DCS: F-5E is the hardest fast jet to fly in bad IMC weather

[Mikaa]

1 hour ago, GGTharos said:

Would you be able to comment on how much more accurate the indicator should be?  My specific question is, if it was to be suggested as a fix for ED (it has been suggested in fact, I'm sure) - by what factor should precession be reduced?

Having never flown the F-5/T-38, I can't comment on the specifics of the model of attitude indicator installed, and would have to dig to source documentation that comments on its (lack of) self-righting capabilities. As for a broad definition of accuracy, FAA documentation (AC 91-46 amongst others) states anything more than 5 degrees pitch and bank during ground ops (taxiing and instrument checks) is considered unreliable. It also states that 'mild precession from horizontal' is possible during sharp turns while taxiing. Idk what the FAA deems 'mild,' but from the wording, one could infer that it's less than 5 degrees. I'd probably start there and set the max precession to about 5 degrees (though in my opinion it should never be more than a few degrees, unless documentation specifically mentions this AI's susceptibility to precessing). Most often, the parallax error to the adjustable 'miniature aircraft reference' is larger than the total precession of the instrument.

From my experience, the most I've gotten an attitude indicator to precess was a few degrees during severe turbulence or hard maneuvering, but once those conditions ceased, it was instantly reading correctly again. If the vacuum pump has failed, or electric power cutoff if elec. operated, then we're talking about a very different scenario, plus you'd get a gyro flag to indicate unreliability. 

The only way I've gotten an attitude indicator to (somewhat) reliably fail is to aggravate a fully developed spin - The gyro backup AI would often tumble at about the fifth or sixth rotation, and even then, I couldn't always get it to tumble, and if it didn't, on recovery it would still be indicating correctly, even after the 4G pullup to level. If it did tumble, usually by the time we'd gotten back in the pattern would it have fixed itself, even if I didn't manually re-cage the AI (maybe 10 minutes). 

1 hour ago, GGTharos said:

Unrealistic ways of 'fixing the issue in flight' aside.  BTW totally agree on where you're going with spatial disorientation here for the suggested solution, I think most people don't realize this.

Agreed. It's a humbling experience first time in IMC. Especially once you realize that your vestibular system is constantly attempting to kill you. 

 

edit: clarity

Edited November 11, 2021 by Mikaa

[GGTharos]

^^^^

Thank you for the detailed reply.  As I have no experience with such systems IRL it's great to get this type of answer, it is instructive and useful. 

I don't know if I'll have time but I'll try to put together a track or two for ED along with suggested solution (constrain the drift - not sure if they even model tumbling in this case) and add self-righting ability.  I am assuming self-righting is due to gravity when flying straight and level.

 

[randomTOTEN]

3 minutes ago, GGTharos said:

I'll try to put together a track or two for ED along with suggested solution

 

[GGTharos]

^^^ Ok perfect!  Thanks

Edit: I see it was reported and locked a long time ago.

Edited November 11, 2021 by GGTharos

[Mikaa]

18 minutes ago, GGTharos said:

^^^^

Thank you for the detailed reply.  As I have no experience with such systems IRL it's great to get this type of answer, it is instructive and useful. 

I don't know if I'll have time but I'll try to put together a track or two for ED along with suggested solution (constrain the drift - not sure if they even model tumbling in this case) and add self-righting ability.  I am assuming self-righting is due to gravity when flying straight and level.

 

Much appreciated. I just dug up the TO 1F-5E-1 AFM, and had a look at the instrumentation section. The description of the ARU-20A (which I believe is the correct AI for our version aircraft) specifically makes mention of the Fast Erect use and gyro stabilization.

TO F-5E-1 pg. 1-73 (Emphasis mine) - "The attitude sphere is stabilized by the displacement gyro (two-gyro platform) powered by the left ac bus and the dc bus. The AHRS rate gyro balances electrical inputs to the displacement gyros so that the attitude sphere maintains position through all aircraft maneuvering. The AI can be tumbled by power interruptions which cause an OFF flag to appear in the lower left of the indicator face. If power failure occurs in any flight condition other than straight and level, the AI may erect to a false vertical when power is returned. The FAST-ERECT switch on the instrument panel next to the AI is provided to expedite gyro erection."

My understanding of the above paragraph is that the FAST-ERECT switch is utilized to recover from power loss based gyro precession (from the gyro spooling down with no power applied to the system). 

Secondly, there is no mention that the FAST-ERECT switch has any relation to the HSI, which currently falsely cages to North when pressed (it shouldn't). 

Finally, I found some relevant information regarding the standby attitude indicator - There's a very specific warning after paragraphs for the ARU-32/A or ARU-42/A-1 standby AI models: - "The indicator may precess following sustained acceleration or deceleration periods and may tumble during maneuvering flight near the vertical." 

That's not terriblly interesting or unexpected for a standby, but what is of interest is the fact that the primary Attitude Indicator ARU-20/A has no such warnings associated with it. I think ED may have accidentally incorrectly interpreted these warnings in the AFM as to apply to all the AI's in the cockpit, not just the standby.  

Edit: just saw the replies. I'd actually be happy if ED fixed the primary to not drift at all unless power is lost, and leave the standby as-is - clearly the AFM makes mention of the standby being susceptible to drift. 

Edited November 11, 2021 by Mikaa

[some1]

The attitude indicator is botched in a way that causes it to show progressive bank if you try to follow its directions. It will right itself after you switch back to flying using the horizon outside. That obviously makes flying in the clouds challenging.

Considering that the fast erect function is not implemented correctly, and that the device shows the same error as standby ADI that is a less complex and completely independent instrument, I doubt anything in DCS F-5 artificial horizon is simulated correctly. As reported here https://forums.eagle.ru/topic/267104-horizon-fast-erect-no-im-not-selling-pills

 

Edited November 11, 2021 by some1

[Baltic Pirate]

On 11/10/2021 at 10:56 AM, Raviar said:

F-5 IS NOT ALLL WEATHER AC, in real life the pilots (the command pilots) will get the weather report prior to flight for whole squadron, even if it is going to be a war the F-5 CAN NOT fly in low visibility weather and they dont, if Pilot do thats a big mistake which easily end up to mishap, 7 to 10 nmi visibility is required ( I cant remember the exact number)

The radar is completely useless they dont even turn it on in real life!

 

Sorry but no. F5 even in DCS has a full IFR panel. It has all the equipment to fly NPA and I do that with it regularly. Now it does lack ILS receiver, something many if not most of this particular -E models came with right off the production line.

How do you think Norwegians, Dutch, Swiss to name a few flew it during the European winters for decades? Only on sunny days?  I've got a news flash. These aircraft were used in all weather conditions. F5's regularly flew and still fly PAR approaches. PAR is a precision approach procedure even down to 100ft ceilings and 1/4 mile visibility. Granted it is not a CATIIIB autoland with 75m RVR and no DH, but it is pretty good for a manually flown aircraft with no HUD.

[=475FG= Dawger]

3 hours ago, Baltic Pirate said:

Sorry but no. F5 even in DCS has a full IFR panel. It has all the equipment to fly NPA and I do that with it regularly. Now it does lack ILS receiver, something many if not most of this particular -E models came with right off the production line.

How do you think Norwegians, Dutch, Swiss to name a few flew it during the European winters for decades? Only on sunny days?  I've got a news flash. These aircraft were used in all weather conditions. F5's regularly flew and still fly PAR approaches. PAR is a precision approach procedure even down to 100ft ceilings and 1/4 mile visibility. Granted it is not a CATIIIB autoland with 75m RVR and no DH, but it is pretty good for a manually flown aircraft with no HUD.

All that is required for a PAR is enough instruments to keep the aircraft upright (A lot less is required to do that than a full panel) and a functioning radio. You can do PAR's in a motorized parachute if you have a radio.

[Baltic Pirate]

25 minutes ago, =475FG= Dawger said:

 You can do PAR's in a motorized parachute if you have a radio.

LOL. Yeah pretty much.

[=475FG= Dawger]

6 minutes ago, Baltic Pirate said:

LOL. Yeah pretty much.

But I do agree with you on the F-5 all weather capability. ILS is not needed in DCS to land in very low weather. You can design TACAN approaches that will get you to below CAT I ILS minimums pretty easily at almost any airport provided you know how to setup a TACAN on the ground in the first place.

[randomTOTEN]

35 minutes ago, =475FG= Dawger said:

You can design TACAN approaches that will get you to below CAT I ILS minimums pretty easily

You don't need to. Non-Precision mins are pretty close anyways, and both can be much lower than the weather minimum for many combat tasks. We're talking 400ft vs 200ft, there's no need to bust mins.

I don't get the combat community's obsession with 0/0 landings. Everybody was all mad at the Hornet because it didn't have ICLS when it first came out, but it had TACAN. On the carrier the TACAN takes you down to 360ft, vs the ICLS 260ft. Everybody was all up in arms about that for months....

...there's no need to try and take a TACAN approach to 200ft.

Edited November 12, 2021 by randomTOTEN

[Cab]

On 11/11/2021 at 3:36 AM, Raviar said:

give you an example, in Iran Iraq war the iranian F-5 pilots never turn the radar on even in A/A engagement (I am not going to the detail about who am I, its just pure fact), they dont use F-5 in IFR condition its suicide! the F-5 which ED delivered is very similar to the IIAF-IRIAF F-5E, some avionics are different but fairly close.

The upgraded F-5 is another story! btw its DCS use what ever its suit you guys, have fun

 

That's sounds more like a limitation of the pilots' training and qualifications than it does of the F-5's IFR equipment. Maybe a case of fair weather pilots flying all weather jets.

[=475FG= Dawger]

2 hours ago, randomTOTEN said:

You don't need to. Non-Precision mins are pretty close anyways, and both can be much lower than the weather minimum for many combat tasks. We're talking 400ft vs 200ft, there's no need to bust mins.

I don't get the combat community's obsession with 0/0 landings. Everybody was all mad at the Hornet because it didn't have ICLS when it first came out, but it had TACAN. On the carrier the TACAN takes you down to 360ft, vs the ICLS 260ft. Everybody was all up in arms about that for months....

...there's no need to try and take a TACAN approach to 200ft.

 

Actually, out in  the real world, the TERPS minimum Height above Touchdown (HAT) for a VOR/DME (TACAN) is 250 feet. It is exceedingly rare to find an actual approach published to those minima but the rules do allow for it. And that 250 foot restriction is because its real life and you can actually die.

You can't actually die in DCS and the weather in DCS is set in stone, literally. Unlike the real world, where the weather changes constantly and can vary greatly by the minute, in DCS it is what it is reported to be and will remain that way. And the obstacles on the ground do not change, also unlike the real world where someone can put up a 100 foot derrick in the middle of the night on the approach path without anyone knowing it.

So if one wants to fly TACAN approaches in 100 foot ceilings, more power to them.

As for the combat realism, there are plenty of early morning weather conditions where one would launch below landing minimums with the expectation that the weather would lift prior to return but may not. In that case you get a PAR or try the TACAN approach.

In any case, visibility is the much more critical minimum for a successful approach. 500 foot overcast with 1/4 mile visibility is much more difficult than 300 ft overcast and 5 miles visibility.

Again, in the real world, reported ceiling below minimums does not restrict you from initiating an approach (for commercial operations) while visibility reported below minimums requires you to refrain from proceeding past the initial approach fix.

 

Of course, military operations have no such restrictions other than their internal regulations, which get tossed out the window in war time.

1 hour ago, Cab said:

That's sounds more like a limitation of the pilots' training and qualifications than it does of the F-5's IFR equipment. Maybe a case of fair weather pilots flying all weather jets.

The F-5 isn't an all weather jet even with a magnificent suite of avionics. It can't deliver ANY ordnance that works in crap weather.

[randomTOTEN]

36 minutes ago, =475FG= Dawger said:

So if one wants to fly TACAN approaches in 100 foot ceilings, more power to them.

True, but the reality is the overwhelming majority don't even know what a TACAN approach is, don't know the concept of an MDA, and think it's impossible to fly an F-5 in the clouds. They're not making a choice to disregard minimums and try their luck. They set 0/0 and try to crash into the AB, when that doesn't work they proclaim the airplane day only.

Intersting remark about TERPS. It looks like you're right about the 250ft. My assumption was 300-1, and the last nonprec I was working with was KLSV and 400-1 3/8 so that's what I used for a comparison.

Edited November 12, 2021 by randomTOTEN

[Cab]

5 hours ago, =475FG= Dawger said:

The F-5 isn't an all weather jet even with a magnificent suite of avionics. It can't deliver ANY ordnance that works in crap weather.

5 hours ago, =475FG= Dawger said:

The F-5 isn't an all weather jet even with a magnificent suite of avionics. It can't deliver ANY ordnance that works in crap weather.

First of all, the OP was talking about flying the jet in IMC conditions, and for that it is well equipped for it's day.

Second, I don't think delivering ordnance in crap weather is the requirement to be considered "all-weather". An airplane that could do that would be an all-weather bomber like the A-6 or F-111. The F-5 on the other is, I believe, considered an all-weather fighter for it's day. It could operated and navigate in IMC conditions, getting to it's target area safely just like other fighters of it's day. It also has radar to find and target aircraft at night an in IMC and could engage enemy aircraft without seeing them. Although I concede that would be very close and sporty. Flying it in all-weather conditions only requires adequate training and a stout heart.

If someone has documentation that says otherwise, I'd be interested to see it.  

[Chuck_Henry]

This discussion went further off the tracks and way more into the weeds than it needed to.

Fact is, the F-5 we have *in DCS* is pretty garbage for flying in IMC, and this is almost entirely because of the attitude gyro discrepancies that have been reported and subsequently ignored by ED for years. 

Lack of ILS is realistic and is still the case for the F-5Ns flown by adversary squadrons such as VFC-13 and VMFT-401. On the off-chance the weather is really bad enough at Fallon or Yuma to require a precision approach, they have very well-trained controllers skilled in letting them down via PARs. Lack of radar altimeter gives some pucker factor, but the F-5s are still allowed to fly in IMC under a CNAF waiver.

That said, this is not an "all-weather" jet by any modern standard and to claim such is ludicrous. Employing it as such is how you die, whether by CFIT or getting shot down.

[Cab]

On 11/12/2021 at 7:51 PM, Chuck_Henry said:

this is not an "all-weather" jet by any modern standard and to claim such is ludicrous. 

Nobody said anything about a "modern standard". The fact is it can operate during day, night, and IMC conditions. It does have limitations because of it's antiquated avionics, but it can be safely in all weather. If you crash it, assuming all the avionics are working properly, then you made a mistake, a bad decision, or the jet was deployed beyond it's capabilities. But that doesn't mean it's not an all weather aircraft.

[TLTeo]

"All weather" here is badly defined. Can it navigate from point A to point B and possibly intercept another aircraft? Sure. Can it hit some ground target with decent precision with no visibility? Obviously not.

[IvanK]

The AI precession is still an issue in the F5E. Needs to be fixed. It makes accurate instrument flight unnecessarily hard ! Real world AI's just don't precess like this !

[Bucic]

8 minutes ago, IvanK said:

The AI precession is still an issue in the F5E. Needs to be fixed. It makes accurate instrument flight unnecessarily hard ! Real world AI's just don't precess like this !

After what we saw happened to the TACAN indicator from Q4 2023 maybe it's better ED just doesn't touch the artificial horizon and leave it be...

[IvanK]

Looks like the latest update has fixed the AI precession issue.

[Bucic]

On 12/7/2024 at 2:36 AM, IvanK said:

Looks like the latest update has fixed the AI precession issue.

That's contrary to what I have reported recently (track included)

https://forum.dcs.world/topic/259460-rapid-drift-of-attitude-indicator/page/3/#findComment-5563920

 

[IvanK]

Hmm watched your vid. The bank error ends up being in the order of 1-2 degrees after flying a similar bank angle for 5 or so minutes.

In my case in regular IF flying using standard bank angles and general manoeuvering I see little to no error. A 1 degree bank error after 5 mins at 2 degrees AOB I would consider insignificant.... and in some installations to be expected. In fact their is references to this in the old USAF TP55 Instrument Flying manuals using MM1 and MM2 Attitude indicators. Not sure if the F5E is an MM2 or MM3 though.

When I was taught IF it was standard caveat not to use low sustained bank angles as that might be below the threshold for the auto erection circuits to cut in. Early generation remote AI systems talk about avoiding sustained turns with turn rates <40deg min (so 0.67 deg sec) for this very reason.

Previously we were getting bank errors as much as 10 degrees.... anyway thats my 2cents worth

Edited December 10, 2024 by IvanK

[Bucic]

@IvanK

 I'm not sure which video/track you watched but the most recent ones focused on 15+ deg bank turns. Also, please make sure you're not noting as the final error the value after an opposite "corrective" turn that in one of my tracks/videos cancelled out the previous error (from the opposite turn). I also noted in my report that trigger values for auto levelling may be exactly what's at play here.

All in all In have delivered tracks with 5-7 degree error after deep turns or aerobatics in the most recent DCS build.

[IvanK]

This video you linked to: