Solvent Welding a Polycarbonate Honeycomb Structure

[StevenRS11]

Hello

I have been working on solvent welding a polycarbonate honeycomb between two polycarbonate sheets, and I have a few questions that maybe someone knowledgeable here could help me with.

On a small scale, (2"x2") I have gotten good results by pressing the polycarbonate honeycomb onto a cotton felt pad, soaked with solvent, then clamping the wet honeycomb onto the plastic sheet. Scaling this up, however, is proving difficult.

Solvents I have used are dichloromethane, ethyl acetate, MEK, and a few others- all honestly perform fairly well. I would like to avoid using DCM if possible, so I have been focusing on ethyl acetate recently. The problem I run into is that applying an even coat of the fast drying solvents over the full 15"x8" surface and then clamping them is very hit or miss. Having a felt pad that large soaked with this volatile solvents is also an issue.

Is there any industry standard way to do this, or does anyone have any brilliant ideas?

My other concern is trapping solvent vapors in the now-closed cellular structure of the resulting composite panel. Is this an actual concern?

I can post pictures of the process and materials if this isn't clear enough (I probably will later, anyway). Thanks in advance.

 

[Compositepro]

Polycarbonate is very sensitive to solvent stress cracking, so you should look into whether that will be a problem. Your part could craze and crack days or weeks after bonding.
As for the actual process, use a paint roller to apply a solvent based adhesive rather than straight solvent. An adhesive will contain some pre-dissolved solvent, which will increase viscosity, improve gap filling, and slow evaporation, so that a roller will work better.
Solvent in the core will slowly diffuse through the plastic and dissipate. There will be a period of time where the solvent will be absorbed by the plastic, which will soften the whole part. So you have to be careful about not applying an excessive amount of adhesive.

 

[StevenRS11]

Since we started annealing the polycarbonate before the welding process, we havent seen any crazing (yet, atleast). Is there a way to ensure that, after bonding, it wont craze?

Elsewhere, I was told heating and holding the welded product to ~100C will help remove solvent from the softened zones around the weld, followed by another annealing cycle. Apparently this prevents crazing, but if I understand how stress cracking works, heating after the process wont help- The residual stresses in the plastic, when it is exposed to the solvent, are what cause cracks to propagate as the solvent induces partial crystallization in the plastic. Heating is, uh, expensive, so I would like to avoid having to do that without concrete proof that it works.

Do you have any advice as to what adhesives would work well for this type of construction? Considering that its essentially a group of butt joints, I guess peel strength is what I need. Originally we where using an epoxy applied to the plastic sheets, but we moved away from that because it wasted a great deal of epoxy and the glue line tended to fail. I have considered a urethane adhesive, but I don't have much first hand experience using them so I wasnt sure where to start.

The tricky thing is that whatever joining method we use, it need to be clear and atleast somewhat aesthetically pleasing- this is for a consumer product, and considering the entire thing is transparent polycarbonate, the glue line needs to look good as well.

 

[MikeHalloran]

When a honeycomb structure is stressed in bending, the core to surface joints are stressed in shear. ... a lot.
We don't know how your assembly is stressed.

I personally have never been happy with solvent bonding polycarbonate.
Can you convert the assembly so the core elements have tabs that snap into countersunk holes in the surface sheets, and just assemble everything mechanically in a press?



Mike Halloran
Pembroke Pines, FL, USA

 

[Demon3]

Can you vibration weld it and avoid all the ESCR and other solvent related problems?

Chris DeArmitt PhD
President - Phantom Plastics LLC

Consulting, ideas and training on plastic materials

http://www.phantomplastics.com

 

[btrueblood]

Good thoughts from the usual culprits above. My only addition is to consider a uv-cure clear epoxy, preferrably in a fairly flexible formulation, though it may not work if the PC is fairly thick (the UV required for cure initiation I think won't be transmitted well by the PC).

Why polycarbonate? Would acrylic work in your application - it has much more forgiving characteristics for solvent bonding.

Does the honeycomb need to be clear? It's going to distort images no matter what level of clarity you achieve...so maybe apply a sheet of clear acrylic adhesive to it?

 

[StevenRS11]

First of all, thanks for the help with this- I have been having issues getting this to work well for a while.


Here are some pictures of a small sample. The AA battery is for scale. As you can see, the honeycomb cells are relatively small, and the final product is a roughly table tennis shaped 'paddle' of this material. The entire thing is about 15" by 9", and the two polycarbonate sheets fit over the honeycomb in a clamshell type arrangement, where the edges actually overlap. If that isn't clear, I can post pictures of that as well.

The plastic sheets are thin, .040 to .030 depending on the specific model.

I don't think that mechanical assembly with tabs/countersunk holes will work, given the large area and number of joints that would be needed. As for vibration welding, is that not generally used for linear joints, that are parallel with the direction of the vibration? I feel like the vibrations would damage the cells of the honeycomb as the sheet material slid back and forth. I can see this working with very high frequency and very low amplitude vibrations, though. Unless under those conditions not enough energy would be carried through the honeycomb and into the sheet due to its flexibility.

The primary design consideration is stiffness with regard to force applied perpendicular to the plas

http://www.eng-tips.com/viewthread.cfm?qid=371115tic

sheets, on the paddle "face". Total deflection under a 3 kilogram load in the middle of the paddle face with supports 5 inches apart needs to stay under 5 thousands of an inch for regulatory reasons.

The second is toughness, hence my choice of polycarbonate sheets. I did some initial prototyping with PETG, which had a tendency to actually shatter. I imagine acrylic this thin would be even worse. This is for a sporting goods product, so it will be subject to repeated impacts with a hard (but light) plastic ball and occasional, accidental impacts with whatever playing surface the players are using- possibly concrete, but usually hardwood. Existing, competing products tend to fail rather spectacularly when damaged like this, with shards of fiberglass and graphite shooting everywhere, sometime into the players hand.

Weight is also an issue- players tend to prefer paddles what weight between 7 to 9 oz. In a nutshell, we need very light, very stiff, and as durable as possible. Fun!

I almost would prefer to move away from solvent bonding, considering what you all have said here, and especially because it might let us use an aluminum honeycomb core as well. The honeycomb itself does not need to be clear, but the plastic shell around it does. I like the idea of UV cure epoxies, and I think I can get a sample delivered pretty quickly for me to test with. Still, solvent bonding has given me far and away the strongest bond yet, and the stiffest resulting construction. Its just very tricky to do on larger surfaces with the evaporation rate. Not enough solvent, no bond at all. Too much, and the core breaks down. Not to mention the stress cracking issues.

Thanks again,
Steven

 

[Stickybeak25]

Hi there, I would go with a UV-cure adhesive (one with a longer wavelength cure up to 420nm) this will allow it to cure through the PC. You will get a much better finished appearance and not have to think about any redesign. You would need to select one with an appropriate viscosity to achieve a small "fillet" or meniscus around each honeycomb tube to increase bond area, strength, cleavage and peel resistance. The other benefits include - non-yellowing, instant cure on demand, reduced chance of air bubbles and voids (unlike solvent welding), no loss of the intrinsic strength of the polycarb through chemical attack and of course HSE benefits of not handling a highly volatile solvent. Hope this helps.

 

[StevenRS11]

So, update time!

As for UV cure adhesives, we ended up running into issues with oxygen inhibition, flexibility, and price. The stronger UV adhesives generally suffered from more severe inhibition, and because the honeycomb structure undergoes not insignificant local deformation when it strikes the ball, the very high shore UV cure adhesives would shatter. They ones that worked (and they worked exceedingly well) cost hundreds of dollars a liter, which was really too much for the quantities we will be needing.

I found a a two part urethane adhesive that we are currently using, and it seems to work very well. We also started using a larger cell size honeycomb, which uses less glue so perfect clarity isnt necessary anymore. As long as the glue is a translucent white, with no off colors, it is visually unnoticeable. The two part urethane is currently being used in our production products.

That said, it isn't *ideal*, so, while it is not as critical, I am still looking for something even better. Should I start a new thread, or is there some way to change the title of this one, considering solvent welding went out the door a long time ago?

 

[MikeHalloran]

Since the product and goal have not changed, please don't start a new discussion; just keep extending this one, so everyone can see the evolution.



Mike Halloran
Pembroke Pines, FL, USA

 

[btrueblood]

What about the urethane is non-ideal?