Air Marshal Philip Rajkumar (Retd)
I joined the Indian Air Force (IAF) in 1962 and flew de Havilland Vampires, French Ouragans, Mystere IV A’s and Russian Mig-21’s for the first nine years accumulating 1600 hours in my log book. In 1972 I graduated as a test pilot from the French Test Pilots School (Ecole du Personnel Navigant d’Essais et de Reception or EPNER) at Istres, France and was posted to the IAF’s Aircraft and Systems Testing Establishment (ASTE) which is their flight test centre at Bangalore. ASTE was located at Bangalore airfield because India’s only aircraft company Hindustan Aeronautics Limited (HAL) owned the airfield and had its design bureau there. In 1975 I was assigned to be the project pilot for a flight trial in a Gnat fighter to prove some electronic warfare equipment developed in India. The trial was to take place at a base near New Delhi in north India and I had to ferry the aircraft there.
The Gnat was a small single seat light weight fighter designed and developed by the Folland Aircraft Company at Hamble in the United Kingdom. In the 1960’s this aircraft was produced under license at HAL, Bangalore. The chief designer of the aircraft was Ted Petter who had become famous as the designer of the English Electric Canberra medium bomber in the second half of the 1940’s. The Gnat had a wing span of only 22 feet. The high wings had a sweep back of 40 degrees and the basic empty weight was 4600 pounds. The maximum all up weight was 8700 pounds with full internal fuel of 1590 pounds plus 2500 pounds of external load. The power plant was a Bristol Siddley Orpheus 701 turbojet which drew air through side mounted air intakes on the fuselage and produced 4850 pounds sea level static thrust. The top speed was 620 knots at sea level and Mach 0.95 could be reached at 42000 feet. The aircraft had a hydraulically powered flight control system consisting of a powered slab tailplane and ailerons. The rudder was manually controlled. Petter had used some innovative features in his design to keep the weight down. When the tricycle undercarriage was lowered the ailerons drooped down by 22 degrees and acted like flaps (flaperons) though roll control was still available with differential movement of the flaperons. Partial lowering of the undercarriage provided the airbrake function. The 1800 psi hydraulic system was powered by a single engine driven hydraulic pump and in case of a hydraulic failure reversion to manual control was possible. The slab tail could be split by pulling a lever in the cockpit into a fixed electrically trimmable tailplane and a manually operated elevator. The ailerons could be operated manually though the roll control forces increased quite a bit. In contrast the manual elevator had very light control forces and flying in manual was tricky because of this total lack of harmony between the roll and pitch axes. The standard emergency actions in case of hydraulic failure was to split the tail, put the hydraulic cock off and exhaust hydraulic pressure trapped in the pressure lines to the aileron power jacks by gently moving the ailerons. For some unknown reason these jacks were called servodynes and I will be using this term in the rest of my story. Exhausting pressure in the sevodynes took about 60 seconds depending on the amount of roll control being used after the hydraulic failure had occurred. Thereafter the undercarriage had to be lowered using an emergency hydraulic accumulator and a landing carried out in manual. The toe brake pedals on top of the rudder pedals had their own independent hydraulic supply sources and were not affected by the failure of the main hydraulic system. There were fairly severe crosswind limitations for landing in manual because of the high roll control forces. The aircraft was fitted with a light weight ejection seat made especially for the aircraft by Follands.
On 2nd August 1975 I had to check general aircraft serviceability of Gnat IE-1071 before the long ferry flight to Delhi. I did not have very many flying hours on the Gnat having converted recently to that type. I, therefore, went across to the office of the Deputy Chief Test Pilot of HAL, Wing Commander Tilak the previous day and discussed Gnat emergencies with him. In particular we discussed emergency actions at length in case of a hydraulic failure. The monsoon was active over Bangalore with complete cloud cover starting at about 5000 feet above ground level all the way up to 30000 feet. The surface winds were about 15 knots gusting to 20 knots but fortunately right down the westerly runway which was in use. After start up at about 1000 hrs I taxied to the active runway but did not feel happy about the way the stick (control column) was self centering when left aileron was applied. This was a standard check carried out by all Gnat pilots because there had been more than one unexplained Gnat crash wherein aircraft had rolled into the ground immediately after getting airborne. The pilots had been killed. There were no crash data recorders fitted in aircraft those days and meaningful investigations to arrive at the precise causes of the accidents were not possible. The Courts of Inquiry had recommended that self centering of the stick be checked at frequent intervals while taxiing to make sure both aileron sevodynes were in power before takeoff. It was also found out by the Courts after discussions with designers both in the UK and Bangalore that when the ailerons were drooped (the configuration during take off until the undercarriage was retracted and on final approach after the undercarriage was lowered) and one aileron reverted to manual it would not be possible for the pilot to control the roll because of excessive control forces.
With these thoughts playing on my mind I lined up for take off and did a final self centering check of the stick before releasing brakes. The stick stayed stubbornly to the left extreme when I did the check. That was it and I decided to abort the sortie and return to the flight line. I reported the matter to the maintenance crew and my immediate superior Wing Commander Babi Dey who was an experienced Gnat pilot having participated in Gnat development trials in the UK in the late 1950’s. In contrast to Babi’s several hundred hours of experience on the Gnat I had barely a dozen hours on the type. I felt his assessment of the problem and diagnosis would be valuable for rectifying the aircraft. Before the maintenance technicians started to investigate the problem Babi said he would like to start up the aircraft and check out the flying controls. So he went and checked out the aircraft’s flying controls thoroughly and said he could find nothing wrong with it. I asked him about the self centering behavior of the stick and he replied that he had found absolutely no evidence of the problem reported by me. He then told me to go ahead and fly the routine air test I had planned. The sortie profile entailed a climb to 40000 feet to check out the cabin pressurization, temperature control and cruise fuel consumption in the clean configuration as I had to ferry the aircraft without the 66 gallon drop tanks. The IAF rarely ever flew the Gnat without drop tanks and these fuel consumption figures were not readily available. I thought it would be a good idea to collect some real world flight test data.
It was around noon when I started up the aircraft for the second time that day and this time the aileron servodynes and the self centering mechanism of the stick behaved normally. I shut the hydraulic cock in the cockpit and reverted the ailerons to manual and reengaged power. Self centering of the stick was normal and I became convinced that both servodynes were functioning normally under hydraulic power.
I taxied out to the active runway all the while checking the self centering behavior of the stick , took off and commenced a full throttle climb to 40,000 feet. The clean Gnat could climb to that height in four minutes from brake release which was pretty impressive for a non afterburning aircraft designed in the mid-1950s. I came out of the overcast at about 30,000 feet and relaxed for a moment to look at the clear horizon, the blue sky above and white cloud cover below. At that moment the control column jerked sharply to the left and the aircraft had almost 90 degrees of bank to the left in an instant. Even though I was in a climbing attitude the nose dropped quite a bit and I lost about 300-400 feet of height. Fortunately I did not enter the clouds. I rolled back to the wings level position, throttled back and discontinued the climb but I could feel that the ailerons were still in power. I realized that one of the ailerons may revert to manual at any moment. With one aileron in power and the other in manual I would have great difficulty in controlling the aircraft. With complete cloud cover just below me any roll control difficulties would make me lose height rapidly and plunge me back into the clouds. If that happened ejection was the only option .I felt that discretion was the better part of valor and decided to pre-empt a possible emergency. I selected the hydraulic cock off, split the tail and started to exhaust hydraulic pressure to the servodynes by gently rolling the aircraft from side to side. After about a tense minute I felt both the ailerons revert to manual with a reassuring jerk. With the hydraulic system inoperative the next action was to get the undercarriage down and locked using the emergency accumulator. I pulled the lever and waited for what seemed a long time and then I had three green lights in the cockpit indicating that the wheels were down and locked. I could not do a very rapid descent through clouds because I was flying in manual with the wheels down and I had to respect the laid down speed limits. I informed the Bangalore air traffic control that I was returning to base for a priority landing and asked about the surface winds. Fortunately for me the winds were still blowing down the runway in use though it was gusting up to 20 knots. I flew a long flattish approach and carried out my first and only landing in manual control up to that time on the Gnat.
Babi Dey was waiting for me when I switched off at the flight line. He started up the aircraft immediately after I got out. This time when he selected hydraulic power on he was able to see the self centering behaviour of the stick which showed that the left servodyne was in manual while the right one was operating with hydraulic power. For the very first time in 17 years of Gnat operating experience in India we had physical evidence available for proper investigation after a serious roll control malfunction had occurred in the air. HAL technicians removed the left and right servodynes and took them for strip examination to the hydraulic shop. I went with them because I was overcome by curiosity. When the hydraulic filter fitted in the body of the left servodyne was removed we could see the filter element covered with a lot of black stuff some of which was in the form of strands. The right servodyne filter also had the same stuff but to a lesser extent. There was obviously a contaminating source in the aircraft’s hydraulic system. Microscopic examination of the material found showed that it was probably paper or cardboard pieces. Chemical analysis revealed that the substance was mainly cellulose.
We went back to the aircraft in the hangar and opened out the hydraulic system and started to look for a common source which could contaminate both servodynes. The system had only one common hydraulic filter for both servodynes and was the probable suspect. When the metal housing of the filter element was opened we found that the filter cartridge made of cardboard and paper had burst and sent a lot of debris into the pressure lines connected to the servodynes. At long last the mystery behind many unexplained rolling flight control problems in the Gnat was solved.
Remedial measures were instituted in the form of reduced shelf life for filter elements and fairly frequent checks of the filter elements fitted in aircraft on the flight line in operational squadrons. There were no further incidents of this type in the squadrons. When HAL modified the Gnat into the Ajeet fighter in the mid 1970’s the common hydraulic filter was replaced by two filters and an electric switch was provided in the cockpit which enabled the pilot to revert both ailerons to manual in an instant. There was no need to put the hydraulic cock off and wait for up to a minute for reversion to occur.
As in all such incidents luck played a major role in letting me bring back the evidence of a flight control system malfunction. If the problem had occurred on takeoff or just after lift off the aircraft would have rolled into the ground giving me little time to eject. Had it happened during the climb through the overcast I would have been disorientated and I would have had to eject. If it had happened when I broke cloud cover I would have plunged back into clouds with the very first sharp roll to the left. Perhaps the most significant factor was my discussion with Wing Commander Tilak about emergency actions in case of hydraulic and control problems The incident took place when I had an even chance of saving the aircraft and I knew exactly how to handle the emergency. Dame Fortune certainly smiled on me that day!