A320 Engine relight in flight procedure

[Maverick2167]

Hi, I know this topic has been discussed in the forum, but I'm unable to find a definitive answer or reply that I can understand. Here is my query

1. What is the difference between all the three procedures i.e. (1) Starter Assisted (2) Windmill Quick Relight (3) Stabilized Windmill Relight when the procedure is same for all 03 as per QRH?
2. What is the significance of the T<20s and N2 >15% for Quick relight and N2>18% for Stabilized (this figures are for PW engines)
Thank you

[Field In Sight]

It's a good question.

The limiting factor appears to be the minimum N2 speed for the start.
2 of the cases seem pretty straight forward.

1. Starter assist to reach the correct N2, i.e. we are only doing what we do on the ground but in the air.
2. Stabilised windmill. The airflow is fast enough to give us a minimum N2 speed.

The 3rd case of windmill quick relight is a bit more vague. From the manuals there is the following information:

From QRH:For an in-flight restart, the FADEC decides whether the rotor speed is sufficient to perform a windmill relight or needs assistance from the starter in view of the current engine parameters and the flight environment parameters. Within the windmill quick relight envelope, if the crew is using the Quick Windmill Relight procedure, the FADEC will restart the engine after the flight crew cycles the ENG MASTER lever from "ON" to "OFF" then back to "ON".

Fron FCOM
WINDMILLING QUICK RELIGHT In case of inadvertent engine shutdown by cycling the Engine Master lever to OFF then ON, in the windmilling quick relight envelope, the FADEC will attempt automatically a relight regardless of the rotary selector position.

My view of it, like a lot of Airbus documentation is that it probably is of no value and is only there to confuse us.
I'm guessing the Windmill Quick Relight is a safety feature and it publishing it on the chart just makes it harder to read.

Happy to be corrected by any engine specialists though.

[tdracer]

WINDMILLING QUICK RELIGHT In case of inadvertent engine shutdown by cycling the Engine Master lever to OFF then ON, in the windmilling quick relight envelope, the FADEC will attempt automatically a relight regardless of the rotary selector position.

My view of it, like a lot of Airbus documentation is that it probably is of no value and is only there to confuse us.
I'm guessing the Windmill Quick Relight is a safety feature and it publishing it on the chart just makes it harder to read.

Happy to be corrected by any engine specialists though.
Quick Windmill Relight was a regulatory reaction to come incidents where a pilot inadvertently shutdown one (or both) engines when they intended to do something else, and quickly returns the fuel switch to 'RUN'. This first became an issue after introduction of the 767 - the supervisory EEC switches were on the aislestand near the fuel switches. There were two cases where - during Takeoff - one of the EECs failed and the pilot reached down to turn them off but instead pushing the pushbutton EEC switches, the pilot moved both fuel switches to CUTOFF:eek:. Quickly realizing their error, the fuel switches were quickly returned to 'RUN' - in the first case (JT9D engines), one engine recovered, the second went into a start stall - they did an immediate return to land. The general consensus was that the pilot was an idiot (push button switches vs. fuel condition levers) and it wouldn't happen again. Then it did :ugh:. This time CF6-80A engines on takeoff out of LAX - this time both engines recovered (but not before they got within a few hundred feet of the Pacific Ocean) and the flight continued on to Chicago (:confused:) - when the passengers disembarked, some were still wearing their lifejackets... Anyway, after the second incident it was obvious we had a serious human factors problem, and the EEC switches were moved from the aislestand to the pilot overhead and to the best of my knowledge it hasn't happened since on any aircraft type (except for a certain Egypt Air crash, where unless you're an Egyptian investigator, it's believed it was an intentional act).
However the feds jumped on this and insisted we needed to demonstrate "Quick Windmill Relight" capabilities on all new engine types. QWR is a very challenging requirement - especially if the engine was at power prior to the inadvertent shutdown. If the engine was at power prior to shutdown, the compressor is hot and heats the incoming air during the relight attempt - which has the effect of 'overfueling' the engine (compared to a normal windmill start) leading to a start-stall. FADEC has the huge advantage of being able to recognized it's a QWR and adjust the starting fuel flow, but it's still a very challenging design requirement (it's also very, very hard on the engine since shutting it down at power usually results in a major compressor rub due to the rapid, differential cooling of the rotor and case).

[Field In Sight]

Good information.

Every day's a learning day.

[Check Airman]

Quick Windmill Relight was a regulatory reaction to come incidents where a pilot inadvertently shutdown one (or both) engines when they intended to do something else, and quickly returns the fuel switch to 'RUN'. This first became an issue after introduction of the 767 - the supervisory EEC switches were on the aislestand near the fuel switches. There were two cases where - during Takeoff - one of the EECs failed and the pilot reached down to turn them off but instead pushing the pushbutton EEC switches, the pilot moved both fuel switches to CUTOFF:eek:. Quickly realizing their error, the fuel switches were quickly returned to 'RUN' - in the first case (JT9D engines), one engine recovered, the second went into a start stall - they did an immediate return to land. The general consensus was that the pilot was an idiot (push button switches vs. fuel condition levers) and it wouldn't happen again. Then it did :ugh:. This time CF6-80A engines on takeoff out of LAX - this time both engines recovered (but not before they got within a few hundred feet of the Pacific Ocean) and the flight continued on to Chicago (:confused:) - when the passengers disembarked, some were still wearing their lifejackets... Anyway, after the second incident it was obvious we had a serious human factors problem, and the EEC switches were moved from the aislestand to the pilot overhead and to the best of my knowledge it hasn't happened since on any aircraft type (except for a certain Egypt Air crash, where unless you're an Egyptian investigator, it's believed it was an intentional act).
However the feds jumped on this and insisted we needed to demonstrate "Quick Windmill Relight" capabilities on all new engine types. QWR is a very challenging requirement - especially if the engine was at power prior to the inadvertent shutdown. If the engine was at power prior to shutdown, the compressor is hot and heats the incoming air during the relight attempt - which has the effect of 'overfueling' the engine (compared to a normal windmill start) leading to a start-stall. FADEC has the huge advantage of being able to recognized it's a QWR and adjust the starting fuel flow, but it's still a very challenging design requirement (it's also very, very hard on the engine since shutting it down at power usually results in a major compressor rub due to the rapid, differential cooling of the rotor and case).

I learned something new today. Thanks

[megan]

td, I still reckon you could fill a book with these gems. :ok:

[Consol]

Great post tdracer.

[vilas]

tdracer
A different question. Dual engine fail say in cruise 350 the A320 procedure apart from other things asks relight attempt once in relight envelope. CFM CEO engines have relight envelope starting at FL250. Using common sense if it's tried outside envelope due to air density the chances of hot start are there and within the envelope it's tested to restart. Any specific technical issues not attempting restart outside envelope?

[Fursty Ferret]

FADEC has the huge advantage of being able to recognized it's a QWR and adjust the starting fuel flow, but it's still a very challenging design requirement (it's also very, very hard on the engine since shutting it down at power usually results in a major compressor rub due to the rapid, differential cooling of the rotor and case).

Rolls Royce list this method as a last-ditch attempt to clear a locked-in surge condition on their larger engines.

[tdracer]

tdracer
A different question. Dual engine fail say in cruise 350 the A320 procedure apart from other things asks relight attempt once in relight envelope. CFM CEO engines have relight envelope starting at FL250. Using common sense if it's tried outside envelope due to air density the chances of hot start are there and within the envelope it's tested to restart. Any specific technical issues not attempting restart outside envelope?
If sort of depends on who you ask. You have to remember how the in-flight start envelope is defined - it means that a healthy engine should always start - 90% isn't good enough. Outside the envelope doesn't mean it won't start, just that it might not start.
There have always been a lot of discussions about in-flight starting, especially with regard to what EGT start limit you should use. The theory is that you can use a higher EGT start limit in-flight for two reasons - first off it's likely an emergency and engine life is secondary, but more importantly there is more cooling air moving through the engine during a windmill relight attempt relative to a normal on-ground start, so the turbine can withstand a higher EGT without experiencing damage. The problem is quantifying what the EGT limit should be. On the 747-400, we simply used the TO EGT limit for in-flight starting (same for the same engines that went on the 767) - starting with the 777 the engine companies put some more thought and analysis into it and came up with an in-flight EGT start limit that was higher than the ground start but less than TO. But I still consider the in-flight start EGT limit to be little better than an educated guess...
All that being said - in my opinion there is little reason to not to attempt an in-flight start even if you're not in the envelope so long as you keep a close eye on it and abort if it starts to go hot (use the ground EGT start limit - so don't depend on autostart). If it hangs or otherwise doesn't work - abort, wait 30 seconds and try again, but if it starts and responds - you're good to go. As far as I'm concerned, the only downside would be crew workload during a potential emergency (having one pilot monitoring the engine start).
During the 747-8 certification, I tagged along on an unrelated flight test in the hope that we could pick up a test condition that I needed (which ultimately didn't happen). IIRC it was an FMC/Auto-Throttle test, and one condition was to do an extended engine-out A/T cruise - we cruised around for 20 or 30 minutes at 37k with one engine shutdown (we were pretty light). When they were done with the condition, one of the pilots turned to me and asked if he could try a re-start -37k being well outside the restart envelope plus a cold-soaked engine (the GEnx engine was a notoriously difficult engine to start - and in-flight starts had been a major problem with multiple revisions to the FADEC in-flight s/w and the procedures required). I responded 'sure - go ahead and try, just keep an eye on it in case it goes hot'. The engine started right up - no problem.:ok: Now, the division of responsibilities meant that, while I was responsible for the FADEC s/w, I wasn't responsible for starting (although I was obviously involved) so the next day I told my good buddy who was responsible for starting about it - and he was horrified :eek:!
"WHAT IF IT HADN'T STARTED -THE FAA WAS ONBOARD!"
I responded 'So what?' - we were outside the envelope - they couldn't hold it against us if it failed to start outside the envelope.
He responded 'oh yes they could - don't do it again!'
:rolleyes:

[FlightDetent]

HF to consider: Who's gonna recall the correct EGT temp limit for an in-flight restart in the heat of the moment?

Why is it (ehm) that FADEC is not allowed to abort the inflight restart .... so why would you.

[tdracer]

HF to consider: Who's gonna recall the correct EGT temp limit for an in-flight restart in the heat of the moment?

Why is it (ehm) that FADEC is not allowed to abort the inflight restart .... so why would you.

It's certainly not clear-cut - especially in an all-engine out situation.
At least on Boeing, EICAS will display the in-flight EGT start limit (and the appropriate start envelope information - including cross-bleed if applicable)(maybe not on a 737 - not sure about that part). The real question is - is it an emergency? If it's an emergency, then by all means use the in-flight start limit and let autostart do its thing. But if you are so high that you're above the in-flight start envelope, I'd argue it's not (yet) an emergency (even all engine out - if you're above 30k - then it's not an emergency although in the heat of the moment you may not be thinking that...).
In a non-emergency situation, using the full in-flight start EGT limit may result in economic turbine damage, and if you're outside the restart envelope - in a non-emergency - and you damage the turbine trying to start it, you may find yourself having a rather unpleasant conversation with management. I'd like to think that most pilots would know what the ground start EGT limit is for their particular engine/aircraft (although automation - in this case autostart - might interfere with that). In a non-emergency situation, respecting the ground start EGT limit during an in-flight start should preclude doing damage.
The other side of that is that pilots have been known to abort what should have been successful starts - even during an all engine out emergency. Back around 1990, a KLM 747-400 flew through a volcanic ash cloud near Alaska and lost all four engines. They got down to ~10k before they got the engines restarted. Looking at the FDR data after the event, it was determined that the crew had repeatedly aborted what appeared to be successful starts during the emergency (the CF6-80C2 engines on the event aircraft had a very good autostart system - unlike the PW4000 and RB211 autostart systems on the 747-400).
Like I said, it's not always clear cut what to do - particularly in an all engine out situation. But if you trash the turbines trying to restart the engines no matter what, you may find yourself praying you're as good (and lucky) as Carlos Dardano when he deadsticked TACA 110 onto a dike outside New Orleans...

[vilas]

Thanks tdracer
I didn't want to include operational situations where no holds are barred. But good to know it can be done if required. About QWR I didn't find many pilots knowledgeable about it.

[Maverick2167]

Hello. Thank you all for the response. Just want to share my understanding of the procedure as described to me by a A320 TRI and do let me know if it is correct

1. Starter assisted relight- If our flight parameters fall under these region during inflight relight scenario, (Light aircraft,descending at Green dot speed due to obstacles , basically obstacle strategy) the start sequence is limited by the starter engagement limitations i.e. if it does not start within 35, starter cooling required . But the limitations on FCOM is with regards to ground start, couldn't understand this part :-)

2. Windmill Quick relight- Governed by the N2 speed and T<20s (PW engines) ,if the engines are not started by within the time T<20s, the master needs to be turned off.?

3. Windmill Relight- Governed by N2 speed and 30s as per the QRH procedure

[mustafagander]

As tdracer says Inflight starts which are proceeding normally after light up are often aborted coz the acceleration of the core is very slow at altitude. My old mob used to ensure that it was demonstrated in the sim during the 3 year cyclic training/checking programme.a few times to remind us.

[tdracer]

As tdracer says Inflight starts which are proceeding normally after light up are often aborted coz the acceleration of the core is very slow at altitude. My old mob used to ensure that it was demonstrated in the sim during the 3 year cyclic training/checking programme.a few times to remind us.
The slow start characteristics is particularly true of the newer generation of engines - big(ger) fan, small(er) core. That trend has caused the in-flight start characteristics to really suffer - particularly windmill relight.
When we did the 747-400 and 767 re-engine, the CF6-80C2 would start pretty much anytime you tried it (the only real limitation being the core windmill speed being high enough to provide sufficient fuel pressure - ~8-10% N2), and the PW4000/94" was nearly as good.
Later programs proved much more difficult to get acceptable in-flight start characteristics. On the 747-8 (GEnx-2B), we actually had to resort to depressurizing the hydraulic pump during in-flight starts in order to get sufficient windmilling N2 to get an acceptable windmill start envelope.

[mustafagander]

I forgot to mention that I was talking about the B744.