PAST & PRESENT
50 years of success for P&WC's PT6 plus a peek into the future
An overwhelming number of turboprop and helicopter OEMs have built winning aircraft using this outstanding powerplant design.
By Mike Venables
Principal, TriLink Technologies Group
Cutaway of a PT6 showing its unique design features, particularly the reverse air flow arrangement to increase efficiency and the folded combustor to reduce overall length and weight.
This year marks the 50th anniversary of the start of full-scale production of the PT6. In one form or another, this engine is used on more than 130 different aircraft types, both fixed and rotary-wing.
It has also been used to power railway locomotives, 2 Indy cars, a Formula One car and functions as an APU for numerous large aircraft. More than 51,000 PT6 engines have been shipped, and together they have accumulated nearly 400 million flight hours.
Doug Millar was part of the team responsible for the design of the PT6's power turbine. During WWII Millar flew Douglas C47s over the Hump into Burma, and after the war, in 1952, he obtained his doctorate from MIT.
After working with Elvie Smith (later to be Pratt & Whitney's president) at Canada's National Research Council, he eventually joined Canadian Pratt & Whitney in Jan 1957 after convincing Smith to take a job there. His first task was to support the design of the JT12, which powered a variety of aircraft, including the Canadair CL41 Tutor (the CT114 of RCAF Snowbird fame), the North American Sabreliner and the Lockheed JetStar.
Millar joined Fernand Desrochers, John Vrana, Elvie Smith, Jack Beauregard, Jim Rankin, Allan Newland, Hugh Langshur, Richard Guthrie, Arthur Goss and Pete Peterson in Jan 1958 to begin work on Design Study 4, or DS 4—the PT6's initial designation. Gordon Hardy, Ken Ellsworth and Fred Glasspoole joined the team a little later.
The maximum theoretical power that can be generated by a turbine engine is governed by the overall pressure ratio, the maximum temperature entering the turbines and the amount of air flowing through the engine.
Losses through a real engine, including compressor and turbine inefficiencies, eat away at that maximum. The diameter (which controls the air flow) and weight of the engine are set by the size of the aircraft that will eventually use that engine.
DS 4 was intended to produce 500 shaft horsepower (shp), measure no more than 2 ft in diameter and weigh about 300 lbs. These design parameters imposed a few obstacles that the team had to overcome, the first of which involved the compressor.
To get the desired pressure ratio, more than 3 axial compressor stages were required and, with the maximum engine diameter set, the blade size for the last stages would have been ridiculously small. The clearance gap at the blade tip would be a significant percentage of the blade length, destroying efficiencies. The solution was to switch to a centrifugal compressor after 3 axial stages.
The turbine design was even more challenging. The decision to have a "free turbine" (see below) had already been made. Using classical design, that would have meant a 2-stage turbine to run the compressor and another 2-stage turbine to power the propeller (through a gearbox)—but that was too heavy. Millar was given the task of doing the job with 1 stage for each.
In front of each stage (or rotor), there is a row of fixed or "stator" blades whose function is to swirl the incoming gases at the correct angle to hit the spinning rotor blades. To do the job with 1 stage for the power turbine required a huge change in the direction of swirl between the exit of the compressor turbine and the entrance of the power turbine, destroying efficiency.
The design team's solution was to have the power turbine rotate in the opposite direction compared with the compressor turbine. This resulted in efficiencies for both turbines on the order of 90%—an excellent outcome.
As with most solutions, though, this caused another challenge. The shaft from the power turbine would have to run inside the shaft connecting the compressor and the compressor turbine on its way to the front of the engine to drive the propeller. The compressor and its turbine rotate at 37,000 rpm, while the power turbine rotates at about 33,000 rpm.
Fernand Desrochers, who was responsible for stress and vibration, was horrified. With the turbines rotating in different directions, the bearing between the 2 shafts would have to run at 70,000 rpm—an impossible task. The design team suggested flipping the engine back to front so that the power turbine shaft could connect directly to the gearbox at the rear (now the front) of the engine. This further lightened the engine, as the shaft from the power turbine would be much shorter.
The first flight was in the nose of a Beech 18 on May 30, 1961.
A final significant challenge arose during icing tests. The engine was very susceptible to icing as the intake screen on the engine itself could become almost totally blocked. Pete Peterson came up with the idea for an inertial particle separator (IPS), which was also described in a previous article. (See Pro Pilot, Feb 2013, pp 84–88.) The reverse flow arrangement of the PT6 allows for 2 small doors, as shown in the diagram on p 74.
When the pilot activates the IPS, the forward door closes, diverting the air and forcing it to make an abrupt 90° turn as it enters the engine. Most, if not all, of the supercooled water droplets cannot make the turn and continue straight on and out the back of the engine through the aft door, which is now open.
The same is true for dust and debris. This greatly reduces the chance of foreign object damage (FOD). For normal operations, the front door is open and the aft door is closed.
The 1st production PT6 was delivered to Beech on Dec 22, 1963. It was installed in a new version of the piston-powered Queen Air, to be called the King Air. With approximately 7000 King Airs produced to date, this represents 14,000 engines in one aircraft family!