Dog Fight!

[Yo-Yo]

I'd like to offer my congratulations on creating a challenging and reasonably convincing A.I., Yo-Yo. In other aerial combat sim/games I've used, the A.I. has been useful for little more than target practice. Here, I'm having more trouble with the A.I. than I've had in DCS with any human opponent, except for one.

 

You probably won't believe but it uses very simple math when fighting... :) but sometimes I wonder how interesting is AI's behaviour as a result of this math.

 

But now I unveil its Achilles heel: as it trying to predict where you will be you can fool AI. If you have 90 degree angle difference while looping (in the same time you have near vertical attitude and AI has near horisontal regardless who are chasing who) you can roll your plane projecting this point far away from the initial one WITHOUT PULLING much G. As the AI opponent has to apply G-load to chase the point it lose more energy than you.

 

Properly done this trick can help.

[Mike Busutil]

Thanks for the tip ;)

[WildBillKelsoe]

2) Set the rudder trim to about 3 degrees before take off

3) Hold the stick back and throttle up slowly increasing the throttle as you speed up and let the stick move to neutral as you speed up.

 

 

I have to disagree. In all my current takeoffs, using rudder pro by CH, I never set rudder trim at all. Nor do I touch the elevator trim. I just align myself at slow speed with the runway, make sure I'll be on it past 90 MPH, then stopp.

 

RPM to full increase, I slowly advance throttle so in 20 seconds, as I get to 90 MPH, I push the stick full forward, the tail rises, and then I floor the throttle. Once the tail is off the ground, no rudder input is required, and before it is off the ground, the torque effect is smoothed with gradual onset of power.

 

This configuration works with 35% to 75% fuel, with 2 bombs or fuel tanks installed only. I have yet to try a full fledged takeoff.

 

And some runways I found in the Italian/Israeli AAF servers made alot of crashes due to their nature. I hitched a Batumi runway server mission and all takeoffs were smooth.

[Echo38]

With stick back all the way, you can go straight to military power in a second or two, and you shouldn't have any problem keeping it reasonably straight with minimal rudder, at least before the tail rises off the ground.

[Echo38]

Killed my first A.I. in a duel! Took me an inordinate amount of time, though. I think the key was careful occasional use of flaps--10 to 30 degrees. I'm still not good enough to tell when my flaps are hurting or helping, but when I was using 30 degrees at the top of combat turns and during low-speed scissors, it seemed to help. Full flaps is definitely not a good idea (or I'll eat my hat).

 

Edit: duuuude! I just watched my track--that was a 22-minute dogfight. Holy mackerel! I'll bet it won't be able to last that long when the engine management's completed, though. I love this sim!

 

To the O.P.: keep trying. Took me about a dozen defeats before I was able to take him down. Don't hesitate to experiment with flaps. You may want to practice gunnery on a low-skill A.I. so that you can better make use of snapshots; I'm certainly considering doing that. I had quite a few shot opportunities that I missed, which could have won me the fight early if my gunnery were better.

Edited September 20, 2012 by Echo38

[Yo-Yo]

Congratulations! The next step will be w/o flaps.

And the tip for TO: as you are rising the tail smoothly add right rudder but as rotation stops IMMIDIATELY release it or even shot left pedal for a moment if the plane starts going right.

 

THe explanation is - when you are rotating, prop gyroscope is rotating the plane left but as you are at two wheels attitude, both P-factor and gyroscopic moment are reduced.

[Tailgate]

With stick back all the way, you can go straight to military power in a second or two, and you shouldn't have any problem keeping it reasonably straight with minimal rudder, at least before the tail rises off the ground.

 

That's the way I do it, back pressure keeps the tail wheel firmly planted while running up to rotation speed.

[OldCrow47]

This from Dudley Henriques, P-51 pilot

 

This was posted over at A2A's forum so please forgive the A2A references.

Mr Henriques is a modern day P-51 pilot and the insight he gives us will be no surprise to many but I am hoping it will benefit others. I find it a very good study. ...Jim

 

"I've been reading a lot of posting concerning the function of rudder trim on the Mustang as that relates to its takeoff behavior. After reading some of the posts I realize that this subject might need some clarification.

I hope the following information proves useful and helps toward a better understanding of what is involved in a Mustang takeoff.

 

There is one overwhelming point about the rudder trim on the 51 that should be completely understood by everyone. That 6 degrees of right rudder trim is NOT pre-set before takeoff to solve the forces acting on the aircraft during the takeoff run. You do that with rudder and aileron used in conjunction with a COMPLETE understanding of airspeed and how increasing airspeed increases dynamic pressure on the control surfaces.

What that 6 degrees right trim ACTUALLY gives you is an optimally trimmed out airplane at 46 inches and 2700RPM which is your first power reduction AFTER takeoff.

 

The takeoff roll in a Mustang is a very complex event control wise. There are several forces in play and they vary in intensity at various airspeeds as your speed increases. To get it all just right in FSX so that the result is a realistic experience for the sim pilot has been a daunting task for the A2A flight model team. In my opinion they did a super job on this!

You begin your takeoff in a Mustang with the stick aft of neutral to lock the tail wheel but this doesn't last a long time because as the airplane accelerates the tail wants to come up naturally and you don't want to keep that stick back too long because at the angle of attack that produces, lift is developing VERY fast as speed increases and you can easily reach a rotation point on the lift curve where the airplane will fly before you have sufficient airspeed. This situation in a Mustang is extremely dangerous as power is high, angle of attack is high, and you don't have enough airspeed yet to develop the dynamic pressure on the control surfaces you need to deal with the tremendous forces attempting to both yaw and roll the airplane. This is the exact situation that causes torque roll and has killed many a P51 pilot who didn't have a working knowledge of this airplane.

 

So pitch wise you have to play your takeoffs by starting with the stick back, then GENTLY FLYING the stick forward EASILY to bleed off some of that angle of attack and keep rotation from occurring before reaching rotation airspeed.

Now while all this is happening in pitch, you have other forces acting on the aircraft as POWER is increased so you have to deal with this AS you are dealing with the pitch.

The 51 is a whole new airplane as it passes 30 inches. In the real Mustang I always treated takeoff in two distinct segments but blended them together smoothly at the transition point. I powered up slowly to 30 inches with the stick back and let the aircraft accelerate. During this period P Factor is the highest due to the 3 point attitude asymmetrical difference between the right and left halves of the prop disk in operation with a relative wind NOT in line with the airplane's fuselage reference line. This is where you begin to feel the need for a right rudder correction. It's important to note that the spiraling slipstream being created by the prop at low airspeed is very tight and maximized on the side of the vertical stabilizer and rudder. As speed increases the slipstream elongates out a bit and the force its producing lightens somewhat…………but not by much!

As the tail begins to feel light and you have some airspeed over the control surfaces you are now ready to begin dealing with gyroscopic precession. This occurs as the propeller disk is in ROTATION in PITCH as the nose comes down. (It also occurs as you rotate the nose UP at rotation but is not as noticeable in reverse due to your higher airspeed). You CAREFULLY allow the nose to come down while ANTICIPATING a right rudder correction for any left yaw. Of course while all this is happening you also have a tight slipstream spiral against the vertical stabilizer helping induce MORE left yaw. Same correction; right rudder.

Now here's the rub. As you get the power on up past 30 inches you begin to pick up TORQUE. Since torque is a ROLL correction NOT a rudder correction, you will need right aileron to take some of that extra load off the left main gear strut as the engine torque tries to flatten the left main gear tire against the runway. Of note here as well is the fact that your left main gear wheel is actually helping you with torque correction. Understanding this has saved a few lives in the 51 when pilots found themselves in a go-around situation running out of runway. "holding it down" with high power applied till the last moment allows that wheel to help the pilot as he claws for needed airspeed before rotating the go-around.

 

The kicker with aileron correction is that you need dynamic pressure on the ailerons to correct for the torque and if you reach a power setting that starts feeding in torque to your takeoff equation that your airspeed in the takeoff run (dynamic pressure AVAILABLE to the ailerons) isn't sufficient to handle, you won't have aileron available to help with the torque. It's notable that the amount of required right aileron for torque correction will vary somewhat in the A2A Mustangs as the inertia forces are changed a bit between the mil and civ versions. So don't just throw in a lot of right aileron. Let the airplane TELL YOU what it needs then give that to the airplane. This goes for BOTH aileron and rudder correction. The AIRPLANE will TELL YOU what it wants!!!!! Fly this way and you'll do just fine.

So as you can see, the name of the game in a Mustang takeoff is complex control input based on you completely understanding the power/airspeed equation.

As you can see, the P51 is NOT your vanilla GA light airplane. The A2A flight model team worked long and hard in seeing to it that all of you who purchased the Mustang were given as close to a realistic experience as possible. Torque roll is modeled so be careful out there. You screw up on takeoff and there just might be a price to pay )))))

 

I hope this small tutorial helps all of you to a better understanding of these highly complex issues. A2A has invested a LOT of effort in your Mustang development. I sincerely hope you use your P51 not only as a fun machine but as well as an educational tool designed to give you a better understanding of the world of high performance flying."

[Tailgate]

Hopefully he'll have a chance to fly DCS and give us a good review, but the forces and corrective measures he describes really rings true for the DCS Mustang as well. Good post.

[Mike Busutil]

To the O.P.: keep trying. Took me about a dozen defeats before I was able to take him down. Don't hesitate to experiment with flaps. You may want to practice gunnery on a low-skill A.I.

 

 

Thank's :thumbup:

 

But getting a Kill is not the issue. I was just pointing out a difference in flight physics between the user Mustang and AI Mustang.

[Yo-Yo]

Emergency landing gear release

[Echo38]

Yo-Yo, I noticed that the P-51D loading screen shows the gear coming down asymmetrically. Under what conditions does this occur?

[Nate--IRL--]

Yo-Yo, I noticed that the P-51D loading screen shows the gear coming down asymmetrically. Under what conditions does this occur?

 

That's the gear coming up I think.

 

Nate

[Yo-Yo]

Yo-Yo, I noticed that the P-51D loading screen shows the gear coming down asymmetrically. Under what conditions does this occur?

 

I think the reason of the difference can be the plane bank as the legs goes down by gravity.

[Tailgate]

Very strange, something's going on with Yo-Yo's posts where the entire page is black, at least on my system. I had to go to the previous page and add this message.. Can't read his post in the Developer's notes either.

[Nate--IRL--]

Your flash player acting up?

 

Nate

[Tailgate]

I was RDP'd into another system, must have been some device sharing issues. It's fine now. Strange that only those specific posts were affected.

[Yo-Yo]

Less than five minute to cook AI.

Who tells they are UFO?

 

K-14 is fixed and helps a lot.

1-18 to 1-36 was a key point

Edited September 21, 2012 by Yo-Yo

[PeterP]

:D

Great Yo-Yo!

And thanks for info.

 

How about fixing the K-14 also for me ?!

Its Friday! - and I feel badly like I have to shoot some from the sky.

Where is the hot-fix?;)

 

 

Edited September 21, 2012 by PeterP

[Tailgate]

Nice! I was thinking you were using "the Force" there without the sight, but finally saw it when you closed in. Is this the current version or the new dev version?

[Nate--IRL--]

Dev version.

 

Nate

[Yo-Yo]

Nice! I was thinking you were using "the Force" there without the sight, but finally saw it when you closed in. Is this the current version or the new dev version?

 

You are right... concerning the version, I mean.

[Echo38]

Am I mistaken, or do I see a new G-blackout effect?

[JimMack]

Less than five minute to cook AI.

Who tells they are UFO?

 

K-14 is fixed and helps a lot.

1-18 to 1-36 was a key point

 

Nice flying!

If you look at the 100s of 2ndWW gun camera videos available, they tended to fire at a closer range.

[Nate--IRL--]

Am I mistaken, or do I see a new G-blackout effect?

 

Yep.

 

Nate