APU air inlet door

[Hardy Heinlin]

Good morning,

a question for the engineers maybe ...

I learned that the EICAS messages "APU DOOR" incorporate a time delay of 60 seconds. But I don't know if the delay refers to the time until the message appears, or the time until the message disappears.

When an automatic APU shutdown is initiated because there is a door-is-closed signal while commanded open, the N1/N2 RPM will drop below 20% in just 25 seconds. When below 20% the door is commanded to close. This means the message can never appear when the door fails in the closed position and if the delay refers to the time until the message appears.

That's why I think the delay refers to the time until the message disappears, i.e. when a disagreement occurs the message appears promptly, and when N1/N2 drops below 20% (in 25 seconds) during auto shutdown, which triggers the door close command (command agrees with position), the message will stay for another 35 seconds (60-25).

Also, I can't imagine when the START selection commands the door to open, and the door will not open, that the message will only appear in 60 seconds.

And as far as I see, neither the advisory nor the status message is latched.


Regards,

|-|ardy

[John H Watson]

When an automatic APU shutdown is initiated because there is a door-is-closed signal while commanded open, the N1/N2 RPM will drop below 20% in just 25 seconds.

So, in your scenario, the APU is running (with the door open) and, due to a fault, the door is seen as closed?

This is a tricky one. You may not even get an APU DOOR message for this, irrespective of timing. There are quite a few position indication switches which control the position of relays and signals to the APUC. I think the APUC only uses a "13 degree open" position switch in its logic (for start and autoshutdown). The other position switches seem to influence the door motor relays and the APU DOOR EICAS message. You may just get an APU EICAS message generated by the APUC, not the APU DOOR message generated by the other position switches.

And as far as I see, neither the advisory nor the status message is latched.

FWIW, the British Airways Boeing Fault Isolation Manual says the APU DOOR Status message can be latched. Other FIMs I have don't seem to list whether (any) EICAS message is latched.

Background: The APU DOOR message is a "disagree" message based on the position of auxilliary contacts on two different relays, R460 "RLY-APU INLET DR CLOSE" and R461 "RLY-APU INLET DR CLOSE" (Ref: System Schematic Manual (SSM) 49-15-01). The time delay is mentioned on the SSM. Basically, the EIU should receive an earth signal from either of these 2 relays. If no earth signal for 1 minute, the message is generated.
(EDIT: See later messages)

The CMC will probably tell the engineers what the cause of the autoshutdown is (as the APUC will be reporting to the CMC).

Cheers

[Hardy Heinlin]

If no earth signal for 1 minute, the message is generated.

Aha, so it's the other way around. This disagrees with my hope; I want to see that message :-)

The Thales 744 sim provides this malfunction:

CAUSE
Mechanical failure of APU door.
EFFECT
The APU door will remain in the open or close position only. EICAS advisory and
status messages will be displayed.
QRH Reference: APU DOOR
MESSAGES
APU DOOR (advisory and status)


In PSX 10.30 there are two new APU malfunctions: "Oil/N2" (Severe: oil leak; else: N2 overspeed) and "inlet door" (Severe: does not open; else: does not close).

There is no problem to get the APU DOOR message when it's open while it should be closed. I think there must also be a way to get the message when it's closed while it should be open. Maybe in reality the 60 sec delay is only applied to the status message, and that may be latched, and perhaps the advisory appears prompty and disappears promptly when the disagreement is gone?


Cheers,

|-|ardy

[John H Watson]

Standby... I see the APUC-controlled fault relay "R7176" can influence the position of the APU DOOR motor relays.... I'll try to figure out how the diagram works in your Thales scenario..

[Hardy Heinlin]

For now I set the delay to 60 sec for both advisory and status, and both are latched until the malfunction is reset on the Instructor page.

The other related messages work fine, and the restart inhibit logic works too ...

[John H Watson]

Basically, the EIU should receive an earth signal from either of these 2 relays. If no earth signal for 1 minute, the message is generated.

Sorry, I have to retract what I said earlier. I'm getting a little rusty at reading these diagrams. I assumed the APU door open relay R461 would be energised with the door open, but it actually relaxes when the actuator motor stops running  (i.e. when the door has moved to the open position). This means that the EIU gets no earth when the door is open (from either relay R461 and R460). This must mean that the EIU is looking for an earth for more than a minute to generate the fault (not an absence of earth)

In the Thales scenario, the actuator is broken (which may mean the door is stuck open). If the APUC gives a door close signal (due to an Autoshutdown fault or whatever), this affects the position of the APU Inlet Door Relay R7176 which in turn energises the close relay R460, generating an earth at the EIU. The relay will remain energised until the door reaches the closed position (but since it will never close due to the broken actuator, the earth will continue to be present at the EIU until the 1 minute timer triggers the APU DOOR message). The message timer starts when the door close command signal is given.

Similarly, if there is a command to open the door, the open relay energises and an earth is sent to the EIU and the timer starts. If the actuator is broken, the door will never reach the open position and the open relay will remain energised.

But I don't know if the delay refers to the time until the message appears, or the time until the message disappears.

The delay will be for appearance. Probably there will only be a second or two for the message to disappear unless latching logic is involved.

[John H Watson]

For now I set the delay to 60 sec for both advisory and status, and both are latched until the malfunction is reset on the Instructor page.

I think Advisory messages can only be latched by devices external to the EIUs.... In this case by the energising of the APU door open or close relays. If the actuator was broken and you tried to open the door, the messages would appear after 60 seconds (I assume both), but you could get rid of the Advisory message by turning APU controller to OFF. This would de-energise the open relay, so the EIU would not longer get the earth.

Standby... I see the APUC-controlled fault relay "R7176" can influence the position of the APU DOOR motor relays.... I'll try to figure out how the diagram works in your Thales scenario..

I just noticed an error in one of my old PSX beta diagrams (the notes I've added aren't correct). It's not the fault relay R7176 which affects the door motor relays, but the Fuel Control Relay R7200. This relay is affected by the APU switch and by the APUC door-remain-open during shutdown logic (until the rpms are low enough).

Further considerations:

If the APU is running on the ground and the aircraft takes off (with the APU running) and the door actuator fails to move the door to correct (more closed) position when the air/ground relays move to AIR, the APU door message probably should be generated 60 seconds after liftoff. Does the QRH have a procedure for this? I can't imagine it would be safe to increase speed with the door open.

[Hardy Heinlin]

I'll probably return to my original idea that goes like this:

Severe: Once closed, the door will not open.
Else: Once open, the door will not close.

This way the failure will only occur during a command change; the malfunction button on the Instructor will not move the door instantly.

I re-arranged the two pairs of malfunctions like this:

APU failure -- oil/door
Severe: Oil pressure below 50 psig due to oil leak.
Else: Air inlet door stuck in current position.

APU failure -- N1/N2
Severe: N1 underspeed (less than 90%)
Else: N2 overspeed (greater than 103.2%)


Cheers,

|-|ardy

[Hardy Heinlin]

For the APU door related air/ground relays, this is the F22 and H8 stuff, right? The same that is used for the APU fire system.* These are relaxed to ground mode when not powered.


By the way, I see on the diagrams that the air mode should close the door, but I think this cannot be the whole truth because **, as we know, the APU will keep running when left on during takeoff, providing bleed air up to 15000 ft (just no electric power).


Another question: In the next PSX update the APU fire will set the restart inhibit flag. As usual, the fire will destroy and open the fire detection loops A and B. Do you think the next restart should clear the inhibit if the fire no longer exists but the APU FIRE LOOP status messages are still displayed? All detector loops can only be repaired by the "Reset malfunctions" button on the Instructor.


Cheers,

|-|ardy


Edit:

* I now think these are those relays which relax to air mode when not powered.

**  I think in air mode the door just cannot be moved to the open position. So it will remain open if it was open during takeoff (and there is no shutdown). And it will remain closed when it was closed during takeoff.

[John H Watson]

By the way, I see on the diagrams that the air mode should close the door, but I think this cannot be the whole truth because **,

Air mode moves the door from its ground position (45deg) to its flight position (15deg).

These are relaxed to ground mode when not powered.

If you have any notes on air/ground relay R263, it may help (although there are additional logic gates to satisfy. Bat switch position and APU fire switch position may affect the air/ground modes). This relay controls the activation of an APU air/ground relay R359 which controls door angles and stops the door opening in the air.

APU Door Schematic

[John H Watson]

Do you think the next restart should clear the inhibit if the fire no longer exists but the APU FIRE LOOP status messages are still displayed?

I think only active fire alarms can shut down the APU. A dual fault shouldn't activate the alarm so there will be no autoshutdowns.
Note that pushbutton fire tests inhibit the shutdown signal going to the APUC.

While there are lots of (active) things which can cause a shutdown, there are only 5 things which set a re-start inhibit in the APUC. These 5 things will allow the APU to start, but will shut down the APU after a 1 minute cooldown period (assuming there are no factors which autoshutdown the APU immediately). If the restart, cooldown and shutdown period is ok, the restart inhibit is removed.

[Hardy Heinlin]

Thanks for the schematic; this shows the details in the M83 box better. So the 15° inflight position is still an "open" position, and therefore will not cause a disagreement.

I've seen the list of causes for a shutdown and implemented some examples, i.e. N1 underspeed, N2 overspeed, low oil pressure, and inlet door closed. Fire induced shutdown was already implemented in PSX 10.0.0. The restart 1-minute test cycle logic is now fully implemented as well.


Cheers,

|-|ardy

[John H Watson]

I assume the message will appear whenever the door is commanded to a new position and doesn't reach that position within 60 seconds. e.g. if the aircraft takes off with the APU running, the close relay R460 needs to be activated for a short period to run the door to 15 degrees (from 45deg). If the actuator is broken, the 15 degree switch in M83 will not change position and the close relay R460 will be permanently energised (there will be a permanent earth signal at the EIU.. and the message will appear).

Here's a diagram showing how (I think) the close relay R460 is energised via R359 and the 15deg "INT" switch in M83 after liftoff. If the actuator is ok, the 15deg switch will switch to the "NO" position at a fraction less than 15 degrees and the R460 solenoid will de-energise, stopping the motor and removing the earth signal from the EIU.

Diagram showing door close command when aircraft lifts off



I've had to make further amendments to the notes on the original diagram. More errors found. R7176 solenoid should get an earth when the APUC is keeping the door open during the cooldown and shutdown period.

[Hardy Heinlin]

Thank you. I'll make 3 instead of 2 door states: 0°, 15°, 45°.

For valve-like parts I usually implement gradual variables, if they move slowly. That APU door, however, seems to move pretty fast, at least when the APU selector is set to START. N2 starts rising after ca. 2 seconds, and I assume it won't do this before the door is open.


|-|


P.S.: Do you know if there'll be a shutdown if the door is stuck at 45° in flight? Why is it moved to 15° in flight? Is it because the 45° angle will affect the aerodynamics of the fuselage? Or to reduce the airflow into the APU turbine?


One more question ... If the APU is normally shutdown, and the door remains open, will it be possible to restart the APU? I don't think so. There might be a dilemma: If it's stuck at 45°, it won't move to 15° after takeoff, and that is to be avoided in advance, so it won't even start. And if it's stuck at 15°, it won't start because on the ground it has to go to 45°.

[John H Watson]

N2 starts rising after ca. 2 seconds, and I assume it won't do this before the door is open.

We may have discussed this before. I don't really know how fast the door opens because you don't normally go straight from OFF to START. The door begins to open when the APU selector is moved to ON.
The 13 degree switch must be activated before the APUC allows the start sequence to commence. The 13 degree switch talks to the APUC directly.

Someone said the door takes less than 30 seconds to fully open (using a ground based observer), but I don't know how much less than this.

We need to have someone see what parameters are available on the EICAS APU Maintenance Page with the overhead Ground Test switch ON and the APU OFF. If the door position is available (showing CLOSED), we need someone to time how long the door indication takes to show OPEN when the APU selector is moved to ON. This will be the 13 degree switch being activated.

Anyone in the sim soon and who knows how to access the APU Maintenance Page?

[John H Watson]

P.S.: Do you know if there'll be a shutdown if the door is stuck at 45° in flight? Why is it moved to 15° in flight? Is it because the 45° angle will affect the aerodynamics of the fuselage? Or to reduce the airflow into the APU turbine?

Probably a combination of both. Also, a fully open door at high speed will probably damage the door mechanism. It may rip off the door or damage the actuator (Perhaps under extreme loads, the actuator won't allow closure).

(EDIT: Actually the DDG allows the aircraft to be flown with the door fully open. There are weight penalties though).

[John H Watson]

If the APU is normally shutdown, and the door remains open, will it be possible to restart the APU? I don't think so. There might be a dilemma: If it's stuck at 45°, it won't move to 15° after takeoff, and that is to be avoided in advance, so it won't even start. And if it's stuck at 15°, it won't start because on the ground it has to go to 45°.

I don't see any aircraft logic which would prevent a restart. If the APU is shut down after a flight (on the ground) because external power and air is available), the message will appear when the door fails to close... and the defect would have to be fixed or the DDG will have to be consulted before the next departure. If it remains open after APU shutdown during taxi out, then I guess the crew will have to decide whether to fly or not when they see the message. But they shouldn't need to restart the APU in that second situation anyway.

[Hardy Heinlin]

N2 starts rising after ca. 2 seconds, and I assume it won't do this before the door is open.

We may have discussed this before. I don't really know how fast the door opens because you don't normally go straight from OFF to START. The door begins to open when the APU selector is moved to ON.

Yes, I remember. Well, anyway, it's just cosmetics as the door angle is not displayed on the normal EICAS. Let's not waste precious sim time to check this out.

[John H Watson]

If it remains open after APU shutdown during taxi out, then I guess the crew will have to decide whether to fly or not when they see the message. But they shouldn't need to restart the APU in that second situation anyway.

Well, I suppose if the aircraft had to return to the gate then the ability to restart would be an asset.

[Hardy Heinlin]

If the APU is normally shutdown, and the door remains open, will it be possible to restart the APU? I don't think so. There might be a dilemma: If it's stuck at 45°, it won't move to 15° after takeoff, and that is to be avoided in advance, so it won't even start. And if it's stuck at 15°, it won't start because on the ground it has to go to 45°.

I don't see any aircraft logic which would prevent a restart.

If the APU is restarted even though the door has been open before, I assume this will happen:

If stuck at 45°, shutdown will occur after takeoff. No shutdown, DDG allows it *.

If stuck at 15°, shutdown will occur some seconds after APU start, or upon landing.


*

the DDG allows the aircraft to be flown with the door fully open