Can jester fly during refueling?

[upyr1]

I've heard it was easier to do in flight refueling from the back seat so it would be cool if we could hand this off to jester

[Rick50]

 

Hmm... that could be a nice workaround for people struggling to do such precision close formation flying! That way even n00bs could participate in complex missions before they gain the highest skill level!

Especially since the Phantom is probably not as easy/stable to fly as say a Hornet... considering no stability assist computers and such.

Good idea, I like it!

[G.J.S]

4 hours ago, Rick50 said:

 

Hmm... that could be a nice workaround for people struggling to do such precision close formation flying! That way even n00bs could participate in complex missions before they gain the highest skill level!

Especially since the Phantom is probably not as easy/stable to fly as say a Hornet... considering no stability assist computers and such.

Good idea, I like it!

Stab Augs similar to the F-14. Hornet flies within what it allows the pilot to do, and will in a sense help the pilot. Phantom is pilot ALL the way, no computer to stop the machine trying to kill you if you mishandle her. 

[Rick50]

So true, making it all the more difficult for virtual pilots.  

[Tiger-II]

Sims are too stable (flight models do not suffer the instability of real aircraft or model the minor perturbations you get from flying through real air), so I'd say it is a non-issue if you are relatively smooth to begin with (a major problem sim pilots suffer is they are not smooth).

If you can refuel the F-18, you should be able to refuel the F-4 (in the sim).

Edited April 4, 2022 by Tiger-II

[Rick50]

And I bet there's some for whom fueling the '18 or '14 and A-10 is too difficult and don't do long ranges because of it.  I'm just saying if we can have a great AI, it might be nice for those people, to be able to let the AI fuel up, and they can sit, watching the visual cues for later when they practice practice practice! And eventually get good at air fueling themselves

 

[sparrow88]

On 4/5/2022 at 1:29 AM, Tiger-II said:

If you can refuel the F-18, you should be able to refuel the F-4 (in the sim).

 

Thats BS... I can easily do AAR in the Hornet or Mirage almost with my eyes closed, but F-14 wobbles all the time making it 10 times more difficult. 

Still I would never use jester to refuel - feels like a cheat. Might as well turn on unlimited fuel in the mission editor.

 

Edited April 6, 2022 by sparrow88

[HEXO]

On 4/5/2022 at 2:29 AM, Tiger-II said:

Sims are too stable (flight models do not suffer the instability of real aircraft or model the minor perturbations you get from flying through real air), so I'd say it is a non-issue if you are relatively smooth to begin with (a major problem sim pilots suffer is they are not smooth).

If you can refuel the F-18, you should be able to refuel the F-4 (in the sim).

 

all of this is bull crap, modules VERY much suffer from instability like the: F-14, MIG-21, MIG-19, MIG-15, F-5, F-86, literally every helicopter but the apache, MIG-29, so please for the love if god stop sharing miss-information.

[Dragon1-1]

You forgot the Ka-50 (which is rock-stable, in fact, a bit too stable if you don't know how the autopilot works...). In general, this depends on whether the aircraft has FBW or not. FBW birds really are that stable IRL. Non-FBW, true to RL, require constant work to keep flying straight. The only difference is, IRL you feel the exact attitude of the aircraft through the seat, so you're constantly making small adjustments with the controls, while in the sim, you make larger, periodic corrections when you notice your flightpath is off (unless you have a motion platform that recreates the feedback, of course).

[Ala13_ManOWar]

2 hours ago, Dragon1-1 said:

You forgot the Ka-50 (which is rock-stable, in fact, a bit too stable if you don't know how the autopilot works...). In general, this depends on whether the aircraft has FBW or not. FBW birds really are that stable IRL. Non-FBW, true to RL, require constant work to keep flying straight. The only difference is, IRL you feel the exact attitude of the aircraft *through the seat, so you're constantly making small adjustments with the controls, while in the sim, you make larger, periodic corrections when you notice your flightpath is off (unless you have a motion platform that recreates the feedback, of course).

*Through the seat of your pants  , I must add

Edited April 11, 2022 by Ala13_ManOWar

[Tiger-II]

On 4/11/2022 at 8:46 AM, HEXO said:

all of this is bull crap, modules VERY much suffer from instability like the: F-14, MIG-21, MIG-19, MIG-15, F-5, F-86, literally every helicopter but the apache, MIG-29, so please for the love if god stop sharing miss-information.

LOL.

Do they? Really?? Do you even know what instability actually is? Two aircraft you really should be discussing didn't even get a mention, which is very telling indeed.

I have nearly every DCS module going, and I can tell you that at the edges of the flight envelope, several of them are quite seriously deficient. They don't perform as the real aircraft would (because the flight models are inherently stable).

MiG-15: if flying at high altitude/high Mach and you get into an accelerated stall, it should get into an aggressive, unrecoverable spin.

F-5: If you do a loop at too low an airspeed, the aircraft should want to pitch NOSE DOWN and enter an inverted spin. It does not. High alpha flight is also too stable, and in fact doesn't remotely reflect the difficulty of its handling. You can roll with ailerons without any care. The real thing would kill you.

F-86 is hard to spin, but the real thing wasn't.

MiG-19 last time I flew it (2020?) had some strange behaviors and didn't behave as reality. It could also climb far above its service ceiling.

F-14: it used to spin well (initial release) but because of "balance" or whatever the justification was, they dumbed down the flight model so now you need to work to spin it (not realistic at all). It was far better at release. The real thing was very prone to spins if mis-handled, especially at low speed.

The two aircraft you didn't mention which should be top of the list, are the F-18 and F-16. These aircraft were designed to be unstable and can't fly without FBW. Did you explore the flight envelopes of either of these? Again: too stable. Don't get me started on the F-16 handling in pitch. It improved recently, but still has a long way to go. The F-16 also doesn't seem to understand what a stall is. It just mushes.

All these problems are the result of the inherent stability of flight models. Instability must be added. Equations are insufficient to capture the instability of their designs as the airflow itself can't be modelled to sufficient detail to do so. It's a combination of aerodynamics and turbulent airflow around them that causes these handling difficulties. Add on top of that the fact that some aspects of the flight models are seriously lacking in fidelity (the LEF of the F-5 for example). Aside from adding a bit of drag and lift, they don't seem to affect the handling of the aircraft at all.

The magic words that hardly any simmer ever has to utter is: "I did a dumb thing and the aircraft departed controlled flight". Inadvertent spins are non-existent in any sim I can name, yet they kill many aviators in real life on an annual basis. The reason is: stability.

Edited April 23, 2022 by Tiger-II

[Dragon1-1]

20 hours ago, Tiger-II said:

Do they? Really?? Do you even know what instability actually is?

It's quite clear that you don't. For the record, stability in an aircraft comes out from how the forces acting on the aircraft are balanced. In a stable aircraft, the center of aerodynamic lift is behind the center of mass, meaning that as the angle of attack increases, the increased lift creates a rotational moment that reduces angle of attack. Thus, a negative feedback loop that brings the wings towards the zero AoA position. The elevator can then generate an additional rotational moment which counters the moment from the lift, allowing the aircraft to maintain a given AoA. Instability is when that's not the case and there's no feedback loop, so the aircraft doesn't try to return to zero AoA, it just, absent other inputs, stays at whatever AoA it currently is. That's it. So don't try to assign new meanings to a very clearly defined word.

Viper and Hornet both have FBW, which is always active. IRL, they are much more "stable" than non-FBW aircraft, in that they're more controllable and need less attention from the pilot. This doesn't mean they're aerodynamically stable, indeed, the F-16 quite famously has neutral stability, in other words, is unstable. That only means that the FLCS has to actively manage the aircraft's AoA so that it doesn't get out of control, and it does exactly that. FBW also smooths out the stall behavior and actively prevents you from departing the aircraft. The Viper has issues with its FLCS, Hornet too, but that's got nothing to do with their stability of lack thereof. Older aircraft like the F-14 also have features to make stalls less likely. 

You are confusing stability with high AoA controllability and instability with stall and departure modeling. Spins kill people flying light aircraft that don't have any stabilization, stall and spin prevention systems. Since aviation got big after WWII, people have invented a great number of devices that reduce this risk. You really do have to try to spin out in a modern fighter aircraft. Early F-14 might have spun easily, but that problem was mitigated in the versions that we have. Oh, and BTW, equations for modeling turbulent flow are very hard, it's a chaotic phenomenon that requires solving differential equations to model. It's possible, but very hard even for a single aircraft. 

I'm not saying spin modeling is perfect in all aircraft that we have, particularly older birds like F-5, the Sabrejet or any of the MiGs, but mumbo-jumbo about "flight models being inherently stable" isn't the reason. The reason is that these modules are old and somewhat neglected by their creators. Modeling performance at the edge of the envelope is hard, but very much possible in DCS.

Edited April 23, 2022 by Dragon1-1

[Tiger-II]

I thought we were discussing it from the pilot perspective rather than an engineering perspective, but you are correct.

The end result for a pilot, is that an unstable aircraft is always trying to end up out of control, and we do not experience this in the sim because the models are stable.

Quote

Oh, and BTW, equations for modeling turbulent flow are very hard, it's a chaotic phenomenon that requires solving differential equations to model. It's possible, but very hard even for a single aircraft.

It's not hard so much as the fact the average desktop computer lacks the processing power required to do so in real time, which we require for any usable flight simulation. Super-computers can about manage it, but anything of any real use is not capable of being computed in real-time.

It's easy to defeat AoA protections of even something like an F-16 - point the nose at ~45 degrees nose-high, wings-level attitude at idle power. By the time the aircraft bleeds off speed, even if the aircraft starts to pitch nose down, gravity will take over and AoA will far exceed 60 degrees AoA. It should settle quite nicely into a pretty interesting situation. The F-16 as modelled right now doesn't seem to do this. It just kind of falls out of low speed flight and re-gains speed without any problem (i.e., as if it never stalled).

PS: 60 degrees AoA isn't a random figure...

My point still stands that the flight models are stable. It's easy to model instability in the aerodynamic sense - you add something that approximates neutral or negative stability, then create a second layer on top (e.g. FBW sim) that corrects it. The point is though, without doing that, and in the absence of real-time calculations for airflow (turbulent or otherwise) the models themselves will not inherently create this behavior.

Inadvertent spin is another side-effect of stability of flight modelling. A spin is due to increasing yaw rate, created by stalling a wing (e.g. slow flight in an un-coordinated turn). One wing will have a higher AoA than the other, causing it to stall. Beyond the stall AoA it creates an increased amount of drag. This drag causes the yaw moment to increase, increasing yaw rate to that side. As the other wing accelerates it generates more lift, and causes the aircraft to roll over (until it, too, stalls - this is why aircraft sometimes enter inverted spins from an upright entry). The yaw rate becomes such that instead of the aircraft continuing to turn, conservation of momentum means that the aircraft stops turning/flying and starts rotating around the yaw axis (linear momentum becomes angular momentum). At this point it has departed controlled flight and unless the pilot does something, will continue into a fully developed spin where the aircraft is no longer flying, but rotating as it falls like a sycamore leaf to the ground.

We do not see this in simulators why? Because the flight models don't or can't compute it. The only way we can get them to spin (usually) is by inputting certain inputs, and only if certain pre-determined conditions are met, will the flight model act like it is spinning. The equations used for normal forward flight are not capable of developing to a spin without additional help.

This seems to be the point that is being missed when the word "stable" is used. It applies not only to situations like the above, but stability in the aerodynamic sense.

Some aircraft have interesting roll stability characteristics. The 747, at roll angles less than 32 degrees, demonstrates positive spiral stability (that means the pilot must hold in some aileron to maintain a constant bank angle). At 32 degrees of roll it will hold the roll angle with neutral aileron input. Beyond 32 degrees of roll, it has negative stability, and will roll tighter into the turn unless you hold some opposite aileron input to prevent the roll angle increasing further.

A simple model of dihedral will not model this particular behavior. It will always be positively stable at all roll angles. It must be augmented by additional calculations to determine the strength and limits of positive, neutral, and negative stability.

Just look at what happens when you get the F-5 into a high AoA situation and add rudder. It yaws, but the reality is the aircraft should roll! Even this is not correct. This is because the calculations being made even in this "simple" situation of high AoA are not adequate.

If the flight models were more sophisticated and modelled the more noticeable effects accurately, we wouldn't even be having this discussion.

7.22 makes for interesting reading for those who are still unconvinced that flight modelling is quite inaccurate. The flight models are failing at smaller (yet significant) details, so why is everyone so convinced they're right on the more important details? Just because something "hits the numbers" doesn't mean a whole lot for how the calculations were made to get there.

https://www.f4phantom.com/docs/F4_Phantom_Guide.pdf

Edited April 24, 2022 by Tiger-II

[Dragon1-1]

On 4/24/2022 at 3:27 AM, Tiger-II said:

The end result for a pilot, is that an unstable aircraft is always trying to end up out of control, and we do not experience this in the sim because the models are stable.

Here's where you're wrong. An unstable aircraft isn't "trying" to end up out of control. It simply stays where you put it, instead of "seeking its trim" like most of them do. You basically have to explicitly command everything, including starting and stopping each motion. When an aircraft acts to amplify each motion, it's negative stability. Aircraft are not designed for negative stability, and these days, usually have limiters to prevent parts of the envelope where that's the case from being reached. Older birds, like the 747, simply put a great big "NEVER ROLL MORE THAN 32 DEGREES!" warning in the manual. 

On 4/24/2022 at 3:27 AM, Tiger-II said:

It's not hard so much as the fact the average desktop computer lacks the processing power required to do so in real time, which we require for any usable flight simulation. Super-computers can about manage it, but anything of any real use is not capable of being computed in real-time.

For a game programmer (which is still the industry DCS people are in), calculus is hard, particularly dealing with anything involving numerical solvers. There may exist advanced numerical integration routines that can actually approximate that in real time with reasonable precision, however, they're pretty hard to understand and implement unless you're a numerical integration specialist. I doubt ED has one of those people.

On 4/24/2022 at 3:27 AM, Tiger-II said:

It's easy to defeat AoA protections of even something like an F-16 - point the nose at ~45 degrees nose-high, wings-level attitude at idle power. By the time the aircraft bleeds off speed, even if the aircraft starts to pitch nose down, gravity will take over and AoA will far exceed 60 degrees AoA. It should settle quite nicely into a pretty interesting situation. The F-16 as modelled right now doesn't seem to do this. It just kind of falls out of low speed flight and re-gains speed without any problem (i.e., as if it never stalled).

PS: 60 degrees AoA isn't a random figure...

You definitely can get into a deep stall with the Viper in DCS, you just have to try (or run out of energy in a dogfight and then try to fight in vertical...). You can even get out if you know the procedure, I learned in back in my FreeFalcon days and it still works here. I've done this once, you need to simultaneously assault two different limiters in order to make it happen. A pirouette with insufficient airspeed tends to do it. In the other Viper sim, throttling down at high AoA also doesn't get you there, IIRC, and Pete Bonanni said as much in the manual. I think there's an explicit rule in the FLCS that bunts the nose down fast enough to prevent this exact scenario. It doesn't feel like it stalled because it didn't, the Viper really does babysit you like that.

On 4/24/2022 at 3:27 AM, Tiger-II said:

Just look at what happens when you get the F-5 into a high AoA situation and add rudder. It yaws, but the reality is the aircraft should roll! Even this is not correct. This is because the calculations being made even in this "simple" situation of high AoA are not adequate.

Don't look at the F-5, it's an old, largely abandoned module. It's unlikely to be correct at the edges, and it's unlikely to gets its quirks simulated. The MiG-19 is also fairly neglected, and I'm not surprised the FM is bad, because so many other things are. If they can't get the lamp test correct, don't expect them to do a good job on the rest. Korean-era jets also aren't getting as much attention as they should. If you try the F-14, it rolls with the rudder perfectly fine, in fact if you try the ailerons at high AoA, you'll fall out of the sky. The modules that the devs care about are modeled better than the neglected ones, it stands to reason.

On 4/24/2022 at 3:27 AM, Tiger-II said:

Just because something "hits the numbers" doesn't mean a whole lot for how the calculations were made to get there.

The FM has to be, first and foremost, correct in the primary parts of the envelope people are going to be fighting in. In fact, that's a crucial element of any sort of model. It doesn't have to be correct everywhere, just in places that you care about. In particular, there's no need to model the real aerodynamics of the Viper, and then emulate FLCS to compensate for them, because in the end, you end up with the same behavior, except maybe for a few edge cases, such as FLCS failure. The new Viper FM fits the curves pretty well, and if they didn't break the deep stalls in the last update, it's OK by me.

As for "how it gets there", there's one thing that truly matters and that is speed. Symplectic integrators are fun, but if you can get the same curves with a bunch of addition and division operations, the latter method is preferable. DCS is bound to a single core, and even if it wasn't, it's got enough things to spend the CPU cycles on. What we need is the fastest model that is accurate enough to replicate the real jet's combat performance to within reasonable margins of error. Lots of compromises there, and there's no escaping that.

Edited April 25, 2022 by Dragon1-1

[sparrow88]

On 4/23/2022 at 2:14 AM, Tiger-II said:

F-14: it used to spin well (initial release) but because of "balance" or whatever the justification was

I assume thats why HB and others have SMEs - to tell them if its realistic or not. Pls show me where did HB mention "balance" in any of the flight model changes.

[draconus]

On 4/5/2022 at 1:29 AM, Tiger-II said:

Sims are too stable...

What most mean by that is "flying on rails" feeling but that's mission creator failure because the turbulence is set to 0 in 99% cases and it's the oppostite IRL.

Edited May 14, 2022 by draconus

[Exorcet]

On 4/23/2022 at 9:27 PM, Tiger-II said:

I thought we were discussing it from the pilot perspective rather than an engineering perspective, but you are correct.

The end result for a pilot, is that an unstable aircraft is always trying to end up out of control, and we do not experience this in the sim because the models are stable.

We don't experience it because no one makes planes that try to kill the pilot. In actual unstable aircraft you have the safety net of the FBW system which prevents you from feeling the instability.

On 4/23/2022 at 9:27 PM, Tiger-II said:

It's not hard so much as the fact the average desktop computer lacks the processing power required to do so in real time, which we require for any usable flight simulation. Super-computers can about manage it, but anything of any real use is not capable of being computed in real-time.

We've had fast enough computers for decades, the ones in the actual FBW systems have to be fast enough to account for instability or making flyable unstable aircraft wouldn't be possible in the first place.

On 4/23/2022 at 9:27 PM, Tiger-II said:

It's easy to defeat AoA protections of even something like an F-16 - point the nose at ~45 degrees nose-high, wings-level attitude at idle power. By the time the aircraft bleeds off speed, even if the aircraft starts to pitch nose down, gravity will take over and AoA will far exceed 60 degrees AoA. It should settle quite nicely into a pretty interesting situation. The F-16 as modelled right now doesn't seem to do this. It just kind of falls out of low speed flight and re-gains speed without any problem (i.e., as if it never stalled).

PS: 60 degrees AoA isn't a random figure...

The F-16 won't let you point it 45 degrees nose up.

On 4/23/2022 at 9:27 PM, Tiger-II said:

My point still stands that the flight models are stable. It's easy to model instability in the aerodynamic sense - you add something that approximates neutral or negative stability, then create a second layer on top (e.g. FBW sim) that corrects it. The point is though, without doing that, and in the absence of real-time calculations for airflow (turbulent or otherwise) the models themselves will not inherently create this behavior.

Why do we need real time calculations for airflow? All you need is the pitch stability derivative. Stability doesn't have anything to do with turbulence.

On 4/23/2022 at 9:27 PM, Tiger-II said:

The equations used for normal forward flight are not capable of developing to a spin without additional help.

 

What are the equations for normal forward flight?

 

  

14 hours ago, draconus said:

What most mean by that is "flying on rails" feeling but that's mission creator failure becuase the turbulence is set to 0 in 99% cases and it's the oppostite IRL.

You aren't wrong necessarily, but my overwhelming focus when setting weather in a mission is to make it so that the AI doesn't take 30 minutes to taxi. Perhaps when we get an ATC update there will be more interest in freely varying the weather.

Edited May 11, 2022 by Exorcet

[JayTSX]

I may be mistaken but aren't GPUs really good at physics related stuff? I mean they have plenty of parallel performance and since physic simulations aren't really something that you need to do in series of each other it seems logical to use them instead of our CPUs.

That being said I don't really see a necessity for airflow calculations since a good flight model will provide a great experience.

Also as a side note the F-14 feels quite good in my opinion moreso since the last overhaul meaning the plane will slowly rotate if not fully leveled. And above mach 1 she's quite shaky. Heatblurs modules seem to be way better than either the F-18 or F-16 (and much better than the mig-21).

[draconus]

4 minutes ago, JayTSX said:

I may be mistaken but aren't GPUs really good at physics related stuff? I mean they have plenty of parallel performance and since physic simulations aren't really something that you need to do in series of each other it seems logical to use them instead of our CPUs.

GPUs in DCS are fully abused when needed. Not optimally though so the Vulkan is WIP.

Otoh our CPUs use only 1 core and a bit of the second so we want to get all the spare cores working, hence Multicore is WIP.

[Dragon1-1]

For physics specifically, doing them on the GPU requires specific technology, such as PhysX or Havoc, which is unsuitable for DCS, since they're designed for tossing boxes around and not serious physics simulation. I don't know if it's even possible to write a custom simulation engine that would work on both Nvidia and AMD cards, but either way, that's way beyond the scope of DCS. 

[JayTSX]

Ok thx for the heads-up. The stuff they teach at college really is more theoretically possible than realistically manageable. And DCSs performance really does have a lot of potential since it's really only grabbing ram, vram and GPU while the CPU is basically bored but that's probably too off topic

[draconus]

6 hours ago, JayTSX said:

CPU is basically bored

You're wrong.

[JayTSX]

18 hours ago, draconus said:

You're wrong.

 I mean mines at 40% and atleast 4 of the 12 cores aren't utilized by DCS

Before you ask it's a Ryzen 5 2600x paired with an RTX 3060 and 32gb of Ram

But I guess our definitions of bored are quite different 

[draconus]

3 hours ago, JayTSX said:

But I guess our definitions of bored are quite different

I refuse to call our CPUs bored when there are multitude of situations in DCS where they are the bottleneck of performance.

[Hammer1-1]

On 4/6/2022 at 5:59 AM, sparrow88 said:

Thats BS... I can easily do AAR in the Hornet or Mirage almost with my eyes closed, but F-14 wobbles all the time making it 10 times more difficult. 

Still I would never use jester to refuel - feels like a cheat. Might as well turn on unlimited fuel in the mission editor.

 

 

I agree. Ive been able to refuel a2a since it was a thing in LOMAC, and the F-14 gave me the ability to flip the biggest FU finger to the tanker and shoot them down. Frustration = 110%.