Advanced Helicopter FDM Feedback

[dr. hladik 231]

I just watched Dsylexci's video again. Watched/watched it. Not just listen/watch. You are correct he basically descended so fast, and attempt to power his way of it and ran out of engine power. That is not the correct way to test VRS. He made a big deal out of it I thought he knew what he was doing, and I was mainly listening to his video not watching. You can easily exceed exceed engine power and hit low RPM in ARMA 3 now. Sad thing is it can happen to you off of a 200KMH dive if you pull up/bank too much.

VRS does happen in DCS, also in XPLANE. In DCS the way to test is just a steep glide path approach and hold try to hold a high descent rate say 1000FPM, what will happen is you are holding that descent rate and out of the blue a sudden increase in descent. I never really noticed control authority loss other than in the collective, but that is probably hard to model or I can't tell given my joystick. Either way if you don't react fast enough that increase is normally fatal if you are too low to the ground. For me it normally happened, if I was to impatient and landed with a tail wind, and did it to steep and to fast. So in reality it is easy to avoid. Maintain a low descent rate, and land into the wind is what I was always told.

So in arma 3, I basically tried to do the same thing. Descend into my down wash, with a high rate and hold it. See if I get an unexplained sudden increase in descent rate, I never got that. So I'm pretty sure VRS is not modeled.

Yes, current version of RotorLib does not simulate VRS.

[.kju 3244]

which version of RotorLib does A3 use currently?

further resources for interested people:

http://www.rtdynamics.com/v2/wp/

http://www.rtdynamics.com/v2/wp/2008/rotorlib-fdm/

http://www.rtdynamics.com/Release/Whitepapers/4.0/RL_FDM_Whitepaper.pdf

unfortunately the page, pdfs, blog and such seems fairly outdated.

the latest version seems to be 4.1

[CaptainObvious 95]

I believe everyone's input is valid. After all, we all play this game for fun (it's not a full blown flight sim meant for training). If you don't find something too fun let the devs hear your input.

Hm, maybe you're right, but still, I'm not comfortable with doing so and will happily let the guys who know this stuff give the feedback so we all get this FM the right way. :)

[worldsprayer 0]

For those discussing VRS and settling with power, the two are effectively one and the same. Settling with power requires three factors: high rates of decent (greater than 300 fpm), low forward airspeed, and 20-100% power remaining with insufficient power for recovery. The reason its called settling with power is because you HAVE power power to use (that 20-100%) but it simply can't accomplish the job. That would be because of the helo falling into the downwash of its own vortex rings...which is why the official name is VRS.

Frankly if the helicopter is not capable of stopping a decent that isn't being effected by VRS, then the pilot did a really bad job either) loading the aircraft's weight (which wold likely be a transmission limitation...it cant move the power) or the aircraft's max power is limited by environmental conditions and the engines hit their upper limit for temperature and wont work any harder. Exceeding the aircraft limitations and settling with power are two different things.

Edited August 11, 2014 by Worldsprayer

[dr. hladik 231]

Can someone test something for me? I've attempted this about 5 times now and its resulted in a crash of the game: I take an mh9, fly at 100kph from the stratis airfield to the nearby port city, at which point I attempt a landing (obviously ie slowed a bit) at a tactical speed on the docks at stratis 28236101. So far, I've blown up e very time, but each time I blow up, the game crashes. I just tried landing on the docks again nice and slow, and if you land on them the helo vanishes and the game crashes. And then again just now I tried to start a mission where I placed a helo on the pier with me in it...and the game crashed. Then I started a mission, helo on pier with me next to it, mission started but crashed as soon as i got in the helo. I've tried now with mutiple helos.

So...can someone verify this? I'd post it now but someone on the bug tracker is annoyed at my volume of posts atm.

---------- Post added at 06:20 ---------- Previous post was at 06:19 ----------

I feel like everything happens too fast in arma3. Some maneuvers should just take some time to happen, not this swinging around all over the place at the drop of a pin. I don't think the game is simulating inertia, which is a fairly important element in helicopter flight.

Should be fixed in rev 126670

[worldsprayer 0]

Should be fixed in rev 126670

Any chance on knowing on what caused the crash? It's such an interesting and specific thing that I'm curious :confused:

[henrique_rp 10]

Just watched the Dsylexci video along with another one posted by Lyndiman (

). I think they were both very accurate in their description and complaints about the flight model.

The flight model RotorLib uses just generates the most bizarre reactions most of the time. In its current state, this flight model is extremely crude and most of the time dangerously unpredictable. Correcting Pilot Induced Oscillation will most of the time lead to your blowing up.

As it is, the flight model is severely unreliable for its strange randomness and wildness of reactions.

When flying DCS, even at times when you're pushing the helo through its flight envelope, you are totally in control. When flying using the RotorLib flight model, I get the impression the helicopter is always trying to kill me. No matter what. Many times it is successful in its mission.

Another thing that should get immense attention from the developers is the ancient axis configuration page. I'm sorry guys, but that is a joke. Axis response configuration is beyond poor in this game. That should get a lot of attention, since people will have to use a joystick and properly configuring your axis is mandatory in a flight simulation.

My initial impressions of the existence of Vortex Ring and Retreating Blade Stall were wrong. :)

[gatordev 219]

VRS does happen in DCS, also in XPLANE. In DCS the way to test is just a steep glide path approach and hold try to hold a high descent rate say 1000FPM, what will happen is you are holding that descent rate and out of the blue a sudden increase in descent. I never really noticed control authority loss other than in the collective, but that is probably hard to model or I can't tell given my joystick. Either way if you don't react fast enough that increase is normally fatal if you are too low to the ground. For me it normally happened, if I was to impatient and landed with a tail wind, and did it to steep and to fast. So in reality it is easy to avoid. Maintain a low descent rate, and land into the wind is what I was always told.

Didn't mean to suggest DCS doesn't have VRS, just pointing out the way it was being tested wasn't accurate. I don't own DCS, but watching Dsylexci's video sure makes it look like it has a pretty sweet FM.

Now after reading, and trying to figure out the different between settling with power, vs VRS. I'm confused. What I described above is another form form of settling with power, but caused by VRS? Basically I have power and it is getting eaten by my downwash? So is that settling with power or VRS?

You are correct, "power settling" is how it's described and VRS is a form of Settling with Power, but the two are not the same thing. VRS is one of the four stages of blade element theory (mentioned above...Normal, VRS, Autorotation, Windmill braking). These are unique conditions of the rotor system during various flight regimes. Settling with power can be caused by many different things, one of them can be VRS, but there are others (mostly boiling down to Pr > Pa).

For those discussing VRS and settling with power, the two are effectively one and the same. Settling with power requires three factors: high rates of decent (greater than 300 fpm), low forward airspeed, and 20-100% power remaining with insufficient power for recovery. The reason its called settling with power is because you HAVE power power to use (that 20-100%) but it simply can't accomplish the job. That would be because of the helo falling into the downwash of its own vortex rings...which is why the official name is VRS.

Again, no they're not. And just because you have the power and it's not enough to accomplish the job does not equal VRS. Also, your numbers may be specific to one particular aircraft, but they are not an end all value. The B206 and Huey use approx. 800 FPM. The H-60 is somewhere between 800-1500 FPM (depending on speed). In my current aircraft, I can pull up to 144% Tq before it tells on me and up to 149.9999% before someone has to look at it (which usually results in a "it's fine").

Frankly if the helicopter is not capable of stopping a decent that isn't being effected by VRS, then the pilot did a really bad job either) loading the aircraft's weight (which wold likely be a transmission limitation...it cant move the power) or the aircraft's max power is limited by environmental conditions and the engines hit their upper limit for temperature and wont work any harder. Exceeding the aircraft limitations and settling with power are two different things.

For the record, you can be doing a very good job managing all those things and still encounter power limiting because of the changing conditions. Again, it also depends on the helo. In the H-60, I am almost never tranny limited, but if I'm too heavy, I can be TGT limited (as you describe), regardless of the outside temperature. And here's an even more crazy point... In one model of -60, I can be engine limited, and in another model with the EXACT SAME tranny and engines, I'm tranny limited. I have yet to figure that one out.

Power Settling and VRS are often debated in the aviation world, so if you have an actual reference that says power settling is the same thing as VRS, I'd love to see it. I've just only heard it used to describe to VRS. At the end of the day, though, you have to go back to the Blade Element Theory to actually explain what's happening, and VRS is how that regime is defined (flow goes down, curls back up around the tips, and then flows back down again, increasing flow each time, and increasing descent rate).

Edited August 11, 2014 by gatordev

[blackpixxel 53]

Does the rotorlob Flightmodel simulate the roll-behaviour at high speeds as a result of different relative speeds of the rotorblades relative to the air?

If yes, then the Hellcat is simulated wrong. At nearly full speed (about 305 km/h) I have to push the cyclic stick a bit to the left to counter a roll movement to the right. But the blades are rotating counter-clockwise. That means that the blade on the right side has a higher relative speed at high forward flying speeds than the left blade, which would result in a left roll movement that I would have to counter with pushing the cyclic to the right.

In addition to that the Hellcat feels very wrong at high speeds (autotrimm off). I have to push the forward stick nearly to the maximum position to keep the thing flying forward. Otherwise it would instantly flip backwards. But aren't we talking about high inertia values here caused by the weight of the helicopter and the spinning of the main rotor? At least R/C helicopters only require a minimal forward input to keep them flying forward. And which force is trying to flip the helicopter back that fast? Could only be the air resistance of tail stabilisator that has a high angle while flying forward. The rotor speed also drops significantly during this backward movement.

I am not a heli expert, but I am interrested in them and especially in R/C helicopters for 7 years now. For me it feels like flight model has some very strange behaviours..

[vicx 10]

BlackPixxel, I assume this page is relevant. https://community.bistudio.com/wiki/Helicopter_Flight_Model_Config_(XML)

A lot of modelling assumptions have to be made even before populating a shipload of look up tables.

Edited August 11, 2014 by vicx

[Wildgoose 1]

I have to push the forward stick nearly to the maximum position to keep the thing flying forward. Otherwise it would instantly flip backwards.

I had this problem until I deleted all other keyboard assignments in my joystick profile settings in game.....ie. remove the "W,A,S,D" for joystick assignments, and just leave the joystick settings.

Not sure why, but all joystick movements have returned to normal once these key settings were removed.

[the_demongod 31]

Interesting, this happens occasionally (randomly) to me as well, but I have already unbound those WASD keys

It's as if the trim is set way back. If you force trim (manual trim set) while flying forward it seems to fix it- until you release the trim of course.

---------- Post added at 20:16 ---------- Previous post was at 20:09 ----------

;2750791']which version of RotorLib does A3 use currently?

further resources for interested people:

http://www.rtdynamics.com/v2/wp/

http://www.rtdynamics.com/v2/wp/2008/rotorlib-fdm/

http://www.rtdynamics.com/Release/Whitepapers/4.0/RL_FDM_Whitepaper.pdf

unfortunately the page' date=' pdfs, blog and such seems fairly outdated.

the latest version seems to be 4.1[/quote']

Interesting, I didn't know that RotorLib was a 3rd party thing. It seems that there is something there called "FixedWingLib" as well. Does this mean we might eventually get a better fixed wing flight model as well?

Edited August 12, 2014 by the_Demongod

[blackpixxel 53]

Wow, fixed wing flightmodel would be awesome. Then we only need a trackedlib drive model for tanks and a multicoreLib advanced performance model. I would pay for each of them.

[the_demongod 31]

Me too.

I personally play arma for the vehicle combat, so having a DLC that upgrades the other two types of vehicles (ground and fixed wing) to the level we're seeing with helicopters would be a dream come true.

BI: please say something so that I can say "shut up and take my money."

[Henry Rollins 10]

Got to watch Dslyecxi's video yesterday and overall, I think he's on point (as usual). However, how he is testing for VRS is not correct (in either sim). People confuse Settling with Power and VRS as the same thing, and it is not. Settling with power can be due to power required exceeding power available, which then causes Nr to droop. VRS is different in that Nr is at 100% the entire time and you lose control authority. In the video, his test to enter VRS is by lowering the collective all the way down and entering an autorotation (as demonstrated by Nr climbing). Entering an auto is the VERY WAY to get OUT of VRS, not to enter it. The 4 states of rotor flow (in order) are Normal, VRS, Autorotation and then Windmill braking. While different aircraft will have some varying procedures, the general idea is the same. Here's one from A1-H60RA-NFM-000

VRS and settling with power are the same thing. Trust me I have been in it several time in real life and it's on both FAA exams.

---------- Post added at 08:43 AM ---------- Previous post was at 07:57 AM ----------

For those discussing VRS and settling with power, the two are effectively one and the same. Settling with power requires three factors: high rates of decent (greater than 300 fpm), low forward airspeed, and 20-100% power remaining with insufficient power for recovery. The reason its called settling with power is because you HAVE power power to use (that 20-100%) but it simply can't accomplish the job. That would be because of the helo falling into the downwash of its own vortex rings...which is why the official name is VRS.

Frankly if the helicopter is not capable of stopping a decent that isn't being effected by VRS, then the pilot did a really bad job either) loading the aircraft's weight (which wold likely be a transmission limitation...it cant move the power) or the aircraft's max power is limited by environmental conditions and the engines hit their upper limit for temperature and wont work any harder. Exceeding the aircraft limitations and settling with power are two different things.

This guys information is correct.

[worldsprayer 0]

Didn't mean to suggest DCS doesn't have VRS, just pointing out the way it was being tested wasn't accurate. I don't own DCS, but watching Dsylexci's video sure makes it look like it has a pretty sweet FM.

You are correct, "power settling" is how it's described and VRS is a form of Settling with Power, but the two are not the same thing. VRS is one of the four stages of blade element theory (mentioned above...Normal, VRS, Autorotation, Windmill braking). These are unique conditions of the rotor system during various flight regimes. Settling with power can be caused by many different things, one of them can be VRS, but there are others (mostly boiling down to Pr > Pa).

Again, no they're not. And just because you have the power and it's not enough to accomplish the job does not equal VRS. Also, your numbers may be specific to one particular aircraft, but they are not an end all value. The B206 and Huey use approx. 800 FPM. The H-60 is somewhere between 800-1500 FPM (depending on speed). In my current aircraft, I can pull up to 144% Tq before it tells on me and up to 149.9999% before someone has to look at it (which usually results in a "it's fine").

For the record, you can be doing a very good job managing all those things and still encounter power limiting because of the changing conditions. Again, it also depends on the helo. In the H-60, I am almost never tranny limited, but if I'm too heavy, I can be TGT limited (as you describe), regardless of the outside temperature. And here's an even more crazy point... In one model of -60, I can be engine limited, and in another model with the EXACT SAME tranny and engines, I'm tranny limited. I have yet to figure that one out.

Power Settling and VRS are often debated in the aviation world, so if you have an actual reference that says power settling is the same thing as VRS, I'd love to see it. I've just only heard it used to describe to VRS. At the end of the day, though, you have to go back to the Blade Element Theory to actually explain what's happening, and VRS is how that regime is defined (flow goes down, curls back up around the tips, and then flows back down again, increasing flow each time, and increasing descent rate).

You're corect, you can exceed limitations even if you are careful, especially considering some of the places we've been flying the last decade. It's why we came out ith the uh-60m and ah-64e, to give the power that simply wasn't there and forced extra conservative (and sometimes not conservative enough) flying. Besides some wikipedia posts and a helicopter flight handbook and my word on what my flight instructors SAY...well I don't have any solid source. The logic given to me that I use is this: When you're light and dropping fast but can't stop, you can enter VRS because you caught up later rather than sooner with those nice big vortices your bladetips made. When you're heavy and dropping slowly but suddenly can't stop, its because you caught up with those vortices sooner rather than later...if that makes sense. End the end though...same difference...helo goes whammo.

---------- Post added at 09:44 ---------- Previous post was at 09:41 ----------

Does the rotorlob Flightmodel simulate the roll-behaviour at high speeds as a result of different relative speeds of the rotorblades relative to the air?

If yes, then the Hellcat is simulated wrong. At nearly full speed (about 305 km/h) I have to push the cyclic stick a bit to the left to counter a roll movement to the right. But the blades are rotating counter-clockwise. That means that the blade on the right side has a higher relative speed at high forward flying speeds than the left blade, which would result in a left roll movement that I would have to counter with pushing the cyclic to the right.

In addition to that the Hellcat feels very wrong at high speeds (autotrimm off). I have to push the forward stick nearly to the maximum position to keep the thing flying forward. Otherwise it would instantly flip backwards. But aren't we talking about high inertia values here caused by the weight of the helicopter and the spinning of the main rotor? At least R/C helicopters only require a minimal forward input to keep them flying forward. And which force is trying to flip the helicopter back that fast? Could only be the air resistance of tail stabilisator that has a high angle while flying forward. The rotor speed also drops significantly during this backward movement.

I am not a heli expert, but I am interrested in them and especially in R/C helicopters for 7 years now. For me it feels like flight model has some very strange behaviours..

To answser your question, no it doesn't simulate the roll from retreating blade stall, though it is effectively simulating the blowback tendency caused by high airspeed which is ALSO a part of retreating blade stall. The stick being all the way forward at full speed is natural in any helo and is why force trim was created, to allow the pilot to work less on holding the cyclic in a certain area and more on making the fine adjustments to the stick wherever it is. Some helos like blackhawks even have a PBA (pitch bias actuator) that advances blade position at higher airspeed to allow the cyclic to have more effect without having to move so far.

[Guest]

Here is an old clip of a pair of 101st blackhawks with the very tips of their rotor blades contact each other.

Neither Aircraft explode though.. Personally on rare occasions in arma 3 I have crashed a helicopter and it has not exploded. I remember one Incident very early into my experience with Arma 3 where i pitched forward from a stationary hover at too low an altitude and swung into a little V shaped valley, hit some trees and came to a sudden sharp halt, the rotors stopped spinning with a loud jamming BUZZING sound and the windows were all broken.. I was like "WOW this is so cool".. (Having not crashed a helo in arma 3 before then i thought they had removed the explode on contact thing finally...)

To me this seems a far more realistic a result for the vehicle in most cases (granted impact should still damage and/or kill crew/vehicle systems) but most helicopter crashes i've seen do not result in the helicopter exploding. Personally i think the methodology in A3 helicopter crash damaged is backwards.. they shouldn't be starting with an exploding vehilce and working towards a point where it doesnt explode, they should be starting with a non exploding vehicle and then finding the right circumstances in which is should explode... I mean imaging how cool crashes would be where you hit the deck, roll, lose your rotors and the reamaining airframe just keeps rolling several times before soming to stop.. maybe not possible in this engine but well, i dunno, aiming for something like this would make for a much more immersive expereince...

[the_demongod 31]

To be honest, if BI removed the exploding from helicopters altogether, it would probably be more realistic. When helicopters crash they seem to just kind of crumple. Most helicopter crash videos show the rotors getting ripped off, the tail booms getting ripped off, and deformation of the hull, but rarely does anything actually explode.

Imagine if whenever you crash your helicopter, if you just sat there on the ground with all the parts red. It would be more realistic.

[Airwolf 50]

First testing of the new FDM I would say the shakiness is too much. Atleast give the option to turn that off. Helicopters shake, yes. But I don't think its necessary to have that in game, it just seems overly forced onto the flight model.

[blackpixxel 53]

Can a real world pilot say something about the delay of the helicopter reaction to the joystick input?

At the moment it feels like this:

I push cyclic to the front, the helicopter waits half a second and starts to move slowly.

I center the cyclic, but the helicopters roll movement accelerates, even if there is no force that could let him accelerate his nick rotation.

Is this how a real helicopter behaves? Every movement feels like half a second delayed. It cannot be correct that the helicopter is accelerating a cyclic movement after there is no cyclic input anymore.

But maybe I am just too familiar with the extreme fast and accurate responsivnes of helicopters in R/C simulations.

[b00ce 160]

First testing of the new FDM I would say the shakiness is too much. Atleast give the option to turn that off. Helicopters shake, yes. But I don't think its necessary to have that in game, it just seems overly forced onto the flight model.

Agreed, the only time it really should shake noticeably is on start-up, being in the ETL transition, overspeeding the aircraft and when the main rotor/tail rotor are damaged.

Can a real world pilot say something about the delay of the helicopter reaction to the joystick input?

At the moment it feels like this:

I push cyclic to the front, the helicopter waits half a second and starts to move slowly.

I center the cyclic, but the helicopters roll movement accelerates, even if there is no force that could let him accelerate his nick rotation.

Is this how a real helicopter behaves? Every movement feels like half a second delayed. It cannot be correct that the helicopter is accelerating a cyclic movement after there is no cyclic input anymore.

But maybe I am just too familiar with the extreme fast and accurate responsivnes of helicopters in R/C simulations.

Yes, remember to make an adjustment and return to center. If you hold the adjustment till you see it happen, you held it way too long.

[Henry Rollins 10]

Can a real world pilot say something about the delay of the helicopter reaction to the joystick input?

At the moment it feels like this:

I push cyclic to the front, the helicopter waits half a second and starts to move slowly.

I center the cyclic, but the helicopters roll movement accelerates, even if there is no force that could let him accelerate his nick rotation.

Is this how a real helicopter behaves? Every movement feels like half a second delayed. It cannot be correct that the helicopter is accelerating a cyclic movement after there is no cyclic input anymore.

But maybe I am just too familiar with the extreme fast and accurate responsiveness of helicopters in R/C simulations.

I can tell you that in Robinson R-22's and R-44's, which is all I have every flown, both the cyclic and collective are VERY responsive. Most of the time if someone is having issues its because the controls are too responsive and they over correct.

[beastcat 14]

Since the Mi-48 is almost impossible to get off the ground when fully loaded, isn't a reworked refueling and rearming system necessary?

Scripts are nice and all, but you can't choose the amount of fuel you want to load or even dump fuel in vanilla or choose what weapons you want to rearm or remove. Scripts and mods should not stand in for such a fundamental feature.

[blackpixxel 53]

Yes, remember to make an adjustment and return to center. If you hold the adjustment till you see it happen, you held it way too long.

I already fly "foresighted" (if that is the right word) So I stop the adjustment before the helicopter achieves the desired position. But for some reason even the ultralight littlebird flies like it would have a heavy input delay

[RuecanOnRails 10]

When landing, and moving inside the covered hangar buildings, the helicopters have a tendency to become damaged and explode. I believe it's caused by skids clipping slightly into the floor although I'm not certain.

Helicopters are still far to prone to explode causing instant death. If from a near stationary position instant death should not happen when the helicopter rolls onto it's side or nose. The main rotor should break and damage should be potentially irreparable, but it the pilot and crew should have a chance of survival. Coming in too hot from a semi failed auto-rotation is bitter sweet as you land hard, but safe, then slide a little bit, and slowly roll causing instant death.

I was greatly disappointed to see early on in this thread that support for Take on Helicopters merging wouldn't happen. As someone who still plays TKoH occasionally, will the updates and improvements to rotorlib find their way into a TKoH update?

First testing of the new FDM I would say the shakiness is too much. Atleast give the option to turn that off. Helicopters shake, yes. But I don't think its necessary to have that in game, it just seems overly forced onto the flight model.

I think what you're describing is the visual representation of "stress damage" which does have an option in the menu. The helicopter doesn't shake when flown correctly and not damaged in my experience at least. During start up, keep the collective lowered and wait for the RPM to reach max speed (you will know it when you hear it), slowly raise the collective until you are in a low hover. Once in a low hover, either move forward and allow translation lift to gain altitude, or increase collective more while watching torque and rpm.

If you are doing a proper take off, I'm not sure what shaking you're referring to.