artificial horizon-always tilts to a side after start up

[9.JG27 DavidRed]

hello guys....

as im totally in love with this plane, i really want to know every little detail about the correct proceedures to handle and master it.

so this time, i have a question about the artificial horizon.

after starting up, and uncaching the horizon, it tends to bank towards one side.i think its when you start to taxi and yaw the plane on the ground, that it then banks to the opposite direction...

i tend to click on the "i dont know what its called horizon button" to center it.:)

but it seems that soon after recentering, the whole process happens again and again....so mostly until im in the air, i have to recenter it for about 4 or 5 times....

i also recognized that there is a wheel where you can adjust the pitch(vertical attitude) of the horizon....

 

so now to my initial question....whats the correct proceedure, what would a RL P51 pilot do, to bring the artificial horizon in its correct position?

[The LT]

Take-off and pull re-cage when you're in level flight. Simple. The gauge accumulates errors over time though. By the way, the manual describes it all in quite adequate detail. Check out the relevant section.

[9.JG27 DavidRed]

thx for the answer...thats usually what i do as well...re-caging it while taxiing and then once more in level flight...

btw, before i opened this thread, i had a look in the manual to see whether i will find the answer myself, twice, but didnt find anything about how to use the horizon correctly....i only found the section where it describes briefly what it does and what the buttons do...but nothing about the correct proceedure.

[MACADEMIC]

This is actually a very realistic instrument behaviour. The attitude indicator is essentially a gyroscopic instrument driven by a vacuum pump. The gyro needs some time to come up to speed, i.e. aquire the RPM necessary to have it showing the correct position in space. But even so, as LT says, this generation of gyros needs realignment from time to time, but error is usually not as great as you experience right after startup.

 

MAC

[-0303-]

10 year anniversary bug, assuming it ever worked correctly.

On 6/2/2013 at 7:35 AM, The LT said:

...re-cage when you're in level flight. .... The gauge accumulates errors over time though.

It does not. It would be useless if it did.

Part of it's fundamental functionality is continuous self correction to the gravity vector using pendelous vanes. Imagine flying hours in  zero visibility. You can't recage. It's dangerously useless.

~

A few years ago, a P-51 having happily crossed the Atlantic, hopping the shortest watercrossings (Iceland, Faroe(?), Scotland). Fat and happy, almost home safe, nearing London, all the real danger is over...

Then, zero visibility, a nailbiting hour or so before, with ground help, finding it's way safely down, grateful for the P-51's range that let him stay up long enough to find a way down.

 

Edited June 11, 2023 by -0303-

[kablamoman]

Yes, would be nice if they fixed this.

[Mobile_BBQ]

On 6/11/2023 at 4:58 PM, -0303- said:

10 year anniversary bug, assuming it ever worked correctly.

It does not. It would be useless if it did.

Part of it's fundamental functionality is continuous self correction to the gravity vector using pendelous vanes. Imagine flying hours in  zero visibility. You can't recage. It's dangerously useless.

~

A few years ago, a P-51 having happily crossed the Atlantic, hopping the shortest watercrossings (Iceland, Faroe(?), Scotland). Fat and happy, almost home safe, nearing London, all the real danger is over...

Then, zero visibility, a nailbiting hour or so before, with ground help, finding it's way safely down, grateful for the P-51's range that let him stay up long enough to find a way down.

 

 

 

If you're Turn/Bank indicator is perfectly centered - both ball and needle, and the Vertical Speed Indicator is also perfectly centered - reading 0 up / 0 down, plus you're flying in this "stabilized" state for at least 20 seconds, then you should be able to cage the horizon without ever having to look outside the airplane or in zero visibility.

More than likely, the pilot's trouble in the example you gave had to do with not being able to navigate properly with no ground reference and having no modern GPS unit installed.  There was most likely a danger of mid-air collision with other aircraft until the ground control could pinpoint his location and make sure his path was clear.

[-0303-]

I found an account of the P-51 trip (this must be the one or a very similar). I pretty sure I read another account where aid from ground and P-51 range was mentioned (commenting had it been a Piper Cub or similar, time would've run out).  I don't remember the King Air escort.

Anyway, my point was that zero visibility is no joke and DCS Artificial Horizon (AH) is bugged and useless for anything more than a few minutes dipping through a cloud. The P-51 AH not used it seems (substituted by escorting  King Air AH I presume). If he lost the King Air he would have needed it.

A P-51B, with 75 gallon drop tanks even. Article copy locked, text copied from https://archive.ph/FRN66 and article photo taken from Twitter.

vintageaviationecho.com/operation-berlin-express/

Quote

 

Crossing the Atlantic Ocean in P-51B Mustang ‘Berlin Express’

... “As the conditions worsened, we deviated further. Every time I thought it couldn’t get worse, it did.

“I went from being able to see the King Air from 50ft to being able to see it from 20ft. At its worst, for seconds at a time my only visual reference of the King Air was its blue wing tip, which sat six feet from my propeller. For about 90 minutes of the two and a half hour flight, I was flying off the King Air. My whole focus was on flying in tight formation – we had drilled it extensively, now we needed to use it. Then Louis Horschel would call, ‘Coming right turn, 30 degrees’ and having no other visual reference, I would have sworn it was a left turn. Then it happened again and I later realised that with nothing but the King Air as a reference point, my mind could not compute that we were turning right. It was a strange phenomenon.” ...

... Had I lost the King Air, we would ensure there was adequate separation and I would have to fly the Mustang by its instruments alone – it wouldn’t have been the end, but it wouldn’t have been a great place to be with no situational awareness, time to navigate or change frequencies.” After nearly three hours and almost 600 miles of constant manoeuvring through weather, the formation finally broke into clear air around 50 miles north of Cambridge.  The extensive deviations had extended the flight time by one hour and almost 100 miles.

...

 

Edited October 12, 2023 by -0303-

[Ala13_ManOWar]

30 minutes ago, -0303- said:

Anyway, my point was that zero visibility is no joke and DCS Artificial Horizon (AH) is bugged and useless for anything more than a few minutes dipping through a cloud.

You guys don't get it at all… I've said many times but I'll one more… THESE AIRCRAFT WEREN'T SUITED FOR INSTRUMENTAL FLIGHT. Do you wonder why?

 

Modern airworthy examples usually don't equip WWII original instruments, they wouldn't get their airworthiness certificate if so, as well as they don't equip old radios, old range finders, old nothing, they are equipped with VOR/ILS, modern radios, modern instruments with no flaws like old ones had. It's no bug, it's how it worked back then whether people like it or not. It's a simulator for a reason. It's the only simulator for a reason…

[kablamoman]

On 10/12/2023 at 11:18 AM, Ala13_ManOWar said:

You guys don't get it at all… I've said many times but I'll one more… THESE AIRCRAFT WEREN'T SUITED FOR INSTRUMENTAL FLIGHT. Do you wonder why?

 

Modern airworthy examples usually don't equip WWII original instruments, they wouldn't get their airworthiness certificate if so, as well as they don't equip old radios, old range finders, old nothing, they are equipped with VOR/ILS, modern radios, modern instruments with no flaws like old ones had. It's no bug, it's how it worked back then whether people like it or not. It's a simulator for a reason. It's the only simulator for a reason…

I think this is just muddying the waters. Modern restorations use updated instruments and avionics for a variety of reasons, but many still do incorporate original flying instruments.

The crux of the issue is that the instrument simply isn't modelled correctly. It's supposed to incorporate pendulous vanes, that with sufficient vacuum pressure, would trend back to finding the horizon. It doesn't do this and it's not the only attitude indicator that seems to have been modelled this way by ED. Almost all the warbirds, but also more modern aircraft like the F-5 suffer from the same issue. 

Here's an excerpt from a description of the AN5736 (the gyro horizon in the p51) in its service manual:

 

 

Nobody is saying that it shouldn't topple or develop errors if left uncaged during a dogfight. But even if it does, it should slowly return back to an accurate horizon due to the mechanism described above. This would also allow us to uncage said gyros on the ground, after the engine (with vacuum pump) is running, as per normal procedure.

You could (and should) still use the occasional cage to quickly correct errors from extreme maneuvering or low suction pressure when running on the ground, but it would not be required in the way that it is now.

As the instruments are currently modelled, you first have to get up in the air, then pick a reference straight & level attitude to uncage before you get accurate indications. This means, as an example, that you can't uncage it on the ground with the attitude nose-high while the aircraft sits on its tailwheel and have the instrument read accurately -- this is nonsense.

Edited October 14, 2023 by kablamoman

[Ala13_ManOWar]

On 10/12/2023 at 8:26 PM, kablamoman said:

I think this is just muddying the waters. Modern restorations use updated instruments and avionics for a variety of reasons, but many still do incorporate original flying instruments.

I'll take it here. Yes, but no. Since in modern day we know the issues, flaws, quirks, and what not from those instruments (but either this or that equipment, system, engine…) so that the modern day operators usually modify this or that little, tiny, stupid, thingy and that instrument, that system, that whole engine which was known at it's time to be unreliable for whatever reason, all of a sudden works flawlessly with no further trouble. But that's due to 80 years of knowledge mates, not because at their own time those whatever hardware it is was flawless, on the contrary back in time those were brand new inventions and many times were flawed by default and by design. It's just we know better nowadays how to make them work properly, not to mention some spare parts, even if it's an original instrument or whatever else piece of hardware, it's rebuilt-refurbished with brand new modern spare parts and those work a zillion times better than original ones. It's like (silly example, but example after all) "look, how good this aircraft starts up now, back it time they weren't able to start it up Ok at once", but it can be related to just the battery is a modern one and it works… If you get my point.


Anyhow, the American instrument should be caged from time to time IIRC, the British instrument (from which American one derives, again IIRC) is the one which should centre itself with just some level flight time but it doesn't quite well. Whatever it is anyway, don't mix the two of them because we're talking P-51 here but I believe they don't work exactly the same way. Anyhow TBH I don't know internal functioning of those to tell it's good or badly modelled internally.

[kablamoman]

Here's the diagram depicting self-erecting mechanism of the P-51's gyro horizon (from the operation and service manual):

 

Some notes on operating limits. Note procedure for uncaging on the ground before takeoff:

 

From the troubleshooting section related to the "Horizon Bar" not settling:

 

Here's a diagram of the pendulum assembly:

A blurb from a section about proper calibration of the self-erecting mechanism:

Table showing the time tolerances for the horizon bar to settle (self-erect):



 

One can perform a pretty simple experiment with our in-game P-51. Cage the gyro, pitch up to 20 degrees (use the info bar), uncage the gyro and return to 0 degrees pitch with the info bar. The gyro will read 20 degrees nose low. Can you guess what the gyro ends up showing after 6 minutes at max continuous (plenty of engine power to drive the vacuum pump)?

0 minutes:

After 6 minutes:

 

At most, it looks like the gyro may have corrected itself by 2 degrees or so. Certainly not the full 20 degrees, back to the level horizon it should be capable of, and by any metric I can find in the maintenance manual for the instrument, it would be considered defective.

Edited October 14, 2023 by kablamoman

[Mr_sukebe]

Would I be correct in assuming that the self correction system in say the Mossie is how we should expect the ADI in the Mustang to perform?

[kablamoman]

Just now, Mr_sukebe said:

Would I be correct in assuming that the self correction system in say the Mossie is how we should expect the ADI in the Mustang to perform?

I don't have the mosquito so I can't vouch for how they've modelled it, but basically there is a constant, self-correcting mechanism in this instrument that seeks to align itself with the horizon based in the acceleration force of gravity acting on the pendulous vanes. If the instrument is uncaged in an unusual attitude, or an attitude slightly off of perfect alignment with the horizon, it should quickly correct itself within the specified timeframe (probably even quicker than the max tolerance allowable). Currently, the tendency to self-correct is either not modelled at all or is so slow that the instrument would be considered unserviceable.

[Kung]

How to win a debate on the accuracy of WW2 warbird systems in a simulator.

1. Post the original (or relatively contemporary) writeups from manuals and maintenance sheets.

2. Perform and document experiments testing the expected performance of those systems.

3. Profit.

Well, I'm firmly on team "modeled incorrectly".

It's a shame, I've been annoyed by the inconsistent behavior if the AH in the Mustang, so went looking for what I was doing wrong. Seems like I'll need to periodically re-calibrate, but also that I shouldn't have to, dernit!  It also seems like a very, very simple thing for ED to fix.