Friday, April 1, 2005
The Reality of Turbine Autorotations
Fortunately, not many of us have real engine failures. We all should be practiced enough to know how to recognize and deal with one, so it should be no big surprise. Instinct takes over and we do what should come naturally. The problem is that, in a lot of turbine helicopters, what we've practiced and learned to judge things by is not always what happens. Unless the engine gives very clear signs of failure, nearly everyone is surprised when it goes. This is a bit different than what we've practiced.
On every training trip or check ride, you know you're going to get engine failures thrown at you. How realistic they are will depend on where you are. Some folks only ever get the milquetoast, "60 kt., lower the collective and roll the throttle off" type of entry into this "emergency." Others will get something more sudden: "Engine failure--Go!" Some outfits will only practice engine failures at an approved landing area. Regardless of how it's done, you are spring-loaded to expect the "failure." That's the first unrealistic thing.
The second is that, unless the instructor has a death wish, you'll be practicing somewhere pretty safe--not inside the height-velocity curve, for example, or over very inhospitable terrain. If you can guarantee you'll always operate in those conditions, this is realistic. If you routinely operate within the HV curve or over inhospitable terrain, you need to be aware that your training may not have prepared you for the real thing.
The third unrealistic thing is the difference in how a failure is simulated vs. how an engine really fails. For nearly every turbine helicopter I've flown, retarding the throttle to idle is the accepted way of simulating a failure. But engines don't quietly run down like this all the time.
For one popular engine, with six axial stages in the compressor, the deceleration is decidedly gentle regardless of how fast you cut the throttle. The fuel control can't chop the fuel too fast or the engine will flame out. Even if the engine did lose fuel, the compressor's deceleration would be slow because of the mass of its parts. While the compressor is still compressing, air under pressure is getting to the power turbine, making deceleration of the rotor slower than it would be if, say, the power turbine blew up. A different model of this engine has a single centrifugal compressor and the power reduction when its throttle is cut is quite dramatically different.
Other things can cause the engine to stop producing useful power. The drive shaft to the transmission can break. Rare, but it has happened. Keep in mind that the rate of power loss in a real failure may be much different than during practice.
These three items are easily understood and the price we pay to simulate things safely. But the next is one that has caught a lot of people by surprise--and it's a nasty surprise.
Even at idle, the power turbine is still producing power. If you don't believe it, figure out how the rotor stays turning. (Fixed-shaft engines like those on the Gazelle and Alouette/Lama are different.) This may not seem like much power (and for that part of the descent in which the rotor is split from the power turbine, it may not seem of much consequence) but it is enough. Now comes the math. A helicopter's rate of descent in autorotation depends on the weight, density altitude and airspeed. If we keep everything constant, we can figure the change in descent rate if the engine isn't producing power.
Rate of descent can be calculated. (A very close approximation is: 33,000 x change in shaft horsepower from that required in level flight, divided by gross weight.) Let's say we weigh 3,300 lb. and we normally require 160 shp. in level flight at 60 KIAS. If the engine is producing 20 shp. at idle, the rate of descent in our practice would be 160 - 20 = 140 x 10 = 1,400 fpm. If the engine were really stopped, we'd be doing 1,600 fpm.--about 15 percent more. Given that helicopter glide angles are already pretty steep, adding to it at the wrong time isn't much fun.
Some may say, "But the rotor is split from the power turbine and is being driven solely by the air passing up through it. Why is the rate of descent more if the engine is really failed?" The answer is that the turbine is still producing power that is going to the transmission--which takes power to turn things like the hydraulic pump. So even if the rotor is split from the power turbine, that turbine is overcoming some of the losses. With the engine stopped, that power has to come from the rotor.
You now have to stop the increased descent rate and steeper glide angle. This takes more energy than it did with the engine on, because the change in flight path is slightly greater. You'll feel like you're falling through a part of the maneuver that used to be pretty straightforward. If you get that right, the touchdown's another unpleasant surprise. The rate of rotor RPM decay at the end of the autorotation will be much higher because you don't have that idling engine keeping the rotor turning. So you may land more heavily than you did in practice.
The astute of you will wonder, "How does this affect the HV curve? Shouldn't it be bigger than you'd suspect?" In reality, the HV curve is a very complex subject and the authorities will note the differences between engine at idle and engine really off. Every HV demonstration for certification I've seen resulted in no damage to the gear or aircraft, which means the rotor had reserve energy left. It would be reasonable to expect that in a real failure the gear might have to soak up more energy.