How many times has SCAS killed you so far?

[Belphe]

Coming back from a sortie, no weapons, no ammo, 1050 fuel left. Cruising 130kts at 15ft above the Syrian plain. Dodging occasional vegetation with smooth, long turns. Keeping the velocity vector centred, flying nose to tail trim. After a 5-second right turn that required a little bit of a pitch up I roll left and follow with a slight left rudder to keep the velocity vector in the centre. Suddenly SCAS starts screaming and I discover that centring the stick does not stop the left roll. I counter with full right roll but it's too late and too low.

Is this normal and if so, what am I doing wrong?

https://www.youtube.com/watch?v=4JcEFcJlwvQ

[Akiazusa]

Personally,I think it was the low G maneuver killed you,（edit,this was wrong) ,                I think it was not the SCAS killed u.

You can give another try,but this time turn off the SCAS first ,then fly fast,keep collective down and push control stick all the way forward and see if the same roll happens

Anyway fly that fast and aggressively is not good for heli

Edited June 17, 2022 by Akiazusa

[Floyd1212]

2 hours ago, Belphe said:

Cruising 130kts at 15ft above the Syrian plain.

130 knots in the Apache is not "cruising", it is speeding like a madman. 

That's like driving your Dodge Caravan on the highway at 120mph and complaining that is gets a little unstable when you change lanes.

Try cruising at 80 knots (especially at 15 ft) and see if that is more stable.

[Moxica]

It went into rocking chair mode. I've had that happen to me on  every flight since last update  occasions. Even with George by the stick.
George should get some lullaby tune in his repertoire. I could go zzzz.

Edited June 17, 2022 by Moxica

[GrEaSeLiTeNiN]

Your roll left momentum carried you to the point of no return. Unexpected 360 rolls happen quite often with the Apache in its current state. It seems to have this top heavy “boat capsize” tendency. Quite annoying and I can’t imagine the real aircraft being this way.

[Catweedle]

On 6/16/2022 at 7:27 PM, Floyd1212 said:

130 knots in the Apache is not "cruising", it is speeding like a madman. 

 

Spec-technically, 130 knots is below its cruise speed.

[Piker]

I don't know if this is real behavior or not, but currently, if you push left rudder and left stick together, you get a roll. It does it also if you push right rudder and right stick. Inputs between rudder and stick in the same direction cause the roll. Since I figured this out, I haven't rolled the Apache.

[DmitriKozlowsky]

On 6/16/2022 at 7:25 AM, Belphe said:

Coming back from a sortie, no weapons, no ammo, 1050 fuel left. Cruising 130kts at 15ft above the Syrian plain. Dodging occasional vegetation with smooth, long turns. Keeping the velocity vector centred, flying nose to tail trim. After a 5-second right turn that required a little bit of a pitch up I roll left and follow with a slight left rudder to keep the velocity vector in the centre. Suddenly SCAS starts screaming and I discover that centring the stick does not stop the left roll. I counter with full right roll but it's too late and too low.

Is this normal and if so, what am I doing wrong?

https://www.youtube.com/watch?v=4JcEFcJlwvQ

That sounds like you exceeded Vne and suffered a retreating blade stall that may or may not have been accompanied by tailrotor over pitch adding to roll. Uncommandedroll at IAS above 130 knots is a sign of Vne excess. Or (perhaps and) you trimmed your cyclic while in substantial bank. Trim should only be set in level stable flight.

When encountering loss of cyclic authority , immediately turn OFF trim, let go of cyclic for a moment  to let cyclic self-center, center pedals, reduce collective. Take back cyclic, after moment, and hold aircraft in straight attitude, as it slows down. Avoid IAS above 135 knots.

Vne-Velocity Never Exceed

Edited June 18, 2022 by DmitriKozlowsky

[Belphe]

1 hour ago, DmitriKozlowsky said:

That sounds like you exceeded Vne and suffered a retreating blade stall that may or may not have been accompanied by tailrotor over pitch adding to roll. Uncommandedroll at IAS above 130 knots is a sign of Vne excess. Or (perhaps and) you trimmed your cyclic while in substantial bank. Trim should only be set in level stable flight.

When encountering loss of cyclic authority , immediately turn OFF trim, let go of cyclic for a moment  to let cyclic self-center, center pedals, reduce collective. Take back cyclic, after moment, and hold aircraft in straight attitude, as it slows down. Avoid IAS above 135 knots.

Vne-Velocity Never Exceed

We can rule out trimming as I only do that when leveled but what you say about the retreating blade stall makes total sense. Unfortunately, at the moment there are both voices saying that it is normal for the Apache at 130+kts as well as that it's the incorrectly modelled FM. Of course the whole flight model is still very much a WIP but I'd love to know if the above behaviour is actually on ED's radar labelled as "incorrect" or not.

Edited June 18, 2022 by Belphe

[Floyd1212]

2 hours ago, Belphe said:

at the moment there are both voices saying that it is normal for the Apache at 130+kts

I'm not sure anyone is suggesting it is "normal" for the the real Apache to behave this way.  It has been suggested in many threads that this behavior above 130 is an issue.

[Scaley]

On 6/16/2022 at 3:25 PM, Belphe said:

Coming back from a sortie, no weapons, no ammo, 1050 fuel left. Cruising 130kts at 15ft above the Syrian plain. Dodging occasional vegetation with smooth, long turns. Keeping the velocity vector centred, flying nose to tail trim. After a 5-second right turn that required a little bit of a pitch up I roll left and follow with a slight left rudder to keep the velocity vector in the centre. Suddenly SCAS starts screaming and I discover that centring the stick does not stop the left roll. I counter with full right roll but it's too late and too low.

Is this normal and if so, what am I doing wrong?

https://www.youtube.com/watch?v=4JcEFcJlwvQ

OK, so having watched you video on 0.25 speed several times here's my thoughts (and much easier to understand since you had the control indicator showing):

That wasn't anything I'd describe as "smooth long turns". The preceding few banks were pretty aggressive. 130kts at 15' is a very low/fast combination, and leaves you zero error margin. For NOE flight aim for 25' at 70-100kts at least to start with. 

The terminal manoeuvre begins as a roll-out of a right bank to level flight. You do this with quite aggressive left cyclic, with some forward cyclic blending in as you reach zero bank angle, and some left pedal. As you pass zero bank angle you try to stop the (now quite rapid) right roll with right cyclic, but also apply full (as in as far as it will go) forward cyclic. This essentially completely unloads the rotor disk. The DCS helicopter flight model REALLY does not like no load on the disk, and treats this as causing reduced control response. At this stage you still have a small amount of left pedal in. At the 15second point as you pass about 80deg of left bank you aggressively apply full right pedal. Since the Apache tail rotor is above the centreline of aircraft it will roll the aircraft in the opposite direction to the applied pedal, so in this case the right pedal increases your left roll rate. Essentially you now have the tail rotor (which is still effective) rolling the aircraft left, and the cyclic (which due to low disk load is not effective) rolling the aircraft right. The tail rotor wins, although even if you had not applied the right pedal I don't think that would have been recoverable with the amount of forward cyclic you had applied.

Summary factors as I see them:

1) Manoeuvring too aggressively at too high an airspeed with too little altitude

2) Unintentional / instinctive forward cyclic application without understanding what this will do to disk loading

3) Unintentional / instinctive right pedal application to try to keep the nose up without understanding what this will do to roll

2 and 3 look like fast-jet reflexes that are probably conditioned in from flying things without rotors.  

Hope that helps! 

Edit: This is not retreating blade stall at 40% torque and 130KTAS. Even with the current DCS flight model you can get a sustained 150kts dive out of the aircraft at lowish collective. 

Edited June 19, 2022 by Scaley

[DmitriKozlowsky]

6 hours ago, Belphe said:

We can rule out trimming as I only do that when leveled but what you say about the retreating blade stall makes total sense. Unfortunately, at the moment there are both voices saying that it is normal for the Apache at 130+kts as well as that it's the incorrectly modelled FM. Of course the whole flight model is still very much a WIP but I'd love to know if the above behaviour is actually on ED's radar labelled as "incorrect" or not.

 

I only had AH-64D for few weeks. Flying it for sim flight training only thus far. There are quirks, but there are also correct real world behaviors that are reproduced in AH-64D. At least as those appear to me. Weakest area of control is yaw authority and oversensitivity of pedal. Transitioning from slow fwd OGE flight into slow/hover IGE is messy. Roll and yaw become noisy.   Since I am throwing out acronyms I add another one. OMG! The task of transitioning from OGE slow flight at 12 knots parallels with Tarawa, to coming into pre-land IGE hover over deck spot is difficult.

All RL main/tail rotor helicopters will suffer twitchiness of of cyclic and pedals when hovering or slow fwd  below ETL flight with tail wind. Which I observe is happening in sim. While I have yet to experience LTE in the sim, with relative wind from aircraft's 10 O'clock, I do have to use more pedal. So my approaches to landing spot, once near ETL limit of 20-25 knots, are done with wind from my 12. That is the safest technique, and one that is taught in RL helicopter schools. I have now watched and rewatched every video from Helicopter Ground School Coffee With Kenny video and Helicopter Lessons in 10 Minutes or Less! on YT.  The guy behind 'Helicopter Lessons in 10 Minutes or Less!' , I beleive, is a former Army rotary aviator.

It feels to me that DCS:AH-64D helicopter FM does not like crosswinds , hates tailwinds, and partially affected by 'danger zone' from 10 o'clock wind requiring stronger left pedal.

[bradmick]

The aircraft in sim was nowhere near VNE. That’s around 180 to 190 knots true airspeed or so. The helicopter is extremely stable at high speeds in real life. You can very safely and easily do low altitude, and aggressive maneuvering with ease, your only concern is over torquing the helicopter with aggressive movements of the flight controls, and failing to establish the nose above the horizon prior to initiating said low altitude and aggressive maneuvers (the nose likes to tuck, or drop in turns) which will catch the unwary pilot flat footed and lead to a CFIT event. Never once worried about LTE in the real thing either. I definitely worried about over torquing because of my he pedals though on hot days with a heavy aircraft. The pedal mapping is actually very close to the real helicopter also, like, very very close. There’s definitely no lack of authority in them either. The real aircraft will also shrug off tail winds < 10 knots with ease, depending on power available. 
 

Remember, it’s all work in progress. 

[DmitriKozlowsky]

5 hours ago, bradmick said:

The aircraft in sim was nowhere near VNE. That’s around 180 to 190 knots true airspeed or so. The helicopter is extremely stable at high speeds in real life. You can very safely and easily do low altitude, and aggressive maneuvering with ease, your only concern is over torquing the helicopter with aggressive movements of the flight controls, and failing to establish the nose above the horizon prior to initiating said low altitude and aggressive maneuvers (the nose likes to tuck, or drop in turns) which will catch the unwary pilot flat footed and lead to a CFIT event. Never once worried about LTE in the real thing either. I definitely worried about over torquing because of my he pedals though on hot days with a heavy aircraft. The pedal mapping is actually very close to the real helicopter also, like, very very close. There’s definitely no lack of authority in them either. The real aircraft will also shrug off tail winds < 10 knots with ease, depending on power available. 
 

Remember, it’s all work in progress. 

Sounds good. If possible, please post a screen grab of curves on pedals. To me pedals and yaw is most problematic.

[Nealius]

Concur with others, this sounds like retreating blade stall. 

The only time I have ever crashed in the Apache (as in non-combat loss) it was from LTE while transitioning to hover in a hot, high, and heavy situation.

[FireHazard]

Great video. I think somebody mistook the Apache for the Viper ...although although I have no frame of reference, as others have mentioned, things get a bit iffy even in straight and level flight 130k + .... in sim I actually seldom fly above 130 as she'll do her own thing too easily. 

[DmitriKozlowsky]

1 hour ago, FireHazard said:

Great video. I think somebody mistook the Apache for the Viper ...although although I have no frame of reference, as others have mentioned, things get a bit iffy even in straight and level flight 130k + .... in sim I actually seldom fly above 130 as she'll do her own thing too easily. 

135 KIAS is fastest I gotten her. Thats with 60% fuel and no stores. I placed engine in MAX performance mode, 99% power over Nevada flat terrain. I would love to see how others get 150 knots indicated out of current build. At above 130, I have a definite deflection on cyclic to the right side, and thats how I trim. Above 130 knots she wants to roll to the left. At that air speed my pedals are neutral. The slipstream is doing the work of anti-torque. So above 130 to 135 knots I feel retreating blade stall coming on, but not there yet. Which is fine. I bet if somebody starts at angels 5 or 7 already at 130, bottoms the collective, and fwd cyclic slowly into 15 deg, dive , like its a fixed wing, she will exceed 150 knots, and depending on air temp and GW, above 150, perhaps at 155, retreating blade will stall hard. The helicopter will roll to the left uncommanded. I think that some don't realize that knots is not MPH. 150 knots is 173 mph. Some view 150 knots as 150 mph. It aint.

Now Apache is not a dinky little Robinson. Just punch up R44 Tragedy on YT. Multiple RL mishaps, where pilot dove the R44, exceeded Vne, causing a retreating blade stall, which then flapped and sliced its own tail boom off. Followed by R44 breaking up in flight with rotors breaking and flying off, long before the fuselage impacted the ground. Lethal every time.

Helicopters cannot be flown like fixed wing aircraft. Helicopters are flown as helicopters.  Except for few crazy pilots flying BO105 and AW Lynx. Doing loops, Immelmanns, snap rolls. Just crazy.

 

[Vakarian]

45 minutes ago, DmitriKozlowsky said:

135 KIAS is fastest I gotten her. Thats with 60% fuel and no stores. I placed engine in MAX performance mode, 99% power over Nevada flat terrain. I would love to see how others get 150 knots indicated out of current build.

 

Get out of Nevada for starters

[Floyd1212]

5 hours ago, DmitriKozlowsky said:

I bet if somebody starts at angels 5 or 7 already at 130, bottoms the collective, and fwd cyclic slowly into 15 deg, dive , like its a fixed wing, she will exceed 150 knots, and depending on air temp and GW, above 150, perhaps at 155, retreating blade will stall hard. The helicopter will roll to the left uncommanded.

I have exceeded 150+ in the sim plenty of times flying down the other side of a mountain in the Caucasus map.  I think I hit 170 briefly last night before leveling out at the bottom of the valley.  You can counter the left roll tendency with a bit of right stick if you are gentle with the controls, and the ground isn't rushing up to meet you.

[admiki]

When it was released, I was able to get to 205 in a dive. It was hard to keep it going where I want, but no RBS. Haven't tested it for some time now, so IDK how it behaves now

[Scaley]

I think there is some confusion here about the difference between retreating blade stall and dissymmetry of lift. As you accelerate obviously the advancing blades (on the right) generate more lift and the retreating blades (left) generate less. To counter this you apply right and forward cyclic and the aircraft stays level.  

Blade STALL is when increasing angle of attack (by cyclic) does not increase lift, but rather decreases it. If you apply right and/or forward cyclic and it works you are not at retreating blade stall, you are just flying. If you apply right or forward cyclic and it produces left roll then you have retreating blade stall.  

[Raptor9]

1 hour ago, Scaley said:

I think there is some confusion here about the difference between retreating blade stall and dissymmetry of lift. As you accelerate obviously the advancing blades (on the right) generate more lift and the retreating blades (left) generate less. To counter this you apply right and forward cyclic and the aircraft stays level.  

Not quite. Dissymmetry of lift is indeed caused by directional airspeed, but this will not lead to left roll in forward flight. Rather, due to gyroscopic precession and the rotor system wanting to "up-flap" over the nose; this will result in the nose wanting to come up and is countered with forward cyclic. Applying forward cyclic causes the airfoil on the right (advancing side) to reduce its angle of incidence (reducing the amount of lift produced), and the left (retreating side) to increase its angle of incidence (increasing the amount of lift produced). This net effect equalizes the lift across the two halves of the rotor system, maintaining the rotor disk angle at the desired tilt relative to the airframe, and counters the associated attitude change.

The aggressive left rolling motion that is occurring in the DCS AH-64D at high speed when forward cyclic is applied to pitch down is not because of dissymmetry of lift nor is it because of retreating blade stall, it's an inaccurate behavior within the flight model. Whether its from an inaccurate calculation of aerodynamic or mass effects in the DCS AH-64D, I have no idea, but it should not be happening, or at the very least, nowhere near the degree it is currently.

On the topic of retreating blade stall...as the various stall regions develop in high speed flight (which happens much higher than 150 knots by the way), what should happen is the nose will pitch up along with left roll and vibrations. The primary cause of the pitch up motion is because of the dissymmetry of lift and gyroscopic precession as explained above. There's some additional factors at play as well, such as blade flapping, but I'm probably not smart enough to go into that level of detail on the topic. But due to the aircraft characteristics, you should have to work very hard to get into retreating blade stall in the AH-64.

Edited June 20, 2022 by Raptor9

[DmitriKozlowsky]

12 hours ago, Vakarian said:

Get out of Nevada for starters

I have similar top end speed in Georgia, PG (over flat desert), over Cyprus , and Normandy. Cyprus becoming my favorite flying area. It seems to me, that turbulent winds around mountains and hills are implemented in this area to higher degree then other theaters. There are powerlines strung across valleys, and makes for challenging RL-like helicopter flight.

[PilotNeedGunner]

Um... I don't know if you've noticed, but your turns were definitely not "smooth", because your stick deflection was dancing all over the input display rectangle. What you've experienced was most likely a low G condition (different from mast bumping) because you pushed the cyclic forward during every turn recovery. The difference between the last turn before the crash and all the previous turns was that you had a rapid left cyclic input then a sudden max forward cyclic deflection without recentering the stick or made any attempt to apply aft cyclic to reload the rotors. Helicopters don't take kindly to wild cyclic deflections.