Jump to content
aluc4rd

Advanced Helicopter FDM Feedback

Recommended Posts

Found this in DCS huey POH. It really feels like this is missed which is odd considering it's suppose to be 2030 something. "Without this characteristic instability would develop as engine output is increased resulting in N1 speed overshooting or hunting the value necessary to satisfy the new power condition." Damn n1 won't keep up! Not only that but it creates WAY too much yaw as RPM constantly fights to keep up.

2.4.5. Droop Compensator

A droop compensator (#1 in Figure 2.8) maintains engine rpm (N2) as power demand is increased by the pilot. The compensator is a direct mechanical linkage between the collective stick and the speed selector lever on the N2 governor. No crew controls are provided or required. The compensator will hold N2 rpm to 6600 rpm when properly rigged. Droop is defined as the speed change in engine rpm (N2) as power is increased from a no-load condition. It is an inherent characteristic designed into the governor system. Without this characteristic instability would develop as engine output is increased resulting in N1 speed overshooting or hunting the value necessary to satisfy the new power condition.

Share this post


Link to post
Share on other sites
Good god what do you fly? I've 60 K/L/M friends who don't hover that high, my 47 friends laugh at our hover requirements...only thing I know of so power limited these days are -58s.

-60R. Without a dipping sonar system (ALFS) or Hellfire, a crew of 3 and a full bag of gas is ~20.3K. With a dipper and/or Hellfire (and GAU/M240, ammo, etc), it goes up towards past 21K and a HIGE of over 100% is normal. The < 80 knot limit is 120% continuous Tq, though, but often Ng is the bigger factor. Honestly, a lot of this could be solved if we kept our tranny and replaced our aircraft with -401D or -701D engines and the Mike blades, but that would probably be too forward thinking.

Found this in DCS huey POH. It really feels like this is missed which is odd considering it's suppose to be 2030 something. "Without this characteristic instability would develop as engine output is increased resulting in N1 speed overshooting or hunting the value necessary to satisfy the new power condition." Damn n1 won't keep up! Not only that but it creates WAY too much yaw as RPM constantly fights to keep up.

2.4.5. Droop Compensator

A droop compensator (#1 in Figure 2.8) maintains engine rpm (N2) as power demand is increased by the pilot. The compensator is a direct mechanical linkage between the collective stick and the speed selector lever on the N2 governor. No crew controls are provided or required. The compensator will hold N2 rpm to 6600 rpm when properly rigged. Droop is defined as the speed change in engine rpm (N2) as power is increased from a no-load condition. It is an inherent characteristic designed into the governor system. Without this characteristic instability would develop as engine output is increased resulting in N1 speed overshooting or hunting the value necessary to satisfy the new power condition.

The MD-5xx helicopters use a Fuel Control Unit (FCU), which measures ambient conditions and schedules the correct amount of fuel for the given power setting. The newer ones may be more advanced than that, but the same basic idea still stands. For the larger helos that use GE engines, they use Engine Control Units (which are now digital and have lots of extra functions) which both tell the HMU (a really complicated FCU) what to do as well as manage other finite engine conditions and greatly reduces breaking engine limits.

Several of the functions (among many) of the DECUs is to both apply a "Limited Droop Improvement" (LDI), which is specifically for low power to high power conditions, as well as prevent hunting (each DECU working with the HMU and talking to one another). The single biggest limitation of the DECUs I've encountered is the wiring when it gets wet (which it will), but beyond that, the two systems working together is pretty impressive, especially coupled with the AFCS to counter large Tq fluctuations.

Share this post


Link to post
Share on other sites
Now something for you real life UH-60 drivers to ponder. I was trying to figure out why the Ghosthawk rolls left during hover. Part of it seems like the neutral stick position is off. Even in straight line flight it rolled left. I can trim that effect out and in forward flight that left roll goes away. But in hover it always wanted to roll left requiring right cyclic. I also noticed applying more left pedal induced more left roll. Then I noticed something. The tail rotor is canted opposite from a real UH-60. If they actually applied that angle to the model, won't it roll left like it does?

What should happen is the tail-rotor should be pushing the tail in a counter-clockwise direction to counter torque. With the torque being counteracted, it should push the aircraft to the right; this is called translating tendency. To counteract this, left stick needs to be applied, causing the left wheel to be low.

For the ghosthawk the end aircraft position is correct, but the stick motion to get there is mirrored. The side that the tail-rotor is on doesn't matter, it is still pushing air the same way. That being said, with the reversal of the tail rotor on the Ghosthawk, it actually takes away lift when you add left pedal because its blowing the tail down... :rolleyes:

Like Gator said though, the cyclic movement for a hover in a hawk, and most helos with counter-clockwise spinning main rotors, is a "J" motion.

Here are some issues I've found:

The main and tail rotor RPM should be mechanically linked. That being said, I do like the fact that the wind turns the blades.

What little rotation you get from the blades turning in the wind is enough to taxi...

And a bonus glitch that may have been because of the wind being at max after a short "ground taxi"

2mR4XFaawbM

Share this post


Link to post
Share on other sites
What should happen is the tail-rotor should be pushing the tail in a counter-clockwise direction to counter torque. With the torque being counteracted, it should push the aircraft to the right; this is called translating tendency. To counteract this, left stick needs to be applied, causing the left wheel to be low.

Yeah, I'm actually in agreement with Worldsprayer, that it is probably a CG thing that is causing the left roll. The tail rotor is probably also off the CG as well inducing more roll as you increase pedal input. Since it is in line with the main rotors it should actually not have a tendency to induce roll with pedal input in hover. So something is off there. Translating tendency does behave correctly if you bring it level with a lot of right cyclic, the Ghosthawk does translate right.

Speaking of drifting right. The hellcat has a weird sideforce causing it to drift right, which you need to counter with a left lean in hover. It's a clockwise rotating helicopter so it should have translating tendency to the left as I give it right pedal to counter torque. But even in forward flight it sideslips right.

Edited by quickvenge

Share this post


Link to post
Share on other sites
-60R. Without a dipping sonar system (ALFS) or Hellfire, a crew of 3 and a full bag of gas is ~20.3K. With a dipper and/or Hellfire (and GAU/M240, ammo, etc), it goes up towards past 21K and a HIGE of over 100% is normal. The < 80 knot limit is 120% continuous Tq, though, but often Ng is the bigger factor. Honestly, a lot of this could be solved if we kept our tranny and replaced our aircraft with -401D or -701D engines and the Mike blades, but that would probably be too forward thinking.

The MD-5xx helicopters use a Fuel Control Unit (FCU), which measures ambient conditions and schedules the correct amount of fuel for the given power setting. The newer ones may be more advanced than that, but the same basic idea still stands. For the larger helos that use GE engines, they use Engine Control Units (which are now digital and have lots of extra functions) which both tell the HMU (a really complicated FCU) what to do as well as manage other finite engine conditions and greatly reduces breaking engine limits.

Several of the functions (among many) of the DECUs is to both apply a "Limited Droop Improvement" (LDI), which is specifically for low power to high power conditions, as well as prevent hunting (each DECU working with the HMU and talking to one another). The single biggest limitation of the DECUs I've encountered is the wiring when it gets wet (which it will), but beyond that, the two systems working together is pretty impressive, especially coupled with the AFCS to counter large Tq fluctuations.

This is further proof that the limit droop improvement is not working properly in the flight model.

Share this post


Link to post
Share on other sites
And a bonus glitch that may have been because of the wind being at max after a short "ground taxi"

2mR4XFaawbM

your video is on privat, so nobody can watch it. you should put it on not listet.

Share this post


Link to post
Share on other sites

I just haven't had the time to plug in my stick and throttle, but it sounds like I need to start doing so and play around. So the following is just clarifying what I said and not meant to argue one way or the other that things are not working...

This is further proof that the limit droop improvement is not working properly in the flight model.

To clarify, even with LDI, you can still get Np and Nr variations going from low power to high power (like at the bottom of an auto, which is pretty energetic). Depending on your auto technique, you can do a very safe and controlled auto and still get Nr to droop in the recovery. Along with that, you'll get some yaw deviations, but they're very controllable.

Again, I can't comment if the game model is way over-modeled compared to this, just giving some data. The spousal unit will start work next week, which I'm hoping will allow for a little more play time on the computer in the evening.

Share this post


Link to post
Share on other sites
Not a bug.

Complex to explain and this is not a correct thread. Send me a PM what exactly do you need to get.

See http://feedback.arma3.com/view.php?id=16115.

Is this to do with AI targeting vehicles? The only reason I can think of why model positions are calculated based on speed is if you need to hit them.

Presumably, the AI need some lead on targets, so you change the relative model pos for AI to be able to shoot at the target. Good to see you guys are putting in the new vector based UI commands so the real positions can be worked out.

Share this post


Link to post
Share on other sites
your video is on privat, so nobody can watch it. you should put it on not listet.

Fixed. Not sure why it was set to private...

Share this post


Link to post
Share on other sites

I would like to draw your attention on two additional things:

Where are the camera shake effects / G-effects?

Months ago you anounced this as a new feature that will get implemented soon. I havent seen it until now, but helicopters could really benefit from the increased immersion.

The pilots view is often blocked by instruments!

Every tried to land the Orca? I cannot do that, because if I am trying to hover I cannot see anything in front of me, or what is under me. It would really be helpfull if the DEVs would adjust the camera positions higher for many of Arma 3's helicopters. At the moment it is all about luck if I don't hit something while landing. (Not even TrackIR would help, since the instrument panels of some helicopters are so huge that you cannot watch over them to see what is in front of you.)

Share this post


Link to post
Share on other sites
I would like to draw your attention on two additional things:

Where are the camera shake effects / G-effects?

Months ago you anounced this as a new feature that will get implemented soon. I havent seen it until now, but helicopters could really benefit from the increased immersion.

The pilots view is often blocked by instruments!

Every tried to land the Orca? I cannot do that, because if I am trying to hover I cannot see anything in front of me, or what is under me. It would really be helpfull if the DEVs would adjust the camera positions higher for many of Arma 3's helicopters. At the moment it is all about luck if I don't hit something while landing. (Not even TrackIR would help, since the instrument panels of some helicopters are so huge that you cannot watch over them to see what is in front of you.)

The camera shake effects are in, I've noticed them after lots of over torque. Maybe not all camera shakes, but they are in. And G-effects really dont apply to helicopters as we tend to keep the G forces around 1 as much as possible. Jets, well I dunno about the G effects there.

As far as the instrument panel goes for the helicopters, I dunno what to tell you. Thats how most helicopters are designed, if you practice and become familiar it becomes much easier. But I don't think its going to change, or should change. Its from the pilots POV and thats what the pilot sees when they fly.

Share this post


Link to post
Share on other sites
I would like to draw your attention on two additional things:

The pilots view is often blocked by instruments!

Every tried to land the Orca? I cannot do that, because if I am trying to hover I cannot see anything in front of me, or what is under me. It would really be helpfull if the DEVs would adjust the camera positions higher for many of Arma 3's helicopters. At the moment it is all about luck if I don't hit something while landing. (Not even TrackIR would help, since the instrument panels of some helicopters are so huge that you cannot watch over them to see what is in front of you.)

fly in the endaproatch a bit sidways and look 45° side and down the cockpit-window so you can see more from the ground and its a bit easier.

trackir or freetrack or this: http://forums.bistudio.com/showthread.php?172297-Head-Tracking-for-under-10%E2%82%AC/page22

can help you alot.

Share this post


Link to post
Share on other sites

So I tried the dev branch today and the new rotorlib for the first time and I must say that I am intrigued. Having just bought DCS Huey and Black Shark 2 and started my heli career there I've found this very interesting. Keep in mind, I don't have many hours of flight in DCS behind me yet, maybe 10 hours or so trying to wrap my head around it, but there are a few differences from the get go that was a bit strange.

First of all, using the trim was rather difficult, much more difficult than DCS due to the lack of visual refrerence (the small hud thing that DCS has as an option is really helpful). Also, all of the choppers beheaved differently than I expected when taking off. I struggled really really hard to achieve a ground hover on ground effect, and the chopper seemed to bank very steeply to the right (way more than expected). I don't know if this is just me, but it felt hard to control. At the same time, the input lag people talk about was difficult to cope with as well.

Flying with autotrim on and using mouse and keyboard was actually much easier. I felt I had better control, and I was able to be more efficient with the Littlebird than I was with the stick. It does take some time to unlearn the old ARMA flight model techniques though, especially the throttle control, but I welcome the challenge. I was rather decent at flying prior to the model change, now I feel like a complete noob again :P

All in all, my impressions is that we are on the right path, but in terms of how it should behave I will leave that to the people here who knows their stuff.

Share this post


Link to post
Share on other sites

I also finally had a chance to mess around with last night, albeit with only a keyboard. b00ce is correct...PLEASE fix the model when you flare. Nr should spike, not droop. It was getting ridiculous in the LB. The Ghosthawk was better, but still painful.

Share this post


Link to post
Share on other sites
I also finally had a chance to mess around with last night, albeit with only a keyboard. b00ce is correct...PLEASE fix the model when you flare. Nr should spike, not droop. It was getting ridiculous in the LB. The Ghosthawk was better, but still painful.

Yep.

VbuVgdyMysE

Share this post


Link to post
Share on other sites

Wow auto-trim does wonders. Turned it on and it filtered some of the really weird behavior on some choppers. Like the Orca doesn't roll left or right heavily due to collective changes. Ghosthawk roll tendency to the left is much less pronounced. Hellcat still drifts right though, and needs a little left roll to counter.

The rotor torque models have been broken since TOH, but on the xH-6 now it is so much more pronounced as it is so easy enter a low RPM state and basically fall out of the sky no matter how fast you are going. I really hope they fix the rotor torque models and not just give it more power so you don't see the RPM drop.

BI brought up that we have a transmission that we can damage/destroy in one of the SITREPs. They should at least model proper ways to damage it, and provide working torque and rpm gauges instead of a blinking indicators. I want the tranny to be rip apart because of a rapid left cyclic input while at high speed when near max torque already for example. But I don't want it damage because I simply pulled out of a dive or executed a g-loaded turn resulting in a weird over torque condition.

Share this post


Link to post
Share on other sites
I would like to draw your attention on two additional things:

Where are the camera shake effects / G-effects?

Months ago you anounced this as a new feature that will get implemented soon. I havent seen it until now, but helicopters could really benefit from the increased immersion.

The pilots view is often blocked by instruments!

Every tried to land the Orca? I cannot do that, because if I am trying to hover I cannot see anything in front of me, or what is under me. It would really be helpfull if the DEVs would adjust the camera positions higher for many of Arma 3's helicopters. At the moment it is all about luck if I don't hit something while landing. (Not even TrackIR would help, since the instrument panels of some helicopters are so huge that you cannot watch over them to see what is in front of you.)

Have you tried holding left control, and using number pad keys 4-5 to move the cockpit camera . Might help some.

Share this post


Link to post
Share on other sites

EXE rev. 126856 (game)

Quick test of the new Ghosthawk. It now has a pronounced right roll instead of the left roll(auto trim off). I think it requires a little less cyclic to correct than the original left roll, but still too much in my opinion. Also it's engines seem to have lost power, and is a lot more anemic. The pedal behavior particularly during forward flight is better in general though, but I think it still has a tendency to create a roll effect in hover. Though that may just be due to the right roll bias.

Edited by quickvenge

Share this post


Link to post
Share on other sites
EXE rev. 126856 (game)

Quick test of the new Ghosthawk. It now has a pronounced right roll instead of the left roll(auto trim off). I think it requires a little less cyclic to correct than the original left roll, but still too much in my opinion. Also it's engines seem to have lost power, and is a lot more anemic. The pedal behavior particularly during forward flight is better in general though, but I think it still has a tendency to create a roll effect in hover. Though that may just be due to the right roll bias.

The roll doesn't feel pronounced enough to me (Maybe one or 2 degrees off), though it is already orders of magnitude more realistic than previous iterations. The pedals seem less squirrely, that being said it is still entirely too sensitive and has way too much authority. Forward and aft cyclic still produce torque (Causing a terrible yaw during ground taxi, which can be catastrophic) from nothing. The previous two issues make roll-on landings damn near impossible using realistic speeds and approach procedures. Below 50% collective the aircraft still rolls backward. When pulling anything from 50-75% collective on the ground, the aircraft will jump and then settle back down to earth, ground effect seems to only take effect at about 10 or so feet when in reality it should be about one rotor disc.

I agree completely about the lack of power, each GE T700-701D engine produces ~1,900 shaft horsepower. Bear in mind that's the current generation, Lima model and Alpha model were ~1,800 and ~1,600 respectively. If they keep the same level of progress the engines in the Ghosthawk should produce anything from 2,100-2,200 SHP on top of the numerous improvements to the transmission that they come up with. At sea level it shouldn't take more than 65% torque. Not sure how that translates in game because we don't have a torque indicator... Suffice to say, the collective setting shouldn't be anywhere near where it is in game.

Share this post


Link to post
Share on other sites
The roll doesn't feel pronounced enough to me (Maybe one or 2 degrees off), though it is already orders of magnitude more realistic than previous iterations.

Wow it really rolls that much? Shouldn't it go away or be less in forward flight though?

Share this post


Link to post
Share on other sites
Wow it really rolls that much? Shouldn't it go away or be less in forward flight though?

You have to create that roll with cyclic motion in a hover to counteract the tail blowing you sideways.

Share this post


Link to post
Share on other sites
You have to create that roll with cyclic motion in a hover to counteract the tail blowing you sideways.

Oh it isn't the left roll that you have to induce to keep the Ghostwak in hover that I was concerned about. It's the helicopter's natural tendency to roll right that I was criticizing. With the cyclic at its neutral position laterally, it wants to roll right with auto-trim off. So not only do you need left cyclic to correct translating tendency in hover, you needed left cyclic to counter its right roll. Even during forward forward flight you need left cyclic to counter its tendency to roll right.

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now

×