flightace37

Yeah, I think you guys are right about the brakes. I noted as much in the youtube description of the video when I first posted it. Does anyone see a tiny little bit of yaw motion when I kick the rudder over? It's hard to tell from that angle. The way I see it, in a fully realistic model, the aircraft obviously wouldn't move around on the ground because of rudder action with the brakes held. It should, however, twist side to side a little bit because of the flexibility of both the tires and the frame. Asking for that to be modeled would be too much of course, but it would definitely be interesting if it was. If you look closely at your aircraft on the ground, you'll actually see that your tires are squished down a little bit where they make contact.

To that end, there's also a cutoff to prevent the generator from hooking into the electrical system until it's generating sufficient power. Until then, the ammeter will read zero. Page 43 of the DCS P-51D Flight Manual:

You'll notice the rudder doesn't function too well. I imagine ED will fix this before release though. It'd be interesting to see if they model the effect between aircraft. It would certainly add a whole new dimension to air combat.

The green line represents "normal operating range", which in this case, is the range of permissible cruise power settings. Just compare it to the charts. Maximum continuous power is 46" and 2700 RPM. As long as you don't exceed that, you shouldn't ever feel guilty! (Okay, maybe just a little...) As far as realistic takeoff procedure goes, here is the relevant section from the P-51 Pilot Training Manual: So, going to full military power is S.O.P. How long you take to get there is determined by how well you can control the aircraft on your takeoff roll (and the length of runway available.) The in-sim training does recommend using 50" for a while, until you get used to the torque. Ignore the final reference about the brakes. I just like to throw in little details when talking about these sorts of things. Anyway, congratulations on your first takeoff, Hitman! You can only improve from there. I have not watched your track yet, but I just want to encourage you to keep at it. It's absolutely possible to do it by the book, given enough practice. The one exception I currently use is a reduction in rudder trim to compensate for the way it is implemented in the sim. By the way, what did you order at the restaurant? Rudder cable spaghetti? :smilewink:

I did a little testing to see how propwash was modeled. This should help answer some of our questions about ground handling. Enjoy the video! pUgDqhPnekw Working on embedding it, but the youtube tags don't seem to be working. EDIT: Thanks Viper! Now I see how the format works. EDIT2: .

The claim is that gun jams were solved in the D model, then retrofitted to B/C. See http://www.mustangsmustangs.com/p-51/p51variants/P-51B.php. This other site, http://www.aerofiles.com/JBnoram-p51bc.html, claims that the guns jammed because they were turned on their side to fit into the thin wing, forcing kinks into the ammo belts as they fed in. It's quite possible that the modifications they made only alleviated the issue, and the guns still experience jams at high G. The best fix is to avoid shooting under those conditions. Rather than pulling G with the target, say in a flat turn, pitch up into a yo-yo and come down with your nose well ahead. Under those conditions, you don't have to pull nearly as much G to match the target, and thus don't have to expose the ammo belts to the adverse condition. There's many different ways to reposition yourself into a better firing position. Think geometry instead of Hulk Hogan.

It's possible that network issues might cause the aircraft to behave differently, but have you checked for differences in the fuel and weapons choices? There is a very pronounced difference between flying a normally fueled Mustang and one that has the auxiliary tank topped off. There's also a huge change in flight dynamics when you're carrying stores.

I agree that prop wash might not be fully modeled. Deflecting the rudder with significant power on the ground seems to have very little effect, and steering authority improves dramatically after you pick up 15-20 mph (tail wheel locked). A lack of a proper caster model would definitely explain the behavior when the wheel is unlocked. There should be a definite tendency to stay centered if you release the lock while rolling in a straight line. It would wobble a little, of course, but it shouldn't go flipping completely around the way it does now. I'll do some testing later on and we can compare notes. @GLock: I've spent about an hour trying to write this post in a way that won't cause friction on the forums, but I'm not sure if I've succeeded. Please understand that I'm doing my best to avoid being offensive. The essence of the problem is that it is exceedingly difficult and time consuming to try to write this material in a way that everyone can understand. We're talking about years or even decades of experience and training. Fortunate or not, there's no way we can impart that knowledge in the span of a few forum posts. In many cases, it's not possible to talk about a subject without both parties having an understanding of the underlying material. You simply have to put in the time, and that involves asking questions when you don't understand something. We're not hateful people, and we don't look down on you for not being able to understand what we're saying. Just like school teachers, we need you to tell us what you're having difficulty with so that we can find the applicable knowledge in the giant pool of goo that we carry around in our heads. :helpsmilie: Here are descriptions of some terms that stand out to me in effte's post (hopefully with the right definitions for the context): Caster - (I had to actually go look this up, as I got caster and camber confused for a while). This is a design concept that causes a wheel to return to center because of friction forces while rolling. See http://en.wikipedia.org/wiki/Caster_angle for a basic discussion. Effte actually described this in his post, although it may have been an edit. Prop wash - The wind coming from the propeller that flies back over the aircraft, causing flow over the wings and control surfaces. Prop wash can produce enough airflow over the rudder and elevator to allow you to swing the tail left and right, or even raise it above the ground. Track - More properly called track width. This is the distance between a pair of wheels, perpendicular to the direction of motion. Wider track widths provide a more stable platform, with less tendency to tip over in a turn. You'll notice that race cars all tend to be very low to the ground, with an extremely wide track width. This makes them resistant to tipping over, and it also allows them to pull 1G or more in a turn. A common high performance road car, by comparison, is only able to achieve about 0.8G. Hopefully that helps clear up any confusion about terms in effte's post. Remember that these are also just very basic descriptions of what these terms are. When I think about all this stuff, I'm actually picturing the physics in my mind, imagining flow fields for the discussion on prop wash, and the way friction and suspension geometry affect a tire in motion. There's a good example of another use of a term that I might need to define. Flow field - A mathematical/visual model describing the way a liquid (air acts like a liquid) moves around a surface. Here's a relevant Wiki article on vector fields. From this post alone, you can see how terms and concepts compound on each other. Having us drill all the way down to the most basic level isn't practical, especially when you might already know some of the material. Take a look at that article on vector fields. There's a bunch of additional terms in there that you would have to know to develop a full understanding of the concept. How far do I drill down? It's much better to start at the top and work down to concepts that you understand than it is to try to write at a very basic level. That saves both parties a lot of time. We'll do our best to provide a simple description when we can, like I did for the terms above, but if there's no way to get into it without dropping into details, you're going to have to bare with us and keep asking more questions. (And yes, going all the way down to vector fields is probably more than pilots receive in flight training. I'm just fumbling for a way to describe the nature of the problem.)

I had about the same experience, perhaps with a little more fuel. All in all, I think ED has done a wonderful job modeling this aircraft, and they say they still have improvements to make!

I took a look at the track and came up with a couple of pointers: a) Remove the weapons load for now. b) Make sure the mission doesn't have any significant cross-wind c) Use only 1-2 degrees right aileron trim, or none at all. d) 2 degrees forward elevator trim e) 0 to 4 degrees right rudder trim (I find 6 is too much) You have an issue with throttle stability for sure, as others have mentioned. Get that ironed out somehow. Calibrating your controllers might do the trick, but if not, make sure they're plugged into a powered USB hub rather than directly into your computer. Your takeoff roll started fine, but then experienced a pretty gradual left pull. Get on those rudders and counteract it. Hold mild back pressure on the stick to help keep the tail wheel on the ground. If you roll off into the grass, do NOT continue takeoff. Come to a stop and try again from the runway. Most grass in this sim does not make a good runway surface because of all the bumps and other variations. My takeoff roll generally goes something like this, after lineup: 1. Hold brakes and run up engine to 20" Hg. This makes sure the suction system is supplying enough vacuum to the instruments. 2. Set the directional gyro, etc. Double check Prop control full forward. Hold stick back a good bit to lock tailwheel, and provide aerodynamic effect to keep tail on ground as speed develops. 3. Release brakes and let the aircraft start rolling on its own. 4. Increase throttle to 30" and leave it there for a while, until you develop about 30 mph. 5. Increase throttle to 45" in small increments until about 50 mph. Start adding in right aileron to counteract wing dip from torque. 6. Continue increasing throttle towards full power as speed develops. 7. Tail wheel will start flying itself off the ground between 80-90 mph if you're only holding a little bit of back pressure. Make sure to adjust rudder accordingly, as there is a difference in behavior without the wheel forces. 8. Continue holding light back pressure until the aircraft rotates to more level attitude, and lift off gently around 120 mph. For heavier loads (like the rockets you were carrying), wait until 140 mph or so. She might get extremely light and try to fly herself off the runway. If that happens, go ahead and help. Continuing with the weight coming on and off the wheels will introduce some really nasty behavior. Lastly, although it may not affect your takeoff roll, make sure to set your RAM Air lever full forward. You did set the hot air lever correctly, but missed the other one. Hope this helps!

I got her up above 40,000 ft, with a little climb rate still remaining. Power fell off quite a bit by then, but I'm pretty sure she can reach her real service ceiling.

Every virtual fighter pilot should give this free book a read: http://web.comhem.se/~u85627360/inpursuit.pdf And then consider buying this book: Fighter Combat: Tactics and Maneuvering by Robert L. Shaw Supplement the drawings with 3d animations like the one Rudel posted, as necessary. Finally, find a good sim pilot and get some 1:1 training with them. That will help you immensely.

If the F2 figure that I used to judge ground speed is in mph, then this P-51D is much slower than it should be at sea level. The cockpit gauge is in mph. I very much believe the F2 view speed display is in knots with the options set to use Imperial units. Am I wrong? EDIT: I did make an error in my original post, claiming 312 KIAS for the 3000 RPM speed. That should be knots GS. My apologies for any confusion.

Of course. I was hoping that a psycho-analysis of AI behavior would help ED make improvements, should it be part of their mission goals. As stated by several others, the AI is quite effective as is; it's just predictable.

Very good point about orange peel, TimmyD. I wonder which technology was better at the time? That has to be what it comes down to in the end. EDIT: Here's an article discussing laminar flow airfoils, and a bit about the situation with the Mustang: http://www.aviation-history.com/theory/lam-flow.htm Enjoy!

There's no feeling about it, leafer. Some kind of head tracker is the most important piece of flight sim equipment you should acquire, besides a decent joystick. I do all kinds of things during a dogfight that involve head movement. I watch the target, I watch all around me for bogeys, I move my head to look around obstructions in the cockpit to see better, and I check a number of instruments for combat-critical information. All this happens within about a 5 second scan pattern, though most of my attention is on the target, depending on conditions. Lose sight, lose fight. Head tracking also frees up a hat switch to use for trim, which is extremely important in this aircraft, both in and out of combat. Anyway, that was a little off topic. Regarding the AI, I've actually found that they behave rather simply. They're just really good at pushing the aircraft right to its physical limits. If you set up a mission in the editor with two Excellent AI Mustangs in a head-on, co-alt, co-speed pass, you'll find that all they manage to do is keep swinging around at each other for another head-on. They rarely get shots off because they have to break immediately to avoid colliding, thus setting up for the next pass. Observing their turning behavior, it becomes fairly obvious that all they care about is getting the nose around on their opponent to set up a shot. They do examine velocity vector and fly to your future position rather than directly at your aircraft, and this is why they are so difficult to deal with at first. The truth is, all they do is roll their lift vector to turn to your future position as quickly as possible and pull, so long as they have energy to permit it. If not, they compromise with a gentler turn, or less use of the vertical. They do seem to have an understanding of maneuvering speed, and they may or may not choose nose-high or nose-low accordingly. The AI doesn't seem to understand the concepts of pursuit curves, or the intricacies of some of the more complex maneuvers like flat/rolling scissors, defensive spiral, and yo-yos. They manage to approximate yo-yo and rope-a-dope maneuvers in the way they bring their nose around on you, but it is done with very little respect for your intentions. That isn't to say they're not difficult to fight. They really are able to push their airframes, and that can make a world of difference, but they are definitely predictable. I'd be happy to hear that the AI is more complex than this. How am I batting?

The USAAF did stop having their fighters delivered with full paint jobs after a while. Saves weight and cost. As long as the paint is smooth, it actually improves flow characteristics over raw aluminum (which would contain surface defects). This fact is documented in a few discussions of the P-51's laminar flow wing, and the lack of paint was one of the reasons why it didn't perform completely up to expectations. Sorry I don't have the sources, but you should be able to dig up something through Google. I'm not sure whether the weight savings or better flow characteristics would lead to higher top speeds. I do tend to lean towards the paint helping performance, rather than hurting it. The 8 mph loss would be better attributed to the switch from the birdcage to the bubble canopy. Just for fun though, I'll run a test. Nothing like W.O.T. four feet off the water. EDIT: British Mustang with full paint vs American Mustang in bare metal, standard 68% default fuel. Fully trimmed for level flight, 2600 RPM W.O.T. about 20 feet off the deck, starting with dive from 3k to maximize starting speed. Both aircraft slowed down to and stabilized at 321 knots ground speed according to F2 view after 2 minutes hard run. Variation +/- 2 KIAS with pitching due to manual control. They showed just under 400 mph IAS on the cockpit gauge. Needless to say, I think the paint effect is not modeled. And yes, I did toy with RPM to find the best setting. For the record, 3000 RPM resulted in the speed dropping off to around 312 KIAS.

Yeah, but I relapsed over the course of another sleeping session. I should quit sleep altogether. :joystick: I'll make sure to post if I learn any more about what happened.

Seems to be a fundamental theme across all flight sims where the .50 cal is modeled. Take your pick: the various aircraft in IL-2, Targetware, etc. The only one I've seen that modeled the .50s with high damage was Warbirds, but then a lot of people complained that they were too strong, and flew like "lasers". It ends up being a very subjective discussion. From what I've seen, most sims model the .50 shooting either Ball or AP ammo, or a combination of the two. This is correct for many of the nations that used this caliber in WW2, but not the USA (among others). We loaded beltings that contained mixes of Ball, AP, HE, and incendiary rounds, as well as APHEI ammunition and tracers. There's a huge difference in the damage that aircraft-mounted .50s can do with this simple change. Of course, there were probably situations and periods where we were only shooting ball or AP, and I bet the pilots noticed this just as much as we do in sims. Now, as far as my experience in DCS: P-51 goes, there's quite a bit of variation in weapons capability. I've had fights where it took my entire belt to shoot down one plane, and others where a tap of the trigger sawed a wing right off, or set the engine on fire instantly. I do have a sneaking suspicion that ED hasn't added the more specialized rounds into the belt yet. If that's still to be implemented, I think we'll see a big improvement in the consistency of our 6 M2s if/when they do. I also hope they model the explosive visual effect when an HE round goes off. It's really impressive. Note: There are also at least two major versions of the aircraft-mounted .50, one which saw service during WW2 (the M2) developing 800 rounds/minute, and one which entered service in 1947 (the M3), developing 1200 rounds/minute.

Battery: ON Guns/Camera: ON (May not be necessary, but you want the sight anyway) Weapon Selector: ROCKETS (Up) RX Release Mode Knob: SINGLE or AUTO RX Counter: 1 (Reset if necessary) Press Weapon Release, as opposed to Trigger (not a Russian bird, after-all).

And here's the copy of the mission I promised. Use runway 30. PoniesAndConvoysV1.miz

Quick check shows no changes to any of the mission settings, and I did already go through the sim-wide settings post patch. It's also still possible that it's a placebo effect and my methods for controlling the aircraft changed in my sleep. There do seem to be plenty of people still having takeoff issues. Anway, the intent of my original post (now buried), was to congratulate ED on a great patch, whether I've taken my pills or otherwise. :thumbup: Oh, I was running tests on a mission design which involved a small nuke going off at both the red and blue bases if the two sides somehow managed to destroy each other's scoring structures at the same time. This was a 5,000,000 volume triggered explosion. It took out all the ground vehicles within about a 1 mile radius, but left the ammo bunkers and quite a few stock structures standing, if not all of them. It also left the P-51D sitting on the ramp perfectly intact. I figure that's something ED will want to look into at some point. The airfields were Kutaisi and Nalchik. There is also a huge frame rate drop for 1-2 minutes after the explosion, and the crater has clipping issues. These last two are perfectly understandable, of course. I just want ED to be aware of all this. No rush. EDIT: A multiplayer test this morning saw the takeoff issues come back. I'm going to check single player from the mission editor, where I was doing some of my past testing... EDIT2: Singleplayer from the ME showed the same results. I must've had an unusually good day yesterday. I'll try a more relaxed takeoff roll to see if that helps. EDIT3: A much much much more gradual than normal power increase helps quite a bit, but still doesn't provide the smooth takeoff that I saw. Either I'm imagining things and comparing two different missions, or it was a pilotage fluke. If anyone wants to give the takeoff a try, I'll attach it in a new post, below.

That's probably it. No changes to the ground handling at all though? I'm serious. I went from taking off once in a blue moon with 10xHVAR+max permissible weight in fuel, to being able to do it, and maintain centerline very easily. There was no practice involved; it was a sudden change after updating to the new patch. Same exact mission.

Not unless there's a bug in the interface, Nate. That and auto-rudder are some of the first things I made sure were turned off, even before my first flight. I always fly full-switch and from the cockpit, although I do turn on externals so I can appreciate the beauty of the model and take some pictures. ;)

No, I'm not sure. It's a gut feeling. Ground handling during takeoff and landing is DEFINITELY better though. Changes in the flight model may have something to do with the bugfixes with weapon stores. We'll have to see if ED says anything about that, otherwise all we can do is guess. How did you find my post in all that mess? Yeesh!