How Realistic is FS2020?

[einherz]

so why i love fighters ac, you just have to flight incorrect/unpredictable way, and stay higher and or faster... i stay hope developers will give more attention to that path of simulation
i think if developers will try to create absolutelly right simulator, they will need quantum neuro web for this, i read that long time ago, what modeling flight is imposible for real, too much math formules work every single Instant even with direct flight in a relatively homogeneous environment, and from time of changing direction everything changing in much much complicated situation, so real flight simulator is oxymoron, but we can hope it will more and more closer to real

[Lewis - A2A]

Website updated here;
https://a2asimulations.com/a2a-simulati ... -aerostar/

thanks,
Lewis

[einherz]

so big path, that's cool...looks like we can download msfs2020 yet, just tryd it, and that's starts downloading, i stoped it, as didn't change place where to install in the windows setings... after reset store(pending after stop downloading), i think i will wait 18august, that's was must be beta or something predirect soft for msfs2020, i don't know
p.s. just watched msfs videos, yes there will exist overspeed buffeting, spins, flat spins, inverted spins

[gpbarth]

Alpha (and now Beta) testers aren't allowed to say anything about the sim. BUT - what I'd like to know, and haven't really seen a positive answer yet, is if A2A is talking to MS about developing aircraft for the sim, and if so, will those of us who have invested heavily in FSX (or P3D or X-plane) get some kind of discount for the FS2020 versions?

I've been flying FSX/SE for a long time now, and never wanted to go the P3D route because they were continually upgrading and updating the app, to where one version wasn't compatible with the next. X-Plane would have just meant doing the FSX trick all over again. So I stuck with FSX, FSUIPC, TrackIR, Active Sky 2016, FSDreamTeam, EZ-Dok, Aivla EFB, PFPX, ORBX, Navigraph, etc. And of course, my A2A and PMDG aircraft.

Now, none of that is needed. It's all in one package. Supposedly, TrackIR is in the works, but aside from that, you have real-time air traffic, real-time weather, real nav data, flight planning, and best of all, real scenery. True 64-bit, and will use all the memory you can throw at it. And on-line flying with your friends. I've got a ton of $$ tied up in FSX add-ons, but right now I'm ready to delete it all in deference to FS2020.

It will be released on August 18th, and they're taking pre-release orders right now. The most expensive version is only $119.99, so it's half of what you'd pay for P3D (legally). Low end is 59.99. But I digress...

I hate giving up my A2A aircraft. I want to keep them and fly them. Right now, FS2020 really has no study aircraft, and if I give up FSX all I'll have left for those planes is DCS. SO - will A2A (hopefully) be porting over to FS2020? I sure hope so.

[patful]

will those of us who have invested heavily in FSX (or P3D or X-plane) get some kind of discount for the FS2020 versions?

I'm gonna guess a No! there. It's an entirely different platform. We don't have to move to the new platform. I bought the new MSFS, but I won't ask for a discount if A2A happens to make an aircraft that I already have in P3D. I'm anxiously waiting to see what's in the queue behind the Aerostar, not interested in a twin.

[DHenriques_]

so why i love fighters ac, you just have to flight incorrect/unpredictable way, and stay higher and or faster... i stay hope developers will give more attention to that path of simulation
i think if developers will try to create absolutely right simulator, they will need quantum neuro web for this, i read that long time ago, what modeling flight is impossible for real, too much math formules work every single Instant even with direct flight in a relatively homogeneous environment, and from time of changing direction everything changing in much much complicated situation, so real flight simulator is oxymoron, but we can hope it will more and more closer to real

Not necessarily true. In creating a flight model (if you are good enough you create the total effect of what the physics are producing without actually having the physics involved acting on the simulated aircraft. In other words the creator of the flight model has to know what physics are involved in order to replicate their effect in 1's and 0's. If the creator is good enough (as A2A is most certainly) what you will see on the screen will be the TOTAL EFFECT of ALL that is acting on the aircraft at any instant in time.
It is the ability to accomplish this that has set A2A's Accusim program apart from the rest of the field when it comes down to accurate simulation. In my opinion there are no finer creative minds in the business than can be found here at A2A. What sets A2A aside is that those involved with our flight models not only know the physics involved but know how to code the flight model to reflect what is happening in real time.
There are several things that MUST be present in a flight model for it to meet A2A standards. Present must be accuracy, immersion, realism, and fidelity. Accusim reflects all these qualities.
Dudley Henriques

[einherz]

Dudley Henriques, well i talking too critically, but i stay thinking it's not can be real, as aircraft geometry same complicated as air condition for be real simulated. i do not say simulation can not be real at all, but it stay far of real, especially when we talking about extrem situations like work with supercritical aoa, different uncontrollable situations(spins, fluctuation, shading of deflecting surfaces, bending of fix surfaces with it... and more things, what i just don't know), and then, we have very complicated prop physic, at different pitch at different air speed and weather condition, so, i sure guys like a2a are real geniuse and do everything they can to accusim will maximum realistic, but it stay more far from real, than we can imagine, i think so
mike

[AKar]

Dudley Henriques, well i talking too critically, but i stay thinking it's not can be real, as aircraft geometry same complicated as air condition for be real simulated. i do not say simulation can not be real at all, but it stay far of real, especially when we talking about extrem situations like work with supercritical aoa, different uncontrollable situations(spins, fluctuation, shading of deflecting surfaces, bending of fix surfaces with it... and more things, what i just don't know), and then, we have very complicated prop physic, at different pitch at different air speed and weather condition, so, i sure guys like a2a are real geniuse and do everything they can to accusim will maximum realistic, but it stay more far from real, than we can imagine, i think so
mike

Speaking generally, a good simulation engine has in its background a solid, real-time, rigid body simulation engine running as a Newtonian physics background. This makes sure that the "large scale motions" are physically plausible and more or less realistic. Fine tuning of the forces acting on a body is generally not simulated via real-time aerodynamics, but by directly applying the force vectors on the body. In where these forces then come from lies the difficulty. Airplanes have various amounts of flight test data, and of various fidelities, available. From there, the forces acting on the body must be guessed as functions of Mach, Q, TAS and body angles, to name the most important ones. CFD can be used to calculate these functions, but this cannot be done in real time. The basics of flight mechanics are fairly well described, and reasons behind various observed behavior can be deduced, and the functions behind the forces tuned accordingly. But the fine tuning of flight dynamics almost always is a matter of manual adjustment. Not all regimes of flight can be simulated accurately, this is where the basic rigid body simulation takes over, providing plausible, if not truly accurate behavior.

In less developed flight dynamics engines one often has a fairly good, even excellent, behavior of an aircraft thorough its usual phases of flight, but the flight dynamics completely break down when something more... adventurous is attempted. This is a regime where that manual adjustment done to make a physically insufficient simulation to act right within partial flight envelope overpowers the true physics, resulting in impossible behavior.

None of this needs to be computationally too intensive, however, generally speaking, a good FDE must run at fairly high rate. That is, at higher rate than where the frame rates in FSX/P3D often dip in.

-Esa

[einherz]

so that's why in the real life every "same" ac have individual alife character, as real life don't work with mathematical body, but with real fixed and deflection surfaces, theoreticaly we can create quantum pc and give codes with exactly same p-39 with all "rivets" give real atmosphere and all air deltas conditions in every sm^3... give all aero(hydro)dynamic formulas... and after all it will really close to real, but... not at all

[AKar]

so that's why in the real life every "same" ac have individual alife character, as real life don't work with mathematical body, but with real fixed and deflection surfaces, theoreticaly we can create quantum pc and give codes with exactly same p-39 with all "rivets" give real atmosphere and all air deltas conditions in every sm^3... give all aero(hydro)dynamic formulas... and after all it will really close to real, but... not at all

I'm more optimistic in this one, for one can get the essence quite right, easily within where one's individual control setups make hugely more important difference than the discrepancies within the flight dynamics. Think it like the JPEG - the bulk of the information that matters can be hugely compressed with little loss in fidelity. The problem is really where either the details are missed (=not known & studied) or where the basic simulation breaks apart.

One of my favorite flight dynamics test maneuvers is the tail slide. No, I don't claim that many simulations provide thoroughly accurate depictions of what to expect when tail sliding an aircraft of a given model, nor do I expect them to. But there you often see whether the flight dynamics completely break apart and do, for instance, some snappy movements often indicative of the dynamics 'gimbal locking'.

[AKar]

so that's why in the real life every "same" ac have individual alife character, as real life don't work with mathematical body, but with real fixed and deflection surfaces, theoreticaly we can create quantum pc and give codes with exactly same p-39 with all "rivets" give real atmosphere and all air deltas conditions in every sm^3... give all aero(hydro)dynamic formulas... and after all it will really close to real, but... not at all

And I must add that the differences between the "same" airplanes primarily come from the tolerances of manufacture and deformations, repairs affecting the aerodynamics, minute differences in the control system, asymmetric weight & balance distributions, different engine performances, particularly in case of multi-engine airplanes and so on. These are all the effects that require no rivet-level aerodynamics, as there are much, much more pronounced causes for these effects. I've flown two gliders of the same make and model, and of very close serial numbers, and have noticed a distinct if somewhat un-quantifiable difference in between them. The aileron-pedal feel was just clearly different in some way, perhaps one being stiffer in another or something. These are often ignored by the flight dynamics developers. I guess, for it would be an ungrateful job to first simulate an aircraft "perfectly" and then add some imperfections on it. Ironically, these imperfections could often be deduced from the very data the flight dynamics were first developed from.

Edit: And of course, including such often results in complaints that the aircraft does not fly perfectly, as it is supposed to...

-Esa

[DHenriques_]

so that's why in the real life every "same" ac have individual alife character, as real life don't work with mathematical body, but with real fixed and deflection surfaces, theoreticaly we can create quantum pc and give codes with exactly same p-39 with all "rivets" give real atmosphere and all air deltas conditions in every sm^3... give all aero(hydro)dynamic formulas... and after all it will really close to real, but... not at all

And I must add that the differences between the "same" airplanes primarily come from the tolerances of manufacture and deformations, repairs affecting the aerodynamics, minute differences in the control system, asymmetric weight & balance distributions, different engine performances, particularly in case of multi-engine airplanes and so on. These are all the effects that require no rivet-level aerodynamics, as there are much, much more pronounced causes for these effects. I've flown two gliders of the same make and model, and of very close serial numbers, and have noticed a distinct if somewhat un-quantifiable difference in between them. The aileron-pedal feel was just clearly different in some way, perhaps one being stiffer in another or something. These are often ignored by the flight dynamics developers. I guess, for it would be an ungrateful job to first simulate an aircraft "perfectly" and then add some imperfections on it. Ironically, these imperfections could often be deduced from the very data the flight dynamics were first developed from.

Edit: And of course, including such often results in complaints that the aircraft does not fly perfectly, as it is supposed to...

-Esa

One factor that has been extremely hard to replicate in a flight model has been the effect with q on control surfaces. Each aircraft will exhibit different responses to dynamic pressure and inertia due to the size of the surface and the airspeed and altitude at which the control surface is deflected and how far in degrees that surface is deflected coupled to the aircraft gross weight which is changing as fuel is used and the cg changes position on the envelope.
It's a VERY difficult formula for the flight model designer to consider. The result is almost always a compromise coded as a scalar effect that reflects magnitude with a vector tagged on to that for direction.
It's the old "how do we make it LOOK like it would look had we coded all the physics involved"?
I agree with you on the tail slide. Very difficult to code that. For me in testing a flight model I use a super slow roll. Any discrepancy in the code shows up immediately (just as it does when doing an acro eval on a pilot in the actual airplane).
Much more to all this as I know you are well aware. . These are just some basic observations!
Dudley Henriques

[DHenriques_]

so that's why in the real life every "same" ac have individual alife character, as real life don't work with mathematical body, but with real fixed and deflection surfaces, theoreticaly we can create quantum pc and give codes with exactly same p-39 with all "rivets" give real atmosphere and all air deltas conditions in every sm^3... give all aero(hydro)dynamic formulas... and after all it will really close to real, but... not at all

And I must add that the differences between the "same" airplanes primarily come from the tolerances of manufacture and deformations, repairs affecting the aerodynamics, minute differences in the control system, asymmetric weight & balance distributions, different engine performances, particularly in case of multi-engine airplanes and so on. These are all the effects that require no rivet-level aerodynamics, as there are much, much more pronounced causes for these effects. I've flown two gliders of the same make and model, and of very close serial numbers, and have noticed a distinct if somewhat un-quantifiable difference in between them. The aileron-pedal feel was just clearly different in some way, perhaps one being stiffer in another or something. These are often ignored by the flight dynamics developers. I guess, for it would be an ungrateful job to first simulate an aircraft "perfectly" and then add some imperfections on it. Ironically, these imperfections could often be deduced from the very data the flight dynamics were first developed from.

Edit: And of course, including such often results in complaints that the aircraft does not fly perfectly, as it is supposed to...

-Esa

One factor that has been extremely hard to replicate in a flight model has been the effect with q on control surfaces. Each aircraft will exhibit different responses to dynamic pressure and inertia due to the size of the surface and the airspeed and altitude at which the control surface is deflected and how far in degrees that surface is deflected coupled to the aircraft gross weight which is changing as fuel is used and as the cg changes position on the envelope.
It's a VERY difficult formula for the flight model designer to consider. The result is almost always a compromise coded as a scalar effect that reflects magnitude with a vector tagged on to that for direction.
It's the old "how do we make it LOOK like it would look had we coded all the physics involved"?
I agree with you on the tail slide. Very difficult to code that. For me in testing a flight model I use a super slow roll. Any discrepancy in the code shows up immediately (just as it does when doing an acro eval on a pilot in the actual airplane).
Much more to all this as I know you are well aware. . These are just some basic observations!
Dudley Henriques

[AKar]

One factor that has been extremely hard to replicate in a flight model has been the effect with q on control surfaces. Each aircraft will exhibit different responses to dynamic pressure and inertia due to the size of the surface and the airspeed and altitude at which the control surface is deflected and how far in degrees that surface is deflected coupled to the aircraft gross weight which is changing as fuel is used and the cg changes position on the envelope.
It's a VERY difficult formula for the flight model designer to consider. The result is almost always a compromise coded as a scalar effect that reflects magnitude with a vector tagged on to that for direction.
It's the old "how do we make it LOOK like it would look had we coded all the physics involved"?

Yeh, indeed to add to the commentary, one thing I think that was often overlooked was the aeroelasticity. As that adds one a rather complex formula, or in fact a set of formulae, into the system, primarily those of mechanical wing twisting moments and those of resulting 'driven inputs' against 'direct inputs'.

Usually these simply need to be "scripted". But often this was only done where an actual reversal was well documented.

Studying airliners with mechanical controls, one often finds they simply resorted in locking the outer ailerons (where the wing twists) at higher Mach numbers, simply relying to the inner flaperon kind of things for inner ailerons and the spoilers for roll control at high speeds. But does the simulation still work with these specific controls disabled? (In early days of FBW, there were reasons for the Airbus to resort to the comparatively stiff wing in the A320 series, that still remains today.)

In fast, maneuvering jets with by the time novel systems of flight controls, the F-18 being my favorite, this specific aileron reversal had to be engineered into the aileron gains as in strongly negative gain in normal CAS, being dependent of quite some factors. As far as I know the type, this resulted in excellent roll behavior thorough the envelope, using very modest computing power within the FCCs.

-Esa

[einherz]

very interesting