Board index FlightGear Support Flying Helicopter

Helicopter realism -- or not? Real pilot tests?

Helicopter flying is completely different from flying a fixed-wing aircraft and thus requires different skills.

Helicopter realism -- or not? Real pilot tests?

Postby flywisher » Thu May 27, 2010 5:22 pm

I have been enjoying FG very much for a couple of years. (A big THANK YOU! to all the developers and modellers! This thing is Wonderful!) I have flown many aircraft quite well. Lately I have been exploring helicopters again. As to their realism: for some models this is plausible. For others, I cannot believe it.

I recently succeeded in getting my joystick configured in FG. That was an interesting story in itself (for another post). Now it works well. Its twist "rudder" ability makes helicopters much more manageable than using only the mouse. (THANK YOU John Check for the "Users Guide to FGInput - Joystick an Keyboard Bindings for FlightGear!")

But...

I find the Bo105, S58 & S76 challenging but manageable. I want to fly the Robinson R-22. Try as I might, this thing is so unpredictable and violently erratic I find it completely unmanageable. The S51 responds more slowly and is much more manageable, but both of them constantly and readily get into a state of spinning around very fast to the right (always to the right), ignoring my hard left "rudder" input, oscillating violently side-to-side, then tumbling around and over until they crash.

Now here's what makes me suspicious of this. I live in San Diego, California. We are surrounded by a dozen airports large and small. Helicopters of all types and sizes, Navy, Marines, civil, Coast Guard, police, news media, LifeFlight (medical), fire-fighting, private and training fly around us every day of the week, and nighttime as well.

Five miles north of us, Corporate Helicopters flies Robinson R-series machines out of Montgomery Field. They do training, so many of these flights are students.

Every helicopter -- every single one -- moves very smoothly. They are very stable. If real helicopters behaved as unpredictably and erratically as the FG models, particularly the Robinson R-22, we would see them moving erratically as the pilots struggled to keep them under control. This does not happen. So I really question whether our R-22 model is properly configured.

This led me to wonder about my joystick. Whether low quality potentiometers were throwing spurious inputs. This hasn't been a problem with fixed-wing craft, but I explored it anyway. I learned about the "tolerance" and "factor" properties in the xml files. Those seem to help a little.

I saw a posting about the UH-1 being one of the most realistic in FG and tested by a real pilot. But in my machine it has a frame rate of 3. That's right -- 3 fps. So it's un-usable for me. Most aircraft run around 30-40 fps. That makes a usable simulation. As long as the simulation is reasonably accurate.

Are there any real helicopter pilots out there who have tried the Robinson R-22 in FG? Can anyone find a pilot and ask them to try it? I would really like to know. For my part, I will see if someone at Corporate Helicopters will try it on my laptop and give an opinion.

Incidentally, for anyone reading this who is near San Diego, C-H has a full-size, sit-in-it FAA-certified simulator that simulates the R-22. (www.corporatehelicopters.com. Release your script blocker and look under "Flight Training / Flight Simulator" for vivid photos of it.) Not sure I can afford it. But that would be a guaranteed realistic simulation.

Thanks.
Erick

My system:
HP Pavilion a375c; Pentium 4, 3 GHz; RAM: 2 GB; Video: GeForce FX5700LE, 256 MB
Windows XP; DirectX 9.0c
Joystick: Interact Raider Advanced FX
flywisher
 
Posts: 133
Joined: Mon Jun 01, 2009 2:37 am
Location: San Diego, CA USA
Callsign: flywisher
OS: XP, Vista, Mint

Re: Helicopter realism -- or not? Real pilot tests?

Postby HHS » Thu May 27, 2010 6:57 pm

Hi Erick,

Indeed FGFS is able to simulate Helicopters much realistic- means the typical flight behaviour of a helicopter is simulated correct.
Torque effect, Transverse lift, rotor stall, Ground effect... all simulated correct.
And of course all different rotor systems- and there are a lot and they even differ in behaviour.

But like all models our helicopter depends a lot on the datas available. To configure a heli is often much more complicated than an aircraft.

For the Bo105, Bell UH-1 and Aircrane we found free available, very detailled flightdatas which we could used.
Bo105 and Bell-UH1 have been already tested, and both comes quite close to realitity.

Helicopters are complicate to fly- they have a natural tendency to fly in any direction they want- unlike aircrafts. It is like balancing a ball on ball, which is sitting on another ball. That's why you need much longer to learn. And the most important instrument which could help is useless: your butt! :wink:

Therefor you need sensitive hardware- good joystick, pedals, throttles. Mouse isn't recommended!
A twist-rudder joystick is the minimum.

Bo105 has a rigid rotorhead- that means the rotor reacts quite fast to the inputs of the pilot (0.2 sec).
This is simulated really good.

The s58 and S76 have fully articulated rotorheads- the rotor reaction is slower, but the heli itself has more weight so it is manageable for beginners.

Regarding R22: well, here are some datas missing. Weight may be right, but balance and the so called tensors ( how sluggish a body is reacting on forces) are certainly not yet right.
Then: the R22 has so called teetering rotorhead - it reacts quite slow to the inputs. Due to the light weight of the R22 it needs small inputs, and heavy use of the pedals.
Even in real life this heli is quite challenging in many ways and only used by beginners due to its cheap price.

Currently I only know one helicopter pilot here, jag007 and he knows the Bell UH-1. Unfortunately no R22-pilot or other. Help for this model with help of real life pilots is highly appreciated! :D .

UH-1:
We need also to know:

- FlightGear Version
- Which version of the UH-1-

the fdm of it is currently in revsion. We had in the last couple of weeks three different fdm's. Two of them- the first one and the lastest seems to make problems on some linux distris- we guess there is a bug somewhere in YASim/ Nasal or in the compilation method.

But when I look at your system specs: Geforce 5700 is quite low, and DirectX isn't needed. You need the latest OpenGL-Driver and 3d-accleration. FGFS 2.0.0 will work better with a dexcent GPU.

Hope I could help a bit
HHS
Up, up and away
User avatar
HHS
Retired
 
Posts: 3624
Joined: Thu Jul 19, 2007 8:09 am
Version: GIT

Re: Helicopter realism -- or not? Real pilot tests?

Postby Groucho » Fri May 28, 2010 2:54 pm

Just a quick pop up by me- my last comment in here is some time ago. Real life has caught me and is still keeping me busy a lot and actually I am grounded so greetings to those who missed me :wink: .

For the R22 I am propably the only one in the FG community who frequently enjoyed flying it and one of the very few guys who can master it.:)
This little bee is rather a big challenge. It suffers from several issues: Extremely light, rather high positioned main rotor, low rotor speeds with long blades and an extreme input sensitivity to control changes. And now additionally think about windy conditions: Welcome to your personal helicopter hell (indeed a real hell-i-copter).

Whether you consider it realistic or not, it is far from being unpredictable but in fact does what you expect from a helicopter behavior and with the right amount of care (and with amount, I mean very small control steps) this thing can be mastered.
A hint every chopper pilot will give you btw: You need a relaxed position for your controlling arm, otherwise it will get tired soon and you will loose sensitivity. And if your stick has a spring to bring it back to mid-position- remove it as soon as possible.

Hope to see you guys some day back in the air.
_____________________________________
Callsign: D-HARP
Flight locations: TNCM, TKPK, EDNY, LOWI
Aircrafts
Helis: EC130, Bo105, UH-1, R22
Twins/Jets: Aerostar, CRJ-200
User avatar
Groucho
 
Posts: 1216
Joined: Sun Jun 17, 2007 10:28 am
Location: Airborne Lake Constance/Germany
Callsign: D-HARP
Version: GIT
OS: Kubuntu

Re: Helicopter realism -- or not? Real pilot tests?

Postby flywisher » Fri May 28, 2010 8:29 pm

Thank you HHS for your comments. You're right.
Thank you Groucho, for yours. Excellent timing. Great to hear from you! (I've been dealing with that "real life" stuff too. Know what you mean. Wish you all the best.)

A big "Oops!" on me. I run FG 1.9.1 on my desktop machine.

Now... Ta Daaa! GREAT NEWS! I DID IT!

Yesterday, I managed to visit Corporate Helicopters and talk with one of the flight instructors. He graciously agreed to look at my laptop FlightGear and share his opinion of it. The laptop currently has FG 1.0.0, slower CPU, much less video RAM, etc. But using the same joystick and config file, the Bo105 and R-22 models behave about the same as on my desktop machine.

This man trains people in Robinson R-22's. He flies them all the time. In fact, he had just landed from giving a lesson in one when we met. With my FG model, he did way better than I ever have. He managed to more or less hover it for a minute or two and keep it pretty much under control. I have so much trouble with it that I was astonished watching this. I did not believe it possible until I saw it. That said, it got away from him too, and crashed.

He was intrigued. He laughed and agreed that the FG model is not quite realistic. It is significantly more unstable and erratic than the real one. For comparison, I had him try the Bo105, which is much better.

Then things got more interesting.

I pointed out to him the R-22 model's habit of tilting to the right on liftoff, and the Bo105's habit of tilting to the left. Guess what? He told me this is actually realistic. Real helicopters DO lift one skid before the other. This is caused by the pressure from the tail rotor, and the direction of tilt depends on which side the tail rotor is on. The Bo105's tail rotor is opposite the R-22's. Also, he said it is quite possible, easy even, for a novice to flip an R-22 over trying to liftoff.

I argued that the R-22 model's nasty habit of spinning violently around to the right regardless of hard left "rudder" is unrealistic. He agreed.

I argued that the R-22 model's habit of oscillating violently left and right in the air is also unrealistic. He surprised me by agreeing that it is exaggerated, but added that if you put anyone un-skilled with helicopters in an R-22 and they succeed in getting it off the ground, they will quickly have it oscillating in a similar manner. Just not as violently. This is due to the beginners' inevitable over-controlling, together with the pendulum effect of the aircraft hanging from its main rotor. In fact, they call it "pendulum-ing," and it's quite common.

He noted that both models are a little overly sensitive in tipping forward and back.

He also observed that, as HHS said, he relies heavily on "seat of the pants" feel in controlling helicopters, so non-motion simulators (including his own expensive FAA-certified version) are harder to fly than the real thing.

Then...

To really make the day a smashing success, I was able to arrange a 15 minute demo session in their FAA-certified, full size, sit-in-it, cockpit-mockup-in-a-miniature-theater training simulator. This one is made and marketed by "Fly-It" (you can see photos and videos on their website http://www.flyit.com). The website shows conventional separate stick models. This one has the same controls as our FG R-22 model. The simulator is in a small trailer, runs Microsoft Flight Simulator X on a PC, and has a screen that covers the entire wall a few inches in front of the cockpit mockup. My instructor said MSFS X's simulation is very realistic. Just ever so slightly harder to handle than the real thing. Of course, MSFS is not free, does not work on Linux, and requires you to keep a CD/DVD in the drive to use it. FlightGear is definitely more affordable and more friendly.

The immersion is so effective that when he demonstrated by lifting the "helicopter" slightly off the "ground" and letting it move forward, I could swear I felt movement pressures in my seat and back. I actually had to ask him whether they had that thing on motion. They don't. It's static.

Then he let me try it.

As I suspected, it is definitely smoother and more stable than any of FG's models. No abrupt tip-over-and-crash. No promptly-stand-on-your-nose-and-crash. No violently tip side-to-side and crash. Just very smoothly rise up and start moving slowly over the "ground." Given that, it was trickier to handle than I expected. Without his guidance (and a little gentle assistance) I would have soon "crashed" it. You definitely play a juggling game with all inputs at once. Left-and-right, fore-and-aft, up-and-down, and "rudder" pedals (they call them "torque" pedals) all going on at the same time and all influencing each other. Completely different than a fixed-wing craft. I can see his point that it's actually easier to learn in a real helicopter where you can feel the movement.

He repeatedly emphasized (as Groucho said above) that flying helicopters, especially the smaller ones, absolutely requires using LIGHT, GENTLE, SMALL actions on the controls. Like Groucho, my instructor told me to anchor my elbow against leg or ribs to allow more precise control with small movements, and to reduce fatigue. This is the number one challenge for learners. It's an exaggerated version of learning to ride a bicycle, unicycle, or skates. We inevitably over-control until, with practice, we develop a "sense" for the new phenomenon we're working with, and develop habits that compensate automatically. We simply cannot "think" and analyze and respond as fast and accurately as conditioned habits can.

He also told me that the FAA offers some downloadable training documents for free on their website, http://www.faa.gov. A pdf for rotorcraft is:
http://www.faa.gov/library/manuals/airc ... 083-21.pdf.

So there you have it. Right from "the horse's mouth," so to speak. Two "horses," actually. (Thanks again, Groucho, for your input.) FlightGear's realism is pretty darn good, but some models like the R-22 need some further adjustments.

Perhaps those of us who are interested can do some more research and help improve Syd Adams' and Maik Justus' lovely R-22 model.

Here's to more fly-ee, less crash-ee,
Erick

FG 1.9.1, WinXP; FG 1.0.0 Ubuntu Linux & Win2K
flywisher
 
Posts: 133
Joined: Mon Jun 01, 2009 2:37 am
Location: San Diego, CA USA
Callsign: flywisher
OS: XP, Vista, Mint

Re: Helicopter realism -- or not? Real pilot tests?

Postby HHS » Sat May 29, 2010 12:09 pm

Hi, some comments from me:

flywisher wrote:Yesterday, I managed to visit Corporate Helicopters and talk with one of the flight instructors. He graciously agreed to look at my laptop FlightGear and share his opinion of it. The laptop currently has FG 1.0.0, slower CPU, much less video RAM, etc. But using the same joystick and config file, the Bo105 and R-22 models behave about the same as on my desktop machine.

He was intrigued. He laughed and agreed that the FG model is not quite realistic. It is significantly more unstable and erratic than the real one. For comparison, I had him try the Bo105, which is much better.

Then things got more interesting.

I pointed out to him the R-22 model's habit of tilting to the right on liftoff, and the Bo105's habit of tilting to the left. Guess what? He told me this is actually realistic. Real helicopters DO lift one skid before the other. This is caused by the pressure from the tail rotor, and the direction of tilt depends on which side the tail rotor is on. The Bo105's tail rotor is opposite the R-22's. Also, he said it is quite possible, easy even, for a novice to flip an R-22 over trying to liftoff.

Nothing what surprise me. Did you already read the article about flying heli in our wiki?

The Bo105 in 1.0.0 and 1.9.1 aren't that realistic like the 2.0.0 model. But the common user won't notice much difference. The 2.0.0 model is based on measurements datas from NASA. The same for the UH-1, Aircrane and with changes the EC135.

I argued that the R-22 model's nasty habit of spinning violently around to the right regardless of hard left "rudder" is unrealistic. He agreed.

I argued that the R-22 model's habit of oscillating violently left and right in the air is also unrealistic. He surprised me by agreeing that it is exaggerated, but added that if you put anyone un-skilled with helicopters in an R-22 and they succeed in getting it off the ground, they will quickly have it oscillating in a similar manner. Just not as violently. This is due to the beginners' inevitable over-controlling, together with the pendulum effect of the aircraft hanging from its main rotor. In fact, they call it "pendulum-ing," and it's quite common.

He noted that both models are a little overly sensitive in tipping forward and back.


I took a closer look into the R22. The airfoil datas seems to be right (But my Yasim.rotor.excel seems to be broken...), but weight and balance seems to be completly wrong. Then I'm not sure if all measurements on the rotor are correct. I will try to check them when I find some time next days. Busy with real life as well.

.... The simulator is in a small trailer, runs Microsoft Flight Simulator X on a PC, and has a screen that covers the entire wall a few inches in front of the cockpit mockup. My instructor said MSFS X's simulation is very realistic.


*cough* Unless they use an external fmd or extended MSFS X's fdm I have to disgree! MSFS X isn't able to calculate the torque and other things for helicopter simulation! He should know this...


FlightGear's realism is pretty darn good, but some models like the R-22 need some further adjustments.

Perhaps those of us who are interested can do some more research and help improve Syd Adams' and Maik Justus' lovely R-22 model.

Here's to more fly-ee, less crash-ee,
Erick

FG 1.9.1, WinXP; FG 1.0.0 Ubuntu Linux & Win2K


I'll take a look into

Cheers and thanks for your interesting report- we need more users loving helis!

@Groucho:
Great to see you here! I hope you are fine!
Up, up and away
User avatar
HHS
Retired
 
Posts: 3624
Joined: Thu Jul 19, 2007 8:09 am
Version: GIT

Re: Helicopter realism -- or not? Real pilot tests?

Postby HHS » Sat May 29, 2010 1:44 pm

I took some minutes and looked at the rotor measurements. Airfoils and other specifications beside weight balances seems absolutly correct.

Only on the rotor measurements I would disagree- if Syd did a good work on the 3d-model then two values (rellenflaphinge and rel-len-blade-start) seems not to be correct. I tried to correct them and indeed the R22 behaves a bit nicer ;-)

What I coulden't change yet is the behaviour in the jaw-axis: I guess that's needs changes in the weight-balances settings, but therefor I need some time.

But you might all try the improved fdm (copy and paste in R22BetaII.xml) and see if this comes closer. For me it does (nearly to easy to fly ;-)):
Code: Select all
<!--
Robinson R22 Beta II specs:
Weights
Gross Weight 1,370 lb
Empty Weight Equipped (w/full oil) 855 lb
Fuel (19.2 gal) 115 lb
Optional Auxiliary Fuel (10.5 gal)    63 lb
MAXIMUM PER SEAT WEIGHT: 240 lbs
MINIMUM SOLO WEIGHT: 140 lb

Powerplant
Lycoming O-360 Four-Cylinder, Air-Cooled
Derated to 131 horsepower at 2652 rpm

Performance
Maximum Airspeed (Vne) 118 mph (102 kts)
Cruise Airspeed @ 70% Power 110 mph (96 kts)
Maximum Range (no reserve) Over 200 miles
Maximum Range (w/aux fuel) Over 300 miles
Average Fuel Consumption 8 to 10 gph
Rate-of-Climb at Sea Level Over 1,000 fpm
Rate-of-Climb at 10,000 Feet Over 600 fpm
Maximum Operating Altitude 14,000 feet
Hover Ceiling IGE @ 1370 GW 9,400 feet

Rotor
18 degree delta3
530 rpm
Airfoil NACA 63-015
teetering hinge ...
-->
<airplane mass="855">

<approach speed="50" aoa="16">
  <control-setting axis="/controls/engines/engine[0]/throttle" value="0.3"/>
</approach>

<cruise speed="102" alt="0">
  <control-setting axis="/controls/engines/engine[0]/throttle" value="1.0"/>
</cruise>


<rotor name="main" x="0.0" y="0.0" z="0.0" nx="0" ny="0" nz="1.0" fx="1" fy="0" fz="0" ccw="1"
  maxcollective="18" mincollective="0.0"
  mincyclicele="-2" maxcyclicele="2"
  mincyclicail="-10" maxcyclicail="10"
  diameter="7.67" numblades="2" weightperblade="15" relbladecenter="0.5"
  dynamic="0.8" rpm="530" rellenflaphinge="0.014" delta3="0.3"
  delta=".15"
  pitch-a="10"
  pitch-b="15"
  flapmin="-7.6"
  flapmax="14.7"
  flap0="-5"
  phi0="0"
  flap0factor="0"
  notorque="0"
  dragfactor="0.30"
  translift-ve="20"
  translift-maxfactor="1.2"
  ground-effect-constant="0.1"
  twist="-8"
  taper="1"
  chord="0.18288"
  number-of-segments="8"
  number-of-parts="8"
  rel-len-where-incidence-is-measured="0.7"
  rel-len-blade-start="0.06"

  airfoil-lift-coefficient="5.5"
  airfoil-drag-coefficient0="0.005"
  airfoil-drag-coefficient1="0.025"
  incidence-stall-zero-speed="19.25"
  incidence-stall-half-sonic-speed="10.5"
  lift-factor-stall="0.26"
  stall-change-over="7"
  drag-factor-stall="9.0"
  cyclic-factor="0.9"
  rotor-correction-factor="0.9"
  downwashfactor="0.25"
>
  <control-input axis="/controls/flight/aileron-trim" control="CYCLICAIL" split="true"/>
  <control-input axis="/controls/flight/aileron" control="CYCLICAIL"
                 src0="-1.0" src1="1.0"
                 dst0="-0.30" dst1="0.20"/>
  <control-input axis="/controls/flight/elevator-trim" control="CYCLICELE"/>
  <control-input axis="/controls/flight/elevator" control="CYCLICELE"
                 src0="-1.0" src1="1.0"
                 dst0="-1.0" dst1="1.0"/>
  <control-input axis="/controls/engines/engine[0]/throttle" control="COLLECTIVE"
                 src0="0.0" src1="1.0"
                 dst0="1.0" dst1="-1.0"/>
</rotor>


<rotor name="tail" x="-4.41" y="-0.2" z="-0.94" nx="0.0" ny="-1" nz="0.0" fx="1" fy="0" fz="0" ccw="1"
  maxcollective="30" mincollective="0"
  diameter="1.0668" numblades="2" weightperblade="2" relbladecenter="0.55"
  dynamic="0.80" rpm="3356" rellenflaphinge="0.01" delta3="1" translift="0" delta="0.5"
  pitch-a="10"
  pitch-b="15"
  airfoil-lift-coefficient="4.0"
  airfoil-drag-coefficient0="0.01"
  airfoil-drag-coefficient1="0.05"
  incidence-stall-zero-speed="19.25"
  incidence-stall-half-sonic-speed="10.5"
  lift-factor-stall="0.26"
  stall-change-over="7"
  drag-factor-stall="9.0"
  notorque="0"
  taper="1"
  chord="0.116"
  number-of-segments="5"
  number-of-parts="4"
  rel-len-blade-start="0.2"
  rotor-correction-factor="0.8"
  downwashfactor="0.3"

>
<control-input axis="/controls/flight/rudder" control="COLLECTIVE" invert="true"
    src0="-1.0" src1="1.0"
    dst0="-0.7" dst1="0.7"/>
</rotor>


<rotorgear
  max-power-engine="125"
  max-power-rotor-brake="10"
  rotorgear-friction="0.5"
  engine-prop-factor="0.005"
  engine-accel-limit="10"

  yasimdragfactor="10"
  yasimliftfactor="180"
>
  <control-input axis="engines/engine[0]/clutch-engaged" control="ROTORGEARENGINEON"/>
  <control-input axis="/controls/rotor/brake" control="ROTORBRAKE"
                 src0="0.0" src1="1.0"
                 dst0="0.0" dst1="1.0"/>
</rotorgear>


<cockpit x="0.5" y=".33" z="-1.2"/>

<fuselage ax="1.348" ay="0" az="-1.574"
  bx="-0.6" by="0" bz="-1.46" idrag="0"
  width="1.1" taper="0.7" midpoint="0.5"
  cy="1.2" cx="1.2"/><!--cabin-->

<fuselage ax="-0.55" ay="0" az="-1.22"
  bx="-4.308" by="0" bz="-0.928" idrag="0"
  width="0.36" taper="0.2" midpoint="0.0"/><!--tail boom-->

<vstab x="-0.01" y="0" z="-0.540" taper="0.7" length="0.716" chord="0.525">
</vstab>

<fuselage ax="0.065" ay="0" az="-1.57"
  bx="-0.464" by="0" bz="-1.57" idrag="0"
  width="0.7" taper="1.0" midpoint="0.5"/><!--engine-->

<fuselage ax="1.5" ay="-0.918" az="-2.495"
  bx="-0.465" by="-0.918" bz="-2.495" idrag="0"
  width="0.05" taper="1.0" midpoint="0.5"/><!--RH skid-->

<fuselage ax="0.9" ay="-0.427" az="-2.07"
  bx="0.9" by="-0.884" bz="-2.455" idrag="0"
  width="0.05" taper="1.0" midpoint="0.5"/><!--RH skid Frtleg-->

<fuselage ax="-0.365" ay="-0.427" az="-2.07"
  bx="-0.365" by="-0.884" bz="-2.455" idrag="0"
  width="0.05" taper="1.0" midpoint="0.5"/><!--RH skid Rrleg-->

<fuselage ax="1.5" ay="0.918" az="-2.495"
  bx="-0.465" by="0.918" bz="-2.495" idrag="0"
  width="0.05" taper="1.0" midpoint="0.5"/><!--LH skid-->

<fuselage ax="0.9" ay="0.427" az="-2.07"
  bx="0.9" by="0.884" bz="-2.455" idrag="0"
  width="0.05" taper="1.0" midpoint="0.5"/><!--LH skid Frtleg-->

<fuselage ax="-0.365" ay="0.427" az="-2.07"
  bx="-0.365" by="0.884" bz="-2.455" idrag="0"
  width="0.05" taper="1.0" midpoint="0.5"/><!--LH skid Rrleg-->


<vstab x="-4.408" y="0" z="-1.493" taper="1" length="1.06" chord="0.22" effectiveness="0.5">
  <stall aoa="16" width="4" peak="1.5"/>
</vstab>

<gear x="1.5" y="0.92" z="-2.558" skid="1" compression="0.10"
spring="1.0" damp="0.2" sfric="0.5" dfric="0.4">
  </gear>
<gear x="1.5" y="-0.92" z="-2.558" skid="1" compression="0.10"
spring="1.0" damp="0.2" sfric="0.5" dfric="0.4">
  </gear>
<gear x="-0.5" y="0.92" z="-2.558" skid="1" compression="0.10"
spring="1.0" damp="0.2" sfric="0.5" dfric="0.4">
  </gear>
<gear x="-0.5" y="-0.92" z="-2.558" skid="1" compression="0.10"
spring="1.0" damp="0.2" sfric="0.5" dfric="0.4">
  </gear>

<!-- tail rotor guard-->
<gear x="-4.47" y="0.00" z="-1.8" skid="1" compression="0.01"
spring="0.1" sfric="1" dfric="1"/>

<tank x="-0.2185" y="-0.2794" z="-1.8" capacity="115.2"/><!--main-->
<tank x="-0.09652" y="0.2718" z="-1.8" capacity="63"/><!--auxilary-->
<tank x="0" y="0.0" z="-1.7" capacity="0"/>

<weight x="0.5" y="0.27"  z="-1.5" mass-prop="/sim/weight[0]/weight-lb"/><!-- left seat-->
<weight x="0.5" y="-0.236" z="-1.5" mass-prop="/sim/weight[1]/weight-lb"/><!-- right seat-->
<!--
<ballast x="-0.5" y="0.15" z="-1.2" mass="150"/>
<ballast x="-0.5" y="-0.15" z="-1.2" mass="150"/>
<ballast x="0.05" y="0.5" z="-1.5" mass="150"/>
-->
<ballast x="-0.4" y="0" z="-1.3" mass="300"/>
</airplane>
Up, up and away
User avatar
HHS
Retired
 
Posts: 3624
Joined: Thu Jul 19, 2007 8:09 am
Version: GIT

Re: Helicopter realism -- or not? Real pilot tests?

Postby flywisher » Sat May 29, 2010 4:41 pm

Thank you HHS! I've copied your changes and will try it shortly.

Here's a very interesting development.

Yesterday I tried the FG R-22 again. This was the day after my brief session with the flight instructor and his big simulator. I tried to mimic what I saw him do with my FG model. Gentle control inputs, tilt left and back before liftoff to compensate for the right-tilt, etc.

I succeeded in hovering it! Wow! I did way better than ever before!

It's still quite challenging for me. I still can't just fly it across town. But now I have a much better understanding of how helicopters work and how to go about learning to handle them. And I'm rapidly getting much better at actually controlling the R-22 and maintaining a stable hover.

So it really is flyable, even in it's present state! Yea!

Erick
flywisher
 
Posts: 133
Joined: Mon Jun 01, 2009 2:37 am
Location: San Diego, CA USA
Callsign: flywisher
OS: XP, Vista, Mint

Re: Helicopter realism -- or not? Real pilot tests?

Postby Thorsten » Sun May 30, 2010 11:23 am

I've been following the thread with some interest, as I tend to have similar questions. I find it very revealing to get some impressions from a real helicopter pilot.

From my side, just a small comment:

Therefor you need sensitive hardware- good joystick, pedals, throttles. Mouse isn't recommended!
A twist-rudder joystick is the minimum.


I find the Bo-105 quite manageable with the mouse (I haven't tried anything else except the UH-1 which crashes Flightgear when I try to load it). I don't do precision flying with it, but that may just be my lack of practice - I can put it down on the Nimitz, or on Alcatraz (the classic...) and on airports roughly in front of the terminal. I don't hit a standard sized helipad yet, but I see that in reach. So I wouldn't claim that one absolutely needs sensitive hardware (although that is an advantage) - but it is definitely possible to just explore what helicopters handle like without getting additional hardware - so that's my encouragement to beginners who just like to try a bit. It's interesting and can be rewarding.
Thorsten
 
Posts: 11191
Joined: Mon Nov 02, 2009 8:33 am

Re: Helicopter realism -- or not? Real pilot tests?

Postby HHS » Sun May 30, 2010 7:18 pm

Thorsten wrote:I find the Bo-105 quite manageable with the mouse (I haven't tried anything else except the UH-1 which crashes Flightgear when I try to load it). I don't do precision flying with it, but that may just be my lack of practice - I can put it down on the Nimitz, or on Alcatraz (the classic...) and on airports roughly in front of the terminal. I don't hit a standard sized helipad yet, but I see that in reach. So I wouldn't claim that one absolutely needs sensitive hardware (although that is an advantage) - but it is definitely possible to just explore what helicopters handle like without getting additional hardware - so that's my encouragement to beginners who just like to try a bit. It's interesting and can be rewarding.


I must admit- I use mouse as well, as I didn't found a good set of controls yet. Maybe soon- we'll see.

So I have to use the mouse. The Default-settings for mouse-control aren't recommended for flying helicopters. I had to increase sensiblity (the ways my hand does with the mouse are nearly as long like on the real control), then I set the rudder-control on the mouse-wheel.
And when once the behaviour from helicopters are understood, with that it is possible to land the Bo105 on a typical mobile-helipad.


But: for smaller helis like the R22 this won't work good enough- and even for biger helis the movements can't be done smooth enough.
And helis can do more than just land on a small spot- ever tried to hover steady in the air above one spot ?
Hover turn steadily about one spot without any for/side/-rearward moving?
Acrobatics a la Rainer Wilke?
And all this with real weather? (Gusts, changing wind direction...)

This needs sometimes often quick movements, and with the mouse you will reach soon the limits!
And: of course you don't get the right feeling to fly a helicopter- I wish to have my hand on the stick and resting it on my knees, feets on the pedals and the left hand on the collective- like in the real thing.
Up, up and away
User avatar
HHS
Retired
 
Posts: 3624
Joined: Thu Jul 19, 2007 8:09 am
Version: GIT

Re: Helicopter realism -- or not? Real pilot tests?

Postby HHS » Sun May 30, 2010 8:04 pm

Thorsten wrote:(I haven't tried anything else except the UH-1 which crashes Flightgear when I try to load it).


Still crashing?

We still need some more informations like OS, Version-date of the helicopter and error-messages...
Up, up and away
User avatar
HHS
Retired
 
Posts: 3624
Joined: Thu Jul 19, 2007 8:09 am
Version: GIT

Re: Helicopter realism -- or not? Real pilot tests?

Postby flywisher » Mon May 31, 2010 11:56 pm

Thank you Thorsten for joining in. Like you and HHS, I did all my FlightGear flying by mouse until just recently. Only last month did I finally understand how to configure my joystick for FG. But FG's ingenious 3-phase mouse mechanism is great. One rarely needs anything else.

I too flew the Bo105 (and the CH-47 Chinook) by mouse successfully and had a lot of fun with it. I got pushed into the joystick direction by my desire to master the R-22.

HHS is right that helicopters benefit from separate "rudder" (torque pedal) control. (Setting rudder control to the mouse wheel? Brilliant, HHS! I never thought of that!) The reason is that with helicopters it's often necessary to apply "rudder" control simultaneously with the other controls. And with a very light touch. And change it quickly. This is very difficult to do any other way. Especially with light helicopters like the R-22. Things just happen too fast.

Also, I just realized one reason FG's helicopters are so challenging. A typical joystick is a handle on top of a pivot. The control stick ("cyclic") of a helicopter is a foot-and-half long. So the same movement of a joystick necessarily produces a much greater control deflection than the long stick of a helicopter. Or putting it another way, the long stick allows much finer control.

And that makes things unavoidably awkward for us. Oh, well. We can still have fun with it.
flywisher
 
Posts: 133
Joined: Mon Jun 01, 2009 2:37 am
Location: San Diego, CA USA
Callsign: flywisher
OS: XP, Vista, Mint

Re: Helicopter realism -- or not? Real pilot tests?

Postby flywisher » Tue Jun 01, 2010 12:22 am

By the way, guys, last year I stumbled upon a possible way that those of us who cannot afford fancier controllers might build our own.

A fellow in England created his own ship simulator and offers it for free. Sort of like FlightGear, but on a smaller scale. He calls it "Bridge Command." I've played with it. It's nice. Like FG, it's meant to be fairly realistic simulator, not a game. No scores, no enemies, etc.

On his website, I found a link to an illustrated page showing how one fellow created his own physical throttles and steering gear for USB connection using some scrap wood and other materials, a few potentiometers and a BU0836 controller chip.

If anyone would like to explore this, here are the links:

On the Bridge Command website:
http://www.bridgecommand.co.uk/Doc/makethrottle.php

The more extensive display of how to create controls:
http://80.95.161.114/shipsim/forum/inde ... pic=4503.0
The top part of the page is about adapting a real boat's throttle control to USB. Scroll down to the lower part for Sam's (a 16 year old) construction of his own. This is the approach we might adapt to FlightGear.

One or both of these have links to the manufacturer of the BU0836.
I can see this working for a homemade floor-mounted control stick (or cyclic), collective and even rudder (or torque) pedals.
I haven't had time to try it myself, but I offer it in case it might help someone else.

Cheers!
flywisher
 
Posts: 133
Joined: Mon Jun 01, 2009 2:37 am
Location: San Diego, CA USA
Callsign: flywisher
OS: XP, Vista, Mint

Re: Helicopter realism -- or not? Real pilot tests?

Postby Thorsten » Tue Jun 01, 2010 7:10 am

Still crashing?


*shrugs* Still same binary and model - no binary CVS snapshots for Linux, so I don't fly CVS/GIT - so same problems.

And helis can do more than just land on a small spot- ever tried to hover steady in the air above one spot ?


Please don't get me wrong - I'm not saying mouse is a suitable controller for a heli, or you could do all things with it which you can do with a full set of hardware. You can't (or at least I can't hover decently).

But I heard quite often one needs a joystick and pedals. So for a long while I thought 'Well, I'm curious, but I'm not going to buy hardware just to see if I find it interesting.' Therefore for a long time I never got into even trying a helicopter at all (yeah, well, I loaded it, tried to take off, lost control within 3 seconds crashed and that was it). But it can be done with some patience. I have fun with helis and mouse, and it taught me a lot about what the issues in controlling a heli are. Using the mouse is never as good as full equipment, there are limits also for planes for example in aerobatics or as far as the forward slip is concerned. But I just want to get the message out that it's not a huge problem with just the mouse - you're not out, you can still fly, even if it's not as good as with full hardware.
Thorsten
 
Posts: 11191
Joined: Mon Nov 02, 2009 8:33 am

Re: Helicopter realism -- or not? Real pilot tests?

Postby Fahim Dalvi » Tue Jun 01, 2010 7:30 am

flywisher wrote:On his website, I found a link to an illustrated page showing how one fellow created his own physical throttles and steering gear for USB connection using some scrap wood and other materials, a few potentiometers and a BU0836 controller chip.


Hey All,

Just jumping in, Maybe a little off topic but still. flywisher, some time back, a guy called sety showed us how to build our own yoke with a mouse, in Just 5 minutes :D http://www.flightgear.org/forums/viewtopic.php?f=3&t=1956

I thought you might be interested, not for helicopter flying, but for other aircraft.

Regards,
Fahim
Fahim Imaduddin Dalvi
Tracker, Flight Gallery
User avatar
Fahim Dalvi
 
Posts: 1286
Joined: Sun Oct 14, 2007 1:08 pm
Location: Qatar
Callsign: Fahim
Version: 2018.2.2
OS: MacOS

Re: Helicopter realism -- or not? Real pilot tests?

Postby HHS » Tue Jun 01, 2010 8:06 am

flywisher wrote:HHS is right that helicopters benefit from separate "rudder" (torque pedal) control. (Setting rudder control to the mouse wheel? Brilliant, HHS! I never thought of that!) The reason is that with helicopters it's often necessary to apply "rudder" control simultaneously with the other controls. And with a very light touch. And change it quickly. This is very difficult to do any other way. Especially with light helicopters like the R-22. Things just happen too fast.

Also, I just realized one reason FG's helicopters are so challenging. A typical joystick is a handle on top of a pivot. The control stick ("cyclic") of a helicopter is a foot-and-half long. So the same movement of a joystick necessarily produces a much greater control deflection than the long stick of a helicopter. Or putting it another way, the long stick allows much finer control.

And that makes things unavoidably awkward for us. Oh, well. We can still have fun with it.


Absolutely right. Other things are that we don't have as an example any force feedback/ artificial force. That makes also a difference. But these are the limits of desktop simulation.
There are already some sim-controls just for helisims- but even the cheapest of them are more expensive then a good set of pedals, joystick and throttles.

Thorsten wrote:But I just want to get the message out that it's not a huge problem with just the mouse - you're not out, you can still fly, even if it's not as good as with full hardware.


You are absolutely right too- my sentences may people prevent from trying out helicopters. On the other side the recommendation was given just to prevent disapointements of people actually having problems trying helicopters. I started FlightGear with a fulminant conflict in the german forum, just because I tried to fly the helicopter with keyboard and mouse but said that's the fdm was wrong. It wasn't - it was just my lack of informations how to fly helis and my not quite suitable hardware.

Thorsten wrote:*shrugs* Still same binary and model - no binary CVS snapshots for Linux, so I don't fly CVS/GIT - so same problems.

Helijah makes always updates on his own page: http://helijah.free.fr/flightgear/tar.gz/uh1-20-05-2010.tar.gz.
As you are Linux-users we would like to know which error messages you get, how do you compiled etc.- there are issues with the model in linux, while it is working on msand mac without andy problems.
So we are looking for the bug....
Up, up and away
User avatar
HHS
Retired
 
Posts: 3624
Joined: Thu Jul 19, 2007 8:09 am
Version: GIT

Next

Return to Helicopter

Who is online

Users browsing this forum: No registered users and 1 guest