When the Test Plan Went Wrong – Fire and Water – Part1

The first the pilot knew there was a problem was that the crew reported a smell or burning from the cabin area and some strange noises.?Nothing was noted on the electronic cockpit displays but he decided to raise the active sonar.?As this was happening there was an indication of hydraulic failure so the pilot elected to leave the hover (the sonar body still not fully raised) before a fire indication for the No2 engine illuminated.?Having turned towards the landing site at the test control station, the pilot continued to try and deal with the now large volume of cautions and associated audio warnings.?Then things started to go out of control and the aircraft yawed, rolled and descended striking the water 90 seconds after the first indication at about 80 knots, 30 degrees right and 10-15 degrees nose down.

The aircraft crashed at about 0950z that Friday morning having caught fire and then lost control on the BUTEC range near Plockton in Scotland.?Luckily everyone escaped physically uninjured and were quickly rescued by the Stornaway SAR helicopter.?I dealt with the immediate fallout (I was waiting for the aircraft in the hangar) but there was much more to it than that.

This is the story of what happened and why, the cause and the missing links in the chain. Strangely the report has never been publicly released by the MOD but there are many lessons to learn about flight test and operating aircraft in general.

The years 1999 and 2000 were a difficult period for the Merlin maritime helicopter programme as it was struggling to deliver the required outputs and meet the Royal Navy’s in service date requirements.?Causes were many and varied but much of it centred on the how ready the mission system and its sensors (active and passive sonar) were to deliver at least the same capability of the in service Sea King.?As a result novel processes were being looked at to increase test output to match an higher tempo of software and hardware releases for the helicopter.?Out team on the Merlin Intensive Flying Trials Unit (IFTU) were therefor tasked to deliver a test of new aircraft and mission systems software in order to move things forward.

We were lucky, in a way because all of us in the test and training office were flight test graduates (test pilots and Aerosystems), so we had the pre-requisite skillsets (or SQEP) to formulate the required test programme and support flight trials.?Information was passed to us by Lockheed Martin (Prime Contractor to Agusta Westland) on the scope of the change and the detailed modification information went through a series of reviews before it was installed by an industry engineer on two Squadron aircraft.

After installation the aircraft were tested, as ever, in accordance with the engineering documentation and then released for service.?Although most of the changes were to do with the mission system and its sensors, the change required the electrical inhibition of the rotor brake system, in the off condition, as the new software was not compatible with its complex operational sequence.

One aircraft showed no anomalies but the other showed a residual rotor brake pressure of up to 10 bars.?This was diagnosed as being residual pressure in the Number 3 hydraulic system and could be bled off by the engineering team to reduce it to zero.?The problem was it kept coming back, so a technical query was lodged with the manufacturer – an answer for which was never received.

The aircraft then departed from the home base, RNAS Culdrose, for the Scottish Western Isles, where the BUTEC range is located, operating from the small airfield at Plockton, close to the Isle of Skye.

Flying commenced with a series of dipping sonar sorties to test the new software and the performance of the sensor.?The weather for late October was fine and the aircrew were spread amongst a range of guest houses in the town driving to the small airport to brief and fly, whilst evening Squadron meetings were organised in one of the the local pubs.?Whilst the testing went well one of the aircraft quickly went down with a problem which I don’t quite recollect but our subject aircraft carried on the only issue being the rotor brake kept developing a residual pressure even though it was not being used.?So we kept bleeding the system and the four pilots on the trial became rather used to it.?

What we had noticed however was that, every time we raised or lowered the active sonar body (which was on a 2000 ft cable), we got a small kick in the yaw channel; no more than a couple of degrees, but always in the same direction. ??I had been planned to fly in the accident sortie but the night before, during a discussion on who would authorise it, I swapped with the Senior Pilot and instead acted as the duty person in the hangar ready to take the aircraft on the next trip.

So what happened?

The modification of the rotorbrake (or more precisely the inhibition) on the 2 aircraft had been completed with them in different conditions – one was folded and one was spread.?In the folded condition the rotorbrake is engaged (a complex and difficult to interpret process) whereby the rotor brake actuator is engaged but no pressure is on the brake pads.?This does however mean that the distance between the brake stator disc and the rotating disc was ?substantially reduced – to a position whereby any pressure to the system is likely to cause rubbing contact, but not brake application.

With the rotor brake hydraulic control module (RHBM) also in the ground position this meant that any pressure fluctuations in the system would lead to contact.

The rotorbrake on the Merlin is fed by the Number 3 system which is also used to power the undercarriage mechanism and the the sonar body winch, the first of which is typically used on every sortie and the second was the subject of the flight trials.?The route to the crash was set.

After the initial bleed of the system every time the undercarriage was raised and lowered, pressure spikes were fed back into the rotorbrake system, building up the residual pressure because of non-return valves and causing disc contact.?Further bleeding of the system served to prolong the process before the system failed.

As the sonar winch system was being used for long periods (up to 2 minutes) the pressure was being applied to the rotor brake disc. The transient yaw noted by the pilots being an indication that the engine and rotor governing system was having to apply more torque to make the rotor speed constant.

Whilst there was clearly damage being done it was not visible to the engineering team since the rotor brake is enclosed in a titanium housing and as a result rubbing damage was being inflicted cumulatively.?In the end the disc increased temperature to over 2000 degrees centigrade before it shattered in a molten condition breaching the titanium housing and cutting through a hydraulic line.

The misted hydraulic fuel ignited immediately, as you would expect and caused the whole top deck of the aircraft to be engulfed by flames, unseen by the crew but positively identified by the range control building.?From then on the aircraft was doomed.

As the pilot accelerated the aircraft from the hover the flames went around the gearbox and triggered the Number 2 engine fire indication, even though there was no fire.?It then started to soften the tail rotor drive shaft and after probably 20 seconds or so caused it to shear under torque whereby it sprang up toward the rotor head and destroyed the lateral control jack. The aircraft was out of control and crashed almost immediately after.

So what were the lessons??Well that comes in Part 2.