Peter Gray/STOCKHOLM
Until recently, civil aviation authorities allowed only limited training credit for using a helicopter full-flight simulator instead of the aircraft. This meant that operators had to carry out substantial pilot training and testing in the aircraft, with all the disadvantages of cost, aircraft availability, air traffic control limitations, airframe wear and tear, potential safety risks and, not least, the environmental issues of noise and pollution.
The limits were imposed because of the lack of simulation fidelity and the absence of an adequate daylight visual system that could be viewed simultaneously from the pilot and co-pilot seats. The lack of realism is the result of a scarcity, or even absence, of data from the helicopter manufacturers. This has forced simulator manufacturers to estimate the aerodynamic and other characteristics.
Given the complexities of the helicopter, it is no wonder manufacturers have been unable to reproduce faithfully all its handling and performance characteristics throughout the flight envelope. But some manufacturers now provide limited data.
Meanwhile, two helicopter simulators with much more realism and greater credits than usual have recently been certified: FlightSafety International's Bell 430 in the USA and SAS Flight Academy's Bell 212/412 in Sweden. Flight International evaluated the Bell 430 last year (16-22 September issue) and was subsequently invited by SAS to test its machine.
Canada's CAE Electronics, manufacturer of the Swedish centre's simulator, has overcome the lack of data to a large extent by using a professional test pilot to alternate between the aircraft and the simulator, to help designers match the device to the helicopter. The machine is based at the Scandinavian Airlines System training centre just a short bus ride from Stockholm's international airport.
Used for training and testing of normal and emergency operating procedures, simulators are also ideal for crew resource management training, the lack of which has contributed to, or even caused, accidents. Sitting behind the pilot and copilot, instructor/examiner can observe and record how they work together as a crew as they go about managing the flight.
The SAS simulator came on line at the end of last year. It has three configurations - Bell 212, 412SP and 412HP. The visual system is a CAE MaxVue with a 210¼ horizontal by 40¼ vertical wraparound field of view. Two chin windows, one for each pilot, also have a visual display.
The simulator has the usual six axes of motion, with an additional separate three-axis cockpit vibration platform. I found this most realistic. I examined the instructor's station as the door behind us closed and the drawbridge came up. It is much more user-friendly than older models. The instructor can pre-program the flight and spend most of his time observing the crew. The program can be stopped in mid-flight and rerun if necessary - particularly important if the pilot gets it wrong or needs practice. The flight parameters can be recorded, printed and used for the debrief.
I strapped myself into the captain's seat on the right-hand side. The cockpit is identical to the real aircraft in every aspect (even the poor lighting). Lennart Samuelsson, the academy's chief Bell flight instructor, had positioned us on the threshold of the local airport's runway. It was a nice sunlit day and the outside view was of a typical scene. I could see the control tower and other buildings and some hills in the background. The runway ahead had rubber scuff marks and there was grass alongside. Alas, neither were in sharp enough focus to allow me to judge just how far I was away from them - this is one of the basic problems with all helicopter simulator visual systems. The view of the outside world through both side windows and my chin window was excellent, however.
Samuelsson gave me both engines and the rotor at flying RPM, ready for take-off. Like other pilots, I have had bad moments with overcontrolling and excessive aircraft movement during first simulator take-offs. Concentrating heavily, I tentatively raised the collective pitch lever and watched and felt for any suggestions of an attitude change or yawing. We broke ground, although I could not tell exactly when, and arrived in space at an undeterminable low hover height. However, although the hover was erratic, I stayed over the piano keys painted on the runway, the aircraft pointing down its length, a much more successful first attempt than in any of the other eight simulators I have flown. Feeling confident, I tried some sideways, backwards and spot turns. I had to work hard to achieve a lower standard than in the aircraft, but then simulators are not intended for teaching such manoeuvres.
Fiords, valleys and mountains
I went into forward flight and joined another helicopter ahead of me in loose formation. SAS has produced a visual database covering an operating area of about 140km² (50 miles²) and based on actual topography. It has an onshore and an offshore area and contains islands, fjords, valleys, mountains and a three-dimensional moving sea with ships, oil platform and liferafts. Other operating areas can be produced at customer request. I descended to low level over the rolling countryside. The aircraft handled realistically, with equally realistic noise and vibration levels. The outside world, although artificial-looking, was good enough to demand caution in avoiding obstacles and looking where I was going, low and fast. This fun ended when a warning light directly in front of me on the instrument panel indicated that a ground radar station had locked on to us (the simulator is available to military customers). I dispensed chaff and jinked, taking care not to dive. The light went to unlocked.
We went offshore to an oil platform. I set up an approach to the helideck, using my North Sea technique of a steep approach. I had difficulty, however, in judging my rate of closure and finished up too high, starting to lose my concept of where we were in space. This illustrates the problems that still beset helicopter simulators. The brain needs to evaluate accurately the size of the objects around it and have the eyes focus and converge on close-up things. This is particularly important with objects passing alongside. While there is no difficulty in assessing size, one's eyes stay focused at infinity in a simulator, destroying the sense of rate of closure. I climbed away and flew another approach, dragging us in almost level and at a hover taxi pace. We arrived over the centre of the deck and I landed. But it was hard work.
I came to the hover and continued the vertical climb, using the chin window and whatever else my eyes could grab to keep me orientated in space, bearing in mind that all I had ahead was the edge of the deck and open sea. I did a circuit and returned. On the next departure, after asking for maximum weight, I requested an engine failure just at the rotation point when I was committed to continue the take-off. This is not a manoeuvre that any sensible operator would try in reality. There was a bang, a yaw and a loss of rotor RPM. I lowered the lever slightly, to maintain the rotor RPM at its critical minimum of 97%, and concentrated on getting forward and upward speed. It worked out well, although I had to be quick to lower the lever to contain the fading RPM. It was a useful exercise.
We made our way to a large boat equipped with a helideck and, during our journey, daylight became dusk and then night. We arrived alongside the boat and, with the whole vessel well illuminated, I was surprised to find that I had no trouble holding a good hover alongside. Samuelsson, meanwhile, had increased the intensity of the simulated waves, which helped.
I did an instrument departure into the inky black of the night and experienced some heavy icing conditions. I had to increase power to keep the same speed. One engine suffered an inlet particle separator failure, which allowed ice to enter the engine. After wild RPM fluctuations, I shut it down. At this point, if I were a long way offshore, I would have had to consider my single-engine diversions. My crew resource management was being exercised.
Next came engine governor/fuel control malfunctions. In the real world, these are fraught with the distinct possibility of the pilot making the wrong diagnosis and taking the wrong action, sometimes with dire results. The flight manual has lots of good advice, but you cannot simulate these malfunctions in the air. The simulator has every known governor/fuel control malfunction available and I spent valuable time analysing and practising some of them. The engine-fire drill in the Bell 212 and 412, as with some other helicopters, is easy to get wrong. Pulling the lit T-handle, among other functions, shuts down the offending engine and if the pilot then closes the wrong throttle, he finishes up with no engines. If he gets it wrong in the simulator, no harm is done. The instructor can put the fire out at this point or keep it going, in which case the pilot must carry out an immediate forced landing. This type of exercise makes the best use of the simulator and gives the pilot invaluable experience.
Samuelsson put me back in the circuit and gave me a double engine failure. The entry and establishment of autorotation was normal, but the critical part would come with judgement of the flare, levelling and landing. It was harder to judge than in reality, but we did a safe running landing, engines off. I heard the skids touch and scrape along the runway, and the subsequent deceleration effects - nice, realistic, touches and good motion system programming. Some operators have done engines-off training and testing in the actual aircraft, but nearly all find that they eventually damage more machines than actually suffered real double engine failures, a rare occurrence.
I then showed Samuelsson my own version of an autorotation to the surface in instrument meteorological conditions, descending initially at normal autorotation speed of 70kt (130km/h) then selecting and holding 5¼ nose up at 600ft. The aircraft decelerated all the way down and I arrived at the bottom with hardly any forward speed, a modest rate of descent and slightly tail down - which is good, especially if landing on water - taking a first bite of the lever at 50ft radar altitude and the rest just before landing and levelling. It is a benign technique that even a below-average pilot can do successfully, but is not to be practiced in the aircraft.
Tail rotor failures
The Bell flight manual contains five pages describing tail rotor failures - excellent stuff. But none of the recommended procedures can realistically be simulated or practiced in the air. One can only digest the contents and hope that, should you have a tail rotor failure, you can remember what to do. This is where simulators come into their own. I tried drive shaft and pitch change failures at various stages of flight, relying on my memory. I could have walked away from some of my arrivals; others I could not have. I asked Samuelsson to show me the procedures for those instances where I had failed.
There continue to be fatal accidents when pilots get themselves into trouble flying in poor visibility by getting too low and hitting the ground or water, or, equally catastrophic, by the aircraft getting into an unusual attitude from which the pilot does not recover. Both these events can be safely addressed in the simulator by practising low level turns and recovering from unusual attitudes.
One of the most confusing of malfunctions in all helicopters is failure of the drive between the powerplant and main gearbox. The situation is exacerbated at high power settings. The engines are suddenly offloaded, so immediately suffer a gross overspeed. The rotor is no longer being driven, so RPM fades rapidly. What the pilot experiences, as well as the noise of the fracture and the screaming of the engines, is excessively high engine RPM and excessively low rotor RPM - an enormous, confusing split of his tachometer needles and a confusing situation.
If he has been trained well, the pilot's first reaction should be to check his rotor RPM, that vital function that keeps the helicopter in the air. This should induce him to lower the lever, enter autorotation and then deal with the overspeeding engines. I asked Samuelsson to give me a shaft failure at high forward speed some time during the flight. The impact when it happened was dramatic and, by the time I had evaluated the situation and entered autorotation, rotor RPM had dropped to 80%: invaluable training.
The tail rotor pitch change mechanism in the 212 and 412 uses power from the number one hydraulic system. Its failure means that the pilot has to absorb and contain the tail rotor feedback forces. There is no problem in cruise flight, but considerably more pedal pressure is required for the approach, hover and landing. Most pilots, given the choice, will do a running landing. Always examining the worst case, I chose to come to the hover and land. The control forces were identical to those of the aircraft. By this time, I noted that my hover was good and I was having to work less hard to achieve it.
The 412 has several Category A take-off and landing profiles. I chose the more difficult for any simulator, the vertical-type take-off where you raise and back off the helicopter to the critical decision point (CDP) of 160ft, keeping the helipad in the same position in the windscreen throughout. It was more difficult because of my lack of awareness of my position in space, but only required extra concentration to make it work. I had an engine failure at the CDP and rejected the take-off. I tried again, with an engine again failing at CDP, and this time I continued the take-off. Most operators, faced with doing this training in the aircraft, will reduce the weight to make the manoeuvre safer and reduce engine wear and tear, thus depriving the pilot of the knowledge of how it really feels. There are no such disadvantages in the simulator and I could to go to full single engine power and droop the rotor RPM to give me that extra momentary increase in lift for the landing.
Into the weather
Finally, I asked Samuelsson to program the weather during our last 10min of airfield work to deteriorate progressively in terms of visibility and cloudbase, to force me to revert to instruments. When the weather got below visual meteorological conditions, I climbed and did a procedural instrument landing system approach. I broke cloud on the approach at 200ft amid the gloom of a typical instrument day, reminiscent of my North Sea experiences. A quick glance outside showed the first few runway lights fading away in the 600m visibility.
Despite the huge amount of research and development that has gone into helicopter simulators, the basic problem remains of their not being as faithful to the aircraft as their fixed-wing brethren. However, these latest simulators I have tested are much more realistic than their predecessors.
Because of the nature of a helicopter, the simulator motion and visual systems must have much faster responses to pilot inputs than is necessary in fixed-wing aircraft simulators. Bigger and faster computers are being used to overcome the tiny delays that reduce fidelity and to put more dots and details into the visual scene. Despite their remaining limitations, the latest helicopter simulators are a formidable training and testing device.
A full-motion, full-visual simulator will cost three times the price of the helicopter, precluding the small operator from acquiring one, but the operating cost is normally less than the helicopter's. The maintenance required can be more than for the actual aircraft, but availability is normally not less than 98%, which no helicopter can achieve.
This is the most in-depth evaluation I have done on any helicopter simulator, and was possible because it allows so many more manoeuvres, especially emergencies and malfunctions, that cannot be done in the aircraft.
Source: Flight International
