PPRuNe Forums - View Single Post - Qantas A380 Update
View Single Post
Old 10th Jan 2011, 07:19
  #36 (permalink)  
rob_ginger
 
Join Date: Mar 2008
Location: Sydney Australia
Age: 73
Posts: 3
Likes: 0
Received 0 Likes on 0 Posts
Test Pilots

In post 26 mmciau said:

More important outcome of the saga will be whether the A380 pilots have full and unreserved confidence in the engines when they are lifting off from Los Angeles Airport with an approx 14 hour flight ahead of them.
while in post 29 Bootstrap1 said:

The EECs have had a software upgrade that monitors the areas in question more closely.
I looked at the AD for the modifications to the EEC software
(EASA Airworthiness Directives Publishing Tool) and it gives me the willies. The relevant part of the AD is:

Rolls-Royce has developed a modification of the Engine Electronic Controller (EEC) software, featuring an IPT Overspeed Protection System (IPTOS). The purpose of the IPTOS functionality is to detect engine conditions that may
potentially lead to an IP turbine overspeed, and shut down the engine before the level of overspeed reaches the disc burst speed.
I know that the A380 has four engines, so it's impossible to lose all four at once. But then that's what the pilot of a Tristar thought, when the same oil seal was left out of all 3 engines. If you want to quibble - yes, one engine was still running on touchdown, but I bet his freckle was puckering a little.

And then of course, this is comprehensively tested software, not unreliable hardware. But, what about the FADEC software on some 777's, that rolled back from the commanded flex thrust ?

As an engineer who's worked on high-availability control system software (SCADA) for 20 years or so, I can tell you that nearly all the easy bugs are found during system testing, while the really curly ones pop up once a complicated system is in service. So if I was the first "test pilot" on a fully-loaded A380 taking off on a hot day on a shortish runway I think my freckle would be puckering just a little. A bit fat alarm is one thing, but shutting down engines seems very drastic to me. What if the protection software falsely triggers due to a combination of factors, say a power bus glitch plus turbulence, or a lightning strike, etc etc. I hope it never happens !
rob_ginger is offline