No complaints, technical question!! Why does the F16 trim for Pitch but not for roll?

[Lee1]

On 7/13/2022 at 9:38 AM, Machalot said:

What do you mean yours trims in roll? Do you have a different Viper model than everyone else? 

 

I mean if I drop a 2000 pounder from the left wing and my Viper rolls right as a result of this I can trim it out with roll trim

 

[Machalot]

24 minutes ago, Lee1 said:

I mean if I drop a 2000 pounder from the left wing and my Viper rolls right as a result of this I can trim it out with roll trim

 

The OP is clearly asking about automatic roll trim. 

[skywalker22]

How about of Roll trim on more modern F-16s, is there any version which possess it?

[ASAP]

I don’t know, this is purely a guess… but I’d guess cost, complexity, and the “worth it” factor came into play when they were designing the jet. Of the trimming a pilot does 95% is elevator trim, 4% is roll, and 1% is rudder. Those numbers are totally made up, but my point is they probably didn’t bother with it because it’s not that big a deal to give a few clicks of roll trim after weapons release vs elevator trim which is a near constant process especially in a jet that’s designed to be aerodynamically unstable. 

[Blaze1]

On 7/11/2022 at 8:13 PM, toilet2000 said:

Since the question was somewhat already answered, I'll use some of my limited control theory experience to speculate on the reasons. I would think the same as you, but I'll add more possible details:

Auto-trim means the control loop requires at least an integrator to get rid of the steady-state error (so that it actually auto-trims). Those are always used cautiously, as they can definitely deteriorate the transient response. From my limited research on the subject, the roll FLCS doesn't contain any feedback loop. This is generally done because as someone else said, the roll is stable but adding a control loop could lead to instability, especially so if for some reason there's an unforeseen delay in the control loop (phase margin) or a big perturbation (gain margin). An integrator makes it simply much more likely to happen as it removes 90 degrees of phase meaning you're already half-way there.

All this to say, it would have drastically complexified the roll control system and made it possibly unstable, whereas it was stable to begin with (which was not the case on the pitch axis). This adds a ton of complexity to the stability analysis, especially given complex maneuvering as is the case for military fighters, and the F-16 was one of the first aircraft with FBW controls. There's generally a saying in control theory that anything straying away from the natural response of a system should not be done except if needed. IMHO this is the case for the roll-axis in the F-16.

 

Nice post toilet

The following is a quote from an FCS developer/engineer (maybe even for the Viper).  The mention of '7x' was a comparison of the Viper's FCS, to that of the Dassault Falcon 7x which does have an integrator on the roll channel:

"It is just a classical closed loop system with gains designed for a desired roll mode. The one bit of an oddity with the F-16 is the lag filter on the roll command path. It uses one time constant for commands "coming in" and a different (faster) one for commands "going out" (stick releases or reversals). That implementation is definitely a player in stopping rolls quickly (i.e. containing bank angle overshoots.) (Apparently, the analog hardware implementation for that bit of logic was a real challenge.)
As to the 7x, that is a bit curious. It may be that it's more of an outer-loop kind of function, more like an autopilot. Or, maybe they are just taking the hit on getting a classical response to pilot inputs. Hard to tell."

So as per your assertion, an integrator in the roll channel can cause response issues.  The Eurofighter Typhoon and Dassault Rafale both have three axis auto-trim, but of course they're much later designs than the Viper's.

[DummyCatz]

6 hours ago, Blaze1 said:

The one bit of an oddity with the F-16 is the lag filter on the roll command path. It uses one time constant for commands "coming in" and a different (faster) one for commands "going out" (stick releases or reversals). That implementation is definitely a player in stopping rolls quickly (i.e. containing bank angle overshoots.) (Apparently, the analog hardware implementation for that bit of logic was a real challenge.)

Great findings. This is also called the Slow in Fast Out circuit by F16 FLCS engineers.

[Blaze1]

48 minutes ago, DummyCatz said:

Great findings. This is also called the Slow in Fast Out circuit by F16 FLCS engineers.

I didn't realise it had been name as such DummyCatz.  Do you have any more information/links about it?

[DummyCatz]

1 hour ago, Blaze1 said:

I didn't realise it had been name as such DummyCatz.  Do you have any more information/links about it?

Just by chatting with some of the guys, they refer me to NASA TP-1538 with logical diagrams. Just beware that this is a pretty old graph that may not be applicable to current-day FLCS. It's just used to illustrate the Slow In Fast Out circuit. I'm trying to explain it without touching the unintelligible side, such as lag filters and washout filters. 

 

Dial it into Matlab and make a comparison with and without such circuit:

 

As you can see from the time response, the one with Slow In Fast Out (SIFO) circuit increases roll rate command a lot slower than the one without. It almost took it 2 seconds for the SIFO one to reach the set command value. But they both took only 0.5 sec to decrease from the commanded value to 0.

I think your quote is a great explanation.

Edited July 30, 2022 by DummyCatz

[Blaze1]

14 minutes ago, DummyCatz said:

Just by chatting with some of the guys, they refer me to NASA TP-1538 with logical diagrams.

Cheers DummyCatz, I appreciate that.

[Lace]

Just curious really, I know it trims in pitch for 1G, why doesn't it trim for roll?  Surely with zero roll demand from the pilot (via the stick), the aircraft FLCS should be able to provide any required roll trim to give a zero-roll rate condition.  Is there a deliberate reason they* didn't code this in?

* I don't mean ED, I mean General Dynamics/Lockheed Martin.  I'm referring to the real-life Viper.

[MAXsenna]

There was just another thread about this were it was explained in great detail.
Tl/Dr the Viper doesn't do it.

Cheers!

Sent from my MAR-LX1A using Tapatalk

[Lace]

7 minutes ago, MAXsenna said:

There was just another thread about this were it was explained in great detail.
Tl/Dr the Viper doesn't do it.

Cheers!

Sent from my MAR-LX1A using Tapatalk
 

I know it doesn't - I was wondering why.  Do you have a link to the other thread by any chance?

Ok, - found it:

 

Edited October 17, 2022 by Lace
link

[MAXsenna]

I know it doesn't - I was wondering why.  Do you have a link to the other thread by any chance?
Ok, - found it:
 

Yeah, sorry. Would have posted it, but I was on my cell and in a hurry.
Glad you found it!

Sent from my MAR-LX1A using Tapatalk

[GafferDCS]

On 10/17/2022 at 8:14 AM, Lace said:

Just curious really, I know it trims in pitch for 1G, why doesn't it trim for roll?  Surely with zero roll demand from the pilot (via the stick), the aircraft FLCS should be able to provide any required roll trim to give a zero-roll rate condition.  Is there a deliberate reason they* didn't code this in?

* I don't mean ED, I mean General Dynamics/Lockheed Martin.  I'm referring to the real-life Viper.

I’m going to hazard a guess here, they don’t have the FLCS auto trim for roll because they want the pilot to have some “feel” when they are operating with an asymmetric load. So it’s designed to make the pilot trim it out. 

Edited November 5, 2022 by GafferDCS

[LowGlow]

Probably a bit OT, but why do I have to manually trim the pitch sometimes, e.g. after emergency jettison a heavy loadout?

[Deano87]

On 11/17/2022 at 6:50 PM, LowGlow said:

Probably a bit OT, but why do I have to manually trim the pitch sometimes, e.g. after emergency jettison a heavy loadout?

I have no idea, you shouldn't have to. I can't remember the last time I trimmed in pitch in the F-16. Why exactly do you feel like you have to trim? 

[Aquorys]

On 11/5/2022 at 12:29 PM, GafferDCS said:

I’m going to hazard a guess here, they don’t have the FLCS auto trim for roll because they want the pilot to have some “feel” when they are operating with an asymmetric load. So it’s designed to make the pilot trim it out. 

 

My somewhat educated guess would be that the answer to the question why the F-16 is not roll stabilized is simply that the F-16 is one of the earliest applications of a fly-by-wire flight control system, and it would have been too complicated to integrate full 3-axis stabilization. For example, even today, integrating new weapons for the Typhoon (aka Eurofighter), which is fully 3-axis stabilized and follows a "carefree handling" philosophy, requires making changes to the flight control software, simply due to the fact that those weapons change the aerodynamics and thereby the behavior of the aircraft in flight. Today, where we have access to more powerful avionics systems, better test equipment, better simulations, better software development environments, and where engineers have a lot more experience with fly-by-wire systems and the software required for it, this is a much more manageable effort than it was in the 1970s.
Back then, overcomplicating the F-16's flight controls would probably have impeded the integration of weapons, stores and other equipment, and as a result, it would possibly have made the F-16 project unfeasible as a whole.

Edited November 24, 2022 by Aquorys

[LowGlow]

Am 21.11.2022 um 02:40 schrieb Deano87:

I have no idea, you shouldn't have to. I can't remember the last time I trimmed in pitch in the F-16. Why exactly do you feel like you have to trim? 

It happened to me bunch of times in DCS after jettison of bombs and tanks that I had to manually trim pitch, somehow the Viper kept nose up, but I couldn't redproduce it right now.

Another thing I just realized, when you fly upside down in DCS, then pitch trim does not work. Is that correct behaviour?

[Deano87]

2 hours ago, LowGlow said:

Another thing I just realized, when you fly upside down in DCS, then pitch trim does not work. Is that correct behaviour?

Huh?

I think you're misunderstanding what the F-16 flight control system does. The F-16 FLCS tries to maintain 1.0G of load on the aircraft when you put no pitch inputs into the stick, NOT going in a dead straight line. This means it should maintain about a level flight when flying level, but If you roll to 90 degrees of bank the FLCS will load the aircraft up to 1G. The same happens when inverted, the FLCS will pull to maintain 1G of load on the aircraft so the nose will be dropping quite quickly, The same as a -2G pushover actually.

All of the above is why you having to re-trim in pitch after jettisoning makes no sense to me.

[LowGlow]

OK, I reproduced the trim issue: I tried the quick start mission "Nevada: HARM on the Range - Russian SAMs". After having shot the 4 HARMs auto trim failed. Nose would always pitch up.

[Deano87]

1 minute ago, LowGlow said:

OK, I reproduced the trim issue: I tried the quick start mission "Nevada: HARM on the Range - Russian SAMs". After having shot the 4 HARMs auto trim failed. Nose would always pitch up.

If you can make it happen repeatedly then save a track and post it here. I haven't seen anything like this. Please check that you don't have any button mappings interfering with pitch trim. Also when it happens please check the Flight Controls panel in the rear left hand side of the cockpit behind the throttle and make sure that things like the Alternate flap extension is set to "Norm" and the Digital FLCS backup is set to OFF. If either of those are sett wrong then you may have a weird binding set somewhere messing with those too.

[LowGlow]

Oh, are these switches really implemented in DCS?

I remember there was once a mishap with a real Viper where the pilot accidently flipped the wrong switch at low altitude and ejected, as the jet no longer behaved as expected by him. And it turned out later it behaved as expected, just that he cut off the FLCS.