technique

CANOPY MOUNT Fitting and Mounting a Bubble Canopy BY ROBERT WAGNER EAA #56437

his is a discussion of fitting and mounting a bubble type s l i d i n g canopy on a twoplace side-by-side aircraft. The type of canopy discussed herein is one that joins a fixed windshield when closed and has its own streamlined recovery shape rather than relying on a turtleback or tail cone for this feature. A prerequisite to f i t t i n g the canopy is that the fixed windshield must first be in place. It will be assumed t h a t t h i s has already been done and this discussion will proceed from there. Sliding canopies are convenient in that they open the entire cockpit for access in one movement. However, they do have the disadvantage of being difficult to seal. An alternative to the sliding type is the clamshell type. The clamshell is easier to seal because it uses mostly compression seals. However, if you, as do I, prefer a sliding canopy anyway, read on. The installation on which this article is based is my Mustang II with a blown acrylic bubble canopy. Most of the techniques discussed are applicable to vacuum formed oval types as well. The canopy on my aircraft was installed once and reinstalled two more times. This article will discuss the progression of methods and materials used, and experiences along the way. To begin with, let me state that a two-place side-by-side bubble is quite different from a single place. It is much more flexible and is very dependent on its framing and method of support to retain its shape in

T

28 MAY 1999

flight. The challenge is to attain adequate rigidity with as little weight and space penalty as possible. Typical components of a sliding bubble canopy are: an acrylic bubble, front bow, side frame to which rollers or slides are attached, rear frame with center guide, compression seal between the canopy and windshield, a latch and an edge seal at the rear perimeter of the canopy. The first version of my canopy used the standard 5/8 in. O.D. steel tube bow welded to a 5/8 in. square tube side and rear frame. It had a 3point roller support system; two rollers on either side in front and a center roller guide on the rear frame.

The bubble was sandwiched between an aluminum side plate and the steel frame on the sides and it was trimmed to just clear the fuselage contour in the rear. Trimming was very difficult because the canopy changed shape each time the temporary support was changed. I was able to fit it fairly close eventually, but to my chagrin, when I flew the aircraft, the canopy took an altogether different shape. The aerodynamic forces opened up side gaps about 1/4 inch on each side. Both the steel frame and the acrylic bubble are really quite flexible, even when attached to each other. My first fix was to install pins and sockets on the sides of the canopy to keep it

Figure 1

ACRYLIC —— CANOPY

1/16" THICK EXTRUDED ALUMINUM BACKING PLATE — STEEL CHANNEL

RIVETS

NYLON ROLLER

.050 2024 T3 SIDE TO MATCH FUSELAGE AND CANOPY CURVATURES

STEEL ANTI-BULGE GUIDE — 3-ELEMENT BALL BEARING SLIDE — INNER RACE OF SLIDE

RIVETS

NO. 8 SCREWS 1/8" THICK EXTRUDED ALUMINUM ANGLE

ALUMINUM SPACER SHAVED TO MATCH SIDE PLATE CONTOUR

SPACER

FUSELAGE SIDE STRINGER NO. 10 BOLTS

Figure 2

Figure 3

from bulging outward. This worked reasonably well but the gaps were still too large for my liking. Also, there is an appreciable reverse air flow through air gaps at the rear of the canopy due to pressure recovery characteristics of the streamline shape. When flying through rain, one can get quite a spray on the back of the neck. In addition to opening gaps, distortion can also reduce clearances between the canopy and fuselage, which, along with turbulence induced vibration, can cause abrasion of the fuselage surface. Also, the joint between my canopy and windshield was not uniform and it leaked water both on the ground and in the air. Not being satisfied with the way the canopy was fitting, I decided to make changes. My first remounting consisted of replacing the side frames with ball bearing slides. The canopy bow then became an independent member no longer welded to the removed side frame. I also installed a fiberglass edging at the rear of the canopy for a closer fit to the fuselage. 1 attempted to match the canopy bow to the windshield by successive measurements, but again the fit was not good enough, nor did it match the windshield shape sufficiently close. A resilient compression seal was adhered to the windshield bow and a flat nosing of body filler was applied to the canopy bow as a mating contact surface. The non-uniform gap between the canopy and windshield again resulted in water leaks. The rear of the canopy fit better but still needed seals. Sponge seals were tried but were not effective. Hard plastic V-type seals were then tried. They worked reasonably well in areas where the gap was u n i f o r m . However, the hard plastic did not adapt well to curved surfaces, and some leakage remained. Also, some air leaks caused the plastic strips to make a noise similar to a party noise maker. At the first opportunity I replaced the plastic seals with the more pliable neoprene V-strips which conformed much closer to the fuselage shape and sealed much better. The ball bearing slides did a much better job of preventing distortion. Resistance to vertical movement is excellent but there remained some SPORT AVIATION 29

ho r i/on t al m o v e m e n t because the slides are not designed for horizontal loads. The ball bearing slides which I used had heavy extruded aluminum

frames and were attached at the bot-

tom to the cockpit side stringers with screws. The slides are actually designed to be side mounted. The bottom

Figure 4

mounting was made to work, but some drilling through the machined ball races was required. All in all, the second installation was an improvement over the first, but it still left a lot to be desired. 1 recently experienced a fuel tank problem which necessitated removal of the tank from my aircraft. Since my plane was going to be down for an extended period of time, I decided to bite the bullet and tackle the canopy again. I changed some materials, but the main thrust of this effort was to make a better fit, improve resistance to aerodynamic distortion and improve the seals. Different ball bearing slides were used. I decided to use ordinary steel ball bearing drawer slides instead of the extruded a l u m i n u m type used previously. The steel slides are narrower, allowing less installation width and are also less expensive. The b a l l b e a r i n g slides were sidem o u n t e d on the v e r t i c a l leg of

aluminum angles, which were, in turn, bolted to the cockpit side stringers. This method of mounting involved an additional part, but proved much easier to install. The new slides, mounting angles and new rear corner

Figure 5

Figure 6 30 MAY 1999

gussets are shown on Figure 1. Side plates made of .050 in. 2024T3 were attached to the outside race of the ball bearing slides with spacers to position the side plates flush with the fuselage side skins. This provided a frame to contain the acrylic bubble. The parts were fitted and drilled undersize and retained with undersized clecos. All parts were cut oversized where possible and later trimmed after more accurate measurements could be made during trial assemblies. A section through the ball bearing slide assembly is shown on Figure 2. Next, a new canopy bow was made. It was made of 7/8 in. O.D. aluminum tube instead of 5/8 in. O.D. steel, as was used in the original. The larger aluminum tube proved to be slightly stiffer than the steel tube it replaced and provided greater area to mount sealing surfaces and the latch.

Bending the bow turned out to be quite a challenge. I first placed a cardboard panel against the windshield and traced the outline. 1 then calculated the position required for the canopy bow considering the slope of the windshield, thickness of the bubble, etc. I could provide the formulas if anyone is interested, but the calculated shape is only a starting point. It can be done almost as well by eyeball. The contour that T developed was used as a pattern for bending the tube. 1 had the first bending done at a commercial bending company. It did not quite fit my pattern, so I attempted to correct it by hand and promptly ruined it. 1 then decided to bend the tube from scratch with the aid of a conduit bender. First, I filled the tube with packed sand and corked both ends. This made the rube less prone to kinking. When using a conduit bender it must be done very gradually. Even with the greatest care, small bulges developed on the inside of the curve, but they are barely noticeable after the canopy is assembled. I eventually mined a second bow also, but finally the third one was a keeper. The techniques I used in making small hand corrections are as follows: Parts that were excessively curved were pressed against the floor and spread by hand. Parts requiring sharper bending were pulled against a vertical post while pushing against the post with a leg and pulling the ends of the tube with both hands. Sometimes it was necessary to place the tube in a vice and pull or push with a slide-on lever when sufficient leverage could not be obtained any other way. All corrections were done gradually and frequent comparisons with the pattern were required. Additional straightening was required to e l i m i n a t e the undesired waviness 90 degrees to the curvature of the bend. The method which I used was to press the bow between support blocks with a C-clamp. After fitting the bow to the pattern, the bow must then be fitted to the actual windshield for final corrections. I trimmed the ends of the tube 1/8 in. short of touching the side rails and used shims to raise the bow to the required elevation. It was necessary to slot the ends of the tubes to clear the ball bearing slide support angles. It is important to position the canopy bow at a uniform distance from the wind-

GLASTAR

HARTZELL AN AIRCRAFT DESIGNED TO DO WHATEVER, AND A PROP THAT WONT ASK QUESTIONS. "Backcountry, cross-country, on floats, or on a trailer, the GlaStar® defines versatility. And a constantspeed Hartzell propeller makes it even more versatile. Unlike fixed-pitch praps which force you to choose between a 'climb' or 'cruise' design, a constant-speed prop enables you to optimize both. Call us today at (360) 435-8533 and ask about our special builder pricing for Hartzell props."

For your FREE booklet "Technical Issues Involved In Selecting a Propeller System For Your Kitplane,"

use SPORT AVIATION'S reader

service card.

I AR TZ E l!I

Bob Cavinsky, President, Stoddard-Hamilton Aircraft, Inc.

For information, UM SPORT AVIATON's Reader Service Card

Figure 7 shield bow. In my case, a 1/4 in. gap was required for the resilient seal used. To accomplish this I used a 1/4 in. thick plywood spacer and clamped it to the windshield bow, then clamped the canopy bow to the plywood. This

method is shown on Figure 3. The bow is now ready to receive the acrylic bubble. The bubble must first be trimmed sufficiently to clear the top of the ball bearing slides and the rear fuselage. A SPORT AVIATION 31

Next, the frame for the rear guide was fitted. I used the old steel frame,

w h i c h was cut from my o r i g i n a l

Figure 8 gap of at least 1/2 in. is desired between the edge of the acrylic material and the rear fuselage surface. A closer fit is done later with a fiberglass edging. When the bubble is placed on the bow, the areas of the bow that require further curvature correction will become apparent. Many more removals of the canopy bow and corrective bendings were necessary to match the canopy to the windshield. The gap between the front edge of the acrylic bubble and the windshield will not be uniform at this point, but trimming of the front edge of the bubble should be deferred as long as possible. The bubble will change shape each time the shape of the bow and/or the mounting position is changed. The bubble should be supported on the bow at the front and on wooden spacer pieces at the rear. The sides of the b u b b l e should be nested w i t h i n the side plates. When alignment is satisfactory the bubble can be drilled and clecoed to the bow and side plates with undersized (3/32 in.) clecos. Next, the side plates are trimmed and formed ( b e n t ) at the rear and front ends to fit the bubble and fuselage curvatures. Figure 4 shows the results of this forming operation. I decided to extend the side plates rearward to meet the rear guide frame to minimize joint stress on the acrylic bubble. This decision resulted in considerable forming difficulty due to the complex compound curvatures at the canopy to fuselage interface aft of the ball bearing slides. 32 MAY 1999

canopy frame, and welded on mounting pads for fitting to the interior of the bubble. I then clamped the frame in place on the fuselage. A gap was left between the inside of the bubble and the mounting pads. This gap was t h e n f i l l e d w i t h reinforced body filler for a uniform contact fit to the inside of the canopy. The pads were later drilled for riveting to the side plates, with the bubble sandwiched in between. On my previous installation, I experienced contact abrasion between the canopy and fuselage due to inflight distortion and turbulence. This time around I decided to install an additional rear guide which slides

Figure 9

Figure 10

into the end of the rear guide rail when in the closed position. T h i s guide was made from high density polyethylene and is shown on Figure 5. Ineorporated in t h i s guide is a

socket for mounting a canopy lock.

The rear guide is encased in the fiberglass edging.

S T E W A B T 51 -t r?VPS

HARTZEL The perimeter of the canopy aft of the slides is fitted with a fiberglass edging which forms the interface between the canopy and fuselage. This IT'S A LITTLE PIECE OF HISTORY. edging extends to and is bonded with the aluminum side plates. Fiberglass IT S A KICK IN THE PANTS. was used for the rear edging to provide a better transition between the canopy and fuselage curvatures. For your FREE booktet "Tlte Stewart S-5W combines two The edging was laid up with the 'Technical Issues Involved American icons of performance canopy in place on the fuselage. The , In Selecting a Propeller For fuselage was protected with polyeth- I the P-51 Mustang and the big block System Your Krtplane," ylene film, and the acrylic bubble was use SPORT Chevrolet V-8. The new Hartzell protected with paper and masking AVIATION'S 4-blade prop was designed especially for the S-51, reader tape. A .050 in. cardboard spacer was used to separate the edging from the service card. it's 70% the size of the original Mustang's - but fuselage to provide a minimum clearthe performance isn't scaled down. Call us at ance for the prevention of abrasion. (561) 778-0051, and ask about our special builder The form for the fiberglass layup was HARTZELL made of modeling clay. If I were to do pricing for Hartzell props." it again I would use plaster. The modJim Stewart, President, Stewart 51 Inc. eling clay did not hold its shape adequately and allowed some of the irregularities in the cardboard underneath to be transferred to the layup. for inbmxilion, use SPORT AVIATON's Reoder Service Cord The cardboard spacer, modeling clay form and initial layup are shown on the bubble. A view of the finished, slight outward movement in flight Figures 6, 7 and 8. A Dremel cut-off wheel was used trimmed rear edging on the bubble is with the previous ball bearing slides, 1 decided to install anti-bulge guides to carefully remove the excess layup shown on Figure 9. Based on having experienced a at the canopy midpoint. The guides down to the masking tape border on consist of channels riveted to the inside of the canopy and tubular pins with nylon rollers welded to brackets and mounted on the side angles. The rollers engage the c h a n n e l s in the closed position. In addition, holes were drilled in the top of the channels for insertion of pins to retain the canopy in a partially open position. This was done because the canopy tends to open fully by gravity when taxiing. A wide open canopy can be very drafty, whereas a partial opening can provide comfortable ventilation. The channels and guide rollers are shown on Figure 10. The guides accomplished their objective, but they incurred a disadvantage in that they restrict the maximum opening of the canopy to 19 inches compared to 21 inches available without the guides. After the bow and bubble were fitted to the framework and aircraft, the Figure SPORT AVIATION 33

strips for both the canopy and windshield bows. The purpose of

these strips is to provide a flat surface for a compression type seal. A view of the capstrip and flat facing strip is shown on Figures 11 and 12. EPDM "P" type

tubular seals were installed on the

canopy facing strip. Accurate fitting of the canopy bow resulted in uniform seal contact all the way

around. The seal was proven by a

hose test to be watertight, except at the very bottom where the fac-

ing strip was discontinued due to interference with the side angles. Wiper type seals were applied in

Figure 12 front edge of the acrylic bubble was trimmed flush with the bow.

The front capstrip was then fit-

ted. The procedure I used in fitting the capstrip was to lay a strip of paper on the front of the canopy and then trim it until it

attained the desired shape. I then

traced the paper pattern on .032 2024T3 sheet a l u m i n u m . The

a l u m i n u m was then cut (oversize, of course) and fitted to the

bubble. With the aid of tape and a hole finder, the holes already drilled through the bubble to the

bow were transferred to the cap-

s t r i p . The capstrip was then irimmed to the finished contour.

Some breaking of the edges was

required to conform the capstrip to the bubble curvature. 1 t h e n f a b r i c a t e d flat f a c i n g

Figure 14 34 MAY 1999

this area, and while not as effective as compression type seals they do an adequate job for this

small area.

Next, V-strip seals were fitted to the underside of the rear fiberglass edging. These seals expand under air pressure to fill the gap between the edging and fuselage. To be effective, the seals must be installed at a uniform

distance from the fuselage skin

so that they just barely contact the skin in the closed position. To accomplish this, a gage was

made of wire with a hollow pop rivet shank attached to one end. Tape with distance graduations marked on was attached to the

wire. This gage was placed under the edging and pulled until the r i v e t shank made contact

with skin and underside of the edging. The distance at the contact point was read off the tape

and marked on the outside surface of the edging. This was done at uniform intervals all around the perimeter. The canopy was then removed from the aircraft and the d i s t a n c e m a r k s were transferred to the underside for use as a guide in locating the Vstrip seals. The V-strips were adhered to the underside of the edging w i t h contact cement. Seals were not applied until after all p a i n t i n g was done. The Vstrip i n s t a l l a t i o n is shown on Figure 13. Flat head 120 degree pop rivets were used e x c l u s i v e l y in assembling the canopy. The mandrels were knocked out and the

shop heads were flattened with a

rivet gun where accessible. Mandrel holes were filled with micro before painting. Pop type rivets were used because the countersink is shallower than the AN rivet and because the acrylic material is subject to less impact force even though the pulled heads were flattened after pulling. Backing strips were used so that no riveting forces were applied directly to the acrylic surfaces. Rivets (1/8 in. size) were used, except 3/32 in. rivets were used to attach the flat WE LEAVE AEROBATIC ADDICTS HANGING. weather strip facing plate to the canopy bow, and 5/32 in. rivets were used to connect the side plates to the rear guide frame mounting pads and for your FREE booklet Aerobatic flight, especially in a Pitts S1-11B, is 'Technical Issues Involved to the front bow. Rivet holes were addictive. 'The Claw' aerobatic propeller by Hartzell In Selecting a Propeller countersunk to match the rivet heads. System For Rivet holes in the canopy capstrip provides fantastic hang time at the top as well as Your Krtplane," were first dimpled with a 3/32 in. use SPORT 1 outstanding overall performance. And its Kevlar " AVIATION'S rivet before countersinking for the 1/8 in. rivet. Rivet holes in the construction is extremely durable, erosion resistant, and reader service card. acrylic bubble under the capstrip has passed bird and lightning strike tests as part of the were countersunk to match the 3/32 in. dimple in the capstrip, then certification process. Call us at (307) 886-3151 and ask drilled oversize with a #28 drill to about our special builder pricing for Hartzc.ll props." ARTZELL provide expansion clearance. The corners of all holes in the acrylic Slu.irt Horn, Piwident, Aviat Aircraft, Inc. material were rounded slightly as a precaution against stress cracks. | Rivet holes in the acrylic material at the side plates were reamed further For informotion, uw SPORT AVIATION'S Reader Service Card oversize and fitted with 1/8 in. I.D. vinyl tube sleeves, a further precaution to avoid stress cracks. methods and procedures used in one namic forces during flight and techp a r t i c u l a r canopy installation. In- niques used in sealing against water cluded are methods of f i t t i n g the and air intrusion. It is recognized that SUMMARY canopy to the aircraft, devices used in other methods may be equally or more The foregoing is a description of resisting distortion due to aerody- successful t h a n those described herein, but these are ottered in the interest of adding to the knowledge base of sliding bubble canopies, and to encourage ftirther improvements by others. The canopy support method described herein is based on the use of ball bearing slides. Advantages of ball bearing slides are: superior vertical rigidity, fair horizontal rigidity and good resistance to air leakage for the length of the slide. A disadvantage is greater weight. Side and rear guides were added to increase resistance to distortion of the canopy in flight. Figures 14 and 15 show photos of the finished canopy in the closed and open positions.

Figure 15

Robert Wagner, EAA 5643 7, resides at 6755 Cambridge, Dearborn Heights, MI 48 J 27. * SPORT AVIATION 35