After a few months in hibernation pending other commitments, I have restarted work on my project, still plagued by the limited and old fashioned physics model that plagued FS of old. Whilst I understand the desire to make the sim backwards compatible, maintaining compatibility with a game long since obsolete makes no sense to me. I am sure there is scope within the sim to have newer aircraft have their own physics entirely, (and not the new calculation model that Asobo goes to great lengths to describe which still relies on basically the old system), but instead completely now phyiscs, supported by the sim natively and not through modding (as I will be forced to do), bodging or other means. Some examples of how limited it is are the following
no option for more than 4 engines, more than the specified controls surfaces not allowed (multiple tails/wings etc), limited numbers of weight stations, fuel tanks, lights ā¦, the options for anything but conventional aircraft are so limited
vtols, helicopters, balloons and airships completely unsupported (weāll wait and see if helicopters are realistic when they arrive)
Will there be any plan to expand the envelope of possible flight regimes from the old ESP, or is this the endgame. This is supposed to be Flight Simulator, not Plane Simulator after all
The fact they reused an at least two decade old system I find frankly shameful. However they didnāt make it possible to just turn it off, and build something custom. SimConnect has its limits and although a custom FM is possible it still has to fight against the in game systems. There are aspects that I would keep, such as the engine systems, calculations of thrust etc, but I would love to be able to calculate the forces on the aircraft correctly and not have to adjust them to fight the in game model. Some areas though are just dumb, like the limits on weight stations, engines, fuel tanks (you canāt even name them), control surfaces etc. Ik that for my project the realm is niche at best, but even so, the current setup is antiquated and hardly modern in any way. Anyone have any thoughts?
What do you mean with modern?
A few days ago I browsed the SDK manual. I was impressed.
Asobo uses the same system of equations of motion aeronautical engineers use today all day in day out, - now since 100 years succesfully.
To make a suggestion:
In a completely open system one would introduce the own equations of motions. With a few years of formal education
in aeronautical engineering and control thatās no hindrance:
Just three equations for force equilibrium and three more for
momentum equilibrium.
But believe me, there are not so many guys who are able to do this by themselve better than the ones used in the sim.
In the very far future your wish comes true
In a software like ANSYS Fluent you then can simulate the
whole airplane behavior without ever having studied anything.
Then just import your CAD model !
Bye, walter
P.S.: Of course the better the model, the more Iād love the sim !!
And, of course, the Asobo guys would love it too, to give us the best models they can.
By modern I mean the principles by which the simulation is grounded, the fact we are still using cfg files to define aircraft parameters is something from the last century. There is so much limiting the choice of the designer. Then we move onto their new system which has a set of pre-defined surfaces to which the aircraft is built around, other than the most conventional layout, this requires bodging.
Believe me I intend to utilise my own, have made an airship work with custom physics calculations in unity (completely from scratch), the only thing that would be nice in the meantime is the ability to decouple the aircraft from the in game physics, remove wings, (will need to simulate my own) etc
As a layman, Iād assume that should be no great programming problem for Asobo.
Yet I guess, then Asobo would breed their own competition.
For the aircraft designer/engineer it would be great if your idea could be realized.
I have Unity on my PC. Never used it.
Did you programm all the equations of motion (in C++? ) ?
Love your airship!
FSX (and P3D) donāt simulate slats either. Maybe thatās why Asobo didnāt bother.
The hair raising wrong FSX like thrust reverser animation is also still there since day one.
By bodging I mean trying to replicate the behaviour of the aircraft within the existing framework, making do. Great examples are all previous airships, any aircraft that use flaps to create vtol etc etc. A total fudge and no way realistic, thought by now we could move past this.
Also ASOBO breeding their own competition, I thought that was the whole point of them being open and constructive with the community
what is that? are you referring to externally injected flight models?
That canāt be right. MSFS 2020ās āmodernā flight model is the best flight model so far that ever existed for home desktops. Itās very high resolution and offers 4D air mass simulation. This is what their own developers said in public. It canāt go back to the old flight model. If it does, itās no longer āmodernā.
Help me understand this. Are you referring to Legacy or Modern flight model?
I believe they are very honest with the community. After five years of MSFS development they have now announced that they are looking for a French-speaking aerodynamicist programmer and thus arenāt ashamed to admit in public, whether directly or indirectly, that they have a severe lack of knowledge in areas concerning aerodynamics. Thatās a good sign.
The introduction of the Volocopter was extremely interesting to me, as it is another configuration of vehicle. I really hope that support for evtols is properly supported. Even the prospect of having a dedicated electric engine backend system can open up some interesting aircraft (like the Pipistrel electric aircraft already on offer, which would be a fairly easy thing to do seeing as the model already exists in the premium version).
Aviation has so many different facets of aviation, and I REALLY hope they are able to design an open flight model that can accommodate them all:
Flight models:
Conventional aircraft
Tailwheel aircraft
Multi-wing aircraft (biplanes/triplanes etc)
Gliders
Swept wing aircraft
Retractable wing aircraft
Helicopters
Variable Angle Rotorcraft (eg: V22)
Swept wing aircraft
Multi rotors
Lighter than air aircraft (helicopters/blimps, airships)
Lifting body aircraft
Simulation of surfaces that exceed the speed of sound (whether itās a wing, or the whole aircraft)
think of a gulfstream jet whoās wing is approaching the speed of sound, propellers whoās tips do exceed the speed of sound
As you can see the aviation scope is massive, and Iām sure thereās a few Iāve missed. In reality we canāt have them all, but a structure that supports a great deal of them will ensure this sim remains viable for a long time.
Aviation is currently beginning a massive shift in concepts and ideas - similar (and probably greater) than what we had when the jet engine was designed. And the boundaries between air and space are again being diminished.
FSX āthinkingā shouldnāt have been continued at all, and many of its ideas and limitations have appeared in MSFS2020. Flight dynamics, whilst better than FSX , are much improved - but we have a very long way to go yet until theyāre reliably realistic.
By the antiquated system I mean that MSFS 2020 still relies on cfg files to define the basic parameters of the airframe, they may use the most modern physics calculations to give a ārealisticā flight experience but the parameters (simvars) which can be set are very heavily limited by a two plus decade old system. It restricts the ability to do any of the following very eloquently put by Parorng above, only a conventional aircraft is currently supported.
Well I hope they are open to tearing out the old as its the only thing holding back what is otherwise a masterpiece of modern gaming. The most annoying part is they couldāve just started from scratch, and had a tandem of both old and new
You cannot define the aircraft as having multiple wings/tails/surfaces, more than 4 engines, different types of engine, exotic lift varieties, variable geometry to name a few examples.
Moved this thread to Wishlist as I think this should have some attention drawn to it. The lack of flexibility from the old ESP CFG file flight model simvar definition system deserved to die a long time ago
