Roee Kalinsky's RV-7A Project

Canopy Latch
Home Up News Specifications

 

Back Up Next

Canopy Latch

Running Total Hours: 0.0

 

2008.02.01: (0.0) I decided to use the after-market "Attaway" canopy latch [at Attaway Air] [at Fairings, Etc.] instead of the stock Van's design.  Why?  Firstly, Van's latch looks and feels cheesy to me.  It's made of 1/8" thick sheet aluminum, has a "clown nose" plastic ball for an interior handle, and a compression spring that gets bent out of shape against the fuselage.  On the exterior side, it is actuated with two tabs that are always hanging out in the breeze making drag.  I'll concede that in this instance Van's design is light, simple, and it does work.  But the look and feel of it really just isn't very nice (shall we say "unrefined"), and the drag seems wasteful.

The "Attaway" latch, designed by RV-6 builder Robbie Attaway, is a variation on Van's design.  The latching mechanism works basically the same way, with very similar geometry.  But instead of 1/8" sheet metal it is machined from 5/8" aluminum, making for something that feels more like a handle, and doing away with the plastic ball.  The compression spring is replaced by an extension spring, improving the look and feel of the latch, and making the whole fully-functional latch assembly self-contained.  Most notably though, the thicker material makes it possible to replace the exterior tabs with a push-botton, making the latch completely flush with the exterior skin, and saving a bit of drag.  It should be noted however that the Attaway latch is heavier.  No free lunch.

Robbie Attaway no longer sells these latches himself, but his friend Bob Snedaker of Fairings, Etc. now does (for about $80).  Both these guys are very friendly, and were happy to answer my many questions before I finally ordered the latch from Bob.

A couple of months after ordering the latch, Bob still hadn't shipped it because he didn't have any of the return springs, and was unable to find them at his usual sources.  He gave it a noble effort, but finally I asked that he just send me the machined latch pieces and forget about the spring, so I could start working on it and worry about finding a spring later.  Note that the springs he normally uses are Primeline SP 9604: 2.4lb extension spring 11/32" x 1-7/8" x 0.025", and at least at some point in the past they were available at general hardware stores like Lowe's.  He finally found a source that carried them, but is still out of stock at this time.  Anyway, I ended up using the same spring that's used in Van's aileron trim kit (C-626/VA-116).  I found that it works just right, and as a bonus that's one less different part that I'll need to find or keep spares of.

When I finally received the latch (minus the spring), there was another problem.  The bolt holes were all drilled too big.  The hole diameters measured approx 0.030" greater than nominal, which is quite a bit of slop.  Yes, the parts are meant to be able to rotate around the bolts, but with that much slop they literally rattle around the bolts.  I spoke with both Bob and Robbie about this.  Robbie confirmed that this wasn't right.  Bob ultimately didn't disagree, but still couldn't offer a replacement because all the parts he already had on hand were like that, and he wasn't about to machine a new batch (he did offer to take it back for a refund).  Bummer...  But I decided that I could turn that lemon into lemonade, and make some bushings.  The bushing idea actually worked out very nicely, better in fact than the original bushing-less design.  I made stainless steel bushings that would be held tightly by the bolts between the mounting brackets, and I reamed out the holes in the latch parts to the diameter of the bushings with minimal clearance (actually, I reamed the holes to nominal diameter, and I narrowed down the bushings to provide a slight clearance).  This scheme is nice not only because the latch parts rotate very smoothly around the bushings with no extra slop, but also because 1) the length of the bushings precisely controls the spread between the angle brackets to achieve very tight clearances between the handles and the openings, and 2) the bolts can be torqued down all the way giving the entire structure substantial rigidity.

Key lock:

I decided to improvise another non-standard feature and add a key lock to the canopy latch.  This is also a controversial issue among plane owners.  Those against having a lock argue that if somebody is going to steal your radios, you're better off making it easy for him, so he doesn't have to break your canopy to get in.  That's a good argument.  But by that same token, you can have a lock, and choose whether to lock it or leave it unlocked on a case-by-case basis, depending on time/place/cirumstances.  That's the approach I'm taking.

I worked out a simply design using a common 5/8" cam lock (the kind used on office furniture, tool chests, public bathroom fixtures, you name it, and incidentally most production piston aircraft) to secure the canopy handle.  Basically, in the locked position, the cam (or "tongue") fits through a slot in the lower mounting brakcet and up into a matching slot in the canopy handle.  This arrangement is actually very very strong, since the locking force is essentially weaker of the shear strength of the cam tongue, or the tensile strength of the mounting bracket or the canopy handle, all very strong.

Now, these cam locks are more-or-less a generic commodity made and sold by a large number of manufacturers and vendors.  However, while all very similar in design, the dimensions from one manufacturer to another vary just enough to make them not interchangeable.  Argh...  So I bought a few different ones and then selected the one with dimensions that were most favorable for this latch.  My final pick was one made by National (now part of CompX), p/n N185-272 VKA825.  These are available through McMaster-Carr, and also locally at Marshall's. 

2008.06.26: (0.0) Something I didn't like about the Attaway latch design was that it uses the airplane's skin as a spring.  If things are fitted such that normally the forward part of the aft handle sits flush on the interior surface of the skin, then it has to actually push further into the skin when the forward handle (i.e. the latch) is being moving in/out of the latched position.  That's the sort of over-centering geometry that keeps everything secure in the closed position.  It works, but I didn't like putting that kind of stress on that small area of the skin.  It takes a lot of force, which doesn't feel right, and I could anticipate that causing deformation or cracks in the skin at the nearby rivets over time.  So my solution was as follows:  1. File down that surface of the aft handle so that it has enough clearance throughout the full range of motion of the latch to not put pressure on the skin.  2. Install a spring plunger at the forward end that will put a controlled amount of pressure on the skin to keep the handle in the nominal closed position.  I used a spring plunger from McMaster-Carr, p/n 84765A51.  This particular one has an initial force of 0.7 lbs. and a final force of 2.3 lbs, which feels about right (I tried a few others, and this was my top choice).  It has a nice feel when operating the latch, and at only 2.3 lbs. will not damage the skin.  Simple refinement, great result!

2008.12.11: (0.0) For a few months now I've had the latch mechanism all fabricated, fitted, and sitting on the shelf waiting for a convenient time to install.  Well today Greg Larson came over to help me do some riveting on the canopy frame, and since he still had some free time, we quickly also riveted the latch brackets to the fuselage.

2008.12.31: (0.0) Progress update (working on this sporadically):

MECHANICAL ACTUATION

I adjusted the latch cams on the WD-617 torque tube weldment (which came from Van's nowhere near symmetric) and I fit the latch lugs to the canopy frame.  One came out perfect and the other one for some reason was off.  But a second attempt with a replacement part was also perfect.  [Note that I did have to put 1/8" shims between the canopy frame and the latch lugs to move them further aft into position.]  With this done, I get nice smooth action on the torque tube, and good solid locking force with the bottom of the side rails spaced 1/8" above the canopy deck.




I intend to attach plastic standoffs to the bottom of the canopy frame aft angles that will rest against the F-757 gussets.  That way, the latch pressure will be directed there, rather than to the back of the plexi against the roll bar.  The standoffs will need to be 0.352" tall.  From top to bottom: 0.62" thickness of the canopy frame side rail, 0.125" spacing from canopy frame side rails to canopy deck, 0.040" thickness of the canopy deck, 0.125" thickness of the longeron.

I fabricated the C-710 pushrod with slightly different dimensions than shown on the plans to accomodate the modified latch handle.  I also used a CM-4M rod end bearing instead of CM-4MS, i.e. one without a mounting stud.  Instead, I will use an AN24-19 clevis bolt and castle nut.  I have to use a slotted head clevis bolt (inserted from the bottom) rather than a standard hex head bolt because the canopy latch handle is counterbored from the bottom for the bolt head or nut, but not wide enough for a socket wrench or any other tool to get a grip on a hex bolt head nut.  A 3/8" spacer is also needed between the canopy latch handle and the rod end bearing in order for the pushrod to clear the latch handle's mounting bracket.

WARNING LIGHT

There have been several reports of pilots accidentally taking off without properly latching the canopy.  While the airplane is still controllable in this condition, it is certainly an undesired situation.  Reportedly, at flight speeds, aerodynamic forces will hold the canopy open several inches, and make it impossible to close in flight.  The pilot would be forced to return to the airport and land in order to close the canopy.  While this occurrence shouldn't constitue an emergency, it does arguably compromise the safety of flight, and could damage the airframe even if the pilot handles the situation appropriately.

So, as an added safety measure, I decided to implement a "canopy open" warning light (and/or audible warning, exact warning logic and presentation TBD).  In terms of sensing, I wanted to detect that the canopy is both closed all the way AND latched.  Now, the canopy frame is sufficiently flexible that it can actually be closed and latched on one side and not the other, so I decided to sense closure on both sides.

I decided to use standard snap action switches (a.k.a. microswitches), same as the one I used for the parking brake valve.  Smaller form factor switches may have made the installation easier due to the cramped space, but the smaller switches are generally less robust physically, and are also less standardized in form factor between manufacturers.  So in the interest of simplifying future maintenance, I stuck with the standard switches.

I played around with many possible geometries for mounting the switches, and quickly narrowed it down to two:

The first approach
involved three switches: one to detect full closure of the canopy on each side, and one to detect that the latch is in the engaged position on the right side only (since left and right are on the same torque tube, actuated from the left).  This had the advantage of having all the switches mounted to fixed parts of the airframe, but had the disadvantage of requiring the addition of several non-trivial mounting brackets in an already cramped area.

The second approach
involved two switches only: one that detects both that the canopy is fully closed and that the latch is actuated on the left, and one likewise on the right.  Basically, the switches are mounted directly to the cam arms on the WD-617 weldment, and sense that the canopy latch lugs are under the cam.  This has the advantage of using only two switches instead of three, and requires no additional mouting brackets -- very clean installation.  It does however require that the lever arms on the switches be modified slightly, and it does involve a wiring run from a moving part (the WD-617) to a fixed part of the airframe.

I decided to go with the second approach.  About modifying the switches:  I had to effectively widen the end of the lever arm to ensure positive contact with the latch lug.  I accomplished this by attaching a 10/32" x 10/32" square of very thin stainless steel sheet to the end of the lever arm.  I used JB weld to glue it -- seems very strong.  I also needed to bend the arm back some at the point just after it clears the casing of the switch.  This bend just made it so I could mount the switch body at a more convenient angle relative to the WD-617 cam arm.  I quickly rigged it up, and it worked very well.

2009.12.19: (0.0) The tip-up canopy incorporates a handle on the inside of the canopy bow that also rotates under the cabin frame to double as a secondary canopy safety latch, or rotated above the cabin frame to hold the canopy slightly open.  Nice little feature, but it needed a little refinement.  In the stock design, the handle rotates freely inside a plastic bushing with no positive retention in any particular position.  So with vibration, etc. the handle can rotate on its own accord.  One very unfortunate scenario that has actually occurred to at least two, is that upon closing the canopy from the outside, the handle rotates slightly, thereby locking the canopy closed from the inside.  There is then no way to get at the handle from outside the aircraft without cutting into something.  Tragic...  To prevent this, I've decided to give the handle some up-down freedom and spring-load it in the up position, so that when unlocked, it can't freely rotate into the locked position.  It has to be forcibly pulled down past the bottom edge of the canopy bow before it can be rotated.  Not easy to explain, but easy to see in the photos below.  Note that I also fabricated an aluminum box to cover up the mechanism.  This serves several purposes: 1. Keeps the spring from hitting the canopy glass if the cotter pin ever fails, 2. Give me a surface I can paint gray like the rest of the cockpit to minimize glare, 3. Clean aesthetics.  Note that I also put a coat of clear UHMW tape over the surfaces of the canopy bow that the handle will rub.

 

Back Up Next

 

 

Send mail to roee@kalinskyconsulting.com with questions or comments about this web site.
Copyright © 2003 Roee Kalinsky
Last modified: December 23, 2010

Visit my consulting web site at www.kalinskyconsulting.com for your engineering needs.