Flight Model feedback from ATAL

[RabidRider]

Real life controls vs CH, Warthog, with or without extension. I think there's something wrong with the FM.

 

Let me just requote the following comment:

 

If you look at this video, it's clear that the pilot is easily able to lift into a stable hover without the slightest hint of a skid forward. If you look closely, you can see that there is very little aft cyclic being applied - nothing like the momentary correction needed in DCS.

 

Aside, (it's already been mentioned) the rotor is clearly turning much faster than the visual model in DCS suggests. In DCS, with the rotor RPM gauge showing 340rpm+ as it should be for flight, the external model seems to rotate at around 150rpm at a rough guess.

 

It's a great module but Polychop do seem to have exercised a little 'poetic licence' when it comes to the flight model in particular. It makes me curious as to how the flight model is being calculated during the lift off phase and the skids interaction with the ground. Something is certainly odd about it.

 

And another thing... Balanced turns appear to be next to impossible if you believe what the slip ball is telling you. If you fly a turn in accordance with the slip ball indications, you end up with a very under ruddered turn, slipping out of the turn direction. You can see this clearly during replays. A balanced turn seems only possible if you apply far more 'in turn' rudder than the ball suggests.

 

 

Another video, min. 2:00. Again, almost no corrections needed.

 

[ame=

]

[/ame]

 

 

I saw a couple of so-called good take-offs in DCS, but they are definitely not the same as RL examples,

much less stable and smooth imo. I can't believe RL pilots would say current state is realistic.

 

Yes in some regimes she flies nice, and she's nimble like the real one, but the flight model definitely needs more work!

 

You can do it Polychop!

Edited May 13, 2016 by RabidRider

[Chelco]

I don't know if it's just me, but in the videos I can see some subtle but crisp corrections. In the last video (02:21 to 02:41) there is noticeable nose-up, nose-down flight during taxi.

 

Cheers,

[hannibal]

yea.. but pilots in the gazelle have a full setup, long shaft cyclic, torque pedals... you cant compare using short joysticks like a TH warthough or x52...

[Chelco]

yea.. but pilots in the gazelle have a full setup, long shaft cyclic, torque pedals... you cant compare using short joysticks like a TH warthough or x52...

 

Understood. But all those have nothing to do with the flight model. All I know is that Polychop recognized that the pedals input was a tad too high.

 

Cheers,

[red777]

Just little remark on the video. there's no co-pilot and no camera installed on top, so a there's a different Weight and balance situation here. Anyway, it will be nice to know what the real pilots will say about that..

[Whisper]

Just little remark on the video. there's no co-pilot and no camera installed on top, so a there's a different Weight and balance situation here. Anyway, it will be nice to know what the real pilots will say about that..

 

That's actually a very good point, the Viviane could well be a game changer. It's that heavy, the gazelle structure had to be reinforced and rotor was changed to support the added weight.

[red777]

That's actually a very good point, the Viviane could well be a game changer. It's that heavy, the gazelle structure had to be reinforced and rotor was changed to support the added weight.

 

 

Plus we all have to remember that in RL you have much more input from the aircraft (to your sense of equilibrium etc...) to just simple visual cue in a simulator like DCS (as good as it can be).

[sorcer3r]

That's actually a very good point, the Viviane could well be a game changer. It's that heavy, the gazelle structure had to be reinforced and rotor was changed to support the added weight.

 

Here is RL footage with Viviane mounted:

 

[shagrat]

Here is RL footage with Viviane mounted:

 

Actually a perfect example of a quick take off to hover, with no balancing the heli light on the skids. So at least we know you can take her off into a stable hover.

 

I had a nice conversation with Olli today and they await feedback from their experts.

A possibility is indeed, that the point of gravity, which is shifted more to the front, through the Viviane mount, compared to the perfectly balanced civilian version, has an effect here.

 

I tried today to reproduce the sliding and it definitely is a noticable issue, but I'm not 100% sure it even is related to the Flight Model, or the Flight Model alone.

 

There is ground collision modeling, ground friction, and wind coming from the DCS environment, which all effect the aircraft in realtime. All or some of these could be the root cause, or a combination/interaction between the FM and these factors.

That is actually one of the cool things about DCS EFM/AFM, as Wags explained a while ago. That the environment is not part of the modeling, and thus independent from generic tables or simple algorithms is what makes this realism possible.

On the backside we have far more interaction and potential interfaces where things can get complicated.

 

One example I notice sometimes is the ground collision modeling making a wrecked aircraft bounce and jump on the ground when it should lie still and not move... That is the realtime calculation at work and I've seen it with multiple aircraft over the years.

 

What the above video showed me, was a pilot quickly(!) getting the bird airborne into a hover, with no loitering "light on the skids", at all... So before we aim fore conclusions, let's wait for the pilot feedback. :)

[Yurgon]

There is ground collision modeling, ground friction, and wind coming from the DCS environment, which all effect the aircraft in realtime.

 

When we looked at it on your rig today, you demonstrated that the effect is noticeable even when turning while on the ground. I'm fairly certain wind does not have a direct influence on the effect.

 

What the above video showed me, was a pilot quickly(!) getting the bird airborne into a hover, with no loitering "light on the skids", at all... So before we aim fore conclusions, let's wait for the pilot feedback. :)

 

I actually think there was quite a pronounced phase of being light in the skids where one could see the strain go away from the entire skid assembly, and then right after coming airborne, the chopper actually pitched nose down a tiny bit. I think this is a perfect showcase of how it should look like in "our" Gazelle, even though the pilot in this case did not prolong the light-on-the-skids phase as we currently do when trying to replicate the effect.

 

BTW, problem persists in 1.5 Release version.

[shagrat]

When we looked at it on your rig today, you demonstrated that the effect is noticeable even when turning while on the ground. I'm fairly certain wind does not have a direct influence on the effect.

I'm not so sure... Wind "shouldn't have" an influence, but if we have a calculation bug somewhere, it could for example move the chopper, when the simulation does not "realize" the Gazelle actually is on the ground.

Imagine the collision detection results say chopper is in the air and then wind calculation adds a bit of momentum to the vector.

That's just one possible root cause that came to my mind.

Or the center of gravity with the ground detection results in a momentum that resembles a slight slope, and friction is not calculated for whatever reason.

For example, I noticed that when you crash not too hard on the FARP matting, you actually "submerge" the skids and/or parrs of the heli into the ground, you still can sometimes elevate the whole chopper out if this situation, most of te time...

 

Just observations and suggestions, on my side. We really need to wait and see what would a real life Gazelle do here. Then we can try to nail, what is causing the different behaviour. And ultimately the Devs can fix it.

 

Though after todays tests, I doubt it is "normal". At least a half to fully pulled back rotordisk should stop any forward movement, right before the nose comes up, tilting on the skids aft end!

If you pull back the cyclic to the point where the nose starts coming up, you should have no more forward movement? :dunno:

[Flagrum]

Wind can not be the (root-)cause of this effect because the issue is present even in no wind conditions.

[hannibal]

Understood. But all those have nothing to do with the flight model. All I know is that Polychop recognized that the pedals input was a tad too high.

 

Cheers,

 

i disagree.

if you have wacky axis curves, thats not going to help precise controls.

supposedly default axis settings should be good enough.

but a extended stick will have more fine movement as you would a joystick.

 

agree that controls have nothing to do with the flight model.

but if you controls arent optimal that is not going to give you the control you want.

[Yurgon]

if you have wacky axis curves, thats not going to help precise controls.

supposedly default axis settings should be good enough.

but a extended stick will have more fine movement as you would a joystick.

 

For starters, I believe most of the Polychop guys themselves use curves, or a reduced saturation, or a combination thereof. And I guess they know better than the rest of us how these affect the input they make. :smartass:

 

Also, I believe the only time we shouldn't use any software modification to the joystick/throtte/whatever input is when we have the exact same controls as the real pilots, and also know for sure that the sim itself does take their input exactly as the real aircraft receives input from the real controls.

 

And even then, if a virtual pilot felt more comfortable with modified inputs, so what?

 

Seeing as most of us don't have hardware that even comes close to the real controls, I'm all for finding the setup that suits everyone's individual needs. For some it will be linear, for some it will be a curve, for some it will be reduced saturation, for some it will be to shelve the product until they get better input devices.

 

I just don't think it's a good idea to promote any particular setting for all devices.

[gospadin]

From the other thread:

 

...

 

Rotor tilted backwards, helicopter sliding forward. This is clearly a bug, it should be obvious to anyone by now.

 

Depends on where the Helo CG is.

[Flamin_Squirrel]

Actually a perfect example of a quick take off to hover, with no balancing the heli light on the skids. So at least we know you can take her off into a stable hover.

 

Nah, if you looks closely you can over the space of a second or two it gently comes off the skids (you can even see the skids flex). Certainly long enough time to slide forward a little in the DCS version.

 

A possibility is indeed, that the point of gravity, which is shifted more to the front, through the Viviane mount, compared to the perfectly balanced civilian version, has an effect here.

 

The drifting forward is caused by a rearward (horizontal) thrust component. Weight acts through the C of G, which has a vertical component. As the forces are perpendicular, one force has no affect on the other.

 

At any rate, when in hover, the Gazelle as it is hovers tail low. This indicates that the C of G is rearward anyway, so arguments about the weight of the Vivian system aren't valid.

 

There is ground collision modeling, ground friction, and wind coming from the DCS environment, which all effect the aircraft in realtime.

 

I setup a scenario with a strong wind and the Gazelle slides forward regardless of pointing into or away from the wind. So if anything, the wind isn't having an effect when it should.

 

Though after todays tests, I doubt it is "normal". At least a half to fully pulled back rotordisk should stop any forward movement, right before the nose comes up, tilting on the skids aft end!

If you pull back the cyclic to the point where the nose starts coming up, you should have no more forward movement? :dunno:

 

Absolutely :thumbup:

[Hook47]

I'm glad to see this feedback from the real pilots. I believed from the start the FM was solid, and have had zero issue flying this chopper from the first minute in the cockpit. Was surprised to see so many people claiming it was uncontrollable, as I feel more in control of this helo than any other in the sim.

 

Awesome work, Polychop

[MBot]

Depends on where the Helo CG is.

 

How is this relevant? Sliding forward requires a force with a forward component. Explain the me where this force is coming from.

[shagrat]

i disagree.

if you have wacky axis curves, thats not going to help precise controls.

supposedly default axis settings should be good enough.

but a extended stick will have more fine movement as you would a joystick.

 

agree that controls have nothing to do with the flight model.

but if you controls arent optimal that is not going to give you the control you want.

Well, the ALAT pilots use axis settings, Poly_Roie does and so does The_Fragger. We talked about the axis settings and he said, the biggest issue with heli controls, is to reproduce the spring force/hydraulic force depending on the mechanism used.

Onr thing that made absolitely sense, was the rotor RPM and resistance.

The more you push the stick out of "center" the more the rotor disk wants to upright itself and gives resistance, depending on the actual system design, e.g. hydraulic dampening.

The only way to reproduce this is to adjust saturation and curves to your personal input devices, unless you have a replica helicopter cyclic.

 

That is why they use saturation and sometimes curves to adjust the input to short throw jetfighter sticks. ;)

[RabidRider]

Nah, if you looks closely you can over the space of a second or two it gently comes off the skids (you can even see the skids flex). Certainly long enough time to slide forward a little in the DCS version.

 

 

 

The drifting forward is caused by a rearward (horizontal) thrust component. Weight acts through the C of G, which has a vertical component. As the forces are perpendicular, one force has no affect on the other.

 

At any rate, when in hover, the Gazelle as it is hovers tail low. This indicates that the C of G is rearward anyway, so arguments about the weight of the Vivian system aren't valid.

 

 

 

I setup a scenario with a strong wind and the Gazelle slides forward regardless of pointing into or away from the wind. So if anything, the wind isn't having an effect when it should.

 

 

 

Absolutely :thumbup:

 

+1 Good points!

[shagrat]

How is this relevant? Sliding forward requires a force with a forward component. Explain the me where this force is coming from.

That is the crux. It might make the nose dip, even give a bit of forward momentum, but that would be after(!) the tail lifts.

The slow creeping movement, with the complete skids on the ground is strange. Especially if we pull the stick aft, such that it immediately after TakeOff nearly flips the Heli and crashes the tail...

Still not sure it is the ground/surface modeling or the FM, or a combination causing this.

[Flamin_Squirrel]

That is the crux. It might make the nose dip, even give a bit of forward momentum, but that would be after(!) the tail lifts.

The slow creeping movement, with the complete skids on the ground is strange. Especially if we pull the stick aft, such that it immediately after TakeOff nearly flips the Heli and crashes the tail...

Still not sure it is the ground/surface modeling or the FM, or a combination causing this.

 

I can't see how it could be anything other that a flight model issue. Where else is the force coming from? The ground isn't pushing it forward!

[shagrat]

I can't see how it could be anything other that a flight model issue. Where else is the force coming from? The ground isn't pushing it forward!

As well, as the ground won't bump and jump a wreck...

 

Bugs are not logical. ;)

[decipher]

From the other thread:

 

 

Rotor tilted backwards, helicopter sliding forward. This is clearly a bug, it should be obvious to anyone by now.

 

i don't see a topic for this specific problem in the dedicated bug report forum. It would probably be a good idea to create one with objective reports and pictures/videos all cleanly sorted. this might help the Devs to replicate/analyze whats going on.

[Mt5_Roie]

I think I mentioned that a couple times at this point...glad someone is paying attention. Start a thread in the bug report and talk about it all your guys want.

 

i don't see a topic for this specific problem in the dedicated bug report forum. It would probably be a good idea to create one with objective reports and pictures/videos all cleanly sorted. this might help the Devs to replicate/analyze whats going on.