Flight director ALT HOLD mode

[Hardy Heinlin]

Hi Fred again,

I'm again sceptical :-) Here's a quote from another book:

Altitude hold control is by three methods:
• Push the ALT HOLD pushbutton on the MCP
• In V/S mode, select a V/S of 0000
• Capture and hold the selected MCP altitude

Altitude Hold With the ALT HOLD Pushbutton
[...]

Altitude Hold with Vertical Speed
When in the V/S mode select a vertical speed of 000. The airplane levels and holds barometric altitude. V/S mode remains engaged and the FMAs show V/S.

Altitude Hold with Altitude Capture
[...]

|-|

[Jeroen Hoppenbrouwers]

This would imply a hidden mode shift when the V/S selection is 0000. A kludge to better conform to some people's idea that VS=0 means the same as ALT HOLD.

This prompts me to suggest flying with VS=100 ft/min for a while. That should behave rather differently from VS=0, as it would not include the artificial ALT HOLD hack. The plane should move steadily, not listening at all to baro changes or altimeter changes.


Jeroen

[Hardy Heinlin]

Reconsidering Fred's points ...

1. After a downwash, V/S 0 won't put the aircraft back to the initial altitude. Climbing would mean airspeed loss and that would be undesired.

-- To me this reason makes no sense. It's a matter of reaction time. If the V/S mode reacts promptly with a pitch up command, no altitude will be lost, and -- that's my point -- during this reaction, airspeed would be lost also.


2. V/S 0 accepts any vertical disturbance. The aircraft will stay at the new altitude the disturbance has put it to.

-- To me, this is exactly the same as flying with the AFDS off. Trim the aircraft and let it fly on its own. When not disturbed, it will maintain a vertical speed of zero. When disturbed, it settles at the new altitude (it takes a long time until the phugoid puts it back). Hence, without V/S 0 mode engaged, the aircraft would be even more stable. Especially in turbulence: Over- und underspeed is compensated by the natural phugoid, while V/S 0 would ignore it. To me, it looks like V/S 0 was a destabilizer if it wouldn't hold any altitude to a certain degree.


Cheers,

|-|ardy

[Hardy Heinlin]

This would imply a hidden mode shift when the V/S selection is 0000. A kludge to better conform to some people's idea that VS=0 means the same as ALT HOLD.

This prompts me to suggest flying with VS=100 ft/min for a while. That should behave rather differently from VS=0, as it would not include the artificial ALT HOLD hack. The plane should move steadily, not listening at all to baro changes or altimeter changes.

In fact, there are various submodes behind the curtain of a mode annunciation. For example, consider the transition phase when approaching an altitude with FLCH; at some point it must stop holding airspeed with elevators, otherwise it would overshoot.


|-|ardy

[Jeroen Hoppenbrouwers]

Correct with those submodes, of course. But I mean here that there is a rather large difference between VS=0000 and VS!=0000 settings, unlike with FL/CH where the submodes kick in for capture etc. and transition by themselves.

VS=0000 seems to mean the same to the machine as pushing ALT HOLD, but the machine lies to the pilot and displays 0000 in an open VS window instead of illuminating the ALT HOLD light.

Of course, it would be even more confusing if it DID illuminate the ALT HOLD light and close the VS window when you put it at 0000 for two seconds.


Jeroen

[Lasse]

Hey All

Hmm I lost a little track of the replies...

However: When an altitude is captured then barometric reference doesn't matter since the ADIRU has locked on the outside air pressure = BARO or STD. Changing the barometric pressure will not cause the A/C to change level. However altitude deviation warning might sound move the ALT TAPE accordingly and show the ALT deviation logic on the ALT TAPE if the BARO/STD deviation or what ever BARO change you make is big enough...
If selecting V/S in level flight A/C will maintain altitude. However if A/C have a vertical rate at the moment V/S is pressed that will show in V/S WINDOW and be kept no matter if moving away from ALT set in the ALT WINDOW.

Mode of capture on other than VNAV is called ALT AQUIRE and is a fine calculation of V/S, SPD and THR reduction...

Hope it helped

Best regards
Lasse

[Hardy Heinlin]

Hi Lasse.

If selecting V/S in level flight A/C will maintain altitude.

With V/S-0000 engaged: When turbulence has pushed the aircraft below the reference altitude, the F/D will command a climb back to that altitude?


Cheers,

|-|ardy

[Lasse]

Hey Hardy

Yes.
Lets say smooth air and level at FL350... V/S is pressed with a A/C vertical rate of 0 (V/S = 0) then later on still in V/S A/C enters a turbulent area then F/D will command A/P to keep current ALT since no input from V/S mode other then 0 has been entered... (The ADIRUs has not been told to exit the outside pressure hold) So the ADIRUs keep the current pressure altitude until a vertical mode has been selected with a valid input or V/S <> 0...

Best regards
Lasse

[John H Watson]

(The ADIRUs has not been told to exit the outside pressure hold) So the ADIRUs keep the current pressure altitude until a vertical mode has been selected with a valid input or V/S <> 0...

Note: The ADIRUs or, more likely, ADCs and IRU's (on the 744) are just sensors. They won't be told to do anything. The FCCs are the computers deciding what data to use from the sensors.

The FCCs receive the following from the ADCs:
Altitude
Altitude rate
Computed Airspeed
Impact pressure
Indicated AOA
Mach
Maximum operating schedule
True airspeed

And from the IRUs:
Pitch attitude
Roll attitude
Pitch rate
Roll rate
Yaw rate
Mag heading
True heading
Mag track
True track
Groundspeed
Inertial altitude
Inertial vertical speed
Accelerations (7)
IRS discretes

Rgds
JHW

[Hardy Heinlin]

I'm sure Lasse knows this :-) It was probably meant in a math-poetic way like saying "the aircraft won't be told to exit ..." and "the aircraft keeps the current ...".

[Hardy Heinlin]

As we all know, when engaging the V/S mode, the aircraft's current vertical speed will be transfered to the MCP V/S window.

As a side effect, in my current code this works not only when engaging the V/S mode, but always whenever the V/S button is pushed while the V/S mode is already engaged. I find this side effect useful. I wonder if the real ship works similarly. In PS1 this is not possible.


Cheers,

|-|ardy


Edit:

I find this side effect useful.

However, when accidentally pressed, it may reset the current desired setting.

Any votes -- pro or contra? Real life would give the answer, but it's difficult to check and not an important feature anyway.


Edit 2: The same applies to HDG HOLD. The current aircraft heading could be transferred to the HDG HOLD mode whenever the HDG HOLD switch has been pressed, not only when HDG HOLD engages, but also when it's already engaged.


Edit 3: ALT HOLD, too. The current baro reference could be transferred to the ALT HOLD mode whenever the ALT HOLD switch has been pressed, no matter if the ALT HOLD mode is already engaged.

[Lasse]

Hey All

@ John: sorry I was sure that the B744 had the integrated ADC IRU (ADIRU) box's in the E&E, like the B737NG and B777. However I can see now that I logged in at Boeing and read the FCOM that its separated ADCs and IRUs. Basically its the same thing in the end... But right you are...
As Hardy wrote then my description was trying to explain that changing pitch mode does not necessarily make the A/C change physical level... And since we do a lot of fun stuff in test flights that is not normal, and I know that the PSX will not be flown 100% passenger friendly every time then I was just trying to eliminate some loop holes... ((E.g FLCH will also keep altitude if the thrust is not altered...)) just to try helping making Hardy's PSX as close to the real thing no matter what the user comes up with of funny ideas...

@Hardy: As I understand the B744 AFDS, however it differs here compared to what I normally fly so I might be a little off, then if deselecting the HDG it reverts to default then its either HDG HOLD and or ATT depending on bank and B744 model. However for ATT it will keep the bank is more than 5 deg and for default HDG HOLD turn wings level and keep the current HDG. For pitch if in default mode and pressing V/S and dials say 2300 ROC into the V/S window and deselects the V/S so it reverts to default then it will keep current V/S of what you have at that specific moment...  Pressing V/S button again it will show that V/S in the window...

Did that make sense? - Please correct me if Im wrong :-)

Best regards
Lasse

[Hardy Heinlin]

Hi Lasse,

not sure if I understood you correctly.

My thought experiment goes like this:

A/P disengaged. Only F/D is engaged. Manual flight. Engage V/S.

Vary your vertical speed with the yoke. Press the V/S switch multiple times while you ride a non-steady vertical path. Will the V/S window value change whenever you press the V/S switch -- or will it change only once upon engaging V/S mode after which the V/S switch will be inactive?

Same principle and questions for HDG HOLD and ALT HOLD. Can the reference values be set only at mode re-engagement or also multiple times after mode engagement (without dis- and re-engaging everytime)?

You just inspired me for another question: When you bank 30° and press HDG HOLD, will the reference heading be set only when you are within 5° bank (can you keep flying circles and the F/D bar will remain centered?) -- or is the reference heading set immediately at mode engagement and the current bank angle is taken into account to compute the final reference? The latter is what I  programmed today:

When pressing HDG HOLD, I add a quarter of the current bank angle to the reference heading.

E.g. when pressing HDG HOLD while heading is 300° and bank is 20°, I take the sum of 300° plus a quarter of 20° -- and that'll be my the reference heading.

300° + 20° / 4 = 305°

If I don't follow the F/D and turn further, the F/D will command me back to 305°. If I do follow, the reference heading will be reached at wings level and no turn back will be required.


Cheers,

|-|ardy

[John H Watson]

I normally fly so I might be a little off, then if deselecting the HDG it reverts to default then its either HDG HOLD and or ATT depending on bank and B744 model.

Hi, Lasse.
By "deselecting", do you mean pushing the "mode" button a second time? Wouldn't this only be possible for armed modes? (LNAV & VNAV on the ground, LOC & G/S before capture, etc). Or is this possible on the NG for engaged modes?

When you bank 30° and press HDG HOLD, will the reference heading be set only when you are within 5° bank (can you keep flying circles and the F/D bar will remain centered?) -- or is the reference heading set immediately at mode engagement and the current bank angle is taken into account to compute the final reference?

Wouldn't this depend on the previous mode prior to pushing HDG HOLD?

According to the manuals, HDG HOLD mode is simply a roll-the-wings-level mode and then follow the heading at wings level.

If no previous selected mode, then ATT mode (common option) will be the starting mode in flight:
First A/P on (no FD on, presumbably) or first FD on in flight: AFDS holds current bank angle if angle between 30~5 degrees. If greater than 30, the command will be to roll to 30 and hold. If less than 5, the command will be to roll to wings level.

Rgds
JHW

[Hardy Heinlin]

Yes, a wings level mode. I fly 30° bank and press HDG HOLD. TOGA or LNAV or ATT or whatever is replaced by HDG HOLD. Now I wonder at what time the reference heading is set. Is it set when wings have reached level -- say, 5 hours later if the pilot levels only 5 hours later. Or is it set promptly and some additional predicted heading tolerance is added to guess the final reference heading which will be reached when the pilot has levelled from, say, 30° bank to the final 0° -- assuming the pilot starts levelling immediately after mode engagement and not only 10 minutes later.

I use the initial bank to estimate an arrival time at which the wings arrive at level. The goal is not ATT mode, but 0° bank.


Cheers,

|-|ardy

[Hardy Heinlin]

P.S.: Bank angle is sort of a heading rate. The steeper the bank angle, the faster the heading rate. The ND trend vector uses this (and groundspeed) for turn radius calculation. I think, instead of the bank angle I could also use the heading rate directly. It would be more precise since the bank angle is not the only variable influencing the heading rate. The elevator becomes increasingly effective with a rising bank angle. On the other hand, this might make the F/D look a bit nervous. This is probably the reason why they only use the bank angle for the ND trend vector. I once used the heading rate for it and the "worm" looked really nervous when moving the elevator at steep bank angles.

[John H Watson]

In PSX, what happens if you are banking left (A/P not engaged, but FD on) then you push HDG HOLD and ignore the FD command? Specifically, what happens after you have flown >180 degrees? Does it continue to show fly right?

[Hardy Heinlin]

It sets the current heading as a reference plus a few more degrees depending on the bank angle. This reference will be commanded then all the time and always for the shortest distance. I'm not happy with that solution. I think it should command wings level all the time until wings level are achieved. And only then -- perhaps when bank <5° -- set the current heading for reference.

[Hardy Heinlin]

I'm now convinced that V/S 0000 is an ALT HOLD mode. An FPM 0 hold mode would be impossible for plain logical reasons. Not only turbulence, but every microscopic altitude deviation would be ignored and would lead to permanent FPM fluctuations.

Example:
At 000 milliseconds: Aircraft fpm: 0.0
At 100 milliseconds: Aircraft fpm: 0.3 (caused by turbulence or A/P yoke)
At 200 milliseconds: Aircraft fpm: 0.0 (corrected by A/P yoke)
At 300 milliseconds: Aircraft fpm: 0.0
At 400 milliseconds: Aircraft fpm: 0.0
At 500 milliseconds: Aircraft fpm: 0.2 (caused by turbulence or A/P yoke)
At 600 milliseconds: Aircraft fpm: 0.0 (corrected by A/P yoke)
At 700 milliseconds: Aircraft fpm: 0.0
At 800 milliseconds: Aircraft fpm: 0.4 (caused by turbulence or A/P yoke)
At 900 milliseconds: Aircraft fpm: 0.0 (corrected by A/P yoke)
etc.

Microscopic fpm fluctuations are corrected by the yoke, but the altitude deviations are not. During these 1000 milliseconds the aircraft loses 0.9 feet of altitude (0.3+0.2+0.4). The average fpm during this time is 0.9 feet per second -- use this technique permanently and it it will be 54 feet in a minute. 54 fpm -- with V/S 0000 engaged.


Cheers,

|-|ardy


P.S.: The above technique is not the only problem. The main problem is this: Any yoke input leads either to oscillations or to a new non-zero vertical speed. If the yoke correction is too weak, a non-zero vertical speed remains. If the yoke correction is overcorrecting, an oscillation starts. Only if the yoke eliminates the error spot on by 100.0%, the vertical speed will be set to 0. Of course, that would be a rare coincidence. In other words, the vertical speed would never reach zero. The aircraft would be in an unstable chaotic state. The yoke would chase back and forth without any dampening reference.

I admit, if we want a non-zero V/S, the V/S will fluctuate also. But that's a different situation. In that case we want to increase the distance between the aircraft and an initial altitude. That's sort of a non-stable condition regarding vertical stability, i.e. we don't want the V/S to come back to zero. Instead, we want a continuous positive or a negative V/S. It doesn't matter if it fluctuates as long as it stays either on the negative or positive side.

While in V/S 0000 mode, we want to get back to V/S zero, this is sort of a stable condition regarding vertical stability. Neither a positive or a negative V/S do we want. We want zero. But that's technically not possible because there are natural fluctuations. Unlike the positive or negative area, zero is an infinetly small area. There is no room for correction as every correction itself is already a non-zero V/S. Hence, we need a fixed altitude reference to get the desired vertical stability with an average of 0 fpm in the long run.