What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Weldon 16 is cracking my canopy. Why???

dwranda

Well Known Member
I had a crack in my canopy and decided to fix it as well as I could using the weldon products. I drilled a hole using a plexi bit to stop the crack. I filled in the crack with the thinner weldon #3 and sanded it down so its smooth. I then filled the stop drill hole with #16. I checked it about an hour later and there were little spiderweb cracks emanating from the hole. I had no idea how they showed up. I drilled those out and just filled those holes with the 16 again. 10 minutes later more cracks showed up. I drilled a bunch of practice holes in some scrap and can't get any cracks to show up in the scrap pieces. Why is my canopy cracking when I put the weldon in the hole?
20191003-123501.jpg
 
Not sure, but my guess is that the weldon is expanding as it drys / cures. Try filling the hole with epoxy or clear silicone .
 
Because of the curvature internal stresses are high so its likely because of the holes you've stop drilled. I'm of the opinion that stop drilling cracks usually makes a crack worse. All it does in truth is shows others that someone has acknowledged a crack. I can't tell you how many times I've seen aircraft skins stop drilled with cracks emanating from stop drilled holes.
 
Weldon 16 is recommended for acrylic products but in fact it is a solvent based product containing a high percentage of MEK. It works by actually dissolving the acrylic. The problem here is that MEK is a double edged sword as it can break down the molecular structure of the acrylic and potentially cause additional stress cracking as you have found. The problem with the Vans canopies is that they are double curvature and the forming process leads to increased levels of internal stresses being locked in. Trying to repair cracks in Vans canopies can therefore be very problematic and is best done by a professional.
My guess is that you are going to end up replacing this canopy or you are going to have to live with something that looks terrible.
Many builders think that using the Sikaflex approach on their canopies will eliminate the problem of cracking but there is mounting evidence that this is not so.
 
Quick Trick/Tip

Used to work in a plastic shop in the 1980's.

Do you have any scrap pieces of the canopy? If you do; you can use a rasp to create little flakes, dissolve them in the Weldon and use as a filler. Thinking it used to work with the #3.

Never had any issues filling gaps or holes that way, thickens up as you add the flakes.

Only fill in a little at a time in thin layers if plugging a hole. Once done sand smooth with super fine sandpaper to get it to almost disappear.

Best regards,
Mike Bauer
 
Last edited:
Just use the search function. There are plenty of threads that detail cracks in Sikaflexed canopies. And this thread is just another one.

While you are correct that there are still cracks when using Sika wasn't there a poll not that long ago that showed the percentage of cracks was still less? My personal theory is that if you keep the primer off of the edge of the canopy you stand a better chance but hey that's just my theory. :D
 
While you are correct that there are still cracks when using Sika wasn't there a poll not that long ago that showed the percentage of cracks was still less? My personal theory is that if you keep the primer off of the edge of the canopy you stand a better chance but hey that's just my theory. :D

This is the most recent Sikaflex poll that I am aware of. The failure rate for Sikaflex canopies turned out to be amazingly similar to the failure rate for canopies attached with fasteners. This, of course, was not what the early adopters of Sikaflex were expecting. Cracking aside, canopies attached with fasteners have a 100% success rate over a 40 year period in terms of structural security while the jury is still out on canopies attached with Sikaflex.

http://www.vansairforce.com/community/showthread.php?t=157159&highlight=Sikaflex+poll

Incidentally, I think your theory about keeping the Sikaflex primer off the canopy edges is logical. The primer contains a lot of really nasty solvents.
In the final analysis however I think that what we are discovering is that cracking is the result of poor canopy construction techniques....regardless of how the canopy is attached. You can get away with a lot of terrible workmanship and riveting on an RV but the canopy is very unforgiving of poor fabrication techniques and impatience. I personally found it to be the most difficult and most demanding part of the build.
 
This is the most recent Sikaflex poll that I am aware of. The failure rate for Sikaflex canopies turned out to be amazingly similar to the failure rate for canopies attached with fasteners. This, of course, was not what the early adopters of Sikaflex were expecting. Cracking aside, canopies attached with fasteners have a 100% success rate over a 40 year period in terms of structural security while the jury is still out on canopies attached with Sikaflex.

http://www.vansairforce.com/community/showthread.php?t=157159&highlight=Sikaflex+poll

Incidentally, I think your theory about keeping the Sikaflex primer off the canopy edges is logical. The primer contains a lot of really nasty solvents.
In the final analysis however I think that what we are discovering is that cracking is the result of poor canopy construction techniques....regardless of how the canopy is attached. You can get away with a lot of terrible workmanship and riveting on an RV but the canopy is very unforgiving of poor fabrication techniques and impatience. I personally found it to be the most difficult and most demanding part of the build.

Ya that's more similar than I thought. Although I am curious as to the percentage of type out there (I would think there are A LOT more riveted than sikad at this point which the poll doesn't suggest, although I could be wrong here). As far as I know the Sika method is 100% as well if it was properly done but I haven't been around long enough to know this for sure. I totally agree with you on constructing the canopy, without out question the hardest part of my build too. I keep it indoors in a locked room in my house because I'm so paranoid of someone hitting it or doing some form of idiocy around it. I think I have about 160 hours into it and I still have to attach the windscreen.
 
As far as I know the Sika method is 100% as well if it was properly done but I haven't been around long enough to know this for sure.

That is the primary reason that Van?s has never endorsed the installation of canopies to flexible frames, and probably never will.
It is a pretty simple process to confirm that a rivet or screw is properly doing what it is supposed to do. An adhesive structural bond... not so much.

I agree that the canopy is the most challenging part of an RV build and everyone at Van?s will typically tell prospective customers that (with the cowling being a close second).
BTW, this is not something limited to building an RV. Most anyone that has built a kit plane design that has a large movable canopy will tell you that was the most challenging part.
 
That is the primary reason that Van’s has never endorsed the installation of canopies to flexible frames, and probably never will.
It is a pretty simple process to confirm that a rivet or screw is properly doing what it is supposed to do. An adhesive structural bond... not so much.

I agree that the canopy is the most challenging part of an RV build and everyone at Van’s will typically tell prospective customers that (with the cowling being a close second).
BTW, this is not something limited to building an RV. Most anyone that has built a kit plane design that has a large movable canopy will tell you that was the most challenging part.

Thanks for the expert comment Scott. I certainly understand Van's stance on this one but not fully. There are quite a few things while building the plane that must be done correctly for safe operation. After attaching my canopy to the frame with Sika I don't see how one could possibly screw it up royally unless a person skipped a major step. Just out of curiosity as I know you guys did some testing on this, did you find that if you skip certain steps such as scuffing, priming, cleaning, etc. that it resulted in sub-par strength? After following Sikas directions (which were pretty simple) it sure is amazing how strong this stuff is. Totally agree with you though that riveting/screws would be more "fool proof" for sure.
 
Last edited:
Well I just did some testing with a scrap piece. I bent it a little bit with some clamps to put some stress on it. Under stress even putting one drop of weld-on on the surface makes it developed cracks. Might be considered crazing? Putting a drop of number 3 on a piece that is not under stress does not create any cracks. I obviously still have some stress in that corner where the crack developed. Who knows how far up from the edge the stress is strong enough to continue to develop cracks. I'm experimenting with the scrap piece to figure out what I can get away with before I do anymore on the actual canopy.
 
Thanks for the expert comment Scott. I certainly understand Van's stance on this one but not fully. There are quite a few things while building the plane that must be done correctly for safe operation. After attaching my canopy to the frame with Sika I don't see how one could possibly screw it up royally unless a person skipped a major step. Just out of curiosity as I know you guys did some testing on this, did you find that if you skip certain steps such as scuffing, priming, cleaning, etc. that it resulted in sub-par strength? After following Sikas directions (which were pretty simple) it sure is amazing how strong this stuff is. Totally agree with you though that riveting/screws would be more "fool proof" for sure.

We have never done any testing of the Sika process on any of the two seat canopy installations.

The only testing done was with the bonding installation process recommended for the RV-10 which is a bit different since it is bonding the acrylic to a composite surface and the adhesive process used is a single part / step.

I don't disagree, that there are many critical construction steps in building an RV but I think that most all of them are very easily inspectable post process, to confirm that they were done correctly. Like my example of fastening something with a rivet vs Sika adhesive..... The rivet joint can easily be inspected by someone else, or even by the builder over time with doing condition inspections. A Sika bonded joint can be inspected to confirm the two surfaces are still joined together but that is about it.

The one exception made in the RV kits to the general line of thinking I mentioned in my other post is the bonding of the RV-10 windows.
I admit it is also process sensitive just like the Sika process, but to a lesser degree in my opinion. Fewer steps in the process mean fewer instances where a critical mistake can be made. Even then, there has been a few bond failures on RV-10's, but if we consider that 933+ RV-10's have been completed and flown, the failure rate appears to be extremely low so I believe that shows that the process is reliable.

If we had more long term service data on canopies installed with the Sika process, we may be able to say the same for it as well. But since my engineer mentality has to always look to data, right now there is not anywhere close to enough data to indicate that.
 
We have never done any testing of the Sika process on any of the two seat canopy installations.

The only testing done was with the bonding installation process recommended for the RV-10 which is a bit different since it is bonding the acrylic to a composite surface and the adhesive process used is a single part / step.

I don't disagree, that there are many critical construction steps in building an RV but I think that most all of them are very easily inspectable post process, to confirm that they were done correctly. Like my example of fastening something with a rivet vs Sika adhesive..... The rivet joint can easily be inspected by someone else, or even by the builder over time with doing condition inspections. A Sika bonded joint can be inspected to confirm the two surfaces are still joined together but that is about it.

The one exception made in the RV kits to the general line of thinking I mentioned in my other post is the bonding of the RV-10 windows.
I admit it is also process sensitive just like the Sika process, but to a lesser degree in my opinion. Fewer steps in the process mean fewer instances where a critical mistake can be made. Even then, there has been a few bond failures on RV-10's, but if we consider that 933+ RV-10's have been completed and flown, the failure rate appears to be extremely low so I believe that shows that the process is reliable.

If we had more long term service data on canopies installed with the Sika process, we may be able to say the same for it as well. But since my engineer mentality has to always look to data, right now there is not anywhere close to enough data to indicate that.

Thanks for clearing that up and giving us a bit of history Scott, many people including myself have wondered about this. Also, I was under the false notion that you guys tested Sika at one point so thanks for setting the record straight. Too bad no one has tracked Sika installations since they came about. If I am remembering my research correctly I want to say it started around 06'-08'. I based my decision in part on how long the Sika method has been around and it seems to be pretty common now with no structural failures to my knowledge. However, I agree that on a company standpoint it is a different ballgame.

Sorry for the thread drift dwranda but hopefully it was worth it to some degree.
 
No problem at all for the drift. In my novice opinion any canopy statistics have to be taken with a grain of salt. We all know these canopies are far from perfect. There is a lot of play in almost every dimension. Too many builder variations could come into play. Walking around Oshkosh I found it very hard to find a canopy that is perfect. There is usually an almost 1/8" gap or overhang somewhere. This is just something we agree to live with to have an awesome airplane in nearly every other aspect. I have a friend building a 14 and can't wait to see his canopy go together. I thought my canopy fit great on my frame when I glued it on but I guess not.
 
Putting a drop of number 3 on a piece that is not under stress does not create any cracks.

It probably is cracking, but at a microscopic level. I did some testing with methylated spirits on formed acrylic some years back. The acrylic looked OK with the naked eye but when I viewed the sample under my Olympus stereoscopic microscope at 30X magnification you could clearly see fine cracks emanating from the affected area. Then, flexing the sample caused the microscopic cracks to propagate by the “work of fracture” principle so that they became larger and visible to the naked eye. The “work of fracture” principle postulates that stresses at the tip of a crack can be extremely high.
In reality acrylic (poly methyl methacrylate) is prone to breakdown at the molecular level by a really wide range of chemical substances, including virtually all solvents.
My best guess is that your particular problem arose because of inadequate treatment of the canopy edge after cutting. The edge was probably not sanded sufficiently to remove all discontinuities at the microscopic level. Most RV builders do not understand the extent of sanding required to remove microscopic discontinuities at edges and penetrations of the canopy. In your case the solvents in the Sikaflex primer probably attacked the acrylic causing a crack to emanate from one of those edge discontinuities.
 
Last edited:
Here are some thoughts, admittedly purely speculative.

Internal stresses in materials are often relieved with “annealing”, which is the application of heat, allowing the molecular structure to “flow”. As the temperature rises, eventually the acrylic will melt. Before that temperature is reached, there may be a point where a slow stress relief occurs, without large scale distortion. If this is true, you would think it would be part of the manufacturing process. But even if so, the NEW stresses caused by installation of the canopy will not be neutralized by the previous annealing. So I’m wondering if an application of heat might anneal the new stresses caused by the installation process. I’m not recommending any particular procedure. Experimentation with scrap materials might shed some light on this. Even if positive results are obtained with experiments, scaling it to a whole installed canopy introduces the problems of temperature control, distribution of air temperature, duration of heating, etc. Perhaps relatively high ambient temperatures should exist during drilling, and installation.
 
Last edited:
My best guess is that your particular problem arose because of inadequate treatment of the canopy edge after cutting. The edge was probably not sanded sufficiently to remove all discontinuities at the microscopic level.

Bob,

What grade of sandpaper do you recommend for the final sanding? 400 wet? Or maybe something else?

Thanks,
 
Played around with some samples. Bent it with clamps so it was under stress. Put 2 drops of weld on 3 on the surface. Within a minute the surface cracks showed up under the drops. After that dried I heated the piece with a heat gun. I moved the gun back and forth about an inch from the sample for about 45 seconds to a minute. The surface was quite hot to the touch. After it cooled I placed 2 more drops on the piece. NO cracking. I must have heated it enough to relieve the stress and there is no distortion visible in the piece.
I will do 1 more sample then if that is also successful I will heat the corner of the canopy which should relieve any stress in that area. Then I should be able to fix the holes with weld on with no further cracking.
Has anyone else taken a heat gun to their canopy?
 
Bob,

What grade of sandpaper do you recommend for the final sanding? 400 wet? Or maybe something else?

Thanks,

I?d use wet and dry throughout. I?d start off with something quite coarse...say 80 grit. I?d use this to completely shape the edge into a continuous radius and to remove enough material to ensure that there are absolutely no little recesses left in the surface. All cutting techniques will leave little discontinuities (pitting) in the acrylic edge and you need to sand down beyond the depth of the deepest of these. Machining and cutting of acrylic also induces localised stresses so if you sand off enough material with the 80 grit you should also be reducing those local stresses. Then come over that surface with progressively finer grit...say 120, then 400, then 800. If you?re really fussy go to 1200. The more glass-like the surface of the edge is the safer it will be. It?s a little bit painstaking but not nearly as painstaking as getting a crack in your canopy and having to replace it.
 
Played around with some samples. Bent it with clamps so it was under stress. Put 2 drops of weld on 3 on the surface. Within a minute the surface cracks showed up under the drops. After that dried I heated the piece with a heat gun. I moved the gun back and forth about an inch from the sample for about 45 seconds to a minute. The surface was quite hot to the touch. After it cooled I placed 2 more drops on the piece. NO cracking. I must have heated it enough to relieve the stress and there is no distortion visible in the piece.
I will do 1 more sample then if that is also successful I will heat the corner of the canopy which should relieve any stress in that area. Then I should be able to fix the holes with weld on with no further cracking.
Has anyone else taken a heat gun to their canopy?

David, annealing of large thermoformed complex acrylic shapes such as glider canopies is common practice and is probably the only way to reduce internal stresses created by the forming process to reasonable levels. I have no idea if Vans canopies are annealed....and it wouldn’t surprise me if Vans don’t know either. At any rate the annealing of aircraft canopies takes a number of hours in an oven at elevated and very controlled temperatures. You will not be annealing your canopy by hitting it with a heat gun for a minute. In fact it is possible that by heating up only a small part of the canopy to a high temperature you might in fact be locking in further stresses.
 
Last edited:
Here are some thoughts, admittedly purely speculative.

Internal stresses in materials are often relieved with ?annealing?, which is the application of heat, allowing the molecular structure to ?flow?. As the temperature rises, eventually the acrylic will melt. Before that temperature is reached, there may be a point where a slow stress relief occurs, without large scale distortion. If this is true, you would think it would be part of the manufacturing process. But even if so, the NEW stresses caused by installation of the canopy will not be neutralized by the previous annealing. So I?m wondering if an application of heat might anneal the new stresses caused by the installation process. I?m not recommending any particular procedure. Experimentation with scrap materials might shed some light on this. Even if positive results are obtained with experiments, scaling it to a whole installed canopy introduces the problems of temperature control, distribution of air temperature, duration of heating, etc. Perhaps relatively high ambient temperatures should exist during drilling, and installation.

John, see my post #21 re annealing. You would think that Vans canopies would be annealed but annealing is expensive because it involves a number of hours of oven time at elevated temperatures....and Vans are very price driven. Maybe Scott can clarify this issue. He has been following and posting on this thread so he will read this.
 
Are Vans canopies annealed to minimise cracking.

I note that Scott did not respond as to whether Vans acrylic canopies were annealed or not after thermoforming to minimise intrinsic stresses that can lead to cracks. As I suggested, they probably don't know if their canopies are annealed or not. Or maybe they looked into it and found out that they have not been annealed. At any rate the silence is a bit deafening.
 
As far as I know the Sika method is 100% as well if it was properly done but I haven't been around long enough to know this for sure.

The Sikaflex method is only as good as the weakest adhesion link in the chain....and the weakest link on the sliding canopies could very easily be Vans powder coated steel. I don't see how builders can control that critical aspect of the adhesion. All I know is that some of the powder coating on my slider frame lifted off because it was not applied properly by Vans subcontractor. I was in Australia and returning the frame to the US was prohibitive so I had to have it chemically stripped and re-powdercoated locally. I believe the Vans powder coat lost adhesion because there was no proper passivation of the steel. Either that or the baking ovens were not running at a sufficient temperature to properly cure the powder coat batch...or the canopy frames were run through too fast and the base metal did not come up to proper temperature. There's lots of ways you can end up with powder coating with poor adhesion. When I asked Vans for a copy of their powder coating specification they told me they did not have one. For that reason alone I would be very reluctant to use Sikaflex on Vans powdercoated steel canopies.
 
Last edited:
Back
Top