|
Post by hanspetter on Nov 25, 2008 10:01:31 GMT -5
Since I bought my yoke I've mainly been flying GA aircraft. My sensitivity settings have worked allright until I tried a passenger jet. I then noticed that I had to make extreme inputs to make it turn. I'm pretty sure it shouldn't take that much.
Actually the first jet I tried was a recently downloaded one, real pilot approved and all. When I checked the MOIs they exceeded the ones by Tom G by one order of magnitude. Allright, 10 times too sluggish -- as real as it gets? However, here's a point we've discussed before -- the choice of controllers are essential and so are the sensitivity settings. The exaggerated MOIs may have passed unnoticed with a different kind of controller even though I doubt it.
Anyway, I started looking at the amount yoke input required to initiate a reasonable turn and "all the way to the right and then wait a bit" seems unrealistic. I ditched the "real pilot approved" flight dynamics and tested out the good old ones. Still sluggish. That's what I'll try to fix by increasing the sensitivity. What I need to know is, how much should I expect to turn the yoke to effect a reasonable bank angle? Can this question be answered at all or is it too vague?
|
|
|
Post by Tom Goodrick on Nov 25, 2008 10:53:08 GMT -5
You'll recall I developed a turn rate test gauge to look at this problem a few years ago. Unfortunately, it does not work with large jets. It does work with small jets, turboprops and piston aircraft. It compares the rate of turn while holding an airspeed and bank angle with the rate of turn predicted by the centrifugal force relation of the physics of the turn. The sim allows people to change the yaw rate damping so that they can make larger aircraft seem more sluggish.
Large aircraft, especially jets with their higher speeds in all phases of flight, must move turn slowly than smaller aircraft. It should NOT be the responsibility of the FD developer to force this. Most real jets have plenty of reaction capability to the controls. This has been demonstrated by the test pilot rolling the 707 and by some accidents in which control surfaces were ripped off the aircraft by extreme pilot reaction. The point is that pilots of large jets are trained to work with smotth and gradual turns and other maneuvers. They don't expect, do not need and do not want to see rapid response to control inputs. Fighter jets are obviously different. Small biz jets can be very responsive although such jets have had crashes related to excessive control.
A turn has two parts: 1) rolling to a bank angle and 2) flying around a circular arc path. The term "rate of trun" applies only to the second part. Rate of turn depends on true airspeed and bank angle. For large airliners, the speed is high and the bank angle is low - seldom more than 30 degrees. Thus the turn rate is low and the radius of turn is large. This is something airline pilots live with. Roll rate as a direct response to aileron input should always be low in airliners. You don't want to get that many passengers mad at you! Assume your boss's mother-in-law is sitting in a window seat in the back row.
In the FS world, you never know what you get when you download an aircraft, as you well know. Much of the FD files are junk. "Real Pilot Approved" means very little. A real Cessna 172 pilot commenting about a Boeing 737 is giving useless information if his comments are based soley on his flight experience. You need commentary from pilots of large jets.
It helps to have standard models of each type of aircraft that you can use for comparison. I have a set of FD files for the default FS2004 Boeingjets - 737, 777 and 747 - in a file called "Big Iron" on my web site. In most cases the FD files are not major departures from the originals. I would recommend that you use those FD files to train yourself as to what performance to expect from the large aircraft. Then, when you download a large aircraft, fly one of the default jets first. Make several circuits around an airport but go high enough and fast enough to get a feel for the problems of climbs and descents with turns.
Consider the passenger in back who has just gotten his favorite liquid refreshment after a long hard business day. Don't spill his drink! Make all banks about 30 degrees or less. Fly smoothly as you change the configuration for climb or for descent and approach. Change the scale of your planning for flights near the airport. Don't wait too long to turn onto final. Don't over-turn and have to turn back to the course you wanted.
Setup your control system sensitivities to work well with my Big Iron FD's. Then don't change them for a new download. You correctly found that the MOI's are off. Change them. I have always found that MOI's off by several orders of magnitude screw things up in many ways. They may solve one problem but the create others. If a roll rate, yaw rate or pitch rate, as direct responses to control inputs is too high, fix it directly in the Flight Tuning section. Reduce the control sensitivity. You can also adjust roll damping directly in the air file (1100 section) but usually that is done to desensitize the response. If you adjust yaw damping or pitch damping you can mess things up and produce very sluggish turns.
As a last resort, describe the problem here in detail and maybe we can help. You may recall we have discussed the Boeing Blended Wing, the Airbus 380, the Boeing 787 and other large aircraft. I see on TV commercials that Boeing is still playing with the Blended Wing showing roll rates that would toss people around in those long rows of seats. I hope their roll controls are limited.
|
|
|
Post by Tom Goodrick on Nov 25, 2008 23:28:12 GMT -5
To refresh my memory, I flew the 737, 777 and 747 today. To check the time to turn 90 degrees I flew the Learjet 45 at the same true airspeed. I am confident these jets fly and respond correctly. They use correct MOI's. I get good response by moving my joystick less than halfway to the stops. But then I have plenty of patience when flying these big aircraft.
When practicing with these jets, you must know the V speeds. I have gauges on the panel that compute these speeds and display them for takeoff and landing. You must also know VMO and MMO as you will often fly these aircraft close to those limits. You can get these speeds from the data in the aircraft.cfg file. For V2 (safe takeoff speed) use 1.3 times the stall speed in takeoff configuration at the weight flown. Do the same for Vref if you don't have my panels and gauges. In general, you should set the fuel weight at half max for practicing as that gets it close to max landing weight.
Starting with the 777, I set a timer before starting to roll into a turn and then timed the turn through roll-out after a 90 degree change in course. This was at about 380-390 KTAS at 15,000 ft where I was doing manual control checks. (But I timed turns done by the autopilot that held speed and altitude much better than I could.) All both the 777 and the 747 took exactly 2 minutes to complete the 90 degree course correction. Then I did the same thing in the Learjet 45 and found its time matched the others even though it has a much faster roll rate. The Learjet 45 will do a nice barrel roll at positive G's in 14 seconds. (It won't spill coffee during the roll. Be careful when a pilot asks if you want a roll with your coffee.) It is true that full stick deflection in roll produces much less motion in a 747 than in a Learjet 45. But that does not mean the 747 is slower to change course. Of course you fly an approach by getting in line with the runway 8 to 10 miles out.
Here are my landing data for the flights: 737 131.1 KIAS -198 FPM at 128056 lb and CG at 18.35% MAC. 777 139.8 KIAS -484 FPM at 538570 lb and CG at 29.54% MAC 747 163.2 KIAS -431 FPM at 687176 lb and CG at 28.21% MAC LJ45 119.8 KIAS -141 FPM at 17980 lb and CG at 14.38% MAC
Vref was 155 for the 777 and 171 for the 747. It was 120 for the LJ45. This explains why the big planes have a sluggish lateral motion on final.
Use these 777 and 747 aircraft from the "Big Iron" download to calibrate your control system. Then fly the unfamiliar airliner.
|
|
|
Post by hanspetter on Nov 27, 2008 5:16:15 GMT -5
Good info However, I'm still curious what criteria I should use for setting my yoke sensitivity. That is, how much should I expect to turn it to reach full aileron response? All the way in either direction means about an 80 degree turn of the yoke and that's something I wouldn't expect to see in a real aircraft. But then, full aileron may be something you rarely see in real life anyway. I'm certainly testing this out on good flight models.
|
|
|
Post by Tom Goodrick on Nov 27, 2008 11:18:45 GMT -5
Doesn't it make a certain amount of sense to have to rotate the yoke to the end of its travel to get "full aileron response?"
What is "full aileron response?"
There are so many ways to adjust this it gets ridiculous fast: 1) In the aircraft.cfg change the aileron control sensitivity in the Flight Tuning section. 2) In the geometry section change the degrees of deflection for the aileron travel. 3) In the .air file change dCl/da to change sensitivity of the roll moment (Cl) to aileron deflection. 4) In the .air file change dCl/dp to change the roll damping.
All this assumes you have the correct MOI's in the file.
Why mess with the yoke calibration and setup if you already have it adjusted for good performance with other aircraft?
Get used to slow response in big aircraft. Don't give it full deflections if it can't respond to that. That is what is neat about flying big aircraft. Get them in line for the runway 8 or more miles out and keep busy maintaining alignment until "over the fence."
|
|