L-39 Flight Model

I don't know if it's already been said, but during testing i found this (just didn't get around to post it straight away)

Flying at VNE in zero wind, pull straight up until just before the yellow marker on the speedometer (just a reference, ok? it can be done at different speeds). keep full throttle still, and try to pull over the top... in some circumstances the aircraft will start to push the nose up and all the way over the top to where you just left off, but in opposite direction. the whole thing should be quite clean, wings level, and pulling full elevator through the whole thing. sometimes it will also start it, even though you are clearly heading down towards earth with a significant angle.. (not random, just some times it's hard to do it right (or wrong.. whatever you'd like to call it)

this will NOT happen if you let of the power..

Being a glider pilot i could possibly not understand the physics of thrust, but is it that because of a negative thrust angle? i don't know, but does that not seem a bit weird that a tiny bit of thrust angle would flip the whole thing around perfectly like that? the engine doesn't have THAT much power after all..

...pull straight up until just before the yellow marker on the speedometer (just a reference, ok? it can be done at different speeds). keep full throttle still, and try to pull over the top... in some circumstances the aircraft will start to push the nose up and all the way over the top to where you just left off, but in opposite direction.

I don't think I follow what you're describing.

...zero wind...

Wind isn't relevant.

On a related note. I was surprised that the Albatros requires very little rudder input. I had to set the curve to 30 and I still have to only touch it very gently in turns. Unless I'm doing something crazy the ball is pretty much centered almost all the time. Did you guys notice ? What you think about that ? I'm no pilot so I'm just asking out of curiosity for other people's opinions.

Hi guys,

Let me start by saying that I am by no means an expert but have read a bit about flying fast jets and my understanding is that the turn and slip indicator (the ball) barely moves during a turn. Fast jet pilots do not use rudders to make co-ordinated turns in the same way light aircraft or glider pilots do. In a fast jet the rudder is only used during a crosswind landing situation where the low approach speed gives it maximum effect.

The rudders on fast jets have very little authority compared to light aircraft and gliders. In FBW aircraft pilots only ever use the rudder for crosswind landings and taxiing.

The effectiveness of the rudder is also a function of the size of the control surface, so the rudders on an A-10 will be much more effective than on an L39.

All that said, reading through the other posts in this thread it seems the ball on the L39 isn't quite behaving the way it should.

:smartass:

RBG

I don't think I follow what you're describing.

I think he's talking about having noticed that the thrust from the L-39 engine is not parallel to the sircrafts axis of flight, so if you add thrust without trimming, the nose will want to drop - he's finding that as he does a loop at full thrust & eases presure off the stick as he comes over tthe top, the nose wants to 'drop' and cause the aircraft to climb out the loop inverted..

"this will NOT happen if you let of(f) the power.."

Thank you for helping me out :)

All in all it's very difficult to simulate something so complex as a flight model properly on a desktop pc.

 

I have experience in 0-, 1-, 2- and 4- engined planes, recips, turboprop and jet. Been flying for just about 30 years now and stopped really counting my hours at 14k ( and that was a while back )

 

One of the most difficult things to model properly in a sim is the transfer of a real life feel and ease of use of systems and flightmodel to a program. Let me give a few examples..

When I program or use the FMS in my "office""my hands and fingers have it's own memory and inputs can be made instinctively and "on the fly". When I do that in a PMDG sim it takes up a a lot more of my resources, attention, and physically hands on then in real life. An A10 pilot will most likely have the same feeling when hitting all the buttons in the sim. It's soo much easier in the real plane.

Same goes for something like using the gear in the Mig 21. Doing it in the plane will be a matter of fact and won't require as much attention as hitting a few keys or trying to hit the right spots with the mouse while trying to keep your TIR stready.

When you talk about flight model there is again a very big discrepancy between sim and rl. Some planes have high control forces or a big difference between roll and pitch inputs. Different feel between high and low speed flight. And ofcourse the lack of physical feedback. A lot of sims model a high speed stall with one wing just dropping violently. Since there is a complete lack of feedback you ( Well me personally.. ;-) ) find yourself too quickly at the limits with a consequent departure of normal flight.. ;-).

 

Since there are a lot of different hardware setups, settings and curves it's very difficult to recreate the right "feel" for everyone. Personally I do have the feeling that the flight-model is acting a bit strange in some parts of the flight enveloppe but since I have no experience in the -39 or any other of the DCS planes I can't really comment on that right now. Since the release of the beta there will be a few more people flying it who do have -39 experience who can hopefully make some educated comments about the flight model.

 

On a personal note, I like where the FC3 planes are going. Proper performance of plane, systems and weapons. Using radar and other systems take the same amount of effort to use as irl. Giving an experience that is comparable to real life. By making the interface to make the systems do what you want them to do easier you counter the extra attention required to do so behind a screen with hotas and mouse.

I am exaggerating now >> Just think of yourself driving a stickshift.. in the car it's done without even thinking about it. If you model it on a pc you need to hit a key for the clutch, then shift , then ease the clutch back up. When you do that with a mouse you will have all your attention on just accomplishing that task. Easing that whole input procedure with a single up and down keystroke or keybinding makes sense to me. To me the FC3 planes are like that.

 

Like I said this is my personal take on how far complex systems should be modelled in the sim. If you want to mouseclick all over the pit to make it do what you want, go for it!

Very interesting thoughts...

I'll bear that in mind

Pman

OR..... you simply build your own homecockpit :D

OR..... you simply build your own homecockpit :D

Or you learn well the keyboard keys that replicates the switch function.:D

Or you learn well the keyboard keys that replicates the switch function.:D

So by learning to type really well you somehow have the "in-plane" experience? Don't think so.

There is no good solution here. It all depends on what you are looking for. I am looking for the in flight eperience. Others get their kick out of hitting each and every button and switch and being able to coldstart every plane. With the advances of technology, Track Ir, and Oculus Rift the sense of being there is getting better and better. So who knows what else will be in store for us in the coming years.

So by learning to type really well you somehow have the "in-plane" experience? Don't think so.

 

There is no good solution here. It all depends on what you are looking for. I am looking for the in flight eperience. Others get their kick out of hitting each and every button and switch and being able to coldstart every plane. With the advances of technology, Track Ir, and Oculus Rift the sense of being there is getting better and better. So who knows what else will be in store for us in the coming years.

I disagree that there is no good solution, and completely agree with your earlier sentiment. I am by no means advocating simplifying the system modeling of our virtual aircraft, but designing a good user interface that will allow most of us to access and use the systems in question with as few keypresses as possible would really go a long way. Even something as simple as using one key to toggle three different flap positions is an improvement over using three different keys to do the same thing. This doesn't have to be mutually exclusive.

VR is coming, keyboards will be useless, mouseclick helps.

The DCS L-39 is very stable which make sense being as its trainer, The stalls are pretty brutal though, I was trying some slow flight (around 160 kilometers p/h) holding level and I got to say it feels like balancing on a needle point at the brink of snap rolling inverted. I even tried an inverted spin and mucked up the recovery and recovered at around 50' down from around 10k, very believable inho even though I'm not sur the little turbine could stay running during a maneuver like this. In the end I find it convincing and fun to fly

Real L-39 stalls like a pussycat... stick talks to you a little bit and let the nose drop and it stops. Deep stall is a lot bumpier - the airplane says "stop doing that to me". But even a deep stall doesn't need much rudder input to keep the ball in the middle.

One thing to note is that the L-39 has very substantial secondary stall characteristics. Stall it a little, let it recover and it will stall again when you pull back on the stick at 20 KIAS higher than the first stall.

When you get into a good stall or spin the L-39 is a real stick snatcher. Both hands on the stick to have good control of it. To properly simulate this the sim should throw your joystick off your desk if you don't hold on tight. :-) The L-39 doesn't like to spin. The L-29 loves it, the L-39 will make you not want to do it a second time.

Also note it's very easy to get on the back side of the power curve during slow flight. Gear and half flaps and when you get a very high alpha it just mushes along. You can have climb power (102%) set and it will plow through the air with the nose up. Also because of this it doesn't pick up airspeed easily when the airplane is dirty. Dropping the nose on landing doesn't do much, you have to add power AND drop the nose, and add power early because of the long spool-up time. Best to set 78% at the perch on a break, leave it there, and manage the descent with angle of bank instead of pitch. There can be some advantage to this - if you are a bit high on the landing you can push the nose down and it doesn't pick up speed. You do not, not, not want to be low and slow on landing in an L-39.

Inverted spin is a prohibited maneuver in the L-39.

AI-TL is very robust engine. As demonstrated by the crash near Boulder City NV, the plane pancaked into the desert, probably 20-30G impact, and the engine was still running. You can tail-slide the airplane, but I wouldn't recommend it.

Note how the ball remains pretty much centered during the entire flight without any rudder input what so ever (only full during takeoff and landing). That doesn't seem right.

Fly the real airplane with your feet on the floor... rudder is for takeoff, landing, and taxi only. :-)

I'm always amused by such discussions. I understand, when there is something obviously wrong. But when there is "something strange" about the FM....it's a little bit weird. I can apply the same for almost any airplane in DCS, because I don't fly these machines and MANY things can feel strange, but I take it as it is, because I'm not a pilot. I can have 100 000 hrs in some Cessna or whatever, but it doesn't give me even the slightest idea of how jet engine with such specific airframe acts. It looks to me, that people enjoy talking about sim more than actually fly it. Just get in the cockpit, learn to fly it...problem solved. It's BETA, so if ED will find some strange behavior (still, they have real pilots to discuss it with), they will update the sim. If You are easy on the stick with flaps extended and on proper speed,...no shaking in the cockpit on my side. People just tend to pull like hell, then, of course, the shaking appears. No offense, but You know this is just a PC sim, right? So enjoy it, fly it and we can get rid of making such weird statements about something, somewhere and not sure if....

Oh geez calm down. As I said in my very first post :

I'm no pilot so I'm just asking out of curiosity for other people's opinions.

No need to start the age old debate "it's just a sim" and whatnot. I was just curious if this the plane really flies this nice. So apparently it does, great. Thanks all.

Did I quote You? It was more like reaction to the origin of this thread. No hard feelings at all...:-)

Real L-39 stalls like a pussycat... stick talks to you a little bit and let the nose drop and it stops. Deep stall is a lot bumpier - the airplane says "stop doing that to me". But even a deep stall doesn't need much rudder input to keep the ball in the middle.

 

One thing to note is that the L-39 has very substantial secondary stall characteristics. Stall it a little, let it recover and it will stall again when you pull back on the stick at 20 KIAS higher than the first stall.

 

When you get into a good stall or spin the L-39 is a real stick snatcher. Both hands on the stick to have good control of it. To properly simulate this the sim should throw your joystick off your desk if you don't hold on tight. :-) The L-39 doesn't like to spin. The L-29 loves it, the L-39 will make you not want to do it a second time.

 

Also note it's very easy to get on the back side of the power curve during slow flight. Gear and half flaps and when you get a very high alpha it just mushes along. You can have climb power (102%) set and it will plow through the air with the nose up. Also because of this it doesn't pick up airspeed easily when the airplane is dirty. Dropping the nose on landing doesn't do much, you have to add power AND drop the nose, and add power early because of the long spool-up time. Best to set 78% at the perch on a break, leave it there, and manage the descent with angle of bank instead of pitch. There can be some advantage to this - if you are a bit high on the landing you can push the nose down and it doesn't pick up speed. You do not, not, not want to be low and slow on landing in an L-39.

 

Inverted spin is a prohibited maneuver in the L-39.

 

AI-TL is very robust engine. As demonstrated by the crash near Boulder City NV, the plane pancaked into the desert, probably 20-30G impact, and the engine was still running. You can tail-slide the airplane, but I wouldn't recommend it.

I can confirm this.:thumbup:

You should add that the ac has a built in mechanical stall warning system (the two little fences in the leading edges). Most people are not going to understand "stick talks to you"... :D

I can confirm this.:thumbup:

 

Most people are not going to understand "stick talks to you"... :D

:D

The L39 is really easy to read. I never seems to do anything unexpected or surprising. You always get some sort of warning before you piss it off.

I have noticed that despite the nose being beyond the vertical - and the aircraft being effectively inverted, a stall always results in the nose falling upwards, and not continuing downwards. The aircraft goes back up to, and and then beyond the vertical, and then into a nose down attitude. Try attempting a loop without sufficent airspeed, and see what I mean. once past the first quarter, after the aircraft nose has gone beyond the vertical, I'd expect a stall to result in the aircraft falling onto it's back, and inverting fully before the nose drops below the horizon. Instead, the nose rises, back beyond the vertical, and into a stall with the aircraft ride side up.

Is the current stall behaviour correct, or would the stall, commencing when partly inverted, result in the aircraft going further inverted, and take the shortest path to a nose down attitude?

I am assuming it is in fact an unfinished aspect to the flight model at the moment?