747-400 RB211 Autostart Windmilling Start
Thread Starter
Join Date: Feb 2004
Location: Australia
Posts: 1,307
Likes: 0
Received 0 Likes
on
0 Posts
747-400 RB211 Autostart Windmilling Start
Folks...
Disregarding QRH procedures for windmilling autostarts .... Is there any airplane logic requirement for the Autostart system to see a transition of the Fuel Control/cutoff Switch (FCS) from OFF to RUN on this aircraft before Autostart begins?
My (engineering) training notes are ambiguous. In one section it just has the logic "FCS in run", but elsewhere, the text says "moving to RUN" (as if a transition is required).
Auto Ignition and Auto Relight are less ambiguous. The RB211 Auto Relight System senses abnormal engine deceleration and, until N2 rpms are below 35%, the Auto Relight logic will override the normal Auto Ignition logic (Auto Ignition logic is N3 <58%, HP fuel valve open).
I guess the only difference you would notice is if the auto ignition and autorelight failed to stop the engine flaming out and the Autostart had problems relighting the engine. The Autostart controls fuel flow: Unsuccessful autostarts see the fuel being cut off prior to successive autostart attempts... although the duration might be short, depending on engine temperature.
Experts?
Disregarding QRH procedures for windmilling autostarts .... Is there any airplane logic requirement for the Autostart system to see a transition of the Fuel Control/cutoff Switch (FCS) from OFF to RUN on this aircraft before Autostart begins?
My (engineering) training notes are ambiguous. In one section it just has the logic "FCS in run", but elsewhere, the text says "moving to RUN" (as if a transition is required).
Auto Ignition and Auto Relight are less ambiguous. The RB211 Auto Relight System senses abnormal engine deceleration and, until N2 rpms are below 35%, the Auto Relight logic will override the normal Auto Ignition logic (Auto Ignition logic is N3 <58%, HP fuel valve open).
I guess the only difference you would notice is if the auto ignition and autorelight failed to stop the engine flaming out and the Autostart had problems relighting the engine. The Autostart controls fuel flow: Unsuccessful autostarts see the fuel being cut off prior to successive autostart attempts... although the duration might be short, depending on engine temperature.
Experts?
Sorry I can't always keep up with the conditions vs control logic in these kind of questions so I'll just keep to some general comments.
If you have a condition where the engine spools down below idle with increasing EGT then I suppose you can assume the auto-relight didn't help. So you have to command the fuel off in order to try a relight-restart. The idea of commanding the fuel off is to give the compressors a chance to once again gulp air without being choked by any burner pressure. Kind of like a food processor that has eaten too many tomatoes and is just swirling the soft stuff around at the bottom without being able to eat anymore of the chunks on top.
Of course if you have commanded the fuel off then you have to command it back on if you want to attempt a restart.
Simple flame-outs are just that (EGT is falling) and it's presumed that the compressors are still able to gulp air and what you now need is ignition with a compressor still spooling.
In all this, the key is what is EGT doing since it defines the failure condition, temporary or not.
Somewhere in all this the auto-logic is working but I'll leave that up to others to contribute
If you have a condition where the engine spools down below idle with increasing EGT then I suppose you can assume the auto-relight didn't help. So you have to command the fuel off in order to try a relight-restart. The idea of commanding the fuel off is to give the compressors a chance to once again gulp air without being choked by any burner pressure. Kind of like a food processor that has eaten too many tomatoes and is just swirling the soft stuff around at the bottom without being able to eat anymore of the chunks on top.
Of course if you have commanded the fuel off then you have to command it back on if you want to attempt a restart.
Simple flame-outs are just that (EGT is falling) and it's presumed that the compressors are still able to gulp air and what you now need is ignition with a compressor still spooling.
In all this, the key is what is EGT doing since it defines the failure condition, temporary or not.
Somewhere in all this the auto-logic is working but I'll leave that up to others to contribute
Thread Starter
Join Date: Feb 2004
Location: Australia
Posts: 1,307
Likes: 0
Received 0 Likes
on
0 Posts
Thanks, Lomapaseo... This make take me a while to digest (especially the tomatoes) :P
I guess it still leaves the possibility that Autostart could take over in this situation prior to fuel cutoff switch cycling (since the EEC is monitoring a host of engine sensors and will be looking for the right conditions to put on fuel and ignition).
I previously assumed cycling was required.... Still unsure...
I guess it still leaves the possibility that Autostart could take over in this situation prior to fuel cutoff switch cycling (since the EEC is monitoring a host of engine sensors and will be looking for the right conditions to put on fuel and ignition).
I previously assumed cycling was required.... Still unsure...
Join Date: Jun 2000
Location: last time I looked I was still here.
Posts: 4,507
Likes: 0
Received 0 Likes
on
0 Posts
It's been many moons & cuba libres ago, but I vaguely remember something in smaller members of the RB211 family about selecting the fuel cutoff to idle to re-set the logic in the control of the variable stator guide vanes. This then helped with better airflow for the restart. It might have had something to do with them being set for a high RPM at failure time and you trying to start them at a lower RPM.
There will be more up to date boffins than I to tell more.
There will be more up to date boffins than I to tell more.
NSEU
"Autostart" is not the same as "Auto-relight" - different logic paths in the FADEC. If the engine flames out or runs down uncommanded, it goes into Auto-relight. In Auto-relight it will obviously turn on ignition and adjust fuel flow for the burner conditions, and if there is evidence of a surge/stall take appropriate action to attempt to clear the surge/stall as well. If the FADEC sees the fuel switch transition it will enter Autostart which will then schedule fuel, ignition, etc. for an in-flight start (potentially with starter assist if the start valve is enabled). While the details vary between engine types (the 747-400/RB211 was a very early autostart implementation - the newer systems are much more sophisticated and complex), there is a major difference between auto-relight and autostart:
Autostart can turn off fuel and will respect EGT limits - auto-relight won't (although the inflight start EGT limit is higher than on-ground, autostart will still turn off fuel to prevent exceeding it).
BTW, the RB211/747-400 isn't really a FADEC - it's an FAFC - "Full Authority Fuel Control". It doesn't control the vanes and bleeds like a true FADEC - the vanes and bleeds are controlled by a fluidic (pneumatic) system that is little changed from what was on the early RB211s.
"Autostart" is not the same as "Auto-relight" - different logic paths in the FADEC. If the engine flames out or runs down uncommanded, it goes into Auto-relight. In Auto-relight it will obviously turn on ignition and adjust fuel flow for the burner conditions, and if there is evidence of a surge/stall take appropriate action to attempt to clear the surge/stall as well. If the FADEC sees the fuel switch transition it will enter Autostart which will then schedule fuel, ignition, etc. for an in-flight start (potentially with starter assist if the start valve is enabled). While the details vary between engine types (the 747-400/RB211 was a very early autostart implementation - the newer systems are much more sophisticated and complex), there is a major difference between auto-relight and autostart:
Autostart can turn off fuel and will respect EGT limits - auto-relight won't (although the inflight start EGT limit is higher than on-ground, autostart will still turn off fuel to prevent exceeding it).
BTW, the RB211/747-400 isn't really a FADEC - it's an FAFC - "Full Authority Fuel Control". It doesn't control the vanes and bleeds like a true FADEC - the vanes and bleeds are controlled by a fluidic (pneumatic) system that is little changed from what was on the early RB211s.
Thread Starter
Join Date: Feb 2004
Location: Australia
Posts: 1,307
Likes: 0
Received 0 Likes
on
0 Posts
Much appreciated, gentlemen.
It looks like the fuel switch is required to transition to Autostart... at least for the RB211.
P.S. Anyone with inside knowledge on the CF6 (with full FADEC)?
Fuel lever required also, for Autostart?
It looks like the fuel switch is required to transition to Autostart... at least for the RB211.
P.S. Anyone with inside knowledge on the CF6 (with full FADEC)?
Fuel lever required also, for Autostart?
P.S. Anyone with inside knowledge on the CF6 (with full FADEC)? 
Fuel lever required also, for Autostart?
Fuel lever required also, for Autostart?
Thread Starter
Join Date: Feb 2004
Location: Australia
Posts: 1,307
Likes: 0
Received 0 Likes
on
0 Posts
The 'difference' between Autostart and Auto-relight I posted above is common design practice for Boeing Autostart installations.
Up until now, I've clearly divided (in my mind) that Auto-Relight (Ignition activation based on N3-deceleration), AutoIgition (ignition based on rpms below a certain value) and Autostart (ignition and fuel application based on complex flow charts in my manuals). I just wasn't sure that Autostart would automatically activate as a result of these flow charts without pilot action.
In my CF6 manuals, I can find no reference to a system similar to the RB211's Auto-Relight system (i.e. rotor deceleration rate-based ignition activation) . This, I assumed, was the big difference in operation between the CF6's AutoIgntion System and the RB211's AutoIgnition System. The RB211 has both types of flameout detection.
With the ignition switch in "NORM", the RB211 relies totally on rpm decel and minimum rpms to trigger ignition. The CF6, as far as I knew (in SINGLE and BOTH) reacted only to minimum rpms and configuration (flaps, etc) to activate ignition automatically.
Are we simply labelling things differently, or am I misunderstanding the operation of the ignition system?
Thanks!
The auto-relight logic on the RB211 is somewhat more robust than the CF6. The Rolls igniter design wears out rather quickly when continuous ignition is ON, so Rolls wanted to certify auto-relight to be equivalent to continuous ignition and get rid of the con ignition (it's been a while, but IIRC they were successful and on at least some Rolls powered 747-400s, they got rid of continuous ignition). There was no attempt on the CF6 to get credit for auto-relight, so the logic is pretty basic - N2 less than the lowest idle N2 for the flight conditions. So on the CF6, con-ignition is commanded whenever flaps are not stowed, anti-ice is ON, or the flight deck switch is selected to Con.
On newer engine installations, the logic is more sophisticated (e.g. rate of decel of N2/N3) and we've certified auto-relight to be equivalent to con ignition for all engine types and gotten rid of the con ignition function entirely.
Does that help?
On newer engine installations, the logic is more sophisticated (e.g. rate of decel of N2/N3) and we've certified auto-relight to be equivalent to con ignition for all engine types and gotten rid of the con ignition function entirely.
Does that help?
Thread Starter
Join Date: Feb 2004
Location: Australia
Posts: 1,307
Likes: 0
Received 0 Likes
on
0 Posts
On newer engine installations, the logic is more sophisticated (e.g. rate of decel of N2/N3) and we've certified auto-relight to be equivalent to con ignition for all engine types and gotten rid of the con ignition function entirely.
Does that help?
Does that help?
By newer engine installations, would this include retrofitting 747-400's .. or is this for next gen aircraft... 747-8, 787, etc.
Much appreciated.
Cheers
Thread Starter
Join Date: Feb 2004
Location: Australia
Posts: 1,307
Likes: 0
Received 0 Likes
on
0 Posts
Much appreciated... Thanks.
I'm also hoping someone has some knowledge about the Pratt & Whitney Autostart system on the 747-400. Perhaps the pilot's FCOMs (or PPRuNe resident experts) can provide some answers.
The Boeing AMM says this about the 747-400 PW4000 Autostart system...
On the ground...
"Starting Anomaly: No Light(up) | Symptom Seen By Autostart Controller: EGT increase less than 28*C after 20 seconds | Autostart Controller Response: Motors the engine for 30 seconds, then starts fuel supply and ignition (using both ignitors). If N2 > 50% (starter cutout speed), the controller stops the start procedure."
It's not clear what happens if the second attempt fails (below 50%N2).
Underneath, it says .. If Starter Duty Cycle Time Exceeds 5 minutes, the starter valve is closed and the controller stops the start procedure.
Does this mean it keeps starting and stopping the engine for 5 minutes?
In flight, the manuals do say the Autostart system will continue indefinitely (no 5 minute limit).
I understand that the AUTOSTART Caution and the AUTOSTART, ignitor 1 and 2 Status messages appear after the second attempt. Is it up to the pilots to stop the autostart after seeing the messages, or will the Supplemental Control Unit abort the start on the second failed start?
Also, I understand it's not standard practice, but it's also not clear what happens if BOTH ignitors are selected for the first start. Does the system not bother a restart (because it already knows dual ignitors won't work).
I'm also hoping someone has some knowledge about the Pratt & Whitney Autostart system on the 747-400. Perhaps the pilot's FCOMs (or PPRuNe resident experts) can provide some answers.
The Boeing AMM says this about the 747-400 PW4000 Autostart system...
On the ground...
"Starting Anomaly: No Light(up) | Symptom Seen By Autostart Controller: EGT increase less than 28*C after 20 seconds | Autostart Controller Response: Motors the engine for 30 seconds, then starts fuel supply and ignition (using both ignitors). If N2 > 50% (starter cutout speed), the controller stops the start procedure."
It's not clear what happens if the second attempt fails (below 50%N2).
Underneath, it says .. If Starter Duty Cycle Time Exceeds 5 minutes, the starter valve is closed and the controller stops the start procedure.
Does this mean it keeps starting and stopping the engine for 5 minutes?
In flight, the manuals do say the Autostart system will continue indefinitely (no 5 minute limit).
I understand that the AUTOSTART Caution and the AUTOSTART, ignitor 1 and 2 Status messages appear after the second attempt. Is it up to the pilots to stop the autostart after seeing the messages, or will the Supplemental Control Unit abort the start on the second failed start?
Also, I understand it's not standard practice, but it's also not clear what happens if BOTH ignitors are selected for the first start. Does the system not bother a restart (because it already knows dual ignitors won't work).
Join Date: Mar 2013
Location: Siliconia
Age: 63
Posts: 44
Likes: 0
Received 0 Likes
on
0 Posts
NSEU
Not current on these types, but Autostart has varied little ...
The starter motor is run continuously up to its duty limit timer (5 minutes).
The initially selected ignition plug is operated until fault condition is detected, turned off for thirty seconds, then selected (along with the alternate plug) until ...
The HP fuel shut-off-valve is opened until fault condition is detected, then closed for thirty seconds, then opened until ...
The control system completes its autostart procedure with the starter motor not running, ignition plugs not operating and HP fuel shut-off valve closed.
The flight crew must deselect the switches that initiated the autostart procedure. The reselection of auotostart procedure by the control switches will then be available. It may be desireable, before trying further engine operation, to allow the starter motor to cool and/or effect some repairs!
I don't know for the PW engine, but I think it's desireable to use one logic (i.e. having two start attempts), because it gives the engine two chances including with the much extended motoring, and it means most (starter and fuel) system control logic only needs one design and test; the ignition plug control and fault reporting may vary by design.
Not current on these types, but Autostart has varied little ...
Does this mean it keeps starting and stopping the engine for 5 minutes?
The initially selected ignition plug is operated until fault condition is detected, turned off for thirty seconds, then selected (along with the alternate plug) until ...
The HP fuel shut-off-valve is opened until fault condition is detected, then closed for thirty seconds, then opened until ...
Is it up to the pilots to stop the autostart after seeing the messages, or will the Supplemental Control Unit abort the start on the second failed start?
The flight crew must deselect the switches that initiated the autostart procedure. The reselection of auotostart procedure by the control switches will then be available. It may be desireable, before trying further engine operation, to allow the starter motor to cool and/or effect some repairs!
Also, I understand it's not standard practice, but it's also not clear what happens if BOTH ignitors are selected for the first start. Does the system not bother a restart (because it already knows dual ignitors won't work).
Thread Starter
Join Date: Feb 2004
Location: Australia
Posts: 1,307
Likes: 0
Received 0 Likes
on
0 Posts
but Autostart has varied little ...
GE: after the 3rd start attempt the engine is motored for two minutes and the starter is shut down. It has a 20 second lightup detector for the first start attempt, then 15 and 15 seconds for the 2nd/3rd attempts. In between start attempts there is a 30 second cooldown timer.
RB: Has two start attempts. After each failed start attempt (due to no lightup) the starter is kept running until the cooldown temperature is achieved (which may be no time at all if there was no lightup). It has a 20 second lightup detection timer on the first and second start attempts.
PW: runs for a total of 5 minutes
Assuming the PW is simply starter runtime based, the GE, in theory, comes close to the PW's 5 minutes, but on the RB (assuming no temperature rise) the start could be aborted 2 x 20 seconds after initial fuel on.
Cheers
NSEU
I did work the PW4000 SCU autostart, but it's been over 25 years and the memory is a bit fuzzy of the specifics. And since I recently retired, I no longer have the option of checking my documentation to make sure (they made me leave all that stuff behind
)
But what I remember is that for the PW4000 if there was no light after the dual ignition, it would cut fuel, motor for 30 seconds, cut the starter and put up the AUTOSTART EICAS message. You are correct that in-flight there is no limit, it'll keep trying indefinitely.
There is a dramatic lack of consistency in autostart between engine types on the 747-400. It was the first time we'd done autostart at Boeing and we sort of made it up as we went. Originally, Pratt didn't believe there was any reason or advantage to having autostart and didn't want anything to do with it, while GE made it basic on the CF6-80C2 FADEC. When Pratt realized that the lack of autostart put them at a serious disadvantage, they had to develop the SCU (Rolls ended up somewhere between Pratt and GE - it didn't take an extra box like the Pratt SCU, but it was still a bit of a last minute add-on). As a result there wasn't as much consistency between engine types as there is on later installations such as the 777.
I did work the PW4000 SCU autostart, but it's been over 25 years and the memory is a bit fuzzy of the specifics. And since I recently retired, I no longer have the option of checking my documentation to make sure (they made me leave all that stuff behind
)But what I remember is that for the PW4000 if there was no light after the dual ignition, it would cut fuel, motor for 30 seconds, cut the starter and put up the AUTOSTART EICAS message. You are correct that in-flight there is no limit, it'll keep trying indefinitely.
There is a dramatic lack of consistency in autostart between engine types on the 747-400. It was the first time we'd done autostart at Boeing and we sort of made it up as we went. Originally, Pratt didn't believe there was any reason or advantage to having autostart and didn't want anything to do with it, while GE made it basic on the CF6-80C2 FADEC. When Pratt realized that the lack of autostart put them at a serious disadvantage, they had to develop the SCU (Rolls ended up somewhere between Pratt and GE - it didn't take an extra box like the Pratt SCU, but it was still a bit of a last minute add-on). As a result there wasn't as much consistency between engine types as there is on later installations such as the 777.
