you look at Kelsall KSS boat building method?

http://www.kelsall.com/UniqueKSS/WhatIsKSS.pdf

[StrongFeather]

Interesting method, OP'. I am a little crazy, but luckily, I don't think I'll be a boat anytime soon. Some of their methods are very similar to my plan, i.e. resin infusion on a flat surface, and taking advantage of the flat surface to achieve a high quality finish straight from the mold. Lucky for me, I don't have to worry about complex curves and extreme pressures that are exerted on boat hulls. One of the reasons that I decided to try a painters canvas style (frame with a skin) panel is to have an area to run wires for lights, etc... a boat builder will tackle this problem much differently.


Steve

[StrongFeather]

Welp, here we go. Test panel #1 is done and although it not perfect, I'm calling it a success. I'm going to drop a few pictures with some notes, now and follow that up with more details on how the test was done.

For this panel, I wanted to tackle several different variables at once, so the skin section is essentially split in half with a different layup on each side, and each of the four sides and one of the four corners was done differently in some way - and the results should be helpful.

I've never vacuum bagged anything (not even fiberglass), let alone vacuum infused or even touched carbon fiber prior to this, so there were numerous "firsts" for me in this test and I've learned quite a lot that can't be learned without actually doing. YouTube is great, but it certainly has its limits.

First up is a picture of the rear.
* I labeled the four sides A-D so that I can reference what was done in the layup and compare it to my notes.
* You can see the delineation between the two halves. Well call them the B-half and the D-half. The B-half has 1 layer of 3k, 5.7oz, 2x2 twill carbon fiber with 2 layers of 7.25 oz, plain weave E-glass fiberglass. The D-half has 2 layers of the 3k CF and 1 layer of 7.25 oz FG. The only real difference between the two hlaves is that you can see more pin holes on the B-half if you hold it up to a light source. Otherwise, they look and feel identical.
* If you look at the inside corners, you can see some darker black areas. This is a result of "bridging", meaning that the vacuum bag was not able to be pulled all the way into the corner when it was under vacuum. This is a result of my inexperience - I cut the bagging film a little too small, and possibly due to the cheap bagging film that I used - it doesn't really have any stretch to it.
* Note: The matte finish on this side is a result of the "peel ply" material that is used to prevent the other consumable materials from becoming part of the finished piece. When peel ply is removed, it leaves behind a perfect surface to mechanically bond to.

DSC00011.JPG (184.12 KiB) Viewed 3016 times

Next up, pictures of the front.
* You'll notice that there are some areas around the edges that didn't fill to a smooth finish. These, I'm pretty sure, were caused by the bridging noted on the back side. Either that or the resin just didn't flow into those areas - but I'm pretty sure it was the bridging. More testing will tell the tale.
* You'll also notice what is likely "print through" around the perimeter of the panel. Print through, in this case, is caused by the wood frame pressing down on the CF while under vacuum. The other possibility is the fact that this cloth was put in and pulled out of the mold several times, which disturbed the twill pattern. Either way, the picture actually makes it look worse than it does in person.

DSC00010.JPG (191.68 KiB) Viewed 3016 times

DSC00008.JPG (123.38 KiB) Viewed 3016 times

In summary, the first test wasn't perfect, but it was a huge success nonetheless. It looks incredible, it's super light (350g or 12oz) and I cannot believe how rigid it is. I might do some destructive testing on this panel at some point just to see how much torture it can withstand, but honestly I'm not sure that'll be necessary. When XPS foam board is bonded into the cavity on the backside and a sheet of whatever (luan, etc...) is bonded across the entire inside, it will be an incredibly strong structure.

I'll work on posting pictures of the layup and infusion process, but it might take me a few days. In the mean time, any questions/comments are welcome.


Steve

[OP827]

Since you are not building a wing structure for a passenger plane, this structure will be plenty strong (may even be an overkill depending how thick the CF layer is) for trailer walls once you glued insulation and structural inner layer. Ideally the system should provide for assembling the panels together by having the perimeter frames. This by itself is a great idea as long as you can make matching panels. This makes me think that once you have walls fabricated, the roof form should be made with walls attached on the sides or at least use them for forming roof curve, so that they match nicely. This is similar to common approach of how Fiber Glass parts are made to fit each other.

Sorry if I missed it, but why you decided to have Carbon Fiber for panels, not a Glass fiber? Carbon fiber is much higher Young Module and Strength, so is it for rigidity and less weight per strength? But is it really needed for this application? CF is a lot of cash if building a good size trailer and more of an airspace and competition boats material when weight and strength are the design drivers, not the cost involved. Now, it also has to be painted white for use in a trailer build, black will be too hot for composite sandwich structure to stay intact. This is the reason all FG composite aircraft are painted white. Looks very good for a first panel experience.

I am eager to see details of your setup for the panels production. I mentioned Kelsall method only because it has similar approach to making flat finished panels on a big flat table, which eliminates traditional method work of fairing before finish paint and/or clear coat. You have the same approach, except you have the perimeter frame, which might make the layup more challenging.

What is your plan for finish/painting the outside surface? I am following with interest. Thanks for sharing.

[StrongFeather]

Ideally the system should provide for assembling the panels together by having the perimeter frames. This by itself is a great idea as long as you can make matching panels.

Thank you, OP’. By design, assembling the panels should be pretty easy for even the most novice builder. Simply put glue (epoxy) on the two panel edges and clamp them together until cured. They could add screws, but they most likely wouldn’t be necessary. Yes, the panels will have to be matched, e.g. all the same height, but that applies to all construction methods. I simply plan on setting my mold up so that the top and bottom pieces remain at a fixed distance ensuring that the height on the wall panels (or length of the roof panels) are all the same – only the width’s will change.

This makes me think that once you have walls fabricated, the roof form should be made with walls attached on the sides or at least use them for forming roof curve, so that they match nicely.

Agreed. Although the first build, which will be for my family of 5, will not be a teardrop, so no curves. If everything goes well, I will probably follow this up by building a teardrop prototype.

Sorry if I missed it, but why you decided to have Carbon Fiber for panels, not a Glass fiber? Carbon fiber is much higher Young Module and Strength, so is it for rigidity and less weight per strength? But is it really needed for this application? CF is a lot of cash if building a good size trailer and more of an airspace and competition boats material when weight and strength are the design drivers, not the cost involved.

I haven’t completely decided on carbon fiber, and I will be testing fiber glass soon. One of the benefits of this method of construction is that the builder will have the option of choosing between the two, or even Kevlar, or any combination of the three. It will also (hopefully) allow them to choose between an in-mold clear finish (for carbon fiber), or a colored gel coat of their choosing. I haven’t tested either yet, but in theory, they’ll come out of the mold smooth and shiny with no post processing needed.

But to answer your question: Why carbon fiber? Well because… Carbon Fiber! If the extra cost is the only deciding factor, why not carbon fiber? Seriously though, fiberglass alone will be plenty strong and only marginally heavier, so the biggest reason to choose one over the other would be cost vs. appearance/uniqueness. Once I’ve done a fiberglass test panel and I have an idea of what the fabric layup will be, I should have a much better idea of how much more CF will be compared to FG. I'll make my decision based on that.

Now, it also has to be painted white for use in a trailer build, black will be too hot for composite sandwich structure to stay intact. This is the reason all FG composite aircraft are painted white.

I’ll have to respectfully disagree with you to some extent on this one, OP’. Yes, black will be hot, and yes, white (or something lighter) is the smarter way to go, but no, it will not have to be painted white. The Tg (Glass Transition Temperature) of the resin that I’m using will be over 150°F. It is absolutely possible, likely in fact, that the surface temp will see 150° or hotter temps, but what happens at Tg? In this case, nothing really. It will lose some strength and flexibility (modulus) while at these temps and then regain those properties when it cools, but it won’t flow, move or come apart while at Tg.

The other consideration would be comfort, i.e. 5 unhappy campers baking in a black sardine can oven. This is a real concern for sure, but for me, one of the very first requirements of this build was to have air conditioning. In my case it will most likely be a 5,000BTU window unit, which should cool 270sq.ft. pretty easily.

I am eager to see details of your setup for the panels production. I mentioned Kelsall method only because it has similar approach to making flat finished panels on a big flat table, which eliminates traditional method work of fairing before finish paint and/or clear coat. You have the same approach, except you have the perimeter frame, which might make the layup more challenging.

I’ll try to get some pictures posted soon.

Steve

[OP827]

I am not as concerned about structural integrity of fiberglass due to surface dark color. Resin will probably handle that heat. But xps foam underneath the fiber may soften and separate from the sandwich at temperatures close to only 60 degree by Celsius or 140 F. Delamination happened to builders here and on expedition portal forums who had other than white surface on their foam fiberglass sandwich build. Just sharing their experience here, I am so far trying to keep my build in shade until it's painted.

Sent from my Moto G Play using Tapatalk

[StrongFeather]

Hmmmmmmm. Interesting. I didn't really consider XPS breaking down. Thanks for the heads up!


Steve

[StrongFeather]

...xps foam underneath the fiber may soften and separate from the sandwich at temperatures close to only 60 degree by Celsius or 140 F. Delamination happened to builders here and on expedition portal forums who had other than white surface on their foam fiberglass sandwich build. Just sharing their experience here, I am so far trying to keep my build in shade until it's painted.

I appreciate you sharing your experience, OP'. I took some time to look into and compare XPS and polyisocyanurate foams. XPS is more rigid and has a higher compression strength, but has a lower heat deflection temperature (HDT). Polyiso' is weaker in all respects, but can handle heat better. Sorry for preaching to the choir.

I then considered the pros and cons of each in respect to this construction method where the foam will be inside a very rigid cavity, bonded and supported on all sides. If the foam is able to actually get to its HDT, I'm not so sure that it would even be able to delaminate. In an extreme case, in theory, it would soften, maybe even slouch a little and then harden again when it cools down - no harm done. With that said, even if it does delaminate, I believe that my panel construction alone (no foam) is more than adequate to stand on its own.

I wasn't able to find any examples here on TNTTT of builders stating that they experienced XPS delamination due to heat - probably because the search function isn't very good. The only statement that did pop up was GPW saying he didn't have an issue after 4 years in Florida, and I believe he also said his foamie is black. I'm wondering what the circumstance were for the ones that you've read about or experienced yourself. Perhaps there are other factors involved other than heat. No matter what, I still plan on looking further into this in the coming days and I'm hoping that you'll share some more with us. Thanks again


Steve

[StrongFeather]

Here are some picture of the panel making process. These are from panel #1, which I posted pictures of last week. Panel #2 is curing right now, but I didn't take any pictures.

The "mold" is a glass mirror with a square wood frame stuck to the glass with gum tape. It's a pretty rudimentary setup for testing purposes, but it worked well. After the "mold" is set up and prepped with mold release, the fabric is laid in and the wooden panel frame is placed. Then fabric is then folded over the frame into the center before the consumable layers are added.

DSC09972.JPG (198.26 KiB) Viewed 2875 times

There's a lot going on in this picture. The consumables mentioned above are peel ply (prevents excess resin on the part and prevents anything above it from becoming part of the part), flow media (a green mesh that allows the resin to flow), vacuum tubing and spiral tubing, and lastly the vacuum bag. The white cup holds the resin, which is drawn in when the clamp is released while the layup is under vacuum - the vacuum line going to the pump is not attached in this picture. The vacuum bag, or bagging film, is taped to the mirror with gum tape. Pleats are added to the perimeter of the bag to allow the film to be worked into the corners of the mold.

DSC09977.JPG (153.25 KiB) Viewed 2875 times

The infusion is in process in this picture. If there ever were an action shot in infusion, this would be it! It's kind of hard to tell, but the bagging film didn't quite make it into the interior corners, which caused the bridging on panel #1. Panel #2 went much better. The tubing around the perimeter would typically be standard spiral infusion tubing, but in this case I chose to go with MTI tubing. MTI has a membrane wrapped around spiral tubing that allows air/gas to pass through, but does not allow resin. Although MTI is considerably more expensive, I believe it's an essential element of this setup because it should force the resin to flow around both sides of the frame, whereas regular spiral tubing, in theory, would just allow the resin to take the path of least resistance. It also decreases the amount of resin needed and allows me to avoid using a catch can, which is a device that prevents resin from being sucked into the pump. Note the vacuum line in the bottom of this picture is the line going to the pump. Another advantage of not using a catch can is that I only use about 1" of tubing for every infusion. I just cut it off at the connector and it's ready for the next one.

DSC09982.JPG (194.72 KiB) Viewed 2875 times

More to follow...

Steve

[StrongFeather]

Here's some post-cure pictures prior to pulling the piece out of the mold.

In this picture, you can really see the green flow media, which again is used to allow the resin to flow through the layup. Reinforcing cloth (CF, FG or Kevlar) is resistant to resin flow, especially while under atmospheric pressure, so without flow media, the resin might start to harden before the entire part is infused, ruining the part. Under the flow media is the peel ply, which is essentially invisible at this point. The extra flow media in the center aids in getting the resin away from the inlet. A long rectangular panel will have flow media running the length of the panel to get the resin to flow to the distant ends. You can also (kind of) see the membrane around the MTI tubing if you look hard enough. In the future, I think I'm going to try using 90° connectors in the corners to make the tubing lay easier around the frame. Since the resin doesn't get past the membrane, the connectors can be reused.

DSC09989.JPG (195.46 KiB) Viewed 2870 times

In this picture, I'm peeling up the consumables from the part. The white material with red stripes is the peel ply. This happens to be the corner that I did not cut the corner out of, so there's a lot of extra material that bulked up on the frame. If you look at the top left corner of the first picture from the previous post, you can see that all of the fabric in that corner ended up in the piece, whereas the other corners were cut out. This was a test, and it turned out pretty much how I expected. The messy bulk was simply cleaned up with a metal file, so really no harm done since it's not in a cosmetic area, but I will be cutting the corners in all future panels.

DSC09998.JPG (194.61 KiB) Viewed 2870 times

With Panel #2, I used one layer of really thick, heavy 18oz fiberglass cloth instead of two 7oz. layers. I thought I could get away with not using flow media on this one because the 18oz cloth has such an open weave - wrong! It did infuse, but it was extremely slow and needed some help from a plastic spatula to get the resin to the edges. Lesson learned. I also brushed in a gel coat on the top half of the panel prior to laying the fabric just to see how it comes out. I also used a fiberglass veil on the left half. The weave pattern of a course fabric, like the 18oz., may show (print) through the finished surface - a veil is a thin matting that is supposed to prevent this "print through".

I'll probably start on #3 tomorrow after I pull #2 from the mold, but I might hold off while I'm waiting on some spray tack to arrive that I ordered today. The spray tack is an aerosol adhesive that I'll use to stick the fabrics into the mold during the dry layup process. This will make the assembly process much easier, and will ensure that all of the fabric ends up where I want it to end up.


Steve