What's new
Van's Air Force

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

Pondering if thermal expansion can break a canopy

Acrylic thickness

A couple more complicating factors:

The thickness of acrylic sheet stock as it comes from the manufacturer can vary slightly across it's surface. Part of our business is making acrylic aircraft windows (including RV12 rear windows for the local builders btw) and for say a typical 3mm thickness we have seen occasional variations from about 2.8mm up to 3.2 in the one sheet. Generally not a problem because many of the windows we make are relatively small so we can avoid any issues. But for the bigger canopies who knows, this might be contributing to the cracking problem.

Something else to consider is that when a canopy is blown, especially the longer narrow types such as the RV8, there is a lot of stretch and thinning of the acrylic happening at the highest point. Starting off with say 3mm acrylic could easily see it reduced to around half that thickness in some cases (think of inflating a balloon). Add in slightly thinner than spec acrylic to start with and this could go some way toward explaining why some installations fair better than others.

I always cringe a little when I think of the compression forces being applied when rivets are used to directly secure acrylic. Not disputing that many get away with it for years on end but it's not ideal. With bolts many hundreds of psi can easily be applied when over-torquing nuts and bolts. That's why good practice is to tighten to a firm fit and then back off a full turn of the nut. It only has to be lightly tightened. Also the hole in the acrylic should be half as big again as the bolt diameter with each bolt centred in the hole i.e. 4mm bolt - 6mm hole. This is another disadvantage of rivets because they try to expand to fill the hole as they are pulled.

Personally I have gone with Sikaflex for my canopy and screen. Trying to give myself the best shot at a trouble free installation. Time while of course tell!

Clive Whittfield
Auckland
New Zealand
 
Fast Back Canopy

Has anyone with the fastback mod had a cracked canopy? This mod would have a shorter canopy and get rid of the point where most cracks seem to occur.
 
I re-read this thread, and have some thoughts.

It seems we've been discussing stress concentration factors, which are typically used when, for example, one places a strip of material containing holes under tension. The stress is "concentrated" around the edges of the holes in the manner depicted here:

xFbW_MleTtCYLdLhorehAwKDjj9MQ5s6sF3Uk1rJ6F3r8lyajZyNHnyosy_6uEkUqXDw5Zs-e3aFeyRqCwYDvJZEydzuEzWgPhdLnKLe9uZxwd302W4j8KVvbtVAyydLnCFJ2IWeZszn-3i9EsFRxBwdb6YUJEHuh8M9qIPdX-U3ySWqiAfEcWf0EHjn2S5ySNc7zvSbC9mcr6faf5drKuAlHwcsZcYcrdReLGgp0IuQYgcw5v8DRkULpfNCyInhcl-MoykAV_qZbXXgMHE8XPeeYZ_HH4YPHD-_4s7rtJJn3Itx3YuivJx_wIo4K4QfGdn9MgGwmjVkw9cLTvlKY4mlg0jApdQndeizATk7tIzEPmw7xDpoP6HFxgTyx_sMKZrzz5gOO9Sj3UiJFtIdIhXeenL2Vx7wkYQ9Lf9SwQJ3mGqcLlaZtcdl-fBK_RMCs5X5fymo_aBU3mKZc2oVVY4ubnkI7LYeYqtKp368DlFUuJTF5permQkIe0sVrp0LxDBZHZfHeD0AtrwPc3Kzlt7kxBp2Ugg-f5peMXVMBAanUE9Ghzs5PVFn0nALamBKX63ZgYZKOjZo5a48ZlU4iGbTGCYqhTUqNVObAtKLzmaT89cbB2YA4R24DYDkEICxPeeC1yvDm349SRLnOvlr6ss4V1MvKd9rWA=w237-h269-no


However, this does not fit our situation entirely. The above depiction assumes the tension is not being applied directly on the edges of the hole. In the case of canopies attached entirely or partially by rivets, once the acrylic has contracted enough to contact the rivets at the extreme ends, the force is applied at the hole. We now have a local contact stress which may be extremely high.

Bill's interesting test of attaching a strip of acrylic to a steel angle might have a different outcome if the strip were wider, yet still attached in the same way. The resulting force on the fastener will be proportional to how wide the strip is. Our canopies are really wide, as analogized to this strip test.

I guess for me, what remains a mystery, is why they don't all crack... When I take mine from a 45F hangar out into 0F or so weather, the clicking it makes is horrifying, once the engine is running and the rapid cooling begins. The clicking is presumably the acrylic shifting relative to the frame, as the mounting scheme was designed to allow. If it stops clicking I need to worry.

It has been doing this every winter for 1600 hours/16 years, and hopefully it will continue to behave. The -8 is obviously worse case, given its length and the apparent mis-match to the curvature near the aft end.
 
My 6A canopy also had a serious mismatch to the frame at the rear, like .625 inches at the middle support. I just put plastic standoffs there and longer bolts thinking it would be a bad idea to cinch it down to the frame.

The "clicking" sound worries you initially but 15 years later with thousands of heating/ cooling cycles, mine is still crack free, rivets, bolts and all.
 
I re-read this thread, and have some thoughts.
I guess for me, what remains a mystery, is why they don't all crack... When I take mine from a 45F hangar out into 0F or so weather, the clicking it makes is horrifying, once the engine is running and the rapid cooling begins. The clicking is presumably the acrylic shifting relative to the frame, as the mounting scheme was designed to allow. If it stops clicking I need to worry.

It has been doing this every winter for 1600 hours/16 years, and hopefully it will continue to behave. The -8 is obviously worse case, given its length and the apparent mis-match to the curvature near the aft end.

Interesting... I have never heard my canopy click in frigid temperatures..I am assuming that is because the hangar is unheated, so things happen at a much slower rate. (or I'm not in the hangar when it's adjusting)

If we look at the current number of flying RV 8s at 1450 and assume 33% are enduring cold winters (I just yanked that number out of nowhere) so about 500. Lets yank another wild number out of the air and say that we have 20 or so cracked canopies (based on the 7 or so reported on this thread).... that gives us 4% failure rate. Not bad. (AND THIS IS ALL PURE CONJECTURE HERE)

If it really was due to thermal / attachment stresses I'm thinking there would be a lot more failures. Like other posters have said, its probably an accumulation of factors... flight loads... RATE of thermal changes, How well the canopy fit in the first place... Hole treatment on install.. etc etc.

I think if we install with care, the odds are with most of us that we wont experience the unhappy event.
 
I re-read this thread, and have some thoughts.

It seems we've been discussing stress concentration factors, which are typically used when, for example, one places a strip of material containing holes under tension. The stress is "concentrated" around the edges of the holes in the manner depicted here:

xFbW_MleTtCYLdLhorehAwKDjj9MQ5s6sF3Uk1rJ6F3r8lyajZyNHnyosy_6uEkUqXDw5Zs-e3aFeyRqCwYDvJZEydzuEzWgPhdLnKLe9uZxwd302W4j8KVvbtVAyydLnCFJ2IWeZszn-3i9EsFRxBwdb6YUJEHuh8M9qIPdX-U3ySWqiAfEcWf0EHjn2S5ySNc7zvSbC9mcr6faf5drKuAlHwcsZcYcrdReLGgp0IuQYgcw5v8DRkULpfNCyInhcl-MoykAV_qZbXXgMHE8XPeeYZ_HH4YPHD-_4s7rtJJn3Itx3YuivJx_wIo4K4QfGdn9MgGwmjVkw9cLTvlKY4mlg0jApdQndeizATk7tIzEPmw7xDpoP6HFxgTyx_sMKZrzz5gOO9Sj3UiJFtIdIhXeenL2Vx7wkYQ9Lf9SwQJ3mGqcLlaZtcdl-fBK_RMCs5X5fymo_aBU3mKZc2oVVY4ubnkI7LYeYqtKp368DlFUuJTF5permQkIe0sVrp0LxDBZHZfHeD0AtrwPc3Kzlt7kxBp2Ugg-f5peMXVMBAanUE9Ghzs5PVFn0nALamBKX63ZgYZKOjZo5a48ZlU4iGbTGCYqhTUqNVObAtKLzmaT89cbB2YA4R24DYDkEICxPeeC1yvDm349SRLnOvlr6ss4V1MvKd9rWA=w237-h269-no


However, this does not fit our situation entirely. The above depiction assumes the tension is not being applied directly on the edges of the hole. In the case of canopies attached entirely or partially by rivets, once the acrylic has contracted enough to contact the rivets at the extreme ends, the force is applied at the hole. We now have a local contact stress which may be extremely high.

Bill's interesting test of attaching a strip of acrylic to a steel angle might have a different outcome if the strip were wider, yet still attached in the same way. The resulting force on the fastener will be proportional to how wide the strip is. Our canopies are really wide, as analogized to this strip test.

I guess for me, what remains a mystery, is why they don't all crack... When I take mine from a 45F hangar out into 0F or so weather, the clicking it makes is horrifying, once the engine is running and the rapid cooling begins. The clicking is presumably the acrylic shifting relative to the frame, as the mounting scheme was designed to allow. If it stops clicking I need to worry.

It has been doing this every winter for 1600 hours/16 years, and hopefully it will continue to behave. The -8 is obviously worse case, given its length and the apparent mis-match to the curvature near the aft end.

Thermal expansion is in all directions and not just linear. The holes do get larger but not enough to account for the lack(in comparison to plastic) of expansion of the aluminum. The canopy would not have stress from simply getting warm and expanding but when it pushes, in a point fashion, against the fastener.

It seems that the canopy needs to float within the mounting system. Maybe using small slots instead of holes. Just commenting from a mechanical point of view and not from an experienced builder.
 
So, how do planes like the P-51 manage the big bubble? Do they float in the frame? I can?t find details on how they are put together. One photo of a NOS canopy shows holes in the front bow area, but I don?t see any along the lower sides.
 
The P-51 - first thing is, it's a 400+Knot airplane and has mush thicker plexiglass, and it might have a different composition for protection. I imagine it's not a 'blown' canopy. It's probably formed in a mold, which ensures a fairly uniform thickness throughout. This also ensures an accurate fit to its canopy frame. I think our problem with RV canopies, particularly RV8 canopies, is twofold. We have a method of attachment that is not ideal. Pulled rivets on plexiglass doesn't allow tempering the stress on the fastener. Secondly, the blown plexiglass canopy is almost never ideal in size and shape, relative to the engineered frame it sits on. There is almost always some stress during attachment to the steel frame, even if you are using sikaflex. This stress seems to be usually concentrated at the rear of the frame on the RV8. Starting the assembly at the rear v.s. the front may eliminate some of this stress. But stresses still exist in this type of assembly. Oversizing the holes will take care of some of this. Using screws that you can gently apply clamping pressure will help too. But the bottom line is, unless you have a plexiglass canopy that is fairly precisely formed in a mold to a frame, steel or otherwise, there's no way can can eliminate stress points, and possible failure due to varying homebuilt construction techniques with this type of structure.
If we were all willing to spend an extra $3K on our engineered canopy, we may have more favorable results. But given that the failure rate is 4%(numbers are suspect....) we are doing pretty good with an affordable canopy on our RV's.
 
The 2 person sailplanes I maintain have huge canopy's on them and they are all rigidly glued to the composite frame with epoxy glue, just like the beautiful fastback kit being produced by Showplanes. I contacted Showplanes to inquire about cracking and this was their response; "There has been over a 100 sold and shipped to all parts of the world. I have not had any customer report back with an issue of cracking from cold or hot temps. " If you study the pictures from their website, you will see that they drill the RV8 canopy for clamping clecos and it looks like larger holes between the clecos to act as epoxy pins when it squeezes out.

Dan mentioned earlier that the epoxy/glass (not carbon) frame has a closer thermal expansion rate to acrylic than the steel frame. Maybe there is something to this??

Phil
 
Good discussion. . . If you oversize the holes and avoid contact and stress between two holes - i.e. the plex has shrunk so that the outsides of those two holes now just touch the rivets or fasteners, the next two holes in each direction have already had that slack taken up and those two holes will be under stress + the subsequent holes even more stress. To never have any loads imposed on the plex, you would have to support it on floating fasteners or something that has the exact same coefficient of expansion. I'm thinking you'd be better off with "tighter" fit between fastener and plastic so stresses are all taken up hopefully evenly and at the same time?

This is certainly a factor combined with the fact all the holes (and restraint) are near the edge.

If your strip has flaws along the edges that are representative of typical canopy edges, this is a useful test. I wonder if there may be an effect of non-uniform stress distribution. Try bonding a big square of plexi along one edge to the steel bar. This replicates the large portion of the canopy that is not strained by the frame.

. . .
Bill's interesting test of attaching a strip of acrylic to a steel angle might have a different outcome if the strip were wider, yet still attached in the same way. The resulting force on the fastener will be proportional to how wide the strip is. Our canopies are really wide, as analogized to this strip test. . .

I really did not polish the edge and had a normal edge distance on one side, but there was 1+" of extra on the opposite side. A bit of stress analysis could say what the width vs force/stress really is, but definitely more than a narrow strip. Just happy I have aluminum frame on my 7.

Alex, nice picture there, I taught an applied mechanics lab as a grad student. I used photosensitive material and stretched the sample then used polarized light to reveal the lines of constant stress. All this way before FEA. The industry is light years more advanced now.

On another thought: the canopy might creep when hot and in compression, thus aggravating the stress increase when colder. This might happen more with steel than aluminum. I still wonder about external loading like latches.
 
Back
Top