Reverse Engineering a Glasair Tailwheel

Home shop machinist.

One mangled Glasair 1 TD tailwheel.

One day last summer, seconds after rolling from the hold short line to depart Las Vegas’ Henderson Airport, a freak wind gust caught David DiGiuseppe’s Glasair 1 TD. In that instant, the plane was lifted up, twirled around and slammed to the ground. A nanosecond later, David knew this airplane isn’t going anywhere today.

The bubbly protrusions inside the steering tube indicate areas where the weld penetrated into the tube (left). Note that the area on the side that separated shows no sign of similar penetration (right).

As the Glasair limped back to the parking area, David knew from the sideways tracking and grinding noise that there had to be some serious damage. At first glance the tailwheel was obviously messed up. Amazingly, a detailed inspection of the main structure revealed nothing: no cracks or telltale signs of delaminating. The mains and engine mount looked good. All that was messed up was the tailwheel and fork weldment. One side was sheared from the weld and the other was twisted so badly that it dragged off half the tire tread during the taxi back from the runway.

I was wrapping up last month’s column when Kevin King called to ask if I knew a welder who might be able to help David remake his tailwheel assembly. One thing led to another and I was looking at my next project for the Home Shop Machinist!

The detached fork leg was flattened so it could be used as a template (left). An alignment punch was used to mark the location of the holes to be drilled to match the radius for the fairing attachment wings (right).

David is not the builder of his Glasair, but he has the plans. He discovered the tailwheel fork was scratch-built based on drawings (as opposed to a factory- or vendor-supplied tailwheel, such as Matco). But the drawings provided only general guidance for dimensions and materials. When we met to discuss the project, I concluded that we didn’t really need a detailed drawing because we had the original which, even in its mangled state, was intact enough to “reverse engineer” (see sidebar below).

An excellent way to produce two identical handmade parts is to tack weld them together and shape them as one. A disc sander was used on the long faces, but the curvy nature of these parts required that most of the shaping be done by hand filing (left). A round file was used to blend the radius to the long edges. Note the discoloration from heat from the tack welds at the front and back (right).
With the shaping complete and the parts still tacked together, the axle hole was drilled through both plates (left). After shaping and drilling, the tack welds were ground down so the parts could be separated for final bending into the required mirror-image forms (right).

The tailwheel, before the incident, worked perfectly, so there was no need to make any changes to the design. One thing I did note about the original was the way the weld failed: The leg that separated didn’t have good weld penetration. The telltale signs were how the weld tore away along the seam—both along the steering tube and along the upper gusset—and how, on the inside of the steering tube, you could see how the weld showed little, if any, penetration.

A hole saw was used to create the inside arc for the gusset plate (left). The band saw was used to rough cut the sides and outside radius (center). Both the gusset plate (shown) and the doubler plate for the tailwheel lock latch required a milled slot (right).
A section of 4130 tubing was faced off and cut to the final dimensions on the lathe. The bung hole for the grease fitting was drilled in the center (left). A threaded bung for the grease fitting was turned on the lathe (right).
After welding, a tap was rethreaded into the bung to clean up any distortion in the threads caused by welding (left). A section of aluminum bar was turned to the exact width of the wheel and bolted to the fork legs (right). This ensured everything was held in exact alignment and spacing for welding. The bar in the foreground was used by master welder Billy Griggs as a hand rest.
One reverse-engineered tailwheel, to go please! Note that the tailwheel lock is engaged.

With that, my task was clear: Make David a new tailwheel weldment, as close to the original as possible but with improved welding. I implemented one change to the design, which was the addition of a threaded bung to better accommodate the grease fitting that was threaded into the steering tube. The original had the fitting threaded directly into the thin-wall tubing, which meant it was barely able to be tightened beyond finger tight. Any tighter and it would have stripped out.


  1. patents stifle innovation, and these days mostly only benefit large corporations not the little guys. As an engineer with over 20 patents as of now and counting, I don’t believe patents should apply to individual noncommercial use (regardless of the law).


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.