Sanding Fiberglass Tube

The Rocketry Forum

Help Support The Rocketry Forum:

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

troj

Wielder Of the Skillet Of Harsh Discipline
TRF Supporter
Joined
Jan 19, 2009
Messages
14,951
Reaction score
1,277
So, I keep seeing instructions that says to sand the mating surfaces on fiberglass tube.

....and I keep asking myself why?!

The inside of the tube, it makes sense -- that's smooth, due to its contact with the mandrel. But the tubes we're using are filament wound, and have already been ground down -- sanding is just cutting more fibers, and repeating what the grinding process already did.

Washing and degreasing the outside of the tube makes complete sense. But why bother to sand it?

-Kevin
 
Does it say to sand the mating surfaces or ALL surfaces?

Sanding the mating surfaces gives the epoxy more grab, but sanding all the surfaces doesn't make much sense.
Ground tubes CAN be sanded to give them a smoother appearance. They arn't perfect right off the grinder. Most people don't bother.

Alex
 
I think they mean where one tube meets another. You sand it to square up the ends, which are often off by a bit (or a lot) depending on how the saw was set when the cut was made.
 
Before answering, go read the instructions for a few fiberglass kits -- you'll find they tell you to sand all surfaces to be bonded. I'm just pointing out that, while this is common, it makes no sense to sand the ouside on tubes that have been ground.

-Kevin
 
Before answering, go read the instructions for a few fiberglass kits -- you'll find they tell you to sand all surfaces to be bonded. I'm just pointing out that, while this is common, it makes no sense to sand the ouside on tubes that have been ground.

Kevin maybe this instruction step is a holdover from the predominant use of G10 from 4 or 5 years ago. Sanding in this case doesn't really compromise the structure and provides bite/texture that is needed with G10. I would agree though, for filament wound tubes this makes little sense.
 
Before answering, go read the instructions for a few fiberglass kits -- you'll find they tell you to sand all surfaces to be bonded. I'm just pointing out that, while this is common, it makes no sense to sand the ouside on tubes that have been ground.

-Kevin


I'm just saying that my limited experience with PR G12 tubes, the ends were not square. In one case it was off by 1/8" which is a lot. That leaves a big gap between the tubes. YMMV...
 
In composites you start to lose surface activation so your bond strength goes down. After 8 hours or so you can lose over 50% of your bond strength. In industry we do what is called a water break test. This can show you if your surface has been properly prepped.


I sent the below in a PM to someone a while back
To answer your question. Yes scoring will give you slightly more surface area but it will not increase your bond strength. When bonding to a composite you want to sand the resin off of the top layer exposing just the tops of the fibers. You are bonding in a molecular sense not a mechanical sense. You want to wipe off the sanded area with acetone until you get a clean rag both before and after you sand. Doing it before will remove mold release and oils from your hands so that you don't sand them into the joint. You want to bond as soon as possible after sanding no more than an hour or so. When you sanded the top of the fibers you excited the atoms you now want to add your epoxy to these excited atoms this will give you your strongest joint.

When you score with a saw blade or use to rough of a grit you will cut fibers. Once a fiber is cut it can no longer carry the load applied to it filling it back in with epoxy wont help either once its cut its cut. It would be like putting cuts in a rope.


Another important thing to remember when bonding is to make sure there are no voids in the bond line. This is done by placing the epoxy so that it is in a triangle cross section. You then clamps the parts together so you get squeeze out on all sides of the joint. you never remove the clamps until the epoxy has cured. Releasing the clamp will allow the parts to spring back and pull air in to the bond line.
 
I never thought of that, really. Of all the glass kits we have, I think only one of them was G10...

That being said, I sand around the fin slots for fillet adhesion, which again, really has no sense to it. I don't think I've ever sanded G12 in preparation for paint. Primer sticks to G12 very nicely.

This is going into the knowledge bank!



Braden
 
In composites you start to lose surface activation so your bond strength goes down. After 8 hours or so you can lose over 50% of your bond strength. In industry we do what is called a water break test. This can show you if your surface has been properly prepped.


I sent the below in a PM to someone a while back

Troj, I think this has something to do with it. Sanding before bonding to a plastic or cardboard airframe adds mechanical strength, but sanding before bonding to a composite tube, when coupled with proper surface cleaning, really ensures you get the full bond strength that your epoxy is spec'ed to, or if some of our crude measurements using ES6209 are accurate, even greater.

For almost every rocket in the hobby, that last 10% of bond strength, or for that matter the last %50 of bond strength, doesn't matter in the slightest, as most hobby rockets are pretty overbuilt.
 
Troj, I think this has something to do with it. Sanding before bonding to a plastic or cardboard airframe adds mechanical strength, but sanding before bonding to a composite tube, when coupled with proper surface cleaning, really ensures you get the full bond strength that your epoxy is spec'ed to, or if some of our crude measurements using ES6209 are accurate, even greater.

For almost every rocket in the hobby, that last 10% of bond strength, or for that matter the last %50 of bond strength, doesn't matter in the slightest, as most hobby rockets are pretty overbuilt.

While that may be true for optimal, the instructions presented with the kits certainly don't provide sufficient detail about cleaning for me to believe that's the case.

I suspect it's more of one of those things that one of the first kits said to do, and it's "just the way we do it". There may be (limited!) cases where it does matter, but in that case, the instructions provided are insufficient.

-Kevin
 
While that may be true for optimal, the instructions presented with the kits certainly don't provide sufficient detail about cleaning for me to believe that's the case.

I suspect it's more of one of those things that one of the first kits said to do, and it's "just the way we do it". There may be (limited!) cases where it does matter, but in that case, the instructions provided are insufficient.

-Kevin

Kevin,

I think part of this is an assumption that people building fiberglass kits know basic building techniques and how a rocket goes together, and won't put too much stock in what the kit instructions say, if they read them at all (e.g., Curtis kits don't come with instructions at all).

As you noted, Curtis grinds his tubes down from the dog-bone shape they are wound in (due to the fiber thread changing directions), and then finish grinds them smooth, far smoother than an 80-or-60-grit finish which is the basic requirement for properly bonding fiberglass parts.

While you could certainly get away with not roughing up the tubes for sub-Mach flights, it does make for a superior bond. In terms of "damaging" the fibers, it's no different than what happens during the initial grinding process; many fibers are ground such that they are non-continuous, which is why the Mongoose tubes have the "rattlesnake" appearance. If you take 80-grit sandpaper to a Curtis filament tube and then wipe it with alcohol, you will be able to feel and see the difference. In my recent no-layup 98mm MD build and flight, which was done to prove that good bonding techniques are all that's necessary if done right, I followed my usual procedure of 80, 60, and then hacking at it with an old blade, as shown in the thread. Having the bond occur at different depths as opposed to on one plane helps tremendously, as does varying the size and direction of the abrasions, with emphasis in the direction of main stress (for fin bonding, east-west across the tube). I'll also cut notches in the fin root, shown in the thread, as well. When roughing up, the order is important -- going from finer to coarser is important because the fine stuff is designed to level out imperfections, so going from blade-60-80 would be running in circles. Remove sheen first, then do the deeper ones. More than a few MD projects in and around this forum have failed in the last year or so due in large part to nuances like this which project hugely into the success or failure of the project.

For G10 sheet, the sheen from the manufacturing process of course mandates this same process to a more extreme degree.
 
Last edited:
Before answering, go read the instructions for a few fiberglass kits -- you'll find they tell you to sand all surfaces to be bonded.

-Kevin


Since I am a guilty party to the above, I will expound! Having written manuals for a couple dozen , here is why :

First I must think of the full range of building skills that the buyers may OR may NOT have. That means many of the buyers will be building their first fiberglass kit & at the other end of the scale those that are seriously experienced, need no instruction and will build it their own way anyhow....... in between is everyone else. They have built some stuff, but need help in the learning curve of highpower and av-bays. Soooo....

I have also seen too many rockets built with very little or NO attention given to proper surface prep, resulting in epoxy going on CR's,fins, vent band etc. with no sanding. Upon first ejection charge going off the CR pulls out with the recovery gear. A hard landing and fillets or fins pop off glassy smooth , and the list goes on. I/we have seen it all. To alleviate this and not have a super lengthily manual that covers every aspect of building, it comes down to: covering the manufacturers butt, give all surfaces that will be bonded with epoxy a good hard sanding.

Remember this too....for many it will be their first experience with epoxy! Good building skills always require 90% preparation and 10% finish work. Really the only surface not requiring sanding is outer tubes. All inner surfaces, and everything made of G-10 plate does, and that is 90% of a standard build. So it is simpler to just say "sand it all"

So yes there is a grey area here that can be left out, but for beginners and those with mid-level skills it is far better to have them sand & prep the Cr@p out of everything, rather than second guessing, missing the boat and screwing up a build.

When one has reached that level of building nirvana where they don't need help or advice, then so be it. They can fend for themselves, reap the rewards or suffer the consequences of their actions. [build skills]

I prefer to error on the side of caution and can safely say: IF you build it, the way I describe in my instructions, you WILL succeed and have a SAFE, durable long lasting project.

Once again I must also add this disclaimer which many of you are familiar with: "there are many ways to do this, this is just mine/ours, and not saying it's the only or the best, but it will get you a quality end result in a simple, straightforward, sensible manner".

Follow that up with the lifetime Guarantee offered on many of these kits & you have a winning combination.

That's my answer to your question & I'm stick'n too it.LOl :smile:

As far as other authors of instructions, I cannot speak for them.

CJ
 
Kevin,

I think part of this is an assumption that people building fiberglass kits know basic building techniques and how a rocket goes together, and won't put too much stock in what the kit instructions say, if they read them at all (e.g., Curtis kits don't come with instructions at all).

As you noted, Curtis grinds his tubes down from the dog-bone shape they are wound in (due to the fiber thread changing directions), and then finish grinds them smooth, far smoother than an 80-or-60-grit finish which is the basic requirement for properly bonding fiberglass parts.

While you could certainly get away with not roughing up the tubes for sub-Mach flights, it does make for a superior bond. In terms of "damaging" the fibers, it's no different than what happens during the initial grinding process; many fibers are ground such that they are non-continuous, which is why the Mongoose tubes have the "rattlesnake" appearance. If you take 80-grit sandpaper to a Curtis filament tube and then wipe it with alcohol, you will be able to feel and see the difference. In my recent no-layup 98mm MD build and flight, which was done to prove that good bonding techniques are all that's necessary if done right, I followed my usual procedure of 80, 60, and then hacking at it with an old blade, as shown in the thread. Having the bond occur at different depths as opposed to on one plane helps tremendously, as does varying the size and direction of the abrasions, with emphasis in the direction of main stress (for fin bonding, east-west across the tube). I'll also cut notches in the fin root, shown in the thread, as well. When roughing up, the order is important -- going from finer to coarser is important because the fine stuff is designed to level out imperfections, so going from blade-60-80 would be running in circles. Remove sheen first, then do the deeper ones. More than a few MD projects in and around this forum have failed in the last year or so due in large part to nuances like this which project hugely into the success or failure of the project.

For G10 sheet, the sheen from the manufacturing process of course mandates this same process to a more extreme degree.


I really don't want to start a pissing match over how to glue toys together As stated earlier most rockets don't need the last 50% of the bond strength. But..... I come from the composite aerospace industry. The advice you are giving is miss guided and incorrect. Im not necessarily trying to change your mind but give every one else the opportunity to decide for them selves. anything rougher than 120 grit is too rough. You are looking for surface activation not mechanical scoring. If you don't believe me take a look at any aircraft manufacture spec or mil handbook 17 you wont find scoring with a saw blade anywhere. The substrate you are destroying with a saw blade has a higher shear strength than a glue you are trying to fill the gaps with. This is why bond line thickness control is so important. Most manufacturer specs will specify a 120-240 grit bond prep. As I stated earlier its a chemical bond you are after not a mechanical bond. This is why bonding as soon as possible after sanding is so important. Don't take my word for it.


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

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
 
I really don't want to start a pissing match over how to glue toys together As stated earlier most rockets don't need the last 50% of the bond strength. But..... I come from the composite aerospace industry. The advice you are giving is miss guided and incorrect. Im not necessarily trying to change your mind but give every one else the opportunity to decide for them selves. anything rougher than 120 grit is too rough. You are looking for surface activation not mechanical scoring. If you don't believe me take a look at any aircraft manufacture spec or mil handbook 17 you wont find scoring with a saw blade anywhere. The substrate you are destroying with a saw blade has a higher shear strength than a glue you are trying to fill the gaps with. This is why bond line thickness control is so important. Most manufacturer specs will specify a 120-240 grit bond prep. As I stated earlier its a chemical bond you are after not a mechanical bond. This is why bonding as soon as possible after sanding is so important. Don't take my word for it.


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

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

I feel like I should be taking your advice. I know I hate it when people argue with me about the industry that I am in and have been involved with for over half my life (metal fab). It really does make sense and I am sure you know what you are talking about but I know in my personal experience sitting there with the parts in front of me even after having read your post in the past, after I scuff eith 120 I think "theres no way the epoxy will grab that" so I want to go to 60 or 80 so I can see some bite. So I can completely understand why people want to scuff the crap out of parts. But if you know the industry and play with the big boys I think I should be listening to you. Afterall I haven't seen many others making tubes looking as nice as your "homemade" tube. So, my ne t high speed project I will go with 120 and thats it. I want to prove to myself thats the way to go. To be clear I'm not trying to say yoyr wrong ij any way just saying how I feel when building
 
Holy CRAP flyinfrog! That's some awesome information you present. I am going to have to study some of those articles a little bit more. If it is of any worth, I've never scuffed with larger than 120 grit, BUT I've also never thought about degreasing much!!!
-Ken
 
But if you know the industry and play with the big boys I think I should be listening to you.
He's right.

Degrease. Scuff. Degrease again. After that, no bare hands. No masking tape. Bond as soon as you can.

Think about the texture that peel ply leaves behind -- that's more than rough enough for bonding.

Bag everything. Cure your layup at the right temperature schedule. Design your layup with materials to handle the dynamic pressure and temperature ranges expected.

This held the fins on a composite rocket at M=4.2. No saw blades, no 60 grit, no notches or stress relief. Just proper layup design, bond prep, cure schedule, and vacuum bags.
 
Interesting reading! Now, I'm really glad I asked. :)

The answer, in summary, seems to be...

1) Yes, it should be sanded, but not with 80 grit -- nothing more coarse than 120.
2) Thorough cleaning and degreasing after sanding is imperative.
3) Sand and clean right before bonding.
4) Clamping is critical.

So, it looks like there's definite room for improvement in the "standard" way things are done in the hobby.

Good info!

-Kevin
 
Interesting reading! Now, I'm really glad I asked. :)

The answer, in summary, seems to be...

1) Yes, it should be sanded, but not with 80 grit -- nothing more coarse than 120.
2) Thorough cleaning and degreasing after sanding is imperative.
3) Sand and clean right before bonding.
4) Clamping is critical.

So, it looks like there's definite room for improvement in the "standard" way things are done in the hobby.
Add a step:

0) Thorough cleaning and degreasing before sanding is imperative.

Otherwise you end up sanding any oil/grease/release/goobers into the fiber matrix, which makes it nearly impossible to get out.

And of course, disclaimer -- if it's not going M2+, you can relax some of these requirements. When I was building my Gizmo, for instance, the free hand I would have used to degrease with was instead occupied by a beer. Which made it much more fun to work on.
 
When I was building my Gizmo, for instance, the free hand I would have used to degrease with was instead occupied by a beer. Which made it much more fun to work on.

I think that's just a. Matter of priorities :)
 
To put it all in one posts I agree David on my gizmo I kinda degreased and kinda scuffed.

Proper bond prep.

Step 1 Clean Part. Acetone wipe until rag comes off clean. Follow with Iso. Alcohol until rag comes off clean. On a side note rags are important under no circumstance should you ever use shop red cloth shop towels. they are coated with some evil oil that is impossible to remove.

Step 2 scuff 120-240 grit just remove the top layer of resin until its not shiny. Sanding should be done in the direction of fibers. (water break test if you are going pro)

Step 3 Clean again. repeat step 1. never touch with your bare hands at this point.

Step 4 bond Put you glue down so that you don't entrap any air then clamp. You should apply sufficient matl. that you get squeeze out with no breaks the full perimeter of you your bond. ( we throw away probably 1/3-2/3 of all glue we mix.) Use a gloved finger or other device to apply a smooth bond radius. With the joints we typically see in rockets a proper radius is probably worth most of your bond strength.

Step 4 cure Follow the directions. Epoxies are meant to have specific conditions when they cure follow them.

Step 5 Unclamp clean up.

Step 6 frosty bevrage
 
To put it all in one posts I agree David on my gizmo I kinda degreased and kinda scuffed.

Proper bond prep.

Step 1 Clean Part. Acetone wipe until rag comes off clean. Follow with Iso. Alcohol until rag comes off clean. On a side note rags are important under no circumstance should you ever use shop red cloth shop towels. they are coated with some evil oil that is impossible to remove.

Step 2 scuff 120-240 grit just remove the top layer of resin until its not shiny. Sanding should be done in the direction of fibers. (water break test if you are going pro)

Step 3 Clean again. repeat step 1. never touch with your bare hands at this point.

Step 4 bond Put you glue down so that you don't entrap any air then clamp. You should apply sufficient matl. that you get squeeze out with no breaks the full perimeter of you your bond. ( we throw away probably 1/3-2/3 of all glue we mix.) Use a gloved finger or other device to apply a smooth bond radius. With the joints we typically see in rockets a proper radius is probably worth most of your bond strength.

Step 4 cure Follow the directions. Epoxies are meant to have specific conditions when they cure follow them.

Step 5 Unclamp clean up.

Step 6 frosty bevrage

Thanks, I will use this method from here on out unless my fins fall off at mach 1.2... JK I bet they will hold up well past that.
 
another thing to keep in mind not all epoxies are equal some are more suited to bonding fins than others. With the larger load bearing fillets you may want to look at something that is filled like a 9394NA. The problem is they can be on the brittle side so maybe something with a little more flex like 9462 might work better. 3m DP430 is also a good choice. On my MD bird I used JB weld.
 
I have used a lot of JB weld in the past. It has worked fine for me up to 1600+ mph. However I have found that it seems to have a problem with hard impacts with the ground. it just cracks down the center. Lately I have been using Proline just because I acquired some in a trade, and I really like working with it and it has been good past mach 2 as well and seems to handle the impacts with the ground better. However I did lose a fin at LDRS when, what I speculate happened, the motor which deployed at apogee attached to the shock cord swung around and knocked off my fin. The fillet on most of the length of one side was just clean stripped off from the fin and still attached to the body, like there was no bond whatsoever to the fin. It did survive the flight of what i would guess was near mach 2 (will never know for sure because my accel data was all garbage for some reason) Now I am out of proline and have placed an order for some of the new rocketpoxy so I will see how that works with your method of prep. Thanks again for your advice.
 
I really enjoyed reading flynfrog's attachments. I'll be honest, a bit was over my head/level/concern for slower flights but I dealt with Young's equation in soil physics regarding water droplets on soil. Never thought that would translate into epoxy on a composite substrate!!! Thanks again Troj for asking the question and for all who have replied with great info.
-Ken
 
With so much bad information out there about surface prep, this really ought to be a sticky.
Everyone needs to know this.

Alex
 
To put it all in one posts I agree David on my gizmo I kinda degreased and kinda scuffed.

Proper bond prep.

Step 1 Clean Part. Acetone wipe until rag comes off clean. Follow with Iso. Alcohol until rag comes off clean. On a side note rags are important under no circumstance should you ever use shop red cloth shop towels. they are coated with some evil oil that is impossible to remove.

Step 2 scuff 120-240 grit just remove the top layer of resin until its not shiny. Sanding should be done in the direction of fibers. (water break test if you are going pro)

Step 3 Clean again. repeat step 1. never touch with your bare hands at this point.

Step 4 bond Put you glue down so that you don't entrap any air then clamp. You should apply sufficient matl. that you get squeeze out with no breaks the full perimeter of you your bond. ( we throw away probably 1/3-2/3 of all glue we mix.) Use a gloved finger or other device to apply a smooth bond radius. With the joints we typically see in rockets a proper radius is probably worth most of your bond strength.

Step 4 cure Follow the directions. Epoxies are meant to have specific conditions when they cure follow them.

Step 5 Unclamp clean up.

Step 6 frosty bevrage

I take it this goes for carbon fiber and fiberglass. Does the same apply with phenolic and quantum tube as well?

Thanks (I'll put down the 80 grit now in favor of 120).
 
I am not sure what works best my guess you fail the tube long before the joint failure. Almost all of the data I have is based on composite to composite or composite to Aluminum.



If I get some time after Hellfire Ill write a proper post with pictures ect for sticky if you guys would like.
 
Agreed. I would have loved to have seen a post like that a few years back when I first got back into the hobby and there was glass and carbon kits readily available. I think it will be very helpful to those just getting into glass and carbon.
 
Back
Top