Case of the Process Variations

After graduating with a BSME, I was working in the seal department of a large jet engine company, where I was assigned the engineering responsibility for all the engine's sump seals. My predecessor, along with two manufacturing engineers, started the process of developing a fiberglass stationary seal at the front of the engine for the first bearing supporting the engine's fan.

Fiberglass (composite) was considered acceptable, as this area was always relatively cool, and a molded seal would cost and weigh less than a metal seal. The ID of this stationary seal provided a cylindrical surface about 50mm (two inches) long for rotating metal "fins" that would run very close, making this a labyrinth seal.

The two manufacturing engineers who designed and procured a mold for this seal called and told me the first piece failed dimensional inspection, and they didn't know what to do. The shape and size of the seal were similar to a five-gallon kitchen pot with a large radius blending the side walls to its flat bottom. Mounted horizontally, the large open diameter held an aluminum flange with a diametric pilot, and a bolt-hole pattern for mounting. Where a pot has a solid bottom, this part had an inside diameter that held the sealing surface of the labyrinth.

Upon studying the inspected dimensions of this prototype, I discovered that the mounting flange was correct, while the sealing part had the proper diameter and length in the proper axial location with respect to the flange. The dimensions that were out of tolerance by 0.5mm (0.020 inches) were those around the large transition radius that defined the shape of the shell. The most likely cause was an improper calculation of how much the composite shell would shrink after curing and cooling to room temperature from its 120C (250F) curing temperature. After checking clearances of the shell with all the other parts when assembled, it was seen that these were in the magnitude of inches -- in other words, plenty of clearance.

So, as the engineer in charge, considering all of the above, including weight, I made the following proposed disposition to manufacturing: I will have the drawing for the part revised so that the dimensions in question matched those of the prototype, along with increasing the tolerances to allow for process variations.

This entry was submitted by Bob Salter and edited by Rob Spiegel.

Bob Salter is a mechanical engineer who works for American Roller Bearing Co. and a dedicated do-it-yourselfer.

Tell us your experience in solving a knotty engineering problem. Send stories to Rob Spiegel for Sherlock Ohms.

Click

here to access the Sherlock archives.