Building a Fuselage Rotisserie

Larger parts and assemblies of an airplane typically require some kind of jigging and support structure. When I built my RV-4, most of this was fabbed from 2x4s and plywood, but for the SR-1 race plane, I’ve used the molds themselves for this function. Normally, a builder of a plastic plane like a Lancair Legacy or Long-EZ doesn’t have this option—they don’t own the molds, or there are no molds (in the case of Rutan method aircraft). And even if you did, molds tend to be an expensive, labor-intensive investment that can be damaged when used as support structure—that’s not really what they are for. That said, the SR-1 is a one-off—there is no plan to go into production, so using the lower fuselage and lower wing molds as jigging/support to assemble those items was part of the original plan.

Once the wing was closed up, there was little remaining work to be done. The molds were put into storage and the wings stored leading edge down on a wing rack. However, a significant amount of work remains to be done on the fuselage following closeout, and leaving the fuselage in the mold was no longer possible since I needed to be able to access the lower half of the fuselage. So, I needed to fabricate some kind of a support.

A fair bit of this remaining work is on the underside or inside of the fuselage, and I decided that a rotisserie would make life a lot easier by allowing me to rotate the fuselage for ease of access. I had considered a rotisserie early in the development of the fuselage, imagining a long aluminum tube that would function both as a rotisserie axle and an alignment jig for bulkheads. In the end, I realized a tube running down the center of the fuselage while building would likely be more headache than advantage.

A couple of years ago, when I was working for a rocket company, I was researching support jigs for welding large-diameter cylinders. Rather than rotate about an axle, wheels support the cylinder externally and rotate, allowing the welding head to remain stationary while the circumferential seam passes beneath it, allowing for a precisely controlled weld line. It occurred to me this would work well for the fuselage: I’d avoid incorporating extra structure into the fuselage, and it wouldn’t interfere with work.

Let’s Get Rollering

First off, let’s define terms to avoid confusion. I call the two large white wheels clamping the fuselage “rollers” and the stationary support on which the rollers rotate the “frame.” The first order of business is to make the rollers. The diameter of the rollers is somewhat arbitrary; however, make sure that the axis of rotation is high enough above the ground so that when you spin the fuselage the tail, canopy, prop, etc., don’t hit the ground. Once the wings are on the plane you of course won’t be able to spin it at all, but having the vehicle high enough to permit the gear to clear the ground may also be a consideration—especially if you have retracts.

The rollers should ideally be located co-planar with a bulkhead to avoid locally stressing the fuselage skin. The forward roller is at the firewall bulkhead. But, because I still needed to tape the external skin seams at the rear bulkhead, I offset the aft roller a few inches from that bulkhead to give me access for taping.

The rollers were routed from melamine-faced particle board using a Shaper Origin. The Origin is a handheld router that combines manual guiding and CNC path correction to follow a CNC router path uploaded to the Origin via USB. This path can be generated in a variety of CAD programs like SolidWorks or Onshape. The Origin orients to the workpiece (i.e., the sheet of particle board) using fiducial markers vis-à-vis a tape applied to the particle board that resembles a line of dominos. The user then digitally “fixes” the outline of the piece to be cut on a digital representation of the workpiece, and off you go. It’s a pretty neat system (you can see a short YouTube video here). It works great for cutting large 2-dimensional pieces like rib or bulkhead molds, but I’ll admit I haven’t used it as much as I originally thought. I actually prefer a traditional router using laser-cut templates, but I was in a hurry and figured the Origin cut parts would be accurate enough.

The Origin worked great—the outside cuts are perfectly circular and roll smoothly. The internal cut line was simply pulled from a cross section of the SolidWorks model of the fuselage and then uploaded to the Origin. An extra half-inch clearance was added to the internal cutout to accommodate a shim layer of Bondo to ensure a perfect fit to the fuselage. The rollers are doubled up to match the thickness of the roller wheel; I bonded mine together with epoxy.

The rollers are split left and right so that I can assemble them on the fuselage. Splicing plates of scrap 1/8-inch aluminum were attached with 1/4-20 bolts. I fabbed the plates by hand, but if I hadn’t been in a hurry, these are the kind of item I’d typically order from an online machine shop like SendCutSend. I’m pretty sure you couldn’t buy the aluminum for cheaper than SendCutSend will cut these plates. But, I was in a hurry, so I knocked them out on the drill press. At this point, I set the rollers aside.

H-Beams and Bondo

The next step is to fabricate a pair of H-beams to hold the rollers parallel to each other and orthogonal to the axis of the rotisserie. While the rest of the rotisserie doesn’t require particularly tight tolerances, these two H-beams need to be as perfectly orthogonal and of equal length as you can reasonably get, otherwise the fuselage will not roll smoothly on the support frame. And remember to locate the H-beams, as well as the splicing plates, far enough from the perimeter of the rollers to avoid interfering with the tracking wheels when rolling!

A quick note on materials here: all of the steel tubing you see is either 1-inch square (1/16-inch wall) or 2-inch square (1/8-inch wall). The 1/8-inch wall is overkill but it was available scrap so that’s what I used. A 1/16-inch wall would be fine. The verticals of the H-beams are 1/8-inch angle. Note that steel is sold in 20-foot sticks, so I typically try to dimension large weld projects like this at lengths of 4/8/12 feet or 5/10 feet to minimize waste. I’ll also note that for items like wheels, I usually only do a few tack welds to hold them in place. They aren’t going to see a lot of load, and a lot of my tooling gets scrapped once I’m finished using it, so I like to make it easy to cut off wheels to reuse on the next project. (Most) everything in a shop should be on wheels!

Set the H-beams aside. Now we’ll prep for the Bondo shim that gives the rollers a nice tight fit to the fuselage. 0.060-inch Freeman adhesive-backed mold wax covered by packaging tape was applied a few inches on either side of the roller’s location on the fuselage as a release (multiple layers of duct tape works well too if you don’t have mold wax). The packaging tape allows you to release the roller from the fuselage once you’ve made the Bondo shim. The 0.060-inch offset is for a layer of 1/16-inch-thick adhesive-backed foam rubber that we’ll attach to the inside of the Bondo shim so the rollers don’t mar the fuselage when they are clamped on. Both the wax and foam rubber are easily sourced online.

The H-beams and rollers are now carefully assembled on the fuselage, which is temporarily supported on sawhorses. Remember that the internal roller cut out is slightly oversize, so there should be a small gap between the roller and the fuselage. I inserted small wedges of wood scrap in a few locations evenly spaced around the perimeter of the roller to center the fuselage inside the roller. I then mixed up Bondo, quickly loaded it into a pastry bag, and squeezed the Bondo into the gap. A tongue depressor was used to make a nice fillet between the roller and fuselage.

A quick side note here: Bondo’s nominal mix ratio is 50:1, but it’s a catalyzed reaction, so it’s very tolerant of off-nominal ratios. That said, because Bondo cures so quickly, and cure rate is temperature dependent, I weigh out my Bondo in order to control working time before cure. When it’s really hot out or I need extra working time, I’ll drop the ratio to 100:1. If it’s cold or I’m being impatient I’ll increase to 25:1. CleanSheets mixing pads are awesome, and I just hole saw through the edge to be able to see the readout on the scale.

Once the Bondo cured, the rollers were gently removed, as well as the mold wax and release. The fillet edges of the Bondo are a bit rough, so those got block sanded down. I applied the adhesive-backed foam rubber to the fay surface of the Bondo and then reassembled the rollers to the fuselage. At this point the two H-beams were clamped to the fore and aft rollers. Once I was sure everything was straight and square, I match drilled the H-beams to the rollers and bolted them in place. I’d recommend labeling your H-beams with Left/Right and Forward/Aft so that you know how to reassemble them in the future.

Frame Welding

Once the rollers/H-beams were finished, I shifted gears and started on the stationary main frame. This is fairly straightforward: It’s simply a pair of 6-inch wheels welded wheels-up to either end of a support arm. A pair of U-shaped arms with smaller 2-inch wheels to keep the frame tracking on the larger wheels is then welded to each end of the support arm. Fabricate two of these frame arms.

A note here: I tweaked (read: whacked with a mallet) the tracking arms so that the distance between the tracking wheels is about 1/16-inch less than the thickness of the roller itself. This keeps the tracking wheels in positive contact with the roller at all times but still allows the roller to be inserted relatively easily into the frame. You can even see in the photos where the tracking wheels have discolored the perimeter of the rollers after a few months of use.

Five-inch locking wheels were then welded to a set of vertical legs; these were then welded to the end of the frame arms. The length of the vertical legs will be determined by the height of your rolling axis. Finally, the two end frames were joined longitudinally with four 8-foot sections of 1-inch tube.

Now for the moment of truth: I dropped the right side of the fore and aft rollers into the stationary frame arms, and bolted their H-beam in place. Next I placed the fuselage on the rollers. Finally I installed the left roller/H-beam, and bolted the roller splice plates together. At this point I was able to roll the fuselage completely around. There was no wobbling or relative movement between the rollers and the frame (other than rolling), so I considered the rotisserie a success.

I’d suggest incorporating a brake of some sort. Mine is simply a strip of 1/8-inch-thick steel that sits just shy of the rear splice plates and is welded to the cross arm below. I can rotate the fuselage to whatever angle and then simply clamp the plate to the roller with a bar clamp to hold that position—easy peasy. I also match drilled a hole in the strip at the location of one of the splice plate bolts while the fuselage was in the zero-degree position. This allows me to rotate the fuselage upright (its default position) and run a bolt through the roller and brake strip for a more permanent/secure brake.

Final Touches and a Three-Month Review

A piece of scrap ply from an RV-4 canopy box was trimmed down to fit the frame and serves as a storage table. The whole assembly was also weighed, with the front and rear wheels as the reference points. This will be useful for determining CG of the finished airplane, since I can load the assembled airplane into the rotisserie, jack the frame up, and lower it onto scales. I’ll subtract the rotisserie weight, and can then determine weight and CG of the airframe. (Final weight of the rotisserie was 150 pounds.) As far as cost, I was able to source a lot of the materials from my scrap bin, but I’d estimate a project cost of around $400.

After spending the better part of a week building the rotisserie, I was wondering if this had been the best use of my time. Now that I’ve been using it for a few months though, I can’t imagine not having it. I love it! It’s made working on the fuselage and engine/firewall-forward so much easier. I’m pretty sure I’ll gain back in productivity the week spent building it and largely avoid the black belt yoga moves some of the fuselage work would’ve required. There’s nothing I’d change based on my experience so far, although you might want to add jack points if you see a future need. I did not bother painting this as it is temporary tooling and will just get chopped up for the next project once it’s served its purpose.