Bonded door skins and nut-plates

[Nev]

I bonded my forward cargo door skin to the frame last week, after much consideration. Doing the rear one first would have been preferable in my mind with the over-lap, but it didn’t work out that way and the rear door is done with countersunk stainless rivets which also gives a very nice finish.

Obviously the failure mode is of prime consideration and in the case of the forward door it would probably be rather rapid, but not catastrophic. I believe the likelihood of this to be low and the finished strength appears to be as strong or stronger than rivets. There’s significantly more surface area attached than with rivets, and little chance of skin cracking etc.

Inside the door I’ve added a bonded grab-latch to assist with opening. This should take as much stress as any of the bonding due to the latch spring strength, and I think will act as a “canary in the coal mine” as the adhesive ages.

For the bonding I’ve used Scotchweld DP190. It has a small flex to it rather that complete rigidity.
Most of my time was spent in researching how secure the installation would be, and how best to prepare the surfaces. Actually once I decided to proceed it was all relatively straightforward. All bonding surfaces were either kept bare or in the case of the door frame, a strip was taken back to bare steel, then scuffed, degreased and cleaned. The DP190 goes on easily, and is able to be handled after around 12 hours. I kept the parts at room temperature (dining room temperature in fact) for several days, full cure takes a week.

I welcome all feedback, concerns or otherwise. As always, if you see something that I don’t, please tell me. If I can find a way of attaching a video later I’ll do that too.

561D0533-1A23-4449-AC7A-99297B24BAC4.jpeg

376FCC29-2926-48B1-851F-F4D8C4711BAA.jpeg

[AKKen07]

Looks great Nev, thanks for posting! I like the door latch details too. Nice.

[Battson]

Sorry to be a 'negative Nancy'....

What led to using glue, instead of riveting like the cargo door? I see risks; struggling to see the advantages.

I glued some components in place on my rear door, like the latch cover. These came loose after about 300 hrs and had to be re-glued, which never works. Vibration beats glue in the "paper-scissors-rock" that is wear and tear.

With door skin - all the forces on that door skin are trying to push and suck it off the fuselage. Particularly once you install gap seals.

If that door skin begins to peel off at the leading edge and projects into the prop wash a little, then the drag could unpeel the rest in a fraction of a second. I have seen this happen in a car I was driving, it happens suddenly and without warning.

If that door comes off, then:

• Best case only your tailplane fabric gets damaged.
• If it folded over the leading edge, then you're going to have a small yaw problem and half your elevator authority gone
• If it folded over the the stabiliser strut, or the flying wires (this is common if you've tried throwing things out the window), then the drag may snap the strut / flying wire
• If it jams between the trim pushrod or in between the horn and stab, then I think you would lose control of the plane

Personally, I would want a row of rivets along the leading edge of the door skin, you don't need many if they are dome head. The rest can be glue.

[Nev]

Jon, thanks - this is exactly the feedback I’m looking for. It’s not being negative, it’s providing a balanced view and there’s a real possibility of learning something I haven’t considered. I understand all the points you make and they’re all relevant. Rivets have been generally proven over time, whereas regarding the Bearhawk, adhesives have not been widely used.

Years ago I flew many hours on a small fleet of Grumman’s - Tigers and Cheetas - an impressive little aircraft and one I loved flying for its handling characteristics and good design. Notably the Grumman’s used a bonded aluminium construction, yet between a number of us flying them full time, I don’t remember any of us ever becoming concerned about it delaminating. which may have catastrophic because the whole plane had the potential to unravel. Yet it didn’t. They were of course being commercially produced to a set standard.

Most of us are fairly comfortable flying with just a single layer of fabric glued to many of our aircraft surfaces. Yet if the fabric or the glue let go, we’re at risk of all the points you make above, and a few more. Should we be replacing fabric with aluminium, and glue with rivets ?

The point I think we’re both making here is that there are very real risks that I need to be aware of. But is it in fact any more dangerous than other parts of the aircraft? Is it higher risk than rivets? Is an adhesive more likely to delaminate, or are we simply more comfortable with the way it’s always been done.

I don’t know the answers to these questions, and if the arguments against the way I’m doing it stack up in my mind to a real and unquantifiable risk, then I’ll indeed add a row of rivets down the front of the door frame.

Thanks again for your reply above and keep the comments coming in please.

[zkelley2]

The fabric that if it comes off causes really bad things to happen is all rib stitched as well as glued. In some LSAs the fabric is actually riveted to the rib instead of stitching. FWIW.

[James]

Hi Nev,

did you research at what temperature the glue softens (if at all)?
All the structural adhesive we use at work (Hysol 9309, 3M 2216, etc etc) noticeably softens with heat, and lets go whenever you put a heat gun on it.
I'd be worried that repeated hot days in Australia, + maybe some reflected sunlight would make it yield enough to cause it to fail like Battson describes.

James

[Nev]

Sorry to be a 'negative Nancy'....

What led to using glue, instead of riveting like the cargo door? I see risks; struggling to see the advantages.

Jon, to answer your question of “why”, it’s a trial run for the main doors. I plan to weld a frame to make sea-plane doors, and bond Lexan to it.

I also plan to bond a one piece skylight to the two longitudinal formers. These are different on the B model. In the skylight case, it will be captured in channel extrusion fore and aft where I plan to use silicon. This should negate any risk of peeling forces. However the full wing doors will be subject to the same forces as the cargo door.

[Nev]

G’day James,

A very valid point. (The “Icarus effect”. I believe it’s struck more than one aviator).

Also relevant for my skylight as mentioned above. Here’s a link to the peel data at different temps (scroll halfway down). It shows a halving of peel strength from 23c to 49c and temps in a parked plane with a skylight could quickly soar to high figures, (and noting that mine wasn’t etched). Presumably in the air the airflow would reduce skin temps.

Hope your build is progressing well.

Nev.

[James]

No worries Nev - I only know enough to ask questions on this forum, not answer them yet :-)

[Steve W]

I have used DP190 on other aircraft with great results. I bonded all three door skins on my Patrol to the steel frames. I had no intention of using rivets until I dropped the baggage door on the floor and struck a corner. The bond separated about an inch in each direction on the corner. I had used a steel prep cleaner and cleaned and etched the aluminum. I replaced the door skin and riveted all three. I didn’t trust the bond to stand up to the forces others have mentioned that are exerted on the skins. I have sealing strips on most of the overlaps and so far so good with the assemblies holding their integrity.

[PaulSA]

I just don't see the point of bonding rather than riveting ......

[svyolo]

I am bonding a few things, mostly plexi, but I agree with you on aluminum door skins. What is the advantage over riveting?

[Battson]

Is an adhesive more likely to delaminate, or are we simply more comfortable with the way it’s always been done.

I think it is more likely, you couldn't delaminate properly formed rivets without a crowbar! See the post above from Steve W.

[Nev]

The bond separated about an inch in each direction on the corner.

Steve, thanks. This is compelling and you’ve convinced me to add a row of rivets to the forward frame of my door.
At this point I am still planning to use DP190 on the skylight longerongs, captured forward and aft in the factory frame with silicon or similar.

I’m very open to ideas now about my main seaplane doors. I’m after a “clean” look, but utilitarian and functional. My plan was to cover them completely in lexan, bonded to the door frame. But with the evidence you’ve provided it would make sense to also reconsider.

[James]

Hi Nev - this thread's got a lot of traction, and I might shamelessly hijack it to ask another question...

I remember a discussion I had with my brother, I said I wanted lexon on the doors to improve protection from penetrating branches etc during a crash, but was worried about how you can kick the doors out post impact when the airframe is bent up. Rivets won't let you do that.

We thought about screwing the lexon to the frames, and slotting each screwhole as a "pull through", but I decided in the end, if you designed the door coverings so that you could kick them out half concussed, surely a branch could also cave it in on impact?

In the end, I'm a long way from the doors, but I'm thinking of designing a "quick release hinge", rather than AN3 bolts to hold the door onto the fuselage?
Any suggestions?

Told ya - I only ask questions mate

James

[Nev]

Hi James, I haven’t got that far yet but to be honest I’m probably not going to worry about it. I had given some thought to getting out quickly in an emergency if needed. There are also the rear doors, and a front door on each side. Many light aircraft only have one door. There’s also an escape area in the cargo area.......a knife through the fabric ceiling ! Or wall, that may be pushing my luck, I need to lose a few pounds. But we have more doors than most. I think the most likely scenario is going to require egress via anywhere other than the roof in a taildragger.

[Steve W]

I also wanted that “clean look”. I wasn’t concerned about de lamination. If you are in a warm climate and won’t be sealing the overlaps of the doors, then I wouldn’t be too worried. BUT if you load/stress the overlaps with seal tape and other similar products and hinge and latch the door under some pressure against the skin, then I think there will be problems down the road. Big deal to repair after paint. If you make your doors and windows close against an inside “lip” and don’t put pressure on the outer skin; other than the airstream trying to rip it off the steel frame, there isn’t any real forces acting on the skin.
And just a heads up on those that are making their windows, every plane I have build has had an issue with air leaking around the front of the frame. I mounted the window latches as far forward as I thought would keep the front tight against the frame but mine leak a little making extra cockpit noise at 150 mph. I have a fairly thick seal at the front that makes the overlap stand off a bit which the air grabs and slightly pulls out. I am still breaking in my engine so am operating at a much higher speed than I would normally cruise. Hopefully it’s not an issue at 23 squared.

[Battson]

I’m very open to ideas now about my main seaplane doors. I’m after a “clean” look, but utilitarian and functional. My plan was to cover them completely in lexan, bonded to the door frame. But with the evidence you’ve provided it would make sense to also reconsider.

I think Simon has done the seaplane doors too, he would be a good one to ask being NZ-based.

[Nev]

I think Simon has done the seaplane doors too, he would be a good one to ask being NZ-based.

Thanks Jon, great idea.

[whee]

I also wanted that “clean look”. I wasn’t concerned about de lamination. If you are in a warm climate and won’t be sealing the overlaps of the doors, then I wouldn’t be too worried. BUT if you load/stress the overlaps with seal tape and other similar products and hinge and latch the door under some pressure against the skin, then I think there will be problems down the road. Big deal to repair after paint. If you make your doors and windows close against an inside “lip” and don’t put pressure on the outer skin; other than the airstream trying to rip it off the steel frame, there isn’t any real forces acting on the skin.
And just a heads up on those that are making their windows, every plane I have build has had an issue with air leaking around the front of the frame. I mounted the window latches as far forward as I thought would keep the front tight against the frame but mine leak a little making extra cockpit noise at 150 mph. I have a fairly thick seal at the front that makes the overlap stand off a bit which the air grabs and slightly pulls out. I am still breaking in my engine so am operating at a much higher speed than I would normally cruise. Hopefully it’s not an issue at 23 squared.

My thoughts are are same as what Steve said. Tight door seals put a lot of pressure on the door skins.

I do think bonding does have the potential to work. Semi-trailers, enclosed trailers, and a myriad of other things are put together using double sided tape. I bought some tape to secure the ribs in the wing tips I was marking but I haven’t gotten very far on that project. How long a bonded joint will last has too many variables for me to be comfortable with on something like my doors but I still think it can work.

I used plexiglass for my doors for two reasons: 1. I couldn’t find UV, scratch resistant, tinted lexan. 2. A guy I thought should know said he would never use lexan on his doors because they would be impossible to kick out. Both doors have cracked in the area where the frame is curved to match the front door former. I wish I had used lexan.

The plexi is secured using the Bob method of threaded holes in the door frame. It is working ok. A few holes have stripped out and on one flight a handful of screws rattled loose and fell out. You can’t tighten the screw very much or you will strip out the threaded holes. I’m still nerding over a solution that I would like but as of now when I replace the plexi with lexan I plan to use double sided tape and screws.

[svyolo]

Can you use nut-certs, or won't they work in round tubing. I haven't tried yet.

[whee]

Nut-certs are the devil so I try to avoid them if at all possible. The do work in this application but I’ll only use them as a last resort.

[JimParker256]

Wish you could "like" a comment... Like Whee, I abhor nut-certs. When they let go (and they will let go, make no mistake on that!), it is absolutely impossible to remove them. Oh, they will spin at will, preventing removal of the screws/bolts without cutting them off, but they won't come out. Such a major PITA!