Bug? A-10C still impossible to trim for level flight

[Fred00]

I wrote about this in a thread in 2020 and coming back to the A-10C it's disappointing to see that this is still an issue.

So in short:
The issue is that the A-10C seems impossible to trim out, most notably in the roll axis. Once it seems that you have managed to trim it out it soon starts to lean one way or the other. Interestingly enough I found out that it leans one way for a while, and then suddenly it starts leaning the other way, without any input other than pitch trim (to keep it somewhat level). So is there some kind of wind effect that affects the trim dynamically or what is causing this? I have attached a track file and this effect is best seen in the last five minutes or so, where the plane is tilting left and then suddenly starts straightening up without any roll trim applied.

 

Trim.trk

Edited December 27, 2023 by Fred00

[King39]

I agree with the difficulty in trimmng the Hog but you might want engaging the AP (LAAP) once you've trimmed it out as best you can. That seems to settle the aircraft down for you too disengage it and fly hands-off but will return to the lazy drunkin turn rate once you start any maneuvers.

[ASAP]

whats the turn/slip indicator showing? Rudder trim is oft forgotten. some load outs could create a lot of yaw moment. If you try to fix it with roll trim you'll get a very unstable aircraft

[ASAP]

What is your loadout as well. This could have an impact on the flight characteristics of the aircraft. 

[Fred00]

ASAP: Sorry for the late reply. I checked the slip indocator, but it's dead center, so no rudder issues from what it seems.

Don't know about the loadout, will check later. This is the instant action mission "Takeoff" for Syria. However, the problem doesn't seem to be loadout since the behavior is present in every mission.

I have also checked my control bindings and there are no conflicts or other issues there.

You can use the Syria "Takeoff" mission to test the behavior. It shouldn't matter which mission, but to be on the safe side and rule out any other differences. After takeoff, trim out the plane and then wait and see what happens. It might fly straight a little while and then suddenly start to roll. Just keep going and it will probably start to roll the other way after a while.

[Avio]

I totally can attest and agree to this roll trim problem, which i had also brought up before years past, more than once.

Just totally impossible to trim wings level, like each trim input is too coarse or something, and there is no setting to adjust the trim amount input with each button press. It would definitely be helpful if ED could refine each trim input a lot more.

On a busy mission looking out, the self rolling off commanded attitude is a serious minus from an otherwise great product.

 

[davidrbarnette]

I know this isn’t the answer being sought, but a couple things:

1) I don’t seem to have trouble trimming the A-10 in DCS; I haven’t flown one in real life, but the trim behavior feels believable to me.

2) Real, non fly-by-wire aircraft are impossible to trim perfectly; due to turbulence, wind, constantly changing fuel states, etc, real planes require constant input from the pilot to fly straight and level. The inputs will be very subtle if trimmed well, but if you were to let go of the yoke or stick on most days without an autopilot engaged, the plane will be banking, climbing, or diving in short order.

[QuiGon]

I'm no authority on this, but I personally don't think this is a bug. I'm no real pilot either, but just like @davidrbarnette said, I can't imagine that it is possibly IRL to trim out a no FBW aircraft 100% perfectly due to how dynamic everything is, starting with constantly changing wind and turbulences.

Edited December 29, 2023 by QuiGon

[ASAP]

I'm a real pilot. Pilots constantly have to adjust elevator trim anytime there is a change to airspeed. Rudder and roll trim on the other hand rarely have to get messed with. Most small aircraft don't have roll trim at all because it's not something that dynamically changes in flight. As long as you don't change the symmetry (or asymmetry) of the stores/fuel in the wings there shouldn't be a need to roll trim continuously in flight. 

[davidrbarnette]

2 hours ago, ASAP said:

I'm a real pilot. Pilots constantly have to adjust elevator trim anytime there is a change to airspeed. Rudder and roll trim on the other hand rarely have to get messed with. Most small aircraft don't have roll trim at all because it's not something that dynamically changes in flight. As long as you don't change the symmetry (or asymmetry) of the stores/fuel in the wings there shouldn't be a need to roll trim continuously in flight. 

I’m also a real pilot. I generally agree with you, however one of the types I’ve personally owned required constant trim and control input in the roll axis due to tanks feeding unevenly during cruise flight. Also, any change in course, regardless of weight imbalances, likely require changes to the control surfaces in the roll axis due to things like wind direction and crab angle, not to mention disturbances from turbulence. 

In most planes I’ve flown, if you take your hands off the controls, it’s just a matter of time until you’re in a bank or turn. It might take a minute or two, but without the pilot intervening, it’ll happen. DCS simulates this in all of the airframes quite well. I’m trying to think of a flight model in DCS that you can just fly along straight and level, and there isn’t one (unless you’re fine with some big altitude and heading excursions).

Also, most small planes don’t have roll trim because it would be an added complexity, more expense, and the pilot is expected to do the flying. Manually adjusted trim tabs on the ailerons are a more common option, but obviously aren’t adjustable in flight. A good example of a small plane with actual aileron trim is the Cirrus SR22.

Again, any non-fly-by-wire aircraft will require the pilot to use control inputs, whether trim or actual stick/yoke input to keep the plane straight and level. Most planes won’t just fly along on their own unless an autopilot or stability augmentation system is engaged.

Edited December 29, 2023 by davidrbarnette

[jaylw314]

1 hour ago, davidrbarnette said:

I’m also a real pilot. I generally agree with you, however one of the types I’ve personally owned required constant trim and control input in the roll axis due to tanks feeding unevenly during cruise flight. Also, any change in course, regardless of weight imbalances, likely require changes to the control surfaces in the roll axis due to things like wind direction and crab angle, not to mention disturbances from turbulence. 

In most planes I’ve flown, if you take your hands off the controls, it’s just a matter of time until you’re in a bank or turn. It might take a minute or two, but without the pilot intervening, it’ll happen. DCS simulates this in all of the airframes quite well. I’m trying to think of a flight model in DCS that you can just fly along straight and level, and there isn’t one (unless you’re fine with some big altitude and heading excursions).

Also, most small planes don’t have roll trim because it would be an added complexity, more expense, and the pilot is expected to do the flying. Manually adjusted trim tabs on the ailerons are a more common option, but obviously aren’t adjustable in flight. A good example of a small plane with actual aileron trim is the Cirrus SR22.

Again, any non-fly-by-wire aircraft will require the pilot to use control inputs, whether trim or actual stick/yoke input to keep the plane straight and level. Most planes won’t just fly along on their own unless an autopilot or stability augmentation system is engaged.

 

Pedantic alert  If your plane requires different roll (or any) trim with changing relative wind, there is something wrong with your plane!

Most small general aviation planes have fairly strong roll stability, so the need for roll trim is pretty minimal.  I'm not sure about the A-10C roll stability and couldn't really find any descriptive data. The weird dihedral suggests good stability, the engines' weight on top suggest low stability, so... 

FWIW I fly a Mooney, which also alternates wing tanks and has a similar mission to the SR22, but has no available roll trim (not sure about rudder trim in newer models).  It has a strong dihedral, though, and I've never felt it's needed roll trim.

Edited December 29, 2023 by jaylw314

[davidrbarnette]

2 hours ago, jaylw314 said:

Pedantic alert  If your plane requires different roll (or any) trim with changing relative wind, there is something wrong with your plane!

Most small general aviation planes have fairly strong roll stability, so the need for roll trim is pretty minimal.  I'm not sure about the A-10C roll stability and couldn't really find any descriptive data. The weird dihedral suggests good stability, the engines' weight on top suggest low stability, so... 

FWIW I fly a Mooney, which also alternates wing tanks and has a similar mission to the SR22, but has no available roll trim (not sure about rudder trim in newer models).  It has a strong dihedral, though, and I've never felt it's needed roll trim.

 

Never flown a Mooney, but I’d like to! Good looking planes, and fast, but I’m getting off topic.

Anyway, I’ll maintain that even the most stable aircraft needs constant input. If you let go of your Mooney’s controls, it’ll just fly along on the same heading and level with no input for an extended period of time? 
 

Just to clarify, I don’t think any small plane NEEDS roll trim, but that’s because the pilot is making constant control inputs to hold heading and keep the wings level. Granted, if you’re trimmed well, they’re small inputs you make with your fingertips, but inputs nonetheless. I think some DCS pilots think you just dial in the trim and go to sleep.

All that to say that the A-10 flight model feels pretty legit to me.

Edited December 29, 2023 by davidrbarnette

[ASAP]

7 hours ago, davidrbarnette said:

In most planes I’ve flown, if you take your hands off the controls, it’s just a matter of time until you’re in a bank or turn. It might take a minute or two, but without the pilot intervening, it’ll happen. DCS simulates this in all of the airframes quite well. I’m trying to think of a flight model in DCS that you can just fly along straight and level, and there isn’t one (unless you’re fine with some big altitude and heading excursions).

The wing tanks should feed equally from both sides unless you have an issue and the pilot isn't monitoring it or correcting it. Even then, the wing tanks are very close to the center line of the aircraft so there isn't a lot of asymmetry in fuel distribution. I'm not saying that you shouldn't have to adjust roll trim at all. but it is very easy to get it pretty much set for your given conditions and until you do something to unbalance the jet (i.e. drop a weapon) there isn't any dramatic need need to adjust based on changing flight conditions. a click or two, maybe. 

Obviously it's not the same as an autopilot and you can't just fall asleep at the flight controls. But you should certainly be able to look down at your knee

board for a couple of seconds without the aircraft making a significant excursion.

There's no reason for run away roll trim issues though like the OP is describing. Relative wind wouldn't be a big deal, the jet can crab just fine. Jet aircraft don't have the same forces from a prop acting on the aircraft that have to be constantly adjusted for out. 

 

5 hours ago, jaylw314 said:

Most small general aviation planes have fairly strong roll stability, so the need for roll trim is pretty minimal.  I'm not sure about the A-10C roll stability and couldn't really find any descriptive data. The weird dihedral suggests good stability, the engines' weight on top suggest low stability, so... 

Everything I've heard is that the is pretty stable and easy to fly.  It does have a tendency to dutch roll, but the Yaw and Pitch SAS compensate for that on their own. 

[jaylw314]

4 hours ago, davidrbarnette said:

Never flown a Mooney, but I’d like to! Good looking planes, and fast, but I’m getting off topic.

Anyway, I’ll maintain that even the most stable aircraft needs constant input. If you let go of your Mooney’s controls, it’ll just fly along on the same heading and level with no input for an extended period of time? 

Nah, there's always some asymmetry, so if you let go, the plane will gradually bank about 5-10 deg to either side, but it will just stay there and be stable.  It's not going to keep on banking more and more.  Sure, roll trim could fix that, but just resting your hand on one side of the yoke is enough to fix it

 

Edited December 30, 2023 by jaylw314

[Avio]

A well trimmed aircraft stays straight and level for a minute or two, but this A10C can do that only a second or two, it seems ! The DCS F15E is able to level trim very well, with each trim input press only inputting a small amount, whereas the A10C per trim input is just too much. Feels like half the time fighting the aircraft rather than the enemy, when looking outside. Even flying straight with just the roll trim seems more economical (and necessary every 2 seconds) than having to move the stick.

Granted, those with a long stick extension probably helps somewhat, but that is more of an exception.

 

[Snoopy]

Having flown the real world sim at Moody and Whiteman and talked to multiple pilots over the years even the real jet requires trim often.

Edited December 30, 2023 by Snoopy

[II.JG1_Vonrd]

19 hours ago, Avio said:

the A10C per trim input is just too much

Another RL Private Pilot here and I agree with the other comments that it's not possible to trim for unlimited hands off flight. The issue with the 'Hog is the over sensitive roll control. There must be an adjustment somewhere in a lua to tone it down. Anyone know where this might be?

[Fred00]

Good discussion guys. However, the fact that the trim is too coarse/insensitive is only part of the problem. An even bigger problem is (like the track shows) that the plane can stay trimmed for a short while and then suddenly start rolling in either direction, and even change direction of roll all of a sudden with out any input. This fact makes the A-10C in DCS quite annoying to fly.

It's a bit strange that the developers can't chime in on this. Maybe there's some kind of purpose behind this behaviour and it would be nice to hear about it (if that's the case).

[Scoop]

Great discussion, but don't forget this is a twin engine plane and the throttles need to be balanced for level flight otherwise thrust will be uneven and a roll will be induced one way or another. If you use a thrustmaster warthog setup with the throttle link pin removed you will see that each throttle, (although you might think you have them together), are actually producing slightly different amounts of thrust, (not much but enough), and so have to be balanced by one being in as lightly further forward or aft position of each other. A good way to see this in action is to conduct a long taxi and have one throttles slightly different to the other and not use any rudder. The plane will go left or right depending on the thrust applied or throttle positions and you can control the direction of taxi by increasing or decreasing the left or right throttle without rudder. Same happens in the air.

[ASAP]

1 hour ago, Scoop said:

Great discussion, but don't forget this is a twin engine plane and the throttles need to be balanced for level flight otherwise thrust will be uneven and a roll will be induced one way or another. If you use a thrustmaster warthog setup with the throttle link pin removed you will see that each throttle, (although you might think you have them together), are actually producing slightly different amounts of thrust, (not much but enough), and so have to be balanced by one being in as lightly further forward or aft position of each other. A good way to see this in action is to conduct a long taxi and have one throttles slightly different to the other and not use any rudder. The plane will go left or right depending on the thrust applied or throttle positions and you can control the direction of taxi by increasing or decreasing the left or right throttle without rudder. Same happens in the air.

Asymmetric thrust would cause a big yaw moment and result in an uncoordinated jet which is slipping or skidding. that will also create a smaller a rolling moment, but doesn't sound like what the OP is describing. Based on what the OP is describing, it sounds to me like it's not an A-10 related issue. It sounds a lot more like some kind of a control setup issue with run away trim. There is no aerodynamic reason that the A-10 would go from rolling strongly to the right, and then switch to a strong left hand roll and then reverse back to right. Having spent a lot of time in the A-10 sim the jet requires very little (if any) roll trim. A few clicks after takeoff to compensate for weapons loadout, and after that you never really need to touch it again. You have to trim elevator with any speed changes or as you burn down gas, but you trim the other axis very seldom. 

[Yurgon]

On 12/27/2023 at 11:08 AM, Fred00 said:

So is there some kind of wind effect that affects the trim dynamically or what is causing this?

I watched your track, but not all the way to the end (there's no way to know ahead of time how long the track is, and the random flying around didn't seem particularly noteworthy).

That region has some mountains and the mission has some wind set. I'm pretty sure that DCS simulates some kind of turbulence close to the ground, to include turbulence in and near mountains. This particular mission doesn't have any turbulence set, but that's no guarantee that DCS doesn't simulate turbulent air according to the general wind settings and the ground.

In a clean test mission over the Black Sea with no wind whatsoever and 20 miles from anything resembling a mountain and using fast forward, I did manage to observe the aircraft rolling right and then left a little. Even so I couldn't tell if maybe that was caused by the tiniest of inputs on my rudder pedals or the stick.

For what it's worth, I flew a couple of hours in the C-101EB today and that aircraft is also impossible to trim perfectly. It requires very regular tiny inputs to keep it flying within the intended parameters. So in the A-10C, while there might be an issue with roll reversing without any input, it's still relatively minor, and more importantly we do have an autopilot that helps us in the A-10C.

Barring real world flying experience, compared to other simulated aircraft the A-10C flight model feels pretty solid to me, and the fact it's not flying on rails and requires regular inputs makes it feel very much alive from my point of view.

In order to call this behavior a bug, we'd need very solid evidence that there's something funky going on, which would probably require disconnecting all analog devices, removing factors like wind, and then showing that the aircraft is still exhibiting a roll-reversal tendency without any inputs and without any notable changes in attitude (as in: constant speed, constant altitude).

As for the sensitivity of trim inputs, yeah, it would be nice if trim could be applied in finer increments. But I can't tell if the DCS A-10C is simply replicating real life behavior (which Snoopy suggests is the case) or if the real aircraft is easier to trim out well. Either way, the autopilot is a great tool to help us, unless we're changing both direction and altitude at the same time.

[ASAP]

46 minutes ago, Yurgon said:

As for the sensitivity of trim inputs, yeah, it would be nice if trim could be applied in finer increments. But I can't tell if the DCS A-10C is simply replicating real life behavior (which Snoopy suggests is the case) or if the real aircraft is easier to trim out well. Either way, the autopilot is a great tool to help us, unless we're changing both direction and altitude at the same time.

From what I've observed in the actual sim the trim is realistic in DCS

[Avio]

After having used the DCS A10C (and a bunch of others) for many years, i am convinced the problem seems very much like due to too coarse roll trim. The pane does not roll in opposite way after without further input like what some mentioned here. Rather, each trim is so coarse that it tilts and roll the plane off what could have been a well trimmed out level flight.

Autopilot is great, but in an intense ground hunt looking out and/or trying to line stuff up in TPod, that uncommanded, untrimmed setting that rolls the jet off intended attitude is just sheer pain that dilutes much of the joy of an otherwise almost real sim.

Youtube videos i seen before in cockpit shows trimming in pitch occasionally yes, but roll trim almost non existent and yet the jet flies level and nice.

As said before, the DCS F15E trims out really, really well in both axes, really fine enough trim input. Each A10C roll trim feels like about 3x that of the eagle, if i have to compare.