The pilot reduced power in an approach to Hyannis, MA airport on Cape Cod. He then attempted to add power and got a nasty surprise: The throttle was frozen and could not be worked free. The pilot went to "Plan B," which was landing on a highway visible from the airplane.
The Scene of the Crime
Unfortunately, the pilot put his landing gear through the wind-shield of an oncoming car, killing a passenger. The aircraft was also damaged in the landing, but the pilot and both passengers emerged unhurt.
The accident occurred only 10 hours after the engine had been overhauled by an aircraft engine shop. The carburetor throttle shaft and associated bushings were replaced as part of the overhaul.
An accident investigation was conducted at a Boston-area testing lab under the auspices of the National Transportation Safety Board. I was not involved in the initial investigation, but had worked with the board on other cases. I always found board members to be highly competent.
The throttle was of the butterfly type. The shaft, which is about a quarter-inch in diameter and 6 inches long, rotates a circular disc (the throttle plate) to control the air intake to the engine and thereby the power. The shaft was so solidly seized in the bushings that it could not be removed by hammering and a vise had to be employed. When the shaft was pressed out, one of the bushings came out fixed to the shaft. The shaft was supposed to turn freely, so something had gone very badly wrong.
The Smoking Gun
Examination of the shaft and bushing interiors showed heavy galling and scoring. The figure gives an overall and a close-up view of the shaft. The damage is clearly visible. The holes through the center of the shaft should have been reamed to remove any burs or lips of metal left by the drilling operation. The reaming had not been done.
It is important that the bushings be "line bored" with a reamer to make sure they and the shaft are precisely aligned. The initial study suggested the shaft was cocked with respect to the bushings, so that line boring had been omitted.
"Tribology" is a fancy word that refers to the inexact science of friction and wear. Nobody really understands tribology, but there are some rules to go by. A cardinal rule is never run like against like. If the bushing is soft, the shaft should be hard. Under no circumstances should surfaces sliding against one another be the same material or of the same hardness. The result will be seizing and galling, as in the present case. The surfaces should have been mirror smooth instead of galled messes.
Both shaft and bushings were of a type of stainless-steel that is available in a range of hardness. The carburetor manufacturer used hard steel for the shaft and soft for the bushings. This combination worked fine. The designer of the kit obtained material specifications by reverse engineering of the original shaft and bushings. He got the composition of the stainless-steel right, but was apparently unaware the alloy came in various hardness. He specified the same soft steel for both shaft and bushing, with disastrous results.
I am still aghast at the wretched job done in rebuilding the carburetor. The kit supplier used incompatible materials and failed to ream the shaft holes. The shop failed to note that the holes in the shaft had not been reamed and rammed the rough shaft through the bushings, dislodging metal and starting the seizing process. The repair shop should have been on the lookout for such problems in a so-called "after-market" product. The repair shop may well have failed to line bore the bushings. Carburetor repair kits available on the Web include reamers for line boring. If the line borer was supplied and not used, the shop is even more at fault.
The production of the kit and its installation were both hack jobs unworthy of even a shade-tree automobile mechanic. The overall level of materials' selection and craftsmanship is reminiscent of myself at about 15. At that age I was hard put to keep a single-speed bicycle in operation.
For more on the Case of the Croupy Carburetor, go to http://rbi.ims.ca/5405-617.