Something wrong with flight dynamics.

[YoYo]

Report your TM Warthog curves and saturation instantly! So that I have something to chew on...:joystick: Thanks!:thumbup:

 

I set up it for 12% X and Y and is better now, throttle as slider, rudder 10% too, but for me FM must be improved also. Regards.

[Whisper]

I set up it for 12% X and Y and is better now, throttle as slider, rudder 10% too, but for me FM must be improved also. Regards.

 

What's the effect of "Throttle as slider" ?

[rrohde]

What's the effect of "Throttle as slider" ?

 

+1

[Flagrum]

What's the effect of "Throttle as slider" ?

Afaik there is some sort of glitch when the axis moves throught 0. It get's bumpy there and fine control is not really possible. This has nothing to do with the hardware, I believe it is a DCS bug. You can observe this when watching the controls indicator closely.

 

This was discovered, iirc, with the Huey: at around 50% (or "0" if not set as a slider) is usually the sweet spot at take off.

[gospadin]

Report your TM Warthog curves and saturation instantly! So that I have something to chew on...:joystick: Thanks!:thumbup:

 

50% cyclic saturation, 0% curves

70% rudder saturation, 0% curves

[Drakoz]

For forward flight turns, I found if I use a little bit of anti-torque (like using the rudder for coordinated flight in an airplane), the turns are much much more stable. But I still have to be very active on the cyclic to keep it stable (see below about kicking the stick). FM is very unstable right now if you are moving the stick at all. But...

 

If you let go of the stick, the DCS Gazelle is so stable, you would think it was on auto pilot. I assume this is the SAS. So when flying straight, I would let go of the cyclic and use trim to adjust pitch and mild turns.

 

 

I found to fly it competently, I have to do very small movements (sensitive stick) during hover, but when flying at speed, I need to do "kick" movements of the cyclic to get movement initiated quickly, and then kick back to stop (e.g. start an aileron roll, and then stop the roll at say 30 degrees - then pull back to make the turn). You can't baby the stick at speed. This is because at speed, the stick is still sensitive, but the DCS Gazelle responds very slowly. Constantly, but slowly. The slightest movement of the stick will cause it to turn or pitch, but then you have to get just the exact right movement back to stop the turn. Since it is so slow to respond, it is difficult to know if you have moved it enough, and hence you oscillate back and forth trying to hold a turn. Let go of the stick, and the oscillations stop and it settles out to whatever the trim settings are.

 

This is using a FFB stick (MS FFB 2). I am unsure exactly how much of the FFB code has been implemented in DCS Gazelle. But I use simFFB for FFB effects which bypasses any FFB control by DCS. The trim functions, however, are being implemented both by simFFB and DCS Gazelle. If set up correctly, simFFB compliments trim in the simulation. I think if you did this with a non-FFB stick, when you trim, and let go of the stick, it would center, and that center position would be correct for the trim. So you should see the same results, FFB or no FFB.

 

So when I say kick the stick, I mean a hard deflection of the cyclic to the side to quickly initiate the roll, and then a hard deflection back to center to stop the roll. Then hard deflections either way to adjust the attitude and control the turn. The same can be said about pitching up or down. You get it to respond quickly to your inputs - quick enough that you can instantly see the result of your input, and then instantly compensate for any over control. Because turning requires both a roll, and then pulling back on the cyclic, this is an active process in both X and Y directions.

 

Though this is a fun challenge, it does not seem how a helicopter should work as it is very unstable. But at least using this technique, I am able to control the helicopter with precision both in mild as well as aggressive maneuvers, and so it is enjoyable now. But only if I can maintain > 20 FPS. Below 20FPS, I can't get fast enough feedback from the sim to see if I need kick it back, and my flights are significantly less stable.

 

I did not see any difference turning off the Auto Pilot functions. So if this is related to SAS, AP has nothing to do with it.

 

Regards,

Michael

[Flagrum]

I did not see any difference turning off the Auto Pilot functions. So if this is related to SAS, AP has nothing to do with it.

I have noticed that the SAS? AP? sort of filters the pilot controls input in some (weired?) way.

 

I wondered why the AP channel gauges do not show any activity during flight - until I experimented with the "Easier Controls" Mode. In that mode, the controls are not so suuuuuuuuuuuper sensitive and your control inputs must/can be more pronounced. And that was when I saw that those AP gauge needles in fact DO move. The AP reacts directly to any input of the controls in some way. This was hardly noticable before because for stick deflections in the range of 1-2 millimeter, the needles also only move about the width of the needle!

 

So, those AP channels seem to only react on pilot inputs (probably even fighting the pilot), but seem to do nothing in regards of actual attitude changes of the helo ...

[Pocket Sized]

[. . .]

If you let go of the stick, the DCS Gazelle is so stable, you would think it was on auto pilot. I assume this is the SAS. So when flying straight, I would let go of the cyclic and use trim to adjust pitch and mild turns.

[. . .]

You can't baby the stick at speed. This is because at speed, the stick is still sensitive, but the DCS Gazelle responds very slowly. Constantly, but slowly.

[. . .]

I did not see any difference turning off the Auto Pilot functions. So if this is related to SAS, AP has nothing to do with it.

 

Regards,

Michael

 

1 & 3: The SAS is definitely on. There is no mechanical feedback mechanism in a helicopter's rotor that will keep it perfectly level. I think the bug is SAS can't be turned off.

 

2: That is also SAS afaik. It tries to hold you upright while hovering and tries to hold attitude while in forward flight. I have managed to get it stable nose down with the stick centered and no trim, which indicates the SAS acts differently at speed.

[Rogue Trooper]

When the roll AP is off I have a nightmare in hover.

Hover is sweet with it on, I do not notice pitch and yaw AP as much as I notice roll... I think.

 

Does anyone notice 1st take off is harder and after landing; the second take off is sweeter... is it me? It could be me.

 

Everything is small movements with this chopper which makes it easier to fly with the fine trim.

[Rongor]

- Can anybody confirm to have managed to keep a HIGE without adjusting the collective pitch by +-5% torque or more? I am unable to achieve that. In most cases the helo remains jumping up and down continously, relatively slowly though.

 

- also regardless how I set curves (TM Warthog), it is impossible not to slide forward on the ground while increasing collective. I end up pulling the cyclic stick so far backwards that my tail boom bumps onto the floor.

Edited May 5, 2016 by Rongor

[Rongor]

when sitting on ground, apply collective.

Notice the helo commencing forward slide.

Reduce torque to 40% and don't touch it anymore.

Pull the stick backwards to counter the forward slide.

Notice how loooong that takes.

Notice how the forward slide only comes to a stop right at that moment your tail boom tilts towards the ground and strikes it.

 

How can this behavior be accurate?

 

No, I never flew a Gazelle in RL. I did some hours on the Gazelle's successor though. The EC-120 also features 3 blades and a fenestron. Also a pretty light weight. Of course this doesn't mean both machines behave similar at all.

 

What I am experiencing here is some behavior that simply doesn't make sense when you think about how you would construct a helicopter.

 

Apparently the DCS Gazelle is quite nose-heavy. I can't tell if it matches the RL Gazelle. I only can say that it is not the way you would expect a helicopter to be.

[exil]

- Can anybody confirm to have managed to keep a HIGE without adjusting the collective pitch by +-5% torque or more? I am unable to achieve that. In most cases the helo remains jumping up and down continously, relatively slowly though.

 

- also regardless how I set curves (TM Warthog), it is impossible not to slide forward on the ground while increasing collective. I end up pulling the cyclic stick so far backwards that my tail boom bumps onto the floor.

 

It takes a bit practise but yes, i am now able to achieve a stable hover in regards to height and heading. The trick is not to use to much pedals because they will influence your power (torque) and therefore your height.

 

In regards to the foreward slip i noticed that if you make her light on the skids and then pull collective a bit faster, she will stabilize and stop the forward tendency. i could imagine that this is a RL behavior.

 

The more i fly with her now, the more i have the feeling that the flight model is pretty accurate.

Turns are still a bit tricky, but overall it's a pretty fantastic helo.

[Rongor]

It takes a bit practise but yes, i am now able to achieve a stable hover in regards to height and heading. The trick is not to use to much pedals because they will influence your power (torque) and therefore your height.

 

In regards to the foreward slip i noticed that if you make her light on the skids and then pull collective a bit faster, she will stabilize and stop the forward tendency. i could imagine that this is a RL behavior.

 

The more i fly with her now, the more i have the feeling that the flight model is pretty accurate.

Turns are still a bit tricky, but overall it's a pretty fantastic helo.

Please confirm you can do that at 3 feet above the ground. And if so, please post your curves.

I keep the pedals pretty stable. So if that is the only trick, it doesn't work for me.

[exil]

Gotta check if it works at 3ft AGL. I think i was at about 5-8 feet most of the time...But i'm going to check that out.

 

I tried different things..for hovering i used at about 65% saturation in pitch and roll axis and about curves set to 2. Yaw axis is about 50% saturation.

 

Currently i'm back to 90% saturation for pitch and roll. I found out that it's more controllable for low level flight with this settings. But makes hovering a lot harder of course.

[Pocket Sized]

- Can anybody confirm to have managed to keep a HIGE without adjusting the collective pitch by +-5% torque or more? I am unable to achieve that. In most cases the helo remains jumping up and down continously, relatively slowly though.

 

- also regardless how I set curves (TM Warthog), it is impossible not to slide forward on the ground while increasing collective. I end up pulling the cyclic stick so far backwards that my tail boom bumps onto the floor.

 

I have, I think. Don't remember any particular difficulties keeping in the ground effect.

 

I have never had this problem. Does anyone have a video of it? Seriously, as long as the SAS is on (it is unless you intentionally disable it) I can lift straight up with no hands on the stick and no trim. My guess is that you're being too cautious lifting off. No reason to stay on the ground after raising the collective unless you feel your engine hasn't gotten it's daily dose of minerals.

[Drakoz]

With practice now, I am able to get into a stable hover with minimal forward sliding, but it doesn't change the fact that the forward slide is still there. Maybe the slide is accurate due to an imbalance from the heavy camera equipment on top of the canopy, and without that, maybe it wouldn't slide forward. But it still seems unsafe for a real helicopter, and I would hope the designers would have counterbalanced the equipment for real.

 

The following link is a recent lift off I recorded. Not perfect, but it shows the point. There is an inset video showing my cyclic and collective, or you can watch the in sim control movements in the lower left corner. I have 70% Saturation on the cyclic X & Y and anti-torque pedals.

 

Here is the same video later, showing another (different) lift off from the external view. The inset video does not correctly show my control inputs, so ignore it. Again, not a perfect lift off. I'm still drifting forward. But it is controllable.

 

If you care to see, the rest of the video shows other examples of how active I have to be on the controls for what should be easy manuvers for a good pilot. It shows what I mean by "kicking" the cyclic and pedals in order to maintain controlled flight. I'm not a good pilot, and of course I would not do most of this in real life. But my point is, after only a few hours of practice in DCS, maintaining this amount of control was relatively easy and the sim is quite fun, regardless of whether the FM is correct or not.

 

Regards,

Michael

[Pat01]

Hi Gents

 

The gazelle is built with a rotor head leaning 3° forward

Then if you don't change anything as soon as you take off it will accelerate until it reaches around 120km/h with a helicopter pitch of 0°

So to hover or take-off straight up you have to counter this, some trim nose up with the china hat may help, but a lot of pilots just counter it by hand

In fact the hover pitch is +3° so the rotor head will be at 0° pitch

 

Take-off trim should be +3° nose up, 1° right roll

Edited May 6, 2016 by Pat01

[Flagrum]

Take-off trim should be +3° nose up, 1° right roll

And if you trim +3.1°, your tail kisses the tarmac ... :cry:

[msalama]

What's with the ground effect, BTW? It's just as if the bird had it in reverse, i.e. when you get near the ground you seem to enter a vortex of sorts (not VRS, mind) that sucks the air from underneath you instead of providing a cushion. Or is it just me?

 

And what's with the weird control response? As has been stated already, it feels like you're fighting a humongous stab augmentation system forcing the chopper to level itself at all times. I'm finding it very hard to believe the helicopter was designed to behave this way - or is it just me again?

 

I realize this is an early access release, so things'll probably change for the better soonish. So take the preceding comment as an opinion only & kudos for a great whirlyburdie too, guys :thumbup:

[gospadin]

And what's with the weird control response? As has been stated already, it feels like you're fighting a humongous stab augmentation system forcing the chopper to level itself at all times. I'm finding it very hard to believe the helicopter was designed to behave this way - or is it just me again?

 

This is what I find odd (been thinking of how to phrase it for a while...) and this is my understanding of what "should" happen:

 

When I apply cyclic in any direction, what the controls should be doing is increasing the effective lifting force on the opposite side of the rotor disc. (ignore precession for now). If I push forward, the swashplate tilts causing the blades in the rear (yes, 90 degrees before the rear with precession) of the disc to increase pitch, to provide more lift on the rear of the disc relative to the front.

 

Assuming I start from a hover, where the lifting force is balanced with the force of gravity through the CG, the above increase in lift at the rear of the disc will apply torque to the body of the helicopter into a nose-down attitude through the rotor mast. This torque will start to tilt the body of the helicopter nose-down, until the movement of the CG to the aft relative to the input vector cancels the asymmetric lifting force, and the body stabilizes at a new nose-down attitude.

 

If I apply more cyclic in the same direction, the force increases, and the balance point of the helicopter's CG relative to the rotor mast should be further away. However, this balance point should still exist. Note that as I pitch over further, I should have to add collective (increase the lift of the entire disc) to stay at the same altitude, because the vertical component of my lift vector is being moved to the horizontal component.

 

For most any input (prior to achieving translational lift), I should have a pretty clear relationship between cyclic input and rotor disc pitch.

 

So I guess the question is, does the gazelle have a FBW system that interprets control input to actually move the mechanical linkages? Is the CG too high (thus making the balance point very hard to find)? On 1.5.3, flying at about 2000m of altitude with 2/3 fuel and only two HOT missiles, the helo was incredibly unstable in the pitch axis, constantly rolling forward and back seemingly independent of my inputs.

 

What I feel in the Gazelle is that a specific amount of cyclic input does NOT correspond to a consistent amount of swashplate movement. I don't know how they model their blade sections and the forces on each piece, but yea, definitely wierd.

 

--gos

 

btw, if my understanding of helicopter physics is wrong, please correct me. My IRL flight experience is limited to fixed wing aircraft.

Edited May 7, 2016 by gospadin

[flyer49]

Hi everyone, I thought I’d throw my two cents in here. I flew the UH-1H and UH-60 for sixteen years in the U.S. Army and have over 2000 hours helicopter time. My take on the Polychops SA342M Gazelle is that it certainly is a handful to fly. My opinion is limited to what I know about the Gazelles flight dynamic, by what I have learned flying the SA342 in DCS, so please keep that in mind while reading this.

 

I consider myself a pretty competent Simulator helicopter pilot (i.e. the DCS Huey) and find the Gazelle is very sensitive to cyclic and collective input but not beyond aircraft that I have flown in the real world. However, the aircraft that displayed this type of behavior were smaller and less sophisticated than the Gazelle. I have never flown the Gazelle (in real life) but would assume that being a warbird, the pilot workload would not require such effort on the pilot’s part, so smooth flight dynamics would have been a primary concern (I may be wrong…but I hope not for the real pilots sake).

 

I tried flying the aircraft with and without the trim system on and I found the Gazelle flew with the same ultra-sensitivity and quirky flight dynamics. I felt no difference when I turned on the trim (I even turned it on during a diving turn, hoping to get some kind of a leveling, yawing,or rolling effect..nothing happened). It is important to point out that I have not tweeked my controls at all so that input reactions are from generic settings on my Thrust Master Warthog. My control tweeking theory is that I shouldn’t have to tweek my controls. I should learn each aircrafts nuances as the developer intended them to be good or bad.

 

I do believe that Polychop Simulation is definitely on the right track. Additionally, I know that they are working on all of the issues that are being discussed. What we have to keep in mind is that this is a Beta (all Beta’s are test beds that use end-users as testers) , and as the name implies, this early release is a bug-finding release and this thread is exactly the right thing to do to get the best product in the end. Keep talking about the issues, the guys at Polychops are reading this stuff, digesting it, and fixing things, so it’s the writers here that are going to get a great helicopter to fly in the end…sorry for being long winded…just my two cents.

Edited May 8, 2016 by flyer49

[Justificus]

…sorry for being long winded…just my two cents.

 

Hey, never apologize! I appreciate you taking the time to give that much for .02 cents! It's worth a lot more than what you'd get with my two cents, that's for sure. ;)

[msalama]

sorry for being long winded…just my two cents.

 

You're not long-winded at all. It's always an education to hear from folks who've extensive RL chopper experience :thumbup: Good stuff mate!

[Mt5_Roie]

I think thats more nickel worth then two cents. Great info!

 

Hi everyone, I thought I’d throw my two cents in here.

[wolle]

Hi Gents

 

The gazelle is built with a rotor head leaning 3° forward

Then if you don't change anything as soon as you take off it will accelerate until it reaches around 120km/h with a helicopter pitch of 0°

So to hover or take-off straight up you have to counter this, some trim nose up with the china hat may help, but a lot of pilots just counter it by hand

In fact the hover pitch is +3° so the rotor head will be at 0° pitch

 

Take-off trim should be +3° nose up, 1° right roll

 

Where are the degrees trim displayed in the cockpit?