When I graduated as a mechanical engineer in 1978, I landed a job in the “cold section” at Pratt & Whitney Canada (PWC) in Montreal. In those days, the “cold section” of turboprop engines included the gearbox — between the power turbine and the propeller — and the accessory drives such as hydraulic pump, starter generator, propeller pitch control, etc. — all driven by gears.
To say that talking about my job in gearing was not really thrilling my friends is a gross understatement. In unison, they would say “why gears”?
My exposure to cylindrical gears had been fairly limited at university, but sufficient to have lengthy discussions with my mentor at PWC who at 60 was about to retire. I ended up in advanced gear engineering, writing gear calculation software, first on a TI-59 hand calculator and then in Fortran, the reference scientific language in the late ‘70s.
Eventually, I decided to undertake graduate studies with a supervisor who had worked at The Gleason Works in 1967-68. My Master’s subject was to model, calculate the kinematics and display spiral bevel gears in 3-D. This is standard fare today, but in 1983-84, nobody was thinking of this as a need in bevel gearing. My Master’s was followed by a Ph.D., still in gear simulation and dealing with load sharing and transmission error in spiral bevel gears, after which I took a tenure track as professor in ME for the following 20 years.
In 1992, one of my graduate students, in an exchange at Prof. Aizoh Kubo’s lab in Japan, got in contact with engineers from Yutaka Seimitsu Kogyo, a daughter company of Toyota, and a manufacturer of hypoid generating machines competing with machines made by Gleason and Klingelnberg.
Unfortunately for Yutaka, and fortunately for me, Yutaka had no bevel gear software of its own and rather had to rely on the Gleason calculations to set up its machines. My own bevel gear software — used for teaching and research needs — was already advanced enough in terms of the simulation of bevel gear manufacturing movements that I could think of replicating Yutaka’s machines, and eventually those of Gleason — which I did, of course.
This work led to the development of HyGEARS V 1.0, unveiled at the 1994 JIMTOF in Osaka, and an intense period of collaboration with Yutaka ensued over which HyGEARS was fully calibrated and started being used by major gear manufacturers in Japan, Korea and the U.S. Nothing being eternal, the collaboration with Yutaka came to an end in 2005.
But the HyGEARS development did not stop there: since 2005, some 15 different gear types were added to HyGEARS, including every face milling processes for spiral bevel and hypoid gears — all of which can be analyzed for TCA and LTCA and can be manufactured on 5-axsis CNC machines with the HyGEARS built-in post-processor. The switch from dedicated to 5-axis CNC machines does not suit every application, but there are a large number of situations where the approach offers significant advantages.
Now, 40+ years after being hired at PWC, I realize that I have been at the very beginning of a vast field of 3-D modeling, numerical methods and simulation in bevel gearing, i.e. — the very basis of how we create gears today.
Over the past decades, straight, spiral bevel and hypoid gear design and manufacturing has evolved from a very nebulous subject — best left to a selected few — into a software-driven, high-tech activity where every micron counts. I like to think that I may have contributed to some aspects to this evolution, which has kept me busy, focused, and mightily interested for four decades.
So, in the end — “why not gears”?
Claude Gosselin is founder, CTO and CEO of Involute Simulation Softwares.
Claude Gosselin, Ph.D., P.Eng.
Involute Simulation Softwares Inc.
1139 des Laurentides
Quebec, QC, Canada, G1S-3C2