PPRuNe Forums - View Single Post - Qantas A380 uncontained #2 engine failure
Old 6th Nov 2010, 03:50
  #456 (permalink)  
Bizman
 
Join Date: Mar 1999
Location: Australia
Posts: 19
Likes: 0
Received 0 Likes on 0 Posts
Rotor Bursts

Have to agree with some observers here that the Signal to Noise ratio on this thread is very low.

Congrats to VR-HEE on a sober observation.

Some years ago, I had responsibility to liaise with a turboshaft manufacturer after one engine on a twin helicopter with 12 souls on board in the South China Sea suffered a rotor burst. Half the turbine disk lodged in the turbine containment wall, the rest went out the pipe. A previous rotor burst on this type, prior to a kevlar containment wall upgrade had resulted in loss of the helicopter.

I learned from that episode that turbine rotor bursts will result from:

1. a failure of the coupling (shaft) connecting the turbine to the compressor/fan (as happened with the helicopter) or intermediate gearing
2. a failure of the splines that lock the rotor disk in place on the shaft
3. consequential damage caused by bearing failure, oil starvation to bearings etc (Trent 700/A330 failure was case in point) where the heat build up or oil fire raises the temperature of rotor material to the point where it can fail
4. Fatigue failure resulting from material damage during assembly of the rotor. This has to be considered an extremely rare event, given quality procedures in every engine shop and manufacturer
5. Fatigue failure due manufacturing defects. Again, on an engine of this type, would have to be considered "impossible" due the strict processes used.
6. Improper parts used during overhaul (as was the case with this helicopter)

Above is not an exhaustive list, but one could surmise the probability that in this case, the problems identified in the AD with the IPT were likely contributory. i.e. some accelerated wear exceedance faster than anticipated by the AD on the faces of the splines of the IP shaft rigid coupling.

The moment the turbine shaft decouples from the compressor load it is driving, the excess energy now available to the turbine will accelerate it so incredibly quickly that centripetal forces will exceed design strength and the rotor disc will fracture and exit the engine. On an engine that size, nothing is going to stop those rotor bits as they contain incredible amounts of kinetic energy. In the case of the turboshaft engine I saw, it was estimated to have burst at around 100,000 rpm. In the case of the Trent it may well have been lower due to size and mass. The video referred to earlier in this thread refers to front fan blades subjected to 7,000 G forces in normal operation.

This may also account for the paxs hearing two separate bangs. The first may have been the splines on the shaft shearing, but still creating enough friction to slow the turbine rotor's acceleration for a finite period until it accelerated to an rpm where it burst.

It is also incorrect to say that un-contained rotor bursts are not designed for. They are. See FAA Advisory Circulars AC25-20 and AC 20-128

The Gulfstream G450 rotor burst plane overlaps the pressurized cabin and does not meet certification criteria for newer generation aircraft above 41,000 ft. Certification to 45,000 ft was achieved through a "grandfather clause." The issue is rapid depressurisation causing insufficient crew LOC time to don rapid oxy masks. The Sino-Swearingen SJ-30 was redesigned for the SJ30-2 to reposition a fuel tank forward to move it out of the rotor burst zone.

In the case of the A380, you can bet Airbus took careful note of what is in the rotor burst zones. See this document, page 15 for considerations:
http://www.bristol.ac.uk/aerospace/m...abindesign.pdf

No question the A380 and the QF crew were vindicated in performing well in this instance. There appears to still have been an element of luck in the direction the rotor took on departure. It could have punctured the fuselage with very bad outcomes for paxs, but it didn't in this case. That would have been an low probability direction compounding an already highly unlikely event.

My point?

RR appears to be suffering a higher rate of rotor bursts than the other two manufacturers, even on mature engines like RB211, and Trent 700, with this 900 failure and the test bed 1000 failure.

Thank fully this rotor burst occurred on climb out on a mostly overland route. Wouldn't fancy a rotor burst mid-pacific with loss of hydraulics and punctured fuel tank.

Just about every other failure of an engine ... loss of blades, ingestion etc ... is "benign" compared to a rotor burst, unless fire occurs. Thankfully, in this case also, no fire resulted. Probably resulting/aided from good thoughtful design on the part of both Airbus and RR.

Me? I would strap on an A380 any day, GP7200 powered of course. With a RR900 engine, I would prefer they first have a definitive answer to why this occurred before flying it again. Whatever it was, it was outside what was anticipated or foreseen. Not a happy situation for A380/RR operators to be in.
Bizman is offline