EC225 crash near Bergen, Norway April 2016

[Concentric]

This is my first posting on this forum, having followed it for several years and learned a lot from some very knowledgeable folks. Since I am an engineer and SLF and not a professional pilot I have been reluctant in the past to comment, but for the subject of this thread I believe there was nothing that pilots could have, or should have, done to prevent the crash. It was clearly a failure of engineering somewhere along the line whether that was design, manufacturing, inspection, materials, maintenance or assembly.

I will not be the only member to have their personal theory as to how the accident occurred but I will make the following simple observations and ask a few questions.

Firstly, with reference to the AIBN Preliminary Report dated 27/5/16, Figure 4 shows the broken upper end of the front suspension bar which has clearly fractured in a ductile manner from a gross tensile overload, evidenced by the elongation of the male eye. Any suggestion that it was not properly pinned at either end or that the lower connecting lug bolts were incorrectly torqued contradicts the evidence. The bar must in fact have been very well restrained and under a tensile force much greater than its design load before it failed in this manner, assuming it was made of the correct material. For anything not designed to, other than the pins and bolts, to have transmitted this force is, frankly, far-fetched. For the front bar at least, it would also appear that the direction of loading was close to its normal arrangement as the pin is not bent and failure was across minimum tensile section.

Secondly, the ATC radar plot (Figure 1) gives a timeline and some indication of groundspeed of LN-OJF. The radar returns are at approximately 4 second intervals and the (straight line) distance covered between 09:54:44 and 09:54:48 is only about 50% of that covered in the preceding 4 seconds. If I have scaled off the map correctly, the initial groundspeed appears to be approximately 148 knots but the average groundspeed for the 4 seconds after 09:54:48 appears to be approximately 74 knots (it could be higher if the aircraft followed an S-turn). The CVFDR is reported to have shown everything ‘normal until a sudden catastrophic failure developed in 1-2seconds’.

The (average) groundspeed after 09:54:48 scales as approximately 52 knots, by which time, taking the wind direction (190 – 200degrees) and final position of the MRH into account, the MRH had probably separated from the aircraft. Thereafter the fuselage would be a ballistic projectile with only air resistance having any effect on its groundspeed. Within the 4 seconds after 09:54:44 the aircraft had decelerated from 148 knots to slightly over 52 knots regardless of flight path. It is even possible it could have been during just the last 2 seconds of that interval.

This would appear to be a very high rate of horizontal deceleration (up to 2.5g) in addition to normal ‘g’ and any additional vertical acceleration and/or rotational acceleration. I will leave it to the professionals to suggest whether this is within normal airframe loading limits and how a helicopter can be made to decelerate this rapidly. Presumably it requires a very severe flare and the rotor to be attached? On top of mast axial loading such a rotation must impart a moment to the top of the gearbox, increasing load on the front suspension bar.

Thirdly, Figure 7 and Figure 9 show how the epicyclic external ring gear burst open across both 1st and 2nd stages around the 4 o’clock position directly adjacent to an M/R servo location. What would be the effect of a planet gear or fragment bursting out and breaking the right hand rear servo?

Fourth, the aft RH and LH suspension bars and all pins appear to have stayed together and the logical conclusion is that the bolts connecting the lugs to the airframe failed. The AH EASB 53A058 requested operators to check, and report back to AH, torque measurements on the airframe lug mounting bolts. Checking the torque is a safety measure; reporting torque figures from across the world could be construed as something else.

Presumably these bolts are designed to take the full design load of the suspension bar, crucially when the geometry is as per the normal arrangement. If designed on the same basis as the front bar the bolts should be capable of resisting failure load of the upper eye. The rear bars did not fail although the LH upper lug looks a little bent. Its lower pin is the only one photographed fitted – does that imply it has bent and could not be removed? Could it have been the last to fail?

If the geometry changes, for example if the front bar has already failed and the mast tilts backwards, the bolts could be subjected to prying action, overloading them in rapid succession. The EASB does not show how long these bolts are. If they are fatigue critical and are required to retain pre-tension they should have as long a length as practical, preferably with L/d> 5 (where L = clamp length; d = diameter). If they are very short they could fail without significant bending of the brackets. As an aside, are there dissimilar materials between the bolts, lugs and airframe? Given the short time since the last MGB and MRH replacement I am not suggesting corrosion would be responsible but it is an interesting design feature.

Failure of the rear bolts, assuming they were new and correctly fitted along with the MGB, is highly suggestive that they failed later in the sequence of events through tensile failure and not fatigue and therefore were not the initial cause.

The AIBN’s latest report of 1 June 2016 points to the very serious possibility of fatigue failure of a 2nd stage planet gear. It is fortunate they recovered 2 parts of this gear with 4 fracture faces. The missing fragments will eventually be found although corrosion may by then have destroyed some of the evidence of beach marks.

I think the AIBN should be commended on how open they have been to date. It would appear that some of their public updates have been forced by the unusual and less than cautious behaviour of Airbus Helicopters in making their own opinions public, independently of the official investigation. I think the AIBN’s open approach will be appreciated by the relatives of those killed in this latest human tragedy and in G-REDL. As an engineer, I know that every day is a school day and I feel for the design engineers at Airbus and maintenance engineers at CHC. What normally works, on this occasion and possibly one previous hasn’t, and they need to understand why that is the case. Above all, it is important to be earnest.

So there you go, that’s my first tuppence worth!