Cabin Altitude (once more)

[Hardy Heinlin]

Back to this older thread started in 2019 ...

There are two subjects in this discussion: The toilet woosh effect and the question whether the CAB ALT display includes the QNH.

The woosh effect is now eliminated in PSX 10.151 (available soon).

What remains is the QNH question ...

Observations on the real 744 have revealed that the captain's baro setting can influence the EICAS CAB ALT value. On the other hand, there are observations (777) where this is not the case.

Could it be that there are two situational modes, maybe one mode for the departure and another mode for the arrival?

I understand that in the departure phase the outflow valves simply start in "fully open" position. For this position the CAB ALT value is irrelevant.

But in the arrival phase, the outflow valves start in a (nearly) closed position, and the pressurization controller needs to plan a linear opening process from cruise altitude to LDG ALT. The EICAS LDG ALT is not indicated as a pressure alitude but as a true altitude above MSL. So the system needs to know the local QNH to determine the true distance between (A) the current aircraft altitude and (B) the runway elevation. Both (A) and (B) must refer to the same measuring reference: Either ISA or QNH. Otherwise the measured distance will be incorrect, and the linear opening process of the outflow valves will cause the CAB ALT to mismatch the LDG ALT shortly before touchdown.

An alternative method could be a non-linear opening process. This method needs no QNH. When the aircraft altitude has reached about half of the distance between the stored CRZ ALT and the RWY elevation, the controller might set the valves to a position that is half between the stored CRZ ALT valve position and fully open. At that point it might plan the same half-way-method for the rest of the descent. And then again, and so on. E.g. starting at CRZ ALT 40000, the half-points would be 20000, 10000, 5000, 2500 etc. This way the last part of the opening process will be very smooth even when the QNH is far from ISA. -- However, I find this method rather unrealistic :-) This exponential profile may cause great CAB ALT changes at the beginning.

Maybe the real system uses a mix of both methods. But even then it needs to know the QNH.

Or it just ignores the QNH and accepts the following effects:

1. At high QNH, the outflow valves will be fully open before landing. I guess that's acceptable.

2. At low QNH, the outflow valves won't be fully open on touchdown. After touchdown the system will slightly close the valves anyway (in this case they are already), and after rollout, they'll open anyway. In this case the opening just takes a few more seconds.

So why is the LDG ALT important and the QNH is not? -- When the runway elevation is high, say, 8000, the profile planning is important. The linear opening process will reach the "fully open" position when the aircraft reaches 8000. This is certainly not desired when the runway elevation is 1000. The valves shouldn't be fully open at 8000 already. Obviously, the QNH can never cause a deviation of several thousands of feet. So we may ignore the QNH. Just set the right LDG ALT. That's enough?

Nevertheless, as I wrote above, observations on the real 744 have revealed that the captain's baro setting can influence the EICAS CAB ALT value. Could this be software version specific?


Regards,

|-|ardy

[John H Watson]

Observations on the real 744 have revealed that the captain's baro setting can influence the EICAS CAB ALT value.

Was this reported or a video? I can't remember. For a Cabin Altitude display and warning test, there is a prerequisite in the Maintenance Manual to set the captain's and F/O's EFIS to STD, then observe that the Cabin Altitude is present (current) field elevation +/-500 feet. I don't know why there is such a big tolerance, but it's a nothing like the range if you alter the baro knob. I suppose if the packs are running, doors mostly closed, there would be a cabin pressure differential, even with the outflow valves fully open.

[Hardy Heinlin]

I can't find a video. I think it was a textual report, about 12 years ago.

As you put the word "warning" in bold font. Wouldn't the warning also refer to the displayed value? At certain values, the display turns amber or red -- no matter whether the value has a QNH additive or not. Perhaps it's just the reference of the test system which includes the QNH?


Question to all pilots and engineers:

Should I remove the QNH additive from the CAB ALT display? (It's just the EICAS display; the internal cabin controller uses STD pressure altitude as usual.)

[John H Watson]

As you put the word "warning" in bold font. Wouldn't the warning also refer to the displayed value?

Ah, of course. But if the font colour change and EICAS warnings are purely linked to fixed numbers, what does it matter what the baro setting is for the test?

I thought this would all be in the schematics, but it doesn't go into enough detail.

My training notes just read:

"Cabin altitude and cabin rate are calculated by the EIU based on a cabin pressure signal received from the Cabin Pressure Controller System (CPCS) [via an ARINC 429 databus]. A backup analog cabin pressure signal is received by the EIU for the cabin altitude and rate calculations in case of a CPCS failure."

I'm not sure if they mean a total failure of the CPC or just the ARINC databus. I can't see the CPC generating a signal without power.

There is a Status message if the two values disagree by 800 feet. I assume the EIU processes the disagreement.

Cabin pressure is sent to the MAWEA from the EIUs for the 10,000' warning and to put the NO SMOKING and FASTEN SEAT BELT signs on in the cabin.

"Differential pressure is calculated by the EIU using cabin pressure and an ambient pressure signal received on an ARINC 429 data bus from the Air Data Computer (ADC)."

Would cabin differential pressure also be STD?

[Hardy Heinlin]

"Differential pressure is calculated by the EIU using cabin pressure and an ambient pressure signal received on an ARINC 429 data bus from the Air Data Computer (ADC)."

Would cabin differential pressure also be STD?

This is all STD in PSX as well. In PSX, when you're winding the EFIS baro knob, the EICAS CAB ALT rate and delta will not change. The baro knob only affects the value on the EICAS screen and the associated alerts. The underlying pressurization control is completely on STD.

E.g. when the aircraft is parked at SECU at 8500 feet, you can wind up the baro knob to increase the CAB ALT display and make it amber and red. But the rate and delta remain zero.

I can disconnect the baro from the EICAS display. But then I don't understand why the maintenance test requires a correct baro setting if it has no influence anyway.

[simonijs]

Could it be that there are two situational modes, maybe one mode for the departure and another mode for the arrival?

Hi,

I do think this is the case. The controller needs to know where a phase starts and the "condition" at the end of that phase. It then plans a schedule based on the end condition of that phase. For the climb it starts with the ambient pressure at the airport of departure (easily converted into Pressure Altitude) and ends at the ISA pressure that goes with the cruising flight level, by definition a pressure altitude. For the descent and landing, it starts at a pressure altitude (easily converted into altitude) and ends at the landing airport field elevation, based on local altimeter setting. Hence, the schedule for the climb phase is based on pressure altitude (STD), the schedule for the descent phase is based on QNH.

Today, I looked into the FCOMs of both the 737 (slightly more detailed than the 747/777 FCOM) and A330. During preflight the pilots of the 737 need to set a cruising altitude in one window of the controller (overhead panel) and the landing field elevation in a second window.
Quoted from the 737 FCOM: "Takeoff airport altitude (actually cabin altitude) is fed into the auto controllers at all times when on the ground". "The descent mode is activated when the airplane descends 0.25 psi below the selected FLT ALT. The cabin begins a proportional descent to slightly below the selected LAND ALT". "The air data inertial reference units (ADIRUs) provides ambient static pressure, baro corrected altitude, non corrected altitude and calibrated airspeed to both automatic controllers. The ADIRUs receive barometric corrections from the Captain's and First Officer's BARO reference selectors".

Quoted from the A330 FCOM: "The controller normally uses the landing elevation and the QNH from the FMGEC, and the pressure altitude from the ADIRS. If FMGEC data are unavailable, the controller uses the Captain BARO Reference from the ADIRS and the LDG ELEV selection." "Pressure rate is optimized so that cabin pressure reaches landing field pressure +0.1 PSI just prior to landing. The cabin descent rate is limited to 750 ft/min." FMGES/C = Flight Management Guidance and Environmental System (Computer).

In the 747, during preflight an initial cruising altitude is selected on the PERF INIT page; let's say: FL310. Before descent, the local QNH is obtained from ATIS and is preselected with the EFIS Baro knob. I think the controller is using this preselected value to plan a descent schedule. While descending the QNH may change by ATC from 1022 to 1021 (or 1023), or from 29,68 to 29,66 (or 29,70). Not a big deal for the controller to adjust...

Hope this makes any sense.

Regards,
Simon

[Jeroen Hoppenbrouwers]

The idea that the controller is either in climb mode (aiming for the known cruise alt, and then adjusting to the set cruise alt whenever it changes), or in descent mode (aiming for the baro corrected field elevation) makes sense. It knows the final target, can calculate a gradient, and then track the gradient against the measured outside pressure.

One thing does not appeal to my engineering brain: that the 747 descent program should rely on the landing field QNH being set before descent. I guess that instead it uses the known landing altitude (from the destination database) all the way down and smooths out the STD --> QNH switchover. By the time the aircraft is about to touch down, the cabin presure should have well caught up with the outside pressure.

It does not need to track altitude. It only needs to know what its final pressure differential should be: probably close to max hull differential at cruise target, and SAP at landing target. Both can be measured, so the final results can be accurately reached, but to assure a smooth gradient, the thing likes to know roughly where it shall end, hence the cruise altitude and landing altitude settings. QNH may even be completely disregarded as the final few thousand feet, either way, it is going to be mostly SAP anyway.

I think.


Hoppie

[Hardy Heinlin]

Having two modes might explain the observation of the QNH influencing the CAB ALT display while other sources suggest this shouldn't be possible. If this explanation is correct, it must have been observed while the system was in arrival mode. It was on a parked aircraft anyway. Probably before FMC init.

As for the STD/QNH switch-over theory: I too can imagine the system uses smooth transitions later in the descent rather than looking at the QNH preset. The lowest TL in the world is probably FL20. E.g. the cabin controller may expect to get a valid QNH not later than circa 2000 pressure feet above LDG ALT, and at that point it might start a smooth transition.


|-|

[John H Watson]

In the 747, during preflight an initial cruising altitude is selected on the PERF INIT page; let's say: FL310. Before descent, the local QNH is obtained from ATIS and is preselected with the EFIS Baro knob. I think the controller is using this preselected value to plan a descent schedule. While descending the QNH may change by ATC from 1022 to 1021 (or 1023), or from 29,68 to 29,66 (or 29,70). Not a big deal for the controller to adjust...

Indeed. But there are probably a few false assumptions regarding how the system handles pressurisation during descent/landing.

Lesser known features of the pressurisation system:

On landing, the outflow valves will depressurise the cabin at the rate of 500fpm, but if the pressurisation system has been programmed with incorrect information and the pressure is not equal within a minute, it will increase the cabin pressure rate to 2000fpm.

If the aircraft is in a shallow descent in VNAV or goes into hold during descent, the cabin pressure will continue to ramp up to the preselected landing field altitude (minus 50 feet) unless the differential pressure reaches the differential pressure memorised at TOD).

The FMC-controlled "external (pressurisation) schedule" is not even used when the A/P is not engaged in VNAV. Handflying to/from cruise altitude is probably not helping passenger comfort. The Cabin Pressurisation Controller "internal schedule" has a default cruise altitude of 39,000 feet.

There seems to be few mode variations:
Full VNAV mode
VNAV mode not engaged, but FMC data available (e.g. landing data)
VNAV mode not engaged and FMC data not available.
Are there others?

Does the pressurisation system use airport general altitude if runway threshold data is not available or does it assume 2000?

[Jeroen Hoppenbrouwers]

If the aircraft is in a shallow descent in VNAV or goes into hold during descent, the cabin pressure will continue to ramp up to the preselected landing field altitude (minus 50 feet) unless the differential pressure reaches the differential pressure memorised at TOD).

Ow. If even these assumptions were already false, we really should stop thinking out a solution that would be easy and logical.

Hoppie

[Hardy Heinlin]

I'm wondering if one of the real 744 pilots on this forum, after flight before leaving the cockpit, could turn the captain's EFIS baro knob a bit up and down to see whether the knob changes the CAB ALT display ...


|-|

[United744]

As I understand it, QNH only affects the display and not the pressure. The systems is always referencing to delta pressure with outside and doesn't care for the "altitude".

The only reason it cares about the landing altitude is because if you took off at sea level (14 psi) and land at 8000 ft (10.91 psi) is it needs to get from whatever the pressure was at cruise giving max delta p down to 10.91 psi for landing. This could be merely 200 ft/min. It tells it the final target and nothing more.

The rate controller is literally that. If the aircraft is climbing stupidly slowly (100 ft/min) the cabin will climb likewise. It's only when the aircraft climb rate exceeds the cabin rate that the controller limits the rate to 500 ft/min for the cabin.

For takeoff, it knows it is takeoff based on TLA angle. That causes the pressure scheduler to prepare for climb. After that, it follows aircraft altitude changes based on static pressure.

Pre-pressurization occurs when the controller sees engine start events on the ground.

After landing, the outflow valves open completely to dump all residual pressure.

I'd fully expect the cabin altitude to respond to baro changes on the altimeter on the ground, but not in flight.

The pressurization controller can know transition altitude from the FMS so I'd expect it to automatically reference standard above that altitude regardless of the altimeter baro setting (or to put it another way, the FMS can signal to the pressure controller that it is to reference standard pressure), but this will not otherwise affect the operation of the system as it is always looking at static pressure to know what to pressurize the cabin to.

[John H Watson]

Pre-pressurization occurs when the controller sees engine start events on the ground.

Engine start only closes the equipment cooling ground exhaust valve (if it isn't already closed due to low temperature). i.e. no air escaping from the Main Equipment Centre.  i.e. any two engines on opposite wings greater than 50%N2.

For takeoff, it knows it is takeoff based on TLA angle.

My books say groundspeed >65kts (from the FMC) for the outflow valves to move towards closed.
This will give a cabin altitude 50 feet lower than airport altitude

As I understand it, QNH only affects the display and not the pressure. The systems is always referencing to delta pressure with outside and doesn't care for the "altitude".

With the a/p in VNAV, the CPC uses:
1) time of climb (predicted by the FMC)
2) cruise differential pressure determined by a table...

...to determine the cabin rate of climb.

To me, it doesn't sound like differential pressures are relevant (except as the final cruise target) or if there is an exceedance due to unexpected climb rates.

[Hardy Heinlin]

Just got a reply from a tester on the real 744: After landing, before engine shutdown, the EICAS CAB ALT display changes with the captain's EFIS baro knob.


P.S.: And not only before engine shutdown; it also changes before takeoff. So, PSX is correct. I'll keep it as is.

[John H Watson]

After landing, before engine shutdown, the EICAS CAB ALT display changes with the captain's EFIS baro knob.

Don't forget that on landing, the other controller is automatically selected (prior to engine shutdown), so it's probably basically the same as pre-flight.

Anyway, thanks for the confirmation. Good to know I wasn't going crazy 

[Hardy Heinlin]

Just wanted to be sure it's not flight phase dependent. If it were so, I guess the flight phase logic would be installed in either controller anyway, so it wouldn't matter which controller was in use.