inigo88
Composite-lover
It's been a running joke on this site for some time that the Garrett TPE-331 direct drive turboprop engine is "all that is man," and is superior to the sissy Canadian Pratt & Whitney free-turbine PT-6. I've read many of the war stories from our resident freight dogs about the exploits of the Garretts on the MU-2s, Metros, Twin Commanders, Conquests, etc.
The idea of the centrifugal "start latches" or "prop locks" was a difficult one to wrap my head around - requiring shutdown in reverse to engage, a bump into reverse before taxi to disengage, and an unfeathering pump should you screw up the shutdown procedure and watch one of your props feather as the oil leaves the hub.
But even more numerous are anecdotes about the Garrett's weird pseudo-autofeather system, called Negative Torque Sensing (NTS) - which upon detecting negative torque in an engine failure, will mostly feather the prop but allow it to continue windmilling until the pilot can either attempt an air-start or manually feather the prop. (See NTS vs. Autofeather)
What's usually glossed over in the NTS discussions though is how the system knows there is negative torque. In the past Boris Badenov, mikecweb and others have attributed it to things like wizardry and PFM, but I've always wondered how the system actually works. So I sat down yesterday googling and geeking out on technical articles, and found the one below.
I believe the NTS Regulator is just a pressure relief valve, allowing oil to escape from the prop hub and activated by differential pressure from the Hydro-Mechanical Torque Sensing System. So to understand how NTS knows when to dump the pressure in the hub, we need to look at the Torque Sensing system as a whole - which also explains how the torque meters work in the cockpit! As usual, I found an outstanding article that explains it better and more articulately than I can:
(Link: http://airsoc.com/articles/view/id/...ed-explanation-of-torque-sensing#.UV94vVebY5Z)
Basically, even though the Garrett is direct-drive it still has two concentric shafts. The larger hollow main shaft is connected directly to the centrifugal compressor and turbine blades. Inside the hollow main shaft is a smaller torsion shaft. On the rear end of the engine, it is splined into the main shaft so that both shafts are always coupled together. At the front of the engine, the front of the hollow Main Shaft has a hollow gear which connects to one end of the Torque Sensor. The front of the Torsion Shaft continues forwards farther into the reduction gear box and connects into both the high speed pinion shaft (which goes through the reduction gearing and connects to the propeller) and the front gear on the Torque Sensor. (This is hard to visualize and I'll try and post a pic later.)
What this means is that even though the engine is direct-drive, the inside Torsion Shaft is what drives the prop, not the main shaft. Since the two are coupled together at the back, they should be spinning at the same speed, however when the Torsion Shaft spins the prop, the reaction force imparted by the prop due to Newton's 3rd Law ("Equal and Opposite") causes that inside Torsion Shaft to twist just a little bit relative to the main shaft. Since both the Main Shaft and Torsion Shaft are connected to the Torque Sensor by gears, that twist between the two shafts causes the Torque Sensor gears (shown in green in the diagram) to move the piston that meters the engine oil through the bleed port.
Finally, the differential oil pressure is measured one of two ways (depending on the generation of the engine). Older Garretts apparently used torque gauges that were mechanical differential oil pressure gauges... so some diaphragm inside would measure the bleed port pressure against the reference pressure (I think this is the "hydraulic compensator" the author refers to). Most later designs use electric torque gauges powered by the so called "Hydro-Mechanical Transducer," which measure that differential pressure using Piezoelectric pressure sensors attached to the accessory gearbox (where this stuff is all located), which take the oil pressure and create a voltage measured at the torque meters. Like the author says, the system creates a positive voltage at idle torque, a 0 voltage at 100% torque and a negative voltage at greater than 100% torque. This is also why the torque meters should read 100% with the battery on and the engines off, since no voltage is being applied to them.
Those two Piezoelectric Pressure Transducers which take the oil pressure and turn it into electrical signals are what makes the whole thing work. The negative voltage at over-torque allows the Torque-Limiter system to work, and a sudden negative torque from an engine failure activates NTS... presumably because as the torque-sensor spins the wrong way all the oil pressure collects at the opposite sensor, which trips the NTS system.
Hope this helps! It was certainly an entertaining way to kill a few hours yesterday.![Smoke :smoke: :smoke:](/forums/images/smiley/smoke.png)
The idea of the centrifugal "start latches" or "prop locks" was a difficult one to wrap my head around - requiring shutdown in reverse to engage, a bump into reverse before taxi to disengage, and an unfeathering pump should you screw up the shutdown procedure and watch one of your props feather as the oil leaves the hub.
But even more numerous are anecdotes about the Garrett's weird pseudo-autofeather system, called Negative Torque Sensing (NTS) - which upon detecting negative torque in an engine failure, will mostly feather the prop but allow it to continue windmilling until the pilot can either attempt an air-start or manually feather the prop. (See NTS vs. Autofeather)
What's usually glossed over in the NTS discussions though is how the system knows there is negative torque. In the past Boris Badenov, mikecweb and others have attributed it to things like wizardry and PFM, but I've always wondered how the system actually works. So I sat down yesterday googling and geeking out on technical articles, and found the one below.
I believe the NTS Regulator is just a pressure relief valve, allowing oil to escape from the prop hub and activated by differential pressure from the Hydro-Mechanical Torque Sensing System. So to understand how NTS knows when to dump the pressure in the hub, we need to look at the Torque Sensing system as a whole - which also explains how the torque meters work in the cockpit! As usual, I found an outstanding article that explains it better and more articulately than I can:
(Link: http://airsoc.com/articles/view/id/...ed-explanation-of-torque-sensing#.UV94vVebY5Z)
From The Shop Floor: A Slightly Simplified Explanation of Torque Sensing
![]()
The torque sensing system as used in the TPE331-5, -10T and -10 engines installed in Turbo Commanders displays the amount of power being generated by the engine on a cockpit gauge marked either as HP or TQ. While there are several means used in the industry to do this, in this case it is performed with a hydro-mechanical system.
These engines have what is essentially a built-in rotating torque wrench to sense output torque. This gives us the ability to adjust both engines to match power output, set the correct amount of power for a given situation, and obtain the maximum power available without damaging the airframe. So, how is this done, you ask? Let’s examine the system.
The TPE331 engine has two shafts, one inside the other. The larger, heavier shaft supports the compressor disks, the turbine disks, and a special splined spacer along with bearings and seals. It also drives a gear train for the accessories. As with any turbine engine, roughly two-thirds of the energy generated by the turbine section is absorbed by the effort required to spin the compressor disks and accessories. Approximately one-third of the produced energy is what is actually used to drive the load – the propeller.
The smaller, lighter, longer shaft assembly is used to deliver the remaining energy to the prop via a two-stage reduction gear train at a roughly 26:1 ratio (26.228:1 to be specific). At 100% RPM the main shaft is spinning at 41,730 RPM and the prop is spinning at 1,591 RPM. The torsion shaft nests inside the main shaft and connects to it by means of a set of splines at the aft end of the engine. This torsion shaft has a unique characteristic in that it twists a known number of degrees relative to the main shaft as the torque changes.
Inside the engine’s gearbox, the two concentric shafts (main and torsion) together spin the torque sensor. The main shaft is connected to the aft gear and the torsion shaft is connected to the forward sliding gear. This pair of gears is coupled by a set of helical gear teeth such that as the torque changes, the forward gear slides a piston in and out of a sleeve to control a bleed port in the chamber used to vary the oil pressure in a separate measurement chamber. The same piston also opens and closes a port for the NTS system (but that’s a subject for another time). So much for the mechanical portion of the system.
Now for the system hydraulics, where working fluid is constantly flowing during engine operation. Look at the graphic and follow the path the oil takes from the engine oil pump as it travels to the positive torque pressure regulator. This regulator drops engine oil pressure to a specified value. The pressure-regulated fluid then travels through a calibrated orifice into a measuring chamber.
As the torque produced by the engine varies, the torque sensor piston varies the opening in the bleed port. It closes the bleed opening as more torque is produced, and opens the bleed port when less torque is produced. So, the pressure in the measurement chamber is controlled by the amount of leakage created by the torque sensor assembly.
In the sketch you’ll notice a plug depicted below the orifice assembly. This is a historical artifact from the days when a “hydraulic compensator” was used; it has since been replaced by an adjustable electronic pressure transducer. In new-production gearbox cases this plug has been removed because it is no longer of use, and removing it simplifies the gearbox casting.
So now we have a working fluid pressure that changes in direct proportion to the output torque of the engine. We need to translate this pressure to something we can use in the cockpit. And, because this engine design operates with a slight vacuum in the gearbox, we also need to take that vacuum into account for the system to indicate correctly.
The gearbox housing has two dedicated ports for this purpose. One port provides the torque pressure, and the other provides the case vacuum pressure. Both ports are connected to an adjustable, electronic differential pressure transducer. The transducer in turn provides the electrical signal necessary for the cockpit gauge to indicate the engine’s power output.
The same signal also is used by the HP/TQ limiting system to prevent overloading the airframe. This electronic transducer is calibrated to provide a +5 VDC signal at what is idle torque, and a +/-0 VDC signal at 100% torque. The signal also goes negative VDC in an over-torque condition – but that’s another subject.
As a side note, does your HP/TQ gage display 100% HP/TQ when power is off, and 0 HP/TQ when aircraft power is turned on with engines off? It should. If it doesn’t, you should get with your favorite Twin Commander Service Center and have it corrected.
To calibrate this torque sensor, the engine is operated with a very expensive, very sensitive, specially calibrated electronic strain gauge type torque sensor mounted between the engine’s output shaft and a load prop. (You’ve probably heard mechanics talk about “the Lebow,” the strain-type torque sensor used for this task.) The engine is warmed up to get the oil to its normal operating temperature, then it is operated at various conditions (a set of specific torque values), followed by a specified torque output value representing the 100% power value for that engine/airframe combination.
The engine torque sensing system pressures (torque signal and case vacuum) at those two data points are noted and recorded on the DSC sheet (Data Sheet Customer) for that engine. Next, a torque transducer is connected to a voltmeter to monitor the output signal, and it is powered and connected to a variable, calibrated, pressure source. It is carefully adjusted to indicate the correct torque at the pressures for the various specified conditions and to indicate 100% torque at the pressure for that point. As a result of this careful calibration the transducer and engine are mated together. If the transducer is to be removed and mated to another engine, it has to be re-calibrated to match that particular engine/airframe combination.
So there you have it – a clever way to give you an indication of the amount of power your engine is providing. Of course, you also use other instrumentation to monitor the engine. And please note that this is not “dial a torque.” The torque transducer output voltage must be adjusted to the values specified in the DSC sheet for that particular engine and airframe combination. If not, you will get a false indication in the cockpit and possibly an incorrect torque limiting system operation.
Mike Grabbe has been in the business of maintaining aircraft since 1970. He worked for a couple of FBOs, and was a Field Service Rep for a turbine engine manufacturer as well as a factory service rep for the Aero Commander Division of Gulfstream Aerospace. He instructed for a number of years at an FAR 147 school, and prior to working at Eagle Creek Aviation Services was the Director of Twin Commander Maintenance Training at FlightSafety International in Bethany, Oklahoma.
Basically, even though the Garrett is direct-drive it still has two concentric shafts. The larger hollow main shaft is connected directly to the centrifugal compressor and turbine blades. Inside the hollow main shaft is a smaller torsion shaft. On the rear end of the engine, it is splined into the main shaft so that both shafts are always coupled together. At the front of the engine, the front of the hollow Main Shaft has a hollow gear which connects to one end of the Torque Sensor. The front of the Torsion Shaft continues forwards farther into the reduction gear box and connects into both the high speed pinion shaft (which goes through the reduction gearing and connects to the propeller) and the front gear on the Torque Sensor. (This is hard to visualize and I'll try and post a pic later.)
What this means is that even though the engine is direct-drive, the inside Torsion Shaft is what drives the prop, not the main shaft. Since the two are coupled together at the back, they should be spinning at the same speed, however when the Torsion Shaft spins the prop, the reaction force imparted by the prop due to Newton's 3rd Law ("Equal and Opposite") causes that inside Torsion Shaft to twist just a little bit relative to the main shaft. Since both the Main Shaft and Torsion Shaft are connected to the Torque Sensor by gears, that twist between the two shafts causes the Torque Sensor gears (shown in green in the diagram) to move the piston that meters the engine oil through the bleed port.
Finally, the differential oil pressure is measured one of two ways (depending on the generation of the engine). Older Garretts apparently used torque gauges that were mechanical differential oil pressure gauges... so some diaphragm inside would measure the bleed port pressure against the reference pressure (I think this is the "hydraulic compensator" the author refers to). Most later designs use electric torque gauges powered by the so called "Hydro-Mechanical Transducer," which measure that differential pressure using Piezoelectric pressure sensors attached to the accessory gearbox (where this stuff is all located), which take the oil pressure and create a voltage measured at the torque meters. Like the author says, the system creates a positive voltage at idle torque, a 0 voltage at 100% torque and a negative voltage at greater than 100% torque. This is also why the torque meters should read 100% with the battery on and the engines off, since no voltage is being applied to them.
Those two Piezoelectric Pressure Transducers which take the oil pressure and turn it into electrical signals are what makes the whole thing work. The negative voltage at over-torque allows the Torque-Limiter system to work, and a sudden negative torque from an engine failure activates NTS... presumably because as the torque-sensor spins the wrong way all the oil pressure collects at the opposite sensor, which trips the NTS system.
Hope this helps! It was certainly an entertaining way to kill a few hours yesterday.
![Smoke :smoke: :smoke:](/forums/images/smiley/smoke.png)