LOW SUCTION/artificial horizon drift (EDIT: tracks added by Aernov)

[Jakey-Poo]

After talking with several members of the community (more knowledgeable than myself) the common consensus for the problem with the artificial horizon drifting constantly is due to low suction in the vacuum system.  After comparing this to readings on the suction gauge during flight this checks out.  The problem then becomes why is the suction consistently below operational range even while in climb power?  Is there a fix for this that I have not stumbled across, or is this a bug?

Edited April 13, 2024 by Jakey-Poo
Updated info

[Nozzle]

The vacuum-based attitude systems on these aircraft were never able to maintain a stable reference for the entire flight even when working perfectly.  Really the best a pilot could hope for was a somewhat stable horizon reference during a cloud layer penetration when absolutely necessary to land.  The expectation was the pilot would have been flying straight and level for enough time before penetration for the vacuum to provide a somewhat reliable reference during the time the pilot had no visual horizon.

 

Important to remember these fighter aircraft were only marginally IFR capable by today's standards.

[Jakey-Poo]

21 minutes ago, Nozzle said:

The vacuum-based attitude systems on these aircraft were never able to maintain a stable reference for the entire flight even when working perfectly.  Really the best a pilot could hope for was a somewhat stable horizon reference during a cloud layer penetration when absolutely necessary to land.  The expectation was the pilot would have been flying straight and level for enough time before penetration for the vacuum to provide a somewhat reliable reference during the time the pilot had no visual horizon.

 

Important to remember these fighter aircraft were only marginally IFR capable by today's standards.

Is it normal then for the suction gauge to be consistently below the operational range, and I'm just tweaking?  I figured there was something that I was missing or operating incorrectly to have these results be the norm.

[grafspee]

Looks like at least  3.8" is required to proper operation of gyro-horizon .

I also has impression that gyro-horyzons drifts a bit too quick, not only in P-47 other warbrids as well. I expect to loose gyro coordination after sudden maneuvers, but i have this problem in level flight as well.

Edited May 19, 2021 by grafspee

[Nealius]

Oddly, I find the P-51 and Spitfire to have fairly reliable gyros. They drift after hard maneuvering, but given a few minutes of stable flight they sort themselves out. The P-47's, however, is a functioning alcoholic. Some consistency needs to be applied here.

[Jakey-Poo]

13 hours ago, grafspee said:

 

Looks like at least  3.8" is required to proper operation of gyro-horizon .

I also has impression that gyro-horyzons drifts a bit too quick, not only in P-47 other warbrids as well. I expect to loose gyro coordination after sudden maneuvers, but i have this problem in level flight as well.

 

 

9 hours ago, Nealius said:

Oddly, I find the P-51 and Spitfire to have fairly reliable gyros. They drift after hard maneuvering, but given a few minutes of stable flight they sort themselves out. The P-47's, however, is a functioning alcoholic. Some consistency needs to be applied here.

 

After flying the P-51 this morning I'm getting more and more the feeling that this has to do with inadequate suction in the vacuum system in the P-47, as opposed to the gyro itself.  While in the P-51 I had no problem keeping the suction gauge within the operational range (although admittedly on the lower side), and, I would assume correspondingly, had no problem with the ADI during "normal" flight.  This is a completely different experience than the Jug.  If this is modeled accurately, and the jug had particular problems with the vacuum system, then fine.  I'm just looking for info.

Edited May 19, 2021 by Jakey-Poo
verbiage

[Jakey-Poo]

Has there been any headway on this?  Haven't flown the jug in a few months, came back and this is still an issue (at least for me).

[Jakey-Poo]

After further testing in Instant Action missions where the suction gauge is reading within normal operating parameters, it’s pretty obvious at this point that the artificial horizon acting up is a symptom of inadequate suction in the vac system.  I’m going to be testing more with the compass, but haven’t noticed anything obvious. 

[NineLine]

Please supply a track of the issue if you are still seeing it. Thanks

[Aernov]

Here are two screenshots, taken in flight after cold start, Channel map:

1) Max. continuous power in climb, suction gauge needle is barely over the bottom red line

2) Max economy in almost level flight, suction is at or slightly below bottom red line

Shouldn't it be firmly in "green" zone even at idle power?

[cw4ogden]

This is still a problem.  Tend to agree with the original poster, the poor gyro is linked to low vacuum.  
As is, the plane would be on a red X, awaiting the crewchief to get his duct tape. 
 

Edited November 16, 2022 by cw4ogden

[Jakey-Poo]

On 11/16/2022 at 6:40 AM, cw4ogden said:

This is still a problem.  Tend to agree with the original poster, the poor gyro is linked to low vacuum.  
As is, the plane would be on a red X, awaiting the crewchief to get his duct tape. 
 

 

If you or anyone else knows how, and would be willing to post a track file for this, it would help me out a ton - I’m computer illiterate; I just fly imaginary airplanes.  This got flagged for missing info, and I’m assuming per NL’s response above that the track files are what they need. 

[Aernov]

P-47D-40 suction on the ground.trk

P-47D-40 suction climb 2700 RPM.trk

Here are two short tracks with P-47D-40 (hot start on parking and in air, I haven't noticed a difference with cold started plane in these tests). At 900 RPM suction stays at about 3 in Hg, with RPM increase it starts to rise (beginning at around 1100 RPM) and gets to about 3.8 (low red line on the gauge) and stops there, never rising above low redline. One test was done on the ground, other - in air at military limit (2700 RPM, 52'' MP), both on Caucasus with "DCS default" weather. Same results were obtained with P-47D-30.
There was an outlier case where suction rose to 4,2 in Hg in flight at high altitude (see screenshot) (P-47D-30 early, Normandy, clod start at Azeville ALG)...

Edit: I've got 4.2 in Hg of suction again with D-30 early on Normandy after an air start, this time at low altitude:

P-47D-30 early suction 4,2.trk

Even if vacuum pump has 1:1 engine-pump RPM ratio (I don't know real ratio, but kinda doubt it is 1:1), it should still provide adequate pressure difference at 2500+ RPM.

Also, aren't external tanks pressurized by vacuum pump exhaust?

Edited March 12, 2024 by Aernov

[Jakey-Poo]

On 3/12/2024 at 6:29 AM, Aernov said:

P-47D-40 suction on the ground.trk 250.36 kB · 10 downloads

P-47D-40 suction climb 2700 RPM.trk 2.52 MB · 18 downloads

Here are two short tracks with P-47D-40 (hot start on parking and in air, I haven't noticed a difference with cold started plane in these tests). At 900 RPM suction stays at about 3 in Hg, with RPM increase it starts to rise (beginning at around 1100 RPM) and gets to about 3.8 (low red line on the gauge) and stops there, never rising above low redline. One test was done on the ground, other - in air at military limit (2700 RPM, 52'' MP), both on Caucasus with "DCS default" weather. Same results were obtained with P-47D-30.
There was an outlier case where suction rose to 4,2 in Hg in flight at high altitude (see screenshot) (P-47D-30 early, Normandy, clod start at Azeville ALG)...

Edit: I've got 4.2 in Hg of suction again with D-30 early on Normandy after an air start, this time at low altitude:

P-47D-30 early suction 4,2.trk 3.39 MB · 12 downloads

Even if vacuum pump has 1:1 engine-pump RPM ratio (I don't know real ratio, but kinda doubt it is 1:1), it should still provide adequate pressure difference at 2500+ RPM.

Also, aren't external tanks pressurized by vacuum pump exhaust?

 

Thanks for the tracks, man!  Hopefully this is what NineLine needed to see to get this issue bumped up. 

[kablamoman]

For what it's worth, this seems to be a problem with all of the artificial horizon gauges in the sim. There are fundamental flaws in how the instrument itself has been modelled, that are clearly separate from any suction pressure that may or may not be modelled into the simulated system (though I think looking at how the suction gauge is being modelled so that it provides realistic readings through various engine power ranges is probably a good idea as well).

Here's another thread for reference, that examines the issue (a lack of self-righting tendency) a bit closer:

 

Edited June 28, 2024 by kablamoman

[Bryster51]

Wonder if this is actually modeled...Kinda doubt it...

 

Yes, the outside air temperature can affect the accuracy of a suction gauge:

• Temperature range shift: The elasticity modulus of the gauge's bourdon changes with temperature, which can cause a shift in the zero set point. This shift can be as much as 1% for every few degrees of temperature change.
• Zero shift: The physical dimensions of the gauge's components change with temperature, which can cause a zero shift. This shift is constant across the entire scale.
• Discoloration: High temperatures can cause the gauge dial to discolor.
• Hardening: High temperatures can cause the gauge's gasket or seal to harden.
• Liquid expansion: If the gauge has a liquid-filled case, the case fill fluid will expand with temperature. 
• Other factors that can impact the accuracy of a vacuum gauge include: Pulsing, Vibrations, Gas type, and Operating pressure range.

Edited October 14, 2024 by Bryster51

[PawlaczGMD]

BTW, this issue is not unique to the P-47. A lot of old gyros drift so much that they are almost useless. Even the Ka-50 backup ADI drifts completely out of true within minutes and never recovers. I was under the impression that the gyro drift correction is not modelled at all TBH. Gyros should drift during maneuvers, but restore their true position in level(stable?) flight.

[Yoda967]

On 10/14/2024 at 4:05 PM, PawlaczGMD said:

BTW, this issue is not unique to the P-47. A lot of old gyros drift so much that they are almost useless. Even the Ka-50 backup ADI drifts completely out of true within minutes and never recovers. I was under the impression that the gyro drift correction is not modelled at all TBH. Gyros should drift during maneuvers, but restore their true position in level(stable?) flight.

I know you're correct on your first point. I've read several books by Spitfire pilots who mention the artificial horizon tumbling during a dog fight.

Over in the Spitfire forum, it's been confirmed that gyro drift is modeled. Gyros do indeed drift during maneuvers, but they don't restore their position after the aircraft returns to level flight. That's the purpose of the adjustment knob on the DG...you check it periodically for agreement with the magnetic compass and adjust the DG to match. 

[PawlaczGMD]

8 minutes ago, Yoda967 said:

I know you're correct on your first point. I've read several books by Spitfire pilots who mention the artificial horizon tumbling during a dog fight.

Over in the Spitfire forum, it's been confirmed that gyro drift is modeled. Gyros do indeed drift during maneuvers, but they don't restore their position after the aircraft returns to level flight. That's the purpose of the adjustment knob on the DG...you check it periodically for agreement with the magnetic compass and adjust the DG to match. 

It depends on the airplane, it might not be on the Spit, but some period gyros have a mechanism which I think is called pendulous vanes, where they can "know" where down is due to a weight bob, and self-align in stable flight.

I'm pretty sure it's on the Mustang, which also has an AH that drifts and doesn't correct.

 The artificial horizon is not the same as the directional gyro, which has the knob you speak of. This instrument doesn't seem to drift at all in the game (on the Spit), and the knob is used to set magnetic variation only.

Edited October 18, 2024 by PawlaczGMD

[Yoda967]

You're right...and my apologies. I misunderstood what you wrote.