Avoiding Canopy Cracking (Long Term)

Last weekend we had our local EAA chapter organized our first annual ?Wings & Wheels? event in Racine, WI. We had a few aircraft fly in, and I had the pleasure of meeting Dick Martin and the equal pleasure of seeing his beautiful RV-8 racing airplane. He pointed out only one problem ? a crack in the canopy about 2? long coming up from the side about halfway down the length.

My question is how many RV8/8A?s are out there with cracked canopies? This number is probably somewhat small with RV-8/8A?s being out there for only about 5 years.

Dick told me to attach it with 4-40 screws tapped into the frame rather than pop rivets. His reasoning that the screws won?t expand into the oversized holes in the canopy, whereas the pop rivets may. Has anybody used screws throughout with success?

I?m starting to consider gluing it on with Sikaflex based on what I?ve been reading on this forum. Are there any long term concerns of the glue degrading over time? I thought I remember reading somewhere that there was someone with a glued canopy for 6 years with no problems. Also, if the canopy is glued to the frame, then is Sikaflex also used to attach the canopy skirt? Should the wind screen be glued, or is the conventional screw method acceptable/desirable?

The goal is to use a method that will last several decades (50 years?) with no issues.

Thanks

glueing

Eric

I have personally talked to 3 RV-8 owners who have had cracks develope since they completed their aircraft. I know that it not a good statistical sample, but definitely help tip the scales to using SikaFlex for me and my partner.

As for longevity. The only downfall of SikaFlex is UV radiation, that is why the black primer. So (IMHO) as long as you account for blocking UV to the adhesive, it should last the life of the plane. It is the same adhesive used on automotive windshields (which by the way are a structural piece of the body), again you will notice the black band around your windshield. I have never heard of any original equipment windshields popping out of an automobile. Again not a scientific survey. In the end you should go with what you are most comfortable with. But, as a compromise, you could also put in a few screws for insurance. I myself am not using any screws, but did glue in a pin, at the very back of the canopy (where the "big pull" will be aerodynamically), in one of my alignment cleco holes.

Good Luck,
Wade Lively

Will Stika Stick?

Some folks wonder if the glue will really hold the canopy on.
I glued my canopy, windscreen, and skirts on. No screws no rivets. Flutter test RV-8 to 215KTAS and I didnt loose it. I only have 40 hours with most of it at 200KTAS. The last thing I am worried about is loosing the canopy.

Only time will tell whether or not the uv will break it down I suppose. Given ALL the cracked canopies Ive seen, this was a no brainer.

Best
Kahuna

Pop rivets do not swell in the middle as solid rivets do. They only expand on the end where the mandrel pulls in. That's why you need a proper sized, round hole when using pulled rivets. The canopy on my -6 has been doing well since 1992 with no cracks. The secret is to make sure the canopy is under NO strain when installed, use aluminum rivets and pull the rivets by hand. You don't want the "shock" of a pneumatic rivet puller on the canopy.
Mel...DAR

I'd think screws are still a better option than pop rivets. As a bonus, they'll be strouger than the afore mentioned aluminum pops.

Edited to correct URL to the correct adhesive ---

SikaFlex sounds like an interesting adhesive. Looking over the product data sheet at http://www.sikaindustry.com/tds-ipd-sikaflex295uv-us.pdf . The 295 does have a low tear propogation resitance of 5 N/mm per the ASTM D624. Converting Newtons to lbs, that is 1.12 lbs to tear a 1mm thick bond. For 2mm it woudl be double the force at 2.24 lbs. This means that IF a tear starts, there is not much strength to prevent it from propogating.

Given that a canopy seperating from the aircraft at high speeds (the most likely failure point) could result in damage to the vertical fin and possible subsequent loss of control followed by in flight breakup from an over loaded structure, it seems prudent to carefully consider how the canopy is attached.

Before relying only on any adhesive to save the airplane, one should have test data on longevity under typical loads, environmental conditions, proper surface prep methods, and on the actual substrates to be used. Otherwise you become the test. Some tests are more expensive than others.

Hi all, first post of what I hope to be many, as a future RV8 builder, with a machinist background.

My experience with plex. canopies is limited to schweizer gliders, and these always seemed to be developing cracks, which had to be stop-drilled pretty frequently. I don't recall how they were attached, but I don't want my RV canopy to exhibit the same.

I think cracks develop from a combination of thermal stresses, an initial attachment stress, then both issues compounded by vibration. I like the concept of a series of 4-40 or perhaps 6-32 (a coarser pitch being less prone to strip in aluminum) button head cap screws, with perhaps a UV resistant rubber grommet installed in a slightly oversized hole in the plexiglass. As the canopy goes through thermal stresses, the holes' interface with the rubber grommets will perhaps absorb these stresses without cracking. Just a thought. With screws, you can also remove the canopy if needed.

Brice

The correct SikaFlex product for the canopy application is 295UV.


Taken From Mickey's website www.RV-8.ch

"Hi Todd

Yes I definitely will try and glue. Matter of fact, talked with SIKA,
the folks mentioned in my previous email and the Sport Aviation article.
(Steve @ 1-888-832-7452).

Very, very nice person. First words were "Sorry, we do not support your
application", right after that, " now that I have said that, how can I
really help you............"

We talked for a while, Steve mentioned that his phone has been ringing
off the hook ever since the Sport Aviation article came out. He kinda
jokingly grumbled that he has had more response to an article SIKA did
not write for, a use SIKA does not support in a magazine SIKA did not
know about than what his total advertisement budget generated last year.

Steve said I should very carefully read the info at
http://www.sikaindustry.com/ipd-marine-window-295. and recommended to use
SIKA 209 to cover the plexi to make it opaque, SIKA 226 for cleaning the
frame and sikaflex 295UV BLACK for gluing the canopy on, trying to get
at least 1/8" space between the canopy and the frame.

He recommends that all structural bonding is done in the BLACK urethane
as it has much better UV properties. if needed, one can make filets with
WHITE urethane over the black stuff, 295UV comes in black and white colors."

How much of each?

RV8RIVETER said:

Steve said I should very carefully read the info at
http://www.sikaindustry.com/ipd-marine-window-295. and recommended to use
SIKA 209 to cover the plexi to make it opaque, SIKA 226 for cleaning the
frame and sikaflex 295UV BLACK for gluing the canopy on, trying to get
at least 1/8" space between the canopy and the frame."

Click to expand...

OK, I'm mostly convinced that I should attempt to SIKA my canopy, windscreen, and skirt. How much of 295, 209, & 226 will I need for everything?

Did you still drill and cleco some holes for alignment reasons? Looking at Mickey's website, I noticed that he glued the canopy on the frame, but did not yet bond the skirts. I wonder if it will be difficult to hold the skirts in place without holes. To those who did this, did you have a gap between the skirts and canopy somehow?

pneumatic vs. pulled

Mel said:

Pop rivets do not swell in the middle as solid rivets do. They only expand on the end where the mandrel pulls in. That's why you need a proper sized, round hole when using pulled rivets. The canopy on my -6 has been doing well since 1992 with no cracks. The secret is to make sure the canopy is under NO strain when installed, use aluminum rivets and pull the rivets by hand. You don't want the "shock" of a pneumatic rivet puller on the canopy.
Mel...DAR

Click to expand...

Interesting thought. Would it be possible to do a poll on those who have had cracked canopies and find out if they used pneumatic pullers?

No cracks in my 8A canopy after 340 hrs(pulled rivets). And it does get a fair amount of vibration from taxiing with canopy open.

Gluing the canopy

ericwolf said:

How much of 295, 209, & 226 will I need for everything?

Did you still drill and cleco some holes for alignment reasons?

To those who did this, did you have a gap between the skirts and canopy somehow?

Click to expand...

I think you can get away with one tube, but I used two, if I recall correctly. If I did it again I'm sure I could do it with one tube.

I didn't drill any holes in the canopy or use any clecos. I didn't really need any. Clamps hold the canopy on fine. Make sure you don't cut the canopy too much. Use the spacers that you will use when gluing when you are trimming the canopy. I used rubber garden hose.

I have not yet glued on the skirt, but I hope to get to it this week. I'm just finishing up the rear part of the skirt. That was a lot of fun! I'll use Scotchweld to glue the skirt to the canopy, and rivets along the canopy frame. No holes in the canopy at all. I'm not using sikaflex on the skirts because sikaflex wants a bit more gap than I want to give it for the skirts. Scotchweld does not need much of a gap.

I drilled 5 holes for clecos, because I wanted to make sure the canopy stayed in the exact same position on the frame as when I dry fitted. It also left me the option of installing a few screws if we felt the need to. I found that the rubber washers that seal the end of a garden hose work great as spacers. I just happened to have a box of them, and they measured out perfectly. I cut them in half and they wrap perfectly around the fram.

I bought 2 tubes, one white and one black. I have since had to purchase one more. I guess I am alot sloppier than Mickey . I used one tube to bond the canopy on everywhere there were no spacers. Then removed the spacers and filled in on the bottom side. I will probably have 3/4 tube left when I am done making it look decent, and I have filled in the entire profile thickness of the frame. One can of primer and cleaner is more than enough. I have not done the skirts yet, but will probabaly SikaFlex them on to the canopy as well, but will not go with the recommended gap, as on the frame. I don't think the gap is relevant here, since the skirts are non structural, flexible and I will be riveting/screwing the skirt to the frame.

rv8ch said:

I think you can get away with one tube, but I used two, if I recall correctly. If I did it again I'm sure I could do it with one tube.

I didn't drill any holes in the canopy or use any clecos. I didn't really need any. Clamps hold the canopy on fine. Make sure you don't cut the canopy too much. Use the spacers that you will use when gluing when you are trimming the canopy. I used rubber garden hose.

I have not yet glued on the skirt, but I hope to get to it this week. I'm just finishing up the rear part of the skirt. That was a lot of fun! I'll use Scotchweld to glue the skirt to the canopy, and rivets along the canopy frame. No holes in the canopy at all. I'm not using sikaflex on the skirts because sikaflex wants a bit more gap than I want to give it for the skirts. Scotchweld does not need much of a gap.

Click to expand...

Mickey,

Thanks for the response. I've been using your website as a guide for gluing the canopy. I just made the first rough cut on the canopy today and will have to decide soon the method of attachment. I'm 98% convinced that gluing is the way to go. Although I'm not much of an experimenter, I think that this is a reasonable risk as long as I keep an eye on it.

I can see how there is no reason to drill holes for aligning and gluing the canopy, but I was mostly concerned with glueing the skirts and keeping them properly aligned. I'll be very interested to see how you do it. I'm also a little concerned that the extra 1/8" gap between the canopy and frame may somehow affect the fitup of the skirts to fuselage. Maybe I'm getting worried over nothing.

Thanks Again!

glue plus metal fasteners ???

I have read all the threads related to glueing the canopy. I am at the point where I must make that decision....in fact, I have already made it (mostly). I have the sika materials ready. Recently I read that both Vans and Sika recommend against installing the canopy with only sikaflex, and recommend that metallic fasteners be used in addition.

Has anyone used some fasteners in addtion to the glue? Would the glue help prevent movement of the canopy which could lead to cracking? Is using metallic fasteners defeating the purpose of using glue?

One approach might be to drill a few holes where clecos could be used to affix the canopy to the frame to hold it in place for the glue process. After the glue sets, pop rivets could be used in those holes, far fewer than Vans process would call for.

Thoughts and opinions would be appreciated.

Chicken Fastners

A common term in the aerospace industry is "chicken fastener". This is for a fastener placed in a bonded joint. It has two purposes, to prevent the initial peel of the joint, since bonded joints don't do well in peel, and to arrest any crack propogation that might already be forming. They act as a failsafe in case the crack forms, surface prep wasn't perfect, bond line thickness/width was not controlled (earlier link to installation instructions recommends at least a 3/16" thickness and a 1.5 to 2" bond width), etc... Its called "chicken", but it is really "prudent engineering".

Take a look at the data sheet at http://www.sikaindustry.com/tds-ipd-sikaflex295uv-us.pdf. It has a peel tear propogation resitance of 5 N/mm or 1.12 lb/mm. So a 1 mm sample of the material will tear with a force of 1.12 lbs. For a 2 mm thick sample, the tear is 2.24 lbs, etc... Not exactly a tough adhesive, which is why they also recommend a 1.5 to 2" bond width.

Using fasteners in the bonded joint will be less likely to crack the canopy because the adhesive is distrubuting the loads more uniformally, whereas fasteners collect the loads locally.

chicken fasteners

Brice,

Thanks for that information. It seems that the use of "chicken fasteners" might be just the ticket. It would by a CYA in response to the CYA of Vans and Sika, in addition to being what seems good engineering.

It seems unlikely that you can get the recommended 1.5 to 2 inch bond width on the canopy frame...probably only .5 inch, if I correctly understand the term "bond width" to be the distance from the top of the joint to the bottom, which would be limited by the width of the canopy frame.

The next question is...what kind of fastener to use? Perhaps the pop rivet method that is normally used? And of course, the process and sequence....glueing first, then drilling through the sika and canopy frame?

CHICKEN FASTENER

I would think that screws would be a better choice than a pop-rivet as "chicken fasteners" in the canopy. Much easier to control stress or clamping force on the plexi, because they are only there in shear to control peel.

test monkey

It looks like I might become the test monkey for the chicken fastener concept. I am willing to do that if you engineer types and guys who have experience with the canopy fastening will help me figure it out.

This is the first time I am feeling the "experimental" part of building an experimental airplane. Perhaps we can forward the work done by the glue pioneers like Mickey Coggins, Eric Wolf et al.

The belt and suspenders approach has some excellent advantages. Mickey and the other guys acknowledge that the glue method is truly experimental and caution should be used in deviating from the plans in this regard. I was prepared to simply glue the canopy, but I now favor the chicken fastener approach. I am happy to hear that it is an accepted method in the aerospace industry. So, now it is time to figure out how to go about it.

RV8riveter, How did the screw method work out? What size screws did you use and what type heads? How did you prepare the plexi for the screws? I assume you used self tapping screws?

Countersinking the plexi would allow the screws to be set flush, but would be a weaker attachment it would seem. If you did not countersink, will the screw heads interfere with the skirt?

The method:

1. Install the screws first, using a standoff that will keep the canopy the correct distance from the canopy frame for the glue. A piece of tubing could be used as a standoff, or perhaps a piece of rubber of the correct thickness. The downside might be that the screw could then work in the plexi since the screw would not be going thru the glue.

2. Glue first, then add screws. This method would not assist in the fixing of the canopy to the frame before the glue is applied, but would have the advantage of having the screw in the glue which would keep it from working??

Brice, perhaps you could tell us the accepted method in the industry? That method is probably well thought out, but might not take into account that the finished surface should be smooth so as not to interfere with the skirt.

If the screw head is only about 1/8 proud of the surface, the skirt can still be attached using glue at the canopy/skirt joint with the skirt standing off the canopy 1/8 inch which facilitates the use of sika.

Any thoughts would be appreciated (except for those that tell me to forget the glue and build per plans)

Tony the Test Monkey

Sika on the canopy and skirt

Greetings,
I used Sika on the canopy and skirt, have 113 hours in all extremes of weather including significant rain with no problems or indications of even the slightest hint of coming loose.
I drilled the canopy and skirt as Van suggested, then followed the Sika instructions exactly to bond the canopy with rubber washer hose (1/8") as a standoff and used spring clamps to hold the canopy to the frame. I made a fillet with a tongue depressor on both sides of the canopy frame to increase the bonded area. I then bonded the skirt to the frame where it overlies the canopy with Sikaflex and used #10 washers as a spacer between the skirt and canopy directly under each cleco used to hold the skirt in position. When the Sika was set, I removed the clecos and pushed the washers out from between the canopy and skirt. I then put the Sika tube up against the hole (through the skirt, canopy and into the frame) and used pressure to extrude the Sika into all the voids left by the washers and push some Sika all the way into the frame. I used pop rivets where the skirt joins directly to the frame below the canopy.
Sounds complicated, but it was easy with masking tape to make very pretty fillets just like doing bathroom caulking - only bigger.
I have no metal fasteners holding my canopy, I would never want to do the canopy over again on this plane - it was painful enough the first time.

Bruce "FM" Edwards
RV-8, flying
200 HP, Dual GRT, Digiflight II VSGV
Flying naked (no paint yet)

If it were me

I have not gotten to that stage and am not building a -8 yet. You do know that if ALL of the designed fasteners were there, the canopy would be secure. We can reaonably assume that adding the Sika will help even out the loads so that they do not collect at the hole, and should help the cracking over just the purely fastened design.

The "industry" would conduct testing to see how few fasteners were needed to arrest crack growth or tearing. If there are too few fasteners and there is a failure, it will not stop until separation.

Then testing is needed to design the correct attachment method and should include the joint design, surface prep issues, standard defects, typical environmental conditions, typical UV exposure, and typical loading. Testing is best accomplished on the ground in the lab.

After the testing, a set of design parameters would be established, and strict manufacturing controls, specifications, and planning would be put in place to ensure that all the "as-built" conditions are covered by the testing. Most critical issues such as bond prep, priming, and bondline thickness and width will have tight controls. Inspection methods would be developed where possible.

Don't forget you might be glueing to the powder coating. Now you have to test the adherance of the Sika to the powder coating, and the powder coating to the frame. The powder coating then becomes a critical process with the same controls needed. Personally, I have seen the powder coating delaminate when riveting, so there is variation in that process.

What happens when the plexi heats up and begins to expand? The coefficient of expansion is much higher than the steel frame, and large shear forces can quickly be generated. The installation instructions warn of this and recommend bonding only small windows with the adhesive.

Consider this. A canopy separation can be fatal. How much are you willing to risk on the my advice and an experimental canopy attachment?

thanks

Thanks Bruce and Brice for your input.

The information about how Bruce attached the skirt is good to know. I like the washer trick.

Brice, you make good points. I had decided to go with glue, after carefully reading all the information I could find on the subject. The information you pointed out in your posts, regarding the minimum bond area etc is of some concern. Having said that, I understand that there are RVs which have flown for years with glued canopies...no failures and no cracks that I know of.

Vans standard method of attachment of the canopy has also produced a canopy that has never departed an RV to my knowledge, but many have cracked. As Bruce pointed out, you don't want to have to do the canopy over.

The extra safety margin of the chicken fasteners would remove any concerns over the long term safety of the canopy attachment, assuming that you used enough of them. Even if you used all that Vans recommends, the benefit of reducing canopy cracking potential is worth the glue IMHO.

I would think that using half the fasteners, or even less, would be reasonable. I am uncomfortable at this point in making the determination of how many to use because any number I came up with would only be a guess, without further information.

I will be removing the powder coat from all metal surfaces to be bonded. I do appreciate the point that all the bonding process must be without fault to create a strong bond. An adequate number of chicken fasteners would be insurance against a faulty bond.

I would like to hear from those of you who have used screws on the canopy...what size, what head, etc.

I just had the "pleasure" of having to removed a damaged canopy off my F1 project which was glued on with Sikaflex. I tried every which way I could think of test test the bonds, one of which was to lay on the cockpit floor and push on the canopy with my feet. I quit with that particular experiment when it became apparent that the frame was bending and the canopy itself was bowing. I don't think there is any need for any additional fasteners.

Regards,
Bob Japundza
RV-6 flying 850 hrs F1 under const.

Cracking the canopy

I am currently going through my A&P classes and realize that I am learning just enough to be dangerous at this point...that said, as I was reading some of the posts here concerning fastening the canopy, I thought I would relay some information we recently learned in class. I have not seen the install instructions that Van's sends, however, seeing that there have been more than a few cracking canopy's, the following might help.

When fastening a plastic to metal you need to understand that the plastic (in most cases) has a thermal expansion rate of two to three times that of steel or aluminum... meaning that if the screws & or rivets are cinched down tight, the plastic will expand at a greater rate than your fastener which will result in the cracks. Rivets are actually thought of as a poor method to fasten a plastic due to the severe compression it applies.
The taught method is to use screws with rubber or a flexable washer that will move with the plastic as it heats up. The book proceedure...tighten the fastener then back it off one turn. I would check w/ Van's to see if this would be accetable in this application.
Hope this helps,
Dan Billingsley
currently...Kitfox IV
Mesa, AZ

Tony

In reply to your questions...

I have not put in screws yet, and at this point hopefully never will. I was considering using countersunk self tapping screws that would be "glued" after drilling a whole in the plexi slightly larger. You need to keep in mind that the screws are like safety wire here and as such even countersunk will be plenty strong enough to keep the canopy from peeling.

After much thought and consideration I have decided not to put any "chicken" fasteners in.
1. In this day of vengefull litigation, there is absolutely no incentive for Van's or Sika to "approve" of the method. So that does not surprise nor bother me.
2. I "tested" a sample to destruction as well as communicated with others and all of their like experiences have been like Bob's above, unless you use a knife the plexi breaks first. While it is not bulletproof strong, it is bonded around the "entire" canopy circumference as well as double surface glued to the skirt, which in my case is glued and riveted to the frame. That is alot of area which adds up to alot of holding force.
3. -8's do not have a huge amount of aerodynamic force trying to blow the canopy off. The canopy is almost impossible to open more than a few inches in flight. Add that to the fact the fairings "overlap" the bubble so no air is being directed under the leading edges of the canopy.
There is no real source for "peel force".

So, I plan on being prudent enough to keep a regular critical eye on the bond and if I ever notice anything unexpencted going on, it will be a relatively simple matter to add the screws then.

DanB said:

When fastening a plastic to metal you need to understand that the plastic (in most cases) has a thermal expansion rate of two to three times that of steel or aluminum...

Click to expand...

This is provided for in the plans. The holes in the plexi are drilled out several sizes larger than the fastener to allow for thermal expansion.

Replace all the pop rivets with #6 screws, either drilled and tapped into the frame or with nuts. You can very easily control the amount of pressure with a screw.

SikaFlex

I used SikaFlex on my RV-8A. Flying fine. I do not think the concern with the peel strength is warranted. First, the loading scenario that would generate a high peel load has not been identified? You could have some fit up loads that would cause peal loads when you clamp the canopy edge to the frame. Mine were very minimal. The largest shear loads would probably come from thermal expansion, but thermal expansion would not generate high peal loads. With the fact that the big cut leading edge of the back bubble tucks under the bow of the front section, there are minimal airloads on the canopy that would put shear loads on the seam, and nothing to put peal loads.

Also, I used two small threaded bolts at the conopy tail to attach the skirt to the plexiglass. All else was bonded sikaflex. I also did not drill the holes, but used a heated small diameter rod to melt the holes. I do not know if this is an acceptable practice, but I have never developed a crack out of a hole I have formed by melting. The melting eliminates all cracks and stress raisers.

Sika and cracked canopy

I think this thread started with the need for statisitics. I meet a very nice guy at Sun N Fun with a beautiful job of sika attachment of his canopy. He develeoped a crack about 6" long just aft of his ear, started at the canopy edge and head straight up. It happened during a SEVERE bump near a storm.

Since I was the originator of this thread a long time ago, I thought I would chime in. I ended up bonding my canopy with Sika with no fasteners and I feel comfortable with this. Thanks to Tony and others searching for better/different ways to do this, this is how progress is made among us homebuilders. I never would have bonded my canopy if it weren't for Mickey's website. (By the way Mickey posted my procedure on his website here)

Anyway, here are a few points that I've made in the past and others have mentioned:
- Sika is strong and has been tested by a few builders by making small test sections
- The bond area is HUGE.
- The canopy is sandwiched between the frame and the skirts. The skirts can be pop riveted to the frame
- If the Sika were to degrade over a long period of time, it could be carefully removed and reapplied in sections without removing the canopy from the frame. This wouldn't be fun, but it would be much easier than replacing a cracked canopy
- The shear loading is kept to reasonable levels because Sika is flexible with the gap between the frame and canopy.
- Gluing the canopy to the frame should be done right to work well
- A perfectly installed riveted canopy will probably have a decent chance of not cracking, but I believe that gluing requires less skill to do an acceptable job.
- I believe the RV-8/8A is the worse case among the RV's since it is the longest. A side-by-side RV is probably less likely to crack if installed with rivets.
- Do what feels right to you.

what model

dbuds2 said:

I think this thread started with the need for statisitics. I meet a very nice guy at Sun N Fun with a beautiful job of sika attachment of his canopy. He develeoped a crack about 6" long just aft of his ear, started at the canopy edge and head straight up. It happened during a SEVERE bump near a storm.

Click to expand...

What model of RV was it...and was it a slider or tip-up.

Head hit canopy?

dbuds2 said:

I think this thread started with the need for statisitics. I meet a very nice guy at Sun N Fun with a beautiful job of sika attachment of his canopy. He develeoped a crack about 6" long just aft of his ear, started at the canopy edge and head straight up. It happened during a SEVERE bump near a storm.

Click to expand...

If his head hit the canopy, that would do it with any attachment method....

gil in Tucson

To Brice

Hi Brice, I sent you a private email via this forum. If you go to the top right corner of the VansAirforce page you can click on "Private Messages" and receive it.

PRC

ericwolf said:

Last weekend we had our local EAA chapter organized our first annual ?Wings & Wheels? event in Racine, WI. We had a few aircraft fly in, and I had the pleasure of meeting Dick Martin and the equal pleasure of seeing his beautiful RV-8 racing airplane. He pointed out only one problem ? a crack in the canopy about 2? long coming up from the side about halfway down the length.

My question is how many RV8/8A?s are out there with cracked canopies? This number is probably somewhat small with RV-8/8A?s being out there for only about 5 years.

Dick told me to attach it with 4-40 screws tapped into the frame rather than pop rivets. His reasoning that the screws won?t expand into the oversized holes in the canopy, whereas the pop rivets may. Has anybody used screws throughout with success?

I?m starting to consider gluing it on with Sikaflex based on what I?ve been reading on this forum. Are there any long term concerns of the glue degrading over time? I thought I remember reading somewhere that there was someone with a glued canopy for 6 years with no problems. Also, if the canopy is glued to the frame, then is Sikaflex also used to attach the canopy skirt? Should the wind screen be glued, or is the conventional screw method acceptable/desirable?

The goal is to use a method that will last several decades (50 years?) with no issues.

Thanks

Click to expand...

Proseal specific for windshields, and safe for acrylic and polycarbonate materials. This is the best "glue" you can buy.

http://www.ppg.com/prc-desoto/pdf/pr1425.pdf

Sealant failure on plexi window

Earlier this year a PAC 750XL turboprop in New Zealand lost the co-pilot's side plexiglass window in flight. The window then hit the tail...but fortunately there was no serious damage to the tail and the crew landed safely. The plane had 1,171 hours on the airframe

The window was held in place only by sealant. It was not the first time the manufacturer of the plane had experienced this problem. They had in fact changed their sealant brand after a previous failure.

The company now uses a mechanical frame to hold the windows in place.

They reported that the sealants failed over time due to UV degradation.

It is relevant that Sikaflex 295 UV is in fact not very UV resistant. But what is more relevant is that the Sika 209 primer for plexi is not UV stable at all and begins to deteriorate almost immediately. When the primer fails, the sealant will delaminate.

Sika's engineers advise builders not to use their product for this application without mechanical fastening. And yet RV builders who have done virtually no research and have no knowledge of mechanical engineering as it pertains to glazing go ahead and use Sikaflex against the manufacturer's recommendation.

Captain Avgas said:

It is relevant that Sikaflex 295 UV is in fact not very UV resistant. But what is more relevant is that the Sika 209 primer for plexi is not UV stable at all and begins to deteriorate almost immediately. When the primer fails, the sealant will delaminate.

Sika's engineers advise builders not to use their product for this application without mechanical fastening. And yet RV builders who have done virtually no research and have no knowledge of mechanical engineering as it pertains to glazing go ahead and use Sikaflex against the manufacturer's recommendation.

Click to expand...

Bob

Obviously, you feel it unwarranted to change Van's canopy assembly method fine. It is your airplane, build it any way you want. But as an unreppentant "gluer" I don't appreciate the comments and please don't presume to know how much research I put into my decisions or tell me how to build my airplane.

Van doesn't recommend Sika. What ever reason would he have to do so, none. To name a few. He also doesn't recommend aluminum gear, any engine other than Lycoming (oops I guess I am somewhat conventional after all), and does recommend and supply a cheesey (IMHO) fuel selector that could easily be turned off.

I guess since they are both windows you can compare a NZ turboprop side window with an RV-8 canopy??? Not to mention the fact that we don't even know what sealant and assembly method was used.

No reason to repeat what has been said in earlier posts. You brought up UV, which is a definite limitation that I am well aware of. If the sealant and bond area has an opaque covering by what mechanism is UV going to cause damage?

Refusing to be shouted down

RV8RIVETER said:

But as an unreppentant "gluer" I don't appreciate the comments and please don't presume to know how much research I put into my decisions or tell me how to build my airplane.

No reason to repeat what has been said in earlier posts. You brought up UV, which is a definite limitation that I am well aware of. If the sealant and bond area has an opaque covering by what mechanism is UV going to cause damage?

Click to expand...

Wade, I would draw your attention to some previous comments you have made on this same thread.

On 9/16/05 you said : "The only downfall of Sikaflex is UV radiation, that is why the black primer."

That is both incorrect and misleading. The black primer Sika 209 is purely to promote adhesion of the Sikaflex 295UV to polymethylmethacrylate (trade name Plexiglass). Sika 209 has in fact very little intrinsic UV stability and in direct sunlight it will start to break down in a few months. When it does so it will delaminate taking the Sikaflex 295UV with it.

On the 9/17/05 you passed on the tip from "Steve" to: "Use Sika 209 to cover the plexi to make it opaque."

People reading this comment would be led to believe (as you obviously did) that the function of the Sika 209 was to prevent UV getting to the Sikaflex 295UV. In fact what is required is a barrier to prevent UV getting to the Sika 209 because the Sika 209 is considerably more vulnerable to UV than the Sikaflex 295UV.

On the 5/8/07 you stated: "-8's do not have a huge amount of aerodynamic force trying to blow the canopy off. The canopy is almost impossible to open more than a few inches in flight. Add that to the fact the fairings "overlap" the bubble so no air is being directed under the leading edges of the canopy. There is no real source for "peel force".

Any "peel forces" or direct tension forces on the sealant due to live loads on the canopy are completely insignificant compared with the shear forces likely to be generated in the sealant by differential expansion. The linear coefficient of thermal expansion of plexiglass is approximately 8 times that of steel. If the sealant bead height is insufficient at all locations to accomodate that differential expansion by lateral flexing of the bead then the shear forces in the sealant will lead to failure at the sealant interface. Over time, and as the sealant becomes less flexible with age, the magnitude of the shear forces necessary to lead to failure will reduce.

The problem we have with the Sikaflex phenomenon is that a number of RV builders who have gone down that path want to "shout down" anyone who expresses reservations about the process. This is despite the fact that there are fundamental engineering problems with the procedure... and both the manufacturer of the aircraft and the manufacturer of the sealant recognise these problems and advise against using sealant as a sole means of canopy attachment.

It is relevant that the window on the PAC 750XL failed at just over 1000 hours flight time. In an RV that might equate to 10 years flying.

There are very few RVs out there with Sikaflex canopies that have many flight hours on them. Builders proposing to go down the Sikaflex path should be fully aware of that.

Not much point in arguing over Sikaflex before taking a look at the PRC offering suggested by Kevin.

The PR-1425 datasheet states "excellent resistance to UV" and lists a peel strength (specific to acrylic) of 214 N/25mm. That is about 70% higher than Sikaflex.

Shore hardness is a bit higher, at 55A vs 35A. Not sure of the ramifications. There is no reference to required bead height/thickness/ in the datasheet. There is no reference to UV problems in the PR-142 adhesion promoter datasheet. PR-1425 is a polysulfide product (think ProSeal family), not polyurethane

It is notable that PRC refers to the product as "windshield and canopy sealant", which is arguably not the same as "adhesive". In any case, I suspect the peel strength spec is a worst case approach. If you take a look at the PRC "Aerospace Sealants Glossary" you'll find the following definition:

54. Peel - A method of separating a bond of two
flexible materials, or a flexible and a rigid material,
whereby the flexible material is pulled from the
mating surface at a 90? or 180? angle to the
plane onto which it is adhered. The stress is
concentrated only along the line of immediate
separation. Strengths are expressed in pounds
per linear inch (pli) or inch width (piw).

I'd submit that the canopy application isn't strictly peel. I'd suggest that what we have is more of a shear or tension issue between two more or less rigid materials (lifting the canopy off the frame). Neither Sika or PRC lists a spec for that. Perhaps we could consider using the peel spec to be a conservative approach and the installed situation better?

Perspective: Using PR-1425, if you assume a canopy to have 10 feet of perimeter and the peel load "linear inch" to be parallel to the perimeter, and the canopy lift load to be evenly distributed, it would take 5760 lbs to pop the top. Yeah, yeah, I know, perfect case, assumptions, yada, yada....

PRC

The PRC product seems worthy of investigation. Does anyone know anything about it other that what has been posted? A google search turns up very little about it.

Whoa there

Bob

First, I am not out to "convert" or convince anyone that they should use any method not blessed by Van's. Since you went back thru you will see I did not say to anyone what they should do. I also, do not have a problem with hearing different opinions, data, facts, or conjecture on any subject and do not believe in "shouting down" as that does not serve any constructive purpose.

I do have a problem when comments are made that such in such is against manuf recommendations, you cannot understand the engineering, and you are going to die. Now yes, that is an exagerration but only to make my position clear. I should also admit I may have been in a bad mood that morning and taken too much offense to the comments, so to that end end I apologize.

As to the transcript. The date listed was almost two years ago and this method was/is new and learning the best way for the process is ongoing. However,at that time Sika literature stated that the 209 primer to promote bonding of the adhesive, yes, yes, and also stated that it would also help shield the adhesive from UV. I did not specify that a coat of exterior UV resistant paint paint suitable for plexiglass be applied to the exterior surface to block UV, but it is clearly in the Sika instructions. I would think it is plain to see, I never intended any of my posts to be step by step installation instructions. Approx 1.5 years ago I called Sika to get there help and information. The first thing the guy said was exactly the same as communicated on the 1st page, "we cannot endorse or recommend our product for this application, now what do you want to know". IMHO that is plain as day lawyer speak CYA. Given the fact that they at best would sell a miniscule fraction of their product to exp a/c builders their risk/reward ratio is so enormous that they have absolutely no logical reason to say anything else.

In my previous post, when I said "If the sealant and bond area has an opaque covering by what mechanism is UV going to cause damage?" I thought that would be sufficient. Since it is not, again as stated above, a band of UV resistant paint the bond length + twice the thickness of the glass should be applied to the exterior surface. Since we are talking about -8's the only exposed glass is the front edge of the bubble. I will also be painting any sealant not otherwise covered by trim.

I stand by my statement about the -8 canopy. Yes shear loads are predominately generated by the difference in expansion. I did not say otherwise. The bead of sealant on my canopy is the entire width of the frame and the entire length of the canopy surface. In addition, the skirt is riveted to the frame and glued to the canopy as well, sandwiching the canopy. I mentioned peel because that would be easier to say than localized simultaneous shear and tensile loads and that is where it seems to take the least force to induce a failure on a piece of test material with the plexi breaking first.

I repeat I am not telling anyone that Van's method is bad or that anyone should choose to glue their canopy. Am also not advocating Sika, it is just the product I found best at the time I glued my canopy. It is a choice I am completely comfortable with.

pr 1425 adhesive

Thanks to Kevin Johnson for pointing out that PR 1425 which is a sealant that is an alternative to sikaflex. I called the manufacturer today and this is what I learned from speaking with an engineer there.

PR1425 is a sealant for aircraft canopies. It is used by Cessna, Boeing, North American, Lockheed and others. It is not sold as an adhesive and is expected to be used with metallic fasteners. It does however possess very strong adhesive qualities.

It is sold in about three formulas, each with a different cure time.

B1- one hour working time
B1 1/2-1/2 one and a half hour working time
B2-2 hour working time

B1 and B2 are usually in stock.

It can be ordered in sizes:

Pint (12oz) about $73.
2 oz about $42
3.5 oz about $47

You also need PR 142 adhesion promoter


You first apply the adhesion promoter......the sealant is applied between 30 minutes and eight hours later.

No specific information was available as to the width of the bond, that is the distance between the canopy frame and canopy, for the best adhesion...remember that the product is not designed to be the adhesive to attach a canopy, only to seal it. The width of the bond, as a sealant, can be from 1/8 inch to 1/2 inch.

The material does not "run" as sikaflex does. This seems both good and bad. You have to push the material into the joint.

It has a peel strength of 5 lbs per linear inch.

It is resistant to water, alcohols, petroleum base and synthetic lub oils.


More information can be found here:

http://www.bergdahl.com/bapr1425.pdf

Sikaflex information can be found here:

http://www.jamestowndistributors.com/userportal/show_product.do?pid=3514&familyName=Sikaflex+295UV

One of the concerns about sikaflex seems to be the UV resistance of both the 295 adhesive/sealer and the 209 primer. However, the published information states that sikaflex 295 has a "high degree of UV resistance".

I asked the PC 1425 engineer about the UV resistance of that product. The answer was not very specific. It passes the ASTM standard test, which requires it to sustain 1000 hours in direct UV without failure. That is about 4 months on a ramp !!! Or 1000 flight hours for a hangared aircraft. I was suprised to learn that there is not meteric for this important characteristic.

I called the Sika engineer, who "knew the drill" and insisted upon talking about "boats" but was very helpful. I mentioned to him the issue of the 295UV product's UV resistance. He says the resistance to UV is excellent, thats why they use it on boats and it lasts out in the sun, all day every day, for years. Thats why it has the UV after its name.

I also asked about the rumor that the 209 primer was not very UV resistant and that it was alleghed that it could cause a failure of the bond. The engineer said he had never heard of such a thing. He also pointed out that in my type of "boat" the bead of product would be protected from the sun by the polycarbonate plexi which will remove about 98% of UV radiation.

I invite and encourage all of you chemical engineer types out there to take a look at both adhesives and render your opinions and observations. I will be making a decision between the two.

tonyjohnson said:

PR1425 is a sealant for aircraft canopies. It is used by Cessna, Boeing, North American, Lockheed and others. It is not sold as an adhesive and is expected to be used with metallic fasteners. It does however possess very strong adhesive qualities.

Click to expand...

If the manufacturer specifically advises against using this product without mechanical fastening....why would anyone ignore this.

tonyjohnson said:

I called the Sika engineer, who "knew the drill" and insisted upon talking about "boats" but was very helpful. I mentioned to him the issue of the 295UV product's UV resistance. He says the resistance to UV is excellent, thats why they use it on boats and it lasts out in the sun, all day every day, for years. Thats why it has the UV after its name.

Click to expand...

What is the name of this "Sika engineer" and what are his qualifications. Are you sure you're not just talking to some young gung ho sales rep with a quota to fill. In cars the edge of the windscreen is coated with a black ceramic strip to protect the polyurethane sealant underneath.

tonyjohnson said:

He also pointed out that in my type of "boat" the bead of product would be protected from the sun by the polycarbonate plexi which will remove about 98% of UV radiation.

Click to expand...

Plexiglass (brand name for acrylic...or polyethylmethacrylate) is not "polycarbonate". And the statement that plexi will "remove about 98% of UV radiation" is a gross oversimplification of a complex phenomenon and quite misleading.

Anyway, that's it from me. I'm going back to building my plane.

ok

Bob,

Please help me gain an understanding of the situation. You mentioned that the sikaflex primer 209 was not very resistant to UV radiation. The sikaflex technical person that I spoke with said that he had never heard that and seemed perplexed by my question.

Can you tell me where you go that information, along with the information that sikaflex 295UV sealant is also not very UV resistant?

tonyjohnson said:

Bob,

Please help me gain an understanding of the situation. You mentioned that the sikaflex primer 209 was not very resistant to UV radiation. The sikaflex technical person that I spoke with said that he had never heard that and seemed perplexed by my question.

Can you tell me where you go that information, along with the information that sikaflex 295UV sealant is also not very UV resistant?

Click to expand...

Tony, my information comes from Sika. And it stands to reason. Single pack polyurethane sealants are orientated towards low price and ease of use. The downside of that it that they are more prone to photochemical damage from radiation with a wavelength less than approx 500 nanometres. This degradation causes the polyurethane to embrittle, which in turn causes it to delaminate over time.

In cars it is not so much of a problem because car windscreens are made of glass and glass has a very similar linear coefficient of thermal expansion to steel. Therefore there is virtually no differential expansion and consequently internal shear stresses cannot build up significantly in the sealant.

I note that in post #39 you referred to your source of information as a Sika "engineer". However in your last post you are now referring to that source as a Sika "technical person". With all due respect (and I mean that) you are talking to a sales rep.

If you wish to be fully informed on this matter I sincerely recommend that you go back to Sika and ask to be referred to a person with scientific credentials. The magnitude of the modification and the consequences of failure demand that you get expert advice.

Tell that person exactly what you intend to do. Tell him what materials you are interfacing (acrylic sheet onto a steel frame) and give him the full linear dimensions (dimensions around the curved perimeter). Ask him if he would recommend the application from a scientific perspective. If he sees merit in the application then ask him to compute the sealant bead height for you.

My rough calcs suggest that bead height minimum will be approx 5/16" (and don't forget we're talking MINIMUM here...3/8" would be safer). From reading other posts on this issue I gather that most adopters have no more than 3/16". In some cases, such as on the RV7 slider, builders let that bead height dwindle to zero at the sides of the canopy due to physical restraints imposed by the way the kit goes together. Those installations will eventually fail. An inadequate bead height at ANY location is like a weak link in a chain. If the bead height is not high enough it does not have the capacity to flex sideways...therefore the differential expansion will be convered into shear forces at the sealant interface.

Potential Sikaflex adopters really need to do more homework before they commit to this very crucial modification. And I don't regard blobbing a bit of Sikaflex onto a scrap of plexi and trying to pull it off with bare hands as "research".

thanks

Bob,

Thanks for the information. I measured the canopy frame and found that the lower tube of the frame, where the canopy attaches is .063 from top to bottom. A bead height of as much as 1/2 inch should be possible.

The frame is 12 feet around. I did not count the bow in the front of the frame. The bow will be under the roll bar fairing and it would seem therefore not exposed to load.

It seems that there are two issues. The first is the size of the bead necessary. It looks like that will work out. The second issue is the degradation of the sikaflex as a result of UV.

Of course, this assumes that the sikaflex is applied properly and is not faulty.

The sales people at sikaflex referred me to the "technical" people. I confess that I assumed that person was knowledgable and did not ask his credentials. It was he who reminded me that the sikaflex would be under the plexi and therefore protected from UV and mentioned the 98% figure, which I also thought was high. When I asked about the various types of UV and their various penetration levels through the plexi, he did not know.

Do you have an estimate of the percentage of UV that would penetrate the plexi? My canopy is from Todd Silver and has a medium tint, if that makes a difference.

Thanks,

Tony

Specs....

tonyjohnson said:

......

Do you have an estimate of the percentage of UV that would penetrate the plexi? My canopy is from Todd Silver and has a medium tint, if that makes a difference.

Thanks,

Tony

Click to expand...

Tony... specs. for Plexiglas (TM) and UV transmission are in Fig. 5 here...

http://www.plexiglas.com/literature/pdf/81.pdf

It cuts out far UV (UV-B and UV-C, and only about half of UV-A)

However, Bob's previous comment of "less than 500 nm" wavelength starts in visible, not UV portion of the spectrum...

I believe the Plexiglas we use is either "G" or "MC", but I have heard of canopy makers using one of the UV grades for sailplane canopies... don't know what Todd"s Canopies uses...

Added...

This guy makes UV blocking sailplane canopies... it can be done...

http://www.thermotecusa.com/UV Block Info.htm

gil in Tucson

tonyjohnson said:

Bob,

Thanks for the information. I measured the canopy frame and found that the lower tube of the frame, where the canopy attaches is .063 from top to bottom. A bead height of as much as 1/2 inch should be possible.

The frame is 12 feet around. I did not count the bow in the front of the frame. The bow will be under the roll bar fairing and it would seem therefore not exposed to load.

Do you have an estimate of the percentage of UV that would penetrate the plexi? My canopy is from Todd Silver and has a medium tint, if that makes a difference.

Thanks,

Tony

Click to expand...

As long as we're on the same wavelength here. The "height" of the bead is the distance between the steel and the plexi. (ie at 90 degrees to the plane of the plexi). The "width" of the bead is the dimension parallel to the plexi. If you can achieve a consistent gap (bead height) between the steel and the plexi of 5/16" (at ALL locations) then I believe that would keep the shear stresses in the sealant within reasonable limits given a broad temperature range as might be experienced by an aircraft.

Also don't make the bead too wide. More is not better. Ask Sika to advise you of the max bead width.

To compute the bead height it will be necessary for Sika to know the four side dimensions (around the curves) of the canopy.

There are MANY different types of plexi with different radiation filtering qualities. If I was you I would not be counting on the canopy to protect the sealant. I would be using an opaque barrier on the outside of the canopy. Every part of the sealant needs to be protected....even edges (particularly edges).

thanks

Bob and Gil,

Thanks for the information. It seems that the sikaflex guy was right about the canopy filtering 95% to 98% of the UV, but only up to about 350nm.

I note that the damaging effect of UV drops rapidly, from a factor of 10 to about .06 as it approaches the visible range, but is still present at a significant level at ranges about 350nm.

Now that I understand the term "bead height". The sika guy said that the minimum bead height was 1/8, but I had intended to go twice that which is still a bit less than Bob suggests as the minimum. That would widen the canopy area by 1/2 inch and I am wondering what affect that would have on the fit of the canopy skirt.

The suggested bead height of 5/16" sounds a little high. Could we explore that number? I'll offer assumptions to get started; the accepted assumptions have a big effect on the outcome. I'm not a trained engineer so maybe our resident pros can throw in here.

A quick look at Machinery's Handbook says thermal expansion of plexi is about 0.000035 in/in/degF. Steels are about 0.000006. The issue is differential thermal expansion, which creates a shear load in the adhesive.

The RV8 canopy is about 6 feet long per side. In a pure bonded application, neither end (or any point) is fixed; the canopy floats on the adhesive. There are no shear loads due to differential thermal expansion at the mid point (3 feet from either end). Max shear is at the ends.

Let's also assume a temperature range of 0F to 120F with initial bonding at 60F...just because it is convenient. Thus the temperature variation from the intial is 60 degrees.

So, 36inches x .000035 = .00126" per degree
.00126 x 60 degrees = .0756" growth in the length of the plexiglass

36" x .000005 = .00018 per degree
.00018 x 60 = .0108" growth in the steel frame

.0756 - .0108 = .0648" difference in thermal expansion.

That's not huge (a fuzz more than 1.5mm), but with zero bead height it would certainly shear the adhesive. The question is how much bead height does it take to keep shear stress to a reasonable level.

BTW, consider that differential expansion and its ramifications for hole stress at solid fasteners.

Comments so far?

Solid fasteners

Dan,

If your figures are correct, the canopy would become 3/4inch longer after max expansion. That seems like a lot. It would seem that the canopy would certainly crack if fastened with pop rivets in that scenario...the first time it experienced thermal expansion.

I once listened to the testimony of a trooper who was testifying about the expansion of roadway materials during the summer in Florida. I asked him how much further it was between Miami and Orlando in the summer than in the winter. I think this may be the same concept. Does the canopy actually enlongate in a direct relationship to its thermal expansion "index"?

Perhaps your hypothesis is correct and the canopy does elongate that much.

I think that we are on the right track here in trying to gather all the available information and explore it and I appreciate your input.

This is a healthy and informative discussion for prosective gluers. My application has a bit of a different twist. I intend to use chicken fasteners, metal screws, though the skirt, probably an aluminum strip, the sika bond, and the canopy frame. If you are correct about the expansion of the canopy being that great, it would seem that those fasteners would cause a problem. They will be embedded in the bond, but will also pass through the canopy.

<<If your figures are correct, the canopy would become 3/4inch longer after max expansion.>>

No. If my figures are correct, the plastic canopy would grow 2 x 0.0756" = 0.1512 longer with a 60F a rise in temperature, or a little less than 5/32".

However, that's not the important dimension. We're concerned with the change in relative lengths of the steel frame and the plastic canopy. Further, consider only half the relative change for purposes of stress to the adhesive; there is no adhesive stress at the midpoint of the canopy length. Max shear is at both the ends.

thanks

Dan,

Thanks for the correction. I am not clear as two why the max stress will be on the ends of the canopy. On the 8, the front end of the slider will be under the fairing of the windscreen and not exposed to wind stress. That would leave the back of the canopy as the primary area of concern.

My plan was to put the chicken fasteners primarily on the sides.

How much would you figure the canopy will move in the area around the side fasteners?

BTW, I am away from home at a hotel, so my responses here may be slow.