Properly bonding composites and what your government doesn't want you to know.

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

flynfrog

Well-Known Member
Joined
Sep 16, 2011
Messages
506
Reaction score
18
How to bond like a pro.


I promised to write this article a few months ago after some conversations in this thread:
https://www.rocketryforum.com/showthread.php?56789-Sanding-Fiberglass-Tube

I didn't start this thread to call out every one who has there own way of bonding. Bonds in rockets generally are not all that stressed. Please don't email me calling me names because you have been bonding rockets together for years and never had one fail. That's probably true but that doesn't mean its optimal.

First a little background on myself. I have been in the composite aerospace industry my entire career building and designing aircraft with fully stressed structural bonds on primary structure. (this means we put glue on important stuff you don't want to fall off) before then I was building composite solar cars and racing them across the country. If there is a wrong way to bond something I have done it.

A little background on bonding. I see many people posting that the only way to get a good bond is to rough up a surface with 30grit paper and a saw blade. Thinking that the increase in surface area will make the bond hold better. This assumes that the bond is purely mechanical. This is a false assumption. The bond we are creating is on a molecular level. We are actually exciting the hydrogen groups and leaving openings for our epoxy to fill creating a bond on an atomic level. Also a highly activated surface will have a much lower surface tension. The lower surface tension will allow for more surface area than a saw blade can dream of. A few other things to note here. Composites get there strength from the fibers running though them. These fibers combined with the resin help the laminate to share the load throughout the part not allowing stress to concentrate in one spot. Now if we cut these fibers they can no longer do there job and are now along as dead weight. By cutting fibers you are reducing your effective panel thickness to the depth of your cut. Another way to think about is what is stronger the epoxy you are smearing on a part or the composite panel with fibers doing the dirty work for the resin. Try breaking some plain resin vs the same resin with fiber in it.


Some more background reading:

https://www.niar.wichita.edu/niarworkshops/Portals/0/Jun17_0200_JimM.pdf

https://www.loctite.sg/sea/content_data/LT4536_TT_Aerospace_Surface_Preparation_Guide.pdf formerly https://www.henkelna.com/us/content_data/14258_LT4536_TT_Aerospace_Surface_Preparation_Guide.pdf

https://www.jams-coe.org/docs/JAMS08_presentations/21.Flinn.pdf

https://www.cozybuilders.org/Oshkosh_Presentations/AbarisCompositeBondingOshkosh2005.pdf

https://www.niar.wichita.edu/agate/documents/materials/dot-faa-ar-01-08.pdf


Now on with the fun stuff how to make the best bond possible.

Rule 1: Be clean.. oils from your hands will compromise bond strength. Mold release on the panel will compromise bond strength. The grease on your sand paper from your last project will compromise your bond strength
Rule 2: Time is of the essence... Molecular activation reduces as time goes by for best results try to have epoxy on your part within an hour of sanding.

Rule 3: Learn your glue not all epoxies are equal. They all have different properties learn what these mean and how they effect your bond.

Rule 4: Air entrapment is the enemy. Your bond strength is measured in PSI if your bond has a huge air bubble in it your SI portion of PSI goes down with your bond strength.

OK now with the actual how too.

Step 1 Gloves:

From this time on thou shall not touch parts without CLEAN gloves. Don't rub your nose, touch anything greasy, touch anything that may be greasy with your gloves on. If you do this change your gloves.

DSC_0251_zps5b7b05a6.jpg


Step 2 cleaning:

We want to clean our pieces before bonding. The reason is when composites are built they are built on molds that have mold release. If we start sanding without removing this we will be grinding mold release into our newly sanded surface making almost impossible to get out. I am using laqurer thinner and paper towels here. In an aerospace setting you would use lint free aviation wipes. They cost way more than my blue towel but ill keep an eye on the lint. In industry we more commonly use acetone and denatured alcohol to clean parts as they tend to evaporate faster and have common specs as to what acetone and denatured alcohol are.

DSC_0250_zps736eaaa9.jpg


DSC_0249_zps8641d489.jpg


DSC_0252_zps69a1f09d.jpg


You want to wipe flipping your towel often until is comes off clean.

DSC_0253_zpsfc7e6f1d.jpg


Step 3 Sanding:

As noted before you don't want the roughest surface you can get. You want an evenly abraded surface without damaging fibers. I would recommend as would most aircraft specs 120-240 grit sand paper or if you have a much fancier garage than mine a grit blast with 200 grit aluminum oxide. We wand to sand in all directions not just one. This ensure we have proper surface activation you can do this by hand I am going to use a DA sander to save time. Be careful you only want to expose the surface of the fiber not go through them. We are trying to remove the top layer of resin leaving the fibers exposed to our epoxy.


220 Grit
DSC_0254_zps6fc295dc.jpg


Remember we want clean sandpaper not one with oil or grease on it.
DSC_0255_zpsf5552783.jpg



After sanding I am doing tool and bag side to show the difference
DSC_0256_zps74774062.jpg


Now give them another wipe down with solvent

DSC_0259_zps068e2f80.jpg


Admire your handy work This is what is know as a water break surface more details at the end.

DSC_0257_zps0c72dc18.jpg


You can see this piece needs a little more. The resin is visible on top of the fibers.

DSC_0260_zps4895223b.jpg


wipe until you get a clean rag.

DSC_0265_zps2f96ab77.jpg




Step 4 epoxy application:
Ideally you will be doing this step immediately after Step 3. Grab your epoxy of choice. I chose this one because it was in the gun. Apply a decent bead of epoxy and spread it on. Its better to have to much than to little. We want enough so that we can clamp and push out all of the air in our bond.

DSC_0266_zps279f85e1.jpg


You can see I should have sanded a little farther on this piece but I was starting to cut fibers and called it good enough
DSC_0267_zps5c81969c.jpg


A good smear on both sides

DSC_0268_zps73e32113.jpg



Place them together
DSC_0270_zps2675918d.jpg


don't push them down

We want to push all of the air out when we clamp. Once a bond has pressure applied to is that pressure can not be released until the bond is cured. Doing so will allow air to get sucked into the bond gap. One thing I have yet to mention here is bond line thickness control some epoxies have a min and max bond gap for optimal strength this can be controlled by adding a small amount of glass beads to the epoxy. These glass beads have an evil down side though they are stress concentrators that can lead to a failure of the joint.

Clamp applied note the squeeze out all around this tells use we had sufficient epoxy in the joint. Now give a smooth radius with your finger or properly calibrated Popsicle stick. The radius is critical we want it to be even and smooth this will allow even stress across the radius avoiding a stress concentrator leading to a crack.

DSC_0271_zps4ceb7e04.jpg



DSC_0272_zps6650ca98.jpg




A bond at 90 degrees for say a min diameter fin done the same way keep in mind your bond area goes way down and the strength of your epoxy becomes critical

DSC_0279_zps6b724107.jpg



Bonus what is water break:

A properly prepped surface will break the water tension on it. See how the water beads on non sanded area but lays flat on the sanded portion. This is a water break

DSC_0283_zpsf0743644.jpg




But flynfrog my tubes came ground doesn't this mean I don't have to sand them? The left side of this tube is as received. The right side is re sanded with 220


DSC_0287_zpsc6166679.jpg
 
Last edited:
Good stuff Ffrog-you have much more patience than I do posting. Umm- pic reversed on post? ^ left is ground, right (on our screens) has been sanded for the break. Nice pics-it's easy to see here. BTW-I used to build the aeroshells (nose cones) for Atlas way back in the day. All those chemicals may explain muy current lack of brain cells. Personal Protective equipment is mandatory here guys! (yes-even sanding-please!)
 
HALLELUJAH! THIS SHOULD BE IN THE BIBLE! Or a sticky at the very least. There is wayyy to much bad information about bonding out there. In almost every kit instruction manual I've seen, its recommended to really rough up both surfaces being bonded, nothing about degreasing or keeping a clean surface.

Alex
 
Good stuff Ffrog-you have much more patience than I do posting. Umm- pic reversed on post? ^ left is ground, right (on our screens) has been sanded for the break. Nice pics-it's easy to see here. BTW-I used to build the aeroshells (nose cones) for Atlas way back in the day. All those chemicals may explain muy current lack of brain cells. Personal Protective equipment is mandatory here guys! (yes-even sanding-please!)


Thanks fixed it. Good point on PPE a n95 dust mask would be a good idea and gloves when handling solvents.
 
Excellent. Now we can just link to this instead of having to write essays in each thread we come across someone epoxying incorrectly.
 
Excellent!

I will only say that for bonding to Quantum tube, some slight differences in surface prep are necessary. Q-tube is not a composite, and in fact does not chemically bond to epoxies. Therefore, the degree of roughness in the bond is crucial as it supplies mechanical obstructions to the joint failing; unlike in composites you really need to rough the joint as much as possible.

However, I imagine the vast majority of rocketeers making bonds that have to be well-done are using some sort of composites (Q-tube being a pretty crappy material for any performance-oriented projects), and your description is excellent and sufficient. =)
 
Another factor you mention that I have previously failed to consider is the importance of keeping bubbles out of fillets. In that respect, thick structural epoxies, especially opaque ones which hide internal bubbles, make it harder to ensure a defect-free fillet.

I guess that's one of the advantages of the runny Proline for fin fillets.
 
if you are using a mix tip you can keep the tip buried in the epoxy so that you don't pull air in.
 
Thank you for explaining in simple and concise terms exactly why I will never build a rocket that uses these types of materials and requiring this type of bonding.

Well done sir!!

Good luck with your future endeavors.
 
This is one of the best write-ups I've seen on best practices for composite bonding.

Thank you for taking the time to do this flynfrog!

Greg
 
Thanks! For a follow up, could you do a post on what things to look for in epoxy spec sheets?

Rule 3: Learn your glue not all epoxies are equal. They all have different properties learn what these mean and how they effect your bond.
 
Thanks! For a follow up, could you do a post on what things to look for in epoxy spec sheets?

If it really matters, it'll be relatively specific to your situation. (e.g. peel strength for surface-mounting fins, neat resin strength for fillets, temperature-versus-strength, pot life at various temperatures, viscosity)
 
Car Vac nailed it. I would also add toughness vs stiffness. Keep an eye on the data sheet when it lists these properties its rarely ever on composites.
 
Do you (or any one else) have data on the change in cured properties when adding fillers to epoxy? For example, what are the differences in cured properties between a fillet made from Aeropoxy structural paste and one made from aeropoxy laminating epoxy thickened with colloidal sillica? What about milled fiberglass? I know that they're "a bad thing", but I haven't ever seen data on how it changes the resultant properties or absolute strength.
 
I don't have any data as to what one is better. They are not bad though in the right volume your ultimate strength goes up. Also the longer the fiber the stronger it will be. Ill see if I can find any white papers on it if I do I will post it here.
 
Fillers are not mentioned in the datasheet, but they are in the MSDS if they are in any sort of volume! Check the MSDS of the paste vs the laminating system, you might find they use the same chemicals with a specific type of filler. Oftimes I've found you can DIY the filler to save buying another big expensive tub of epoxy ;) This of course goes out the window if the basic chemical structure of the epoxies is different...
 
x2, Great post Flynfrog

Do you (or any one else) have data on the change in cured properties when adding fillers to epoxy? For example, what are the differences in cured properties between a fillet made from Aeropoxy structural paste and one made from aeropoxy laminating epoxy thickened with colloidal sillica? What about milled fiberglass? I know that they're "a bad thing", but I haven't ever seen data on how it changes the resultant properties or absolute strength.

Nathan
 
Last edited:
Heat gun can be a win/lose proposition. One can briefly flash some epoxy systems and not hurt the final cure properties too much, but I wouldn't do it. The standard method (that I was taught anyway) believe it or not is to use a torch! The torch puts out enough heat that a quick pass over a tooling coat for instance will pop the surface bubbles, but without transferring much heat to the coat. A heat gun doesn't put out nearly the heat, so paradoxically one ends up raising the temperature of the coat to a greater degree since it must linger longer at any location to raise the surface temperature sufficiently to pop the bubbles.

But if you apply any sort of serious heat for any duration you will be messing with the properties, and not in a favorable direction. It will get weaker among other things. I do not know if this is as true with high temp systems such as cotronics though. If one can keep the temp below Tg then the effects may not be as bad. I haven't seen any data and I'm just speculating on that, so YMMV.

Have a vacuum system handy? One could vacuum degass if the working time is sufficient, the same as one would do with a propellant. Or one could do ultrasonic degassing if one has the equipment. Both of these would be done before one applies, for instance, a fillet. I would expect mixing in any additive would carry a great deal of air into the mix, weakening the final product. One could make some cured samples and get densities to find out how much air is included.

Gerald
 
With fillets, would using a heat gun to warm up the epoxy help in getting the bubbles out?

Probably not most adhesive epoxies we use are filled and have to high of a viscosity. If they are thin enough heat may just make them run.


G_T. There are many epoxies that require heat to reach the full properties
 
With fillets, would using a heat gun to warm up the epoxy help in getting the bubbles out?
It depends on the type of resin you’re using. When building experimental aircraft, all layups are accomplished by using a combination of heat and squeegees to “flow” the resin into the cloth while reducing weight. As flynfrog pointed out, the strength doesn’t come from the resin.

A word of caution:

If you use alcohol to clean your parts, test an area or piece of scrap first. Alcohol will attack some resins. That’s why we don’t use certain fuels in many experimental aircraft.

Really, you should test any cleaning agent prior to overall use.

When building with composites, it’s best to peal ply the location of joints. This provides an ideal bonding surface.
 
... G_T. There are many epoxies that require heat to reach the full properties

Of course, but there is usually a major difference between the temperature in the initial cure, and the temperature cycle for post-cure.

Gerald
 
Thanks for a very detailed explanation of composite / composite bonding - I'm sure this will be invaluable to all pure performance builders and provide a sound basis for sport flyers.

As CCotner observed, the bond process for Quantum (and to a lesser degree paper / card / plywood where epoxy is still a common, if in most cases inappropriate, adhesive) will be rather different and more mechanical bonding preparation is required. If you can find information on this type of bonding it would be of great benefit to everyone, though I doubt that there is much available in professional literature on bonding to Quantum type materials.

I notice the link to the original thread starting this post didn't work so corrected here, in itself it was an interesting read - glad to see there is now a sticky for this so it isn't lost.
 
The bond we are creating is on a molecular level. We are actually activating the outer electron shell atoms causing some to jump up a layer and leaving openings for our epoxy to fill creating a bond on a subatomic level.

This sounds very strange to me.
What is an "outer electron shell atom" ?
I am fully convinced that epoxy does no "subatomic bonding".

Nevertheless your tutorial seems helpful.
 
Two items come to mind when reading flynfrog's write up.

1. The question of how strong is the bonded joint should be considered. In the US aviation industry, for a composite bonded joint to pass tensile strength qualification, either of the two bonded substrates must fail before the bonding joint for the materials and process used to be considered successful and acceptable.

2. Industry has found that surface moisture has a very large effect on the final bond joint achieved. Most hobbyists probably won't try to control this, but keep in mind that large variability in achieved bond strengths are affected by the surface moisture on the two bonded materials. If you are using a non-polar solvent to clean a surface to be bonded, make sure it is dry solvent.
 
I am fully convinced that epoxy does no "subatomic bonding".
Yes, I agree that the statement is incorrect and/or confusing. Presumably he meant sharing of electrons (in the "outer shell"), which is characteristic of chemical bonds.

However, I'm still not convinced that epoxy makes a chemical bond with fiberglass (silica). It might, but it needn't; the epoxy creates a matrix in which the fibers are embedded. Any chemists here able to clarify?

And in general, I think the original post has a lot of value. I've been sanding with finer grit and cleaning the surfaces before sanding also as a result of reading it. Things I've learned here are reflected in my recent Epoxy Basics video.
 
Last edited:
Back
Top