Advanced Flight Model - (AFM). What actually is it?

AFM seems to get banded around this forum a lot. And i can't see a definition or explanation to what is actually means.

With news that LockOn/Flaming Cliffs/2/3 airframes will be sold later this year as separate modules seemingly identical to FC3 versions but with AFM added, it would be great if we could get an explanation of what this meant.

What is AFM and how will it affect my experience as an end user?

Thanks.

The best way to see this would be to compare the flight modelling of the Su-25T you get for free with DCS World with the FC3 aircraft, if you have those. The Su-25T uses the advanced flight model, to give more realistic and enjoyable flight dynamics.

The simple flight model is, well, simpler. It doesn't give as good a feeling of flight, and I think most noticeably the transition between ground and air dynamics (e.g. when landing) is a lot nicer with the AFM.

AFM seems to get banded around this forum a lot. And i can't see a definition or explanation to what is actually means.

 

With news that LockOn/Flaming Cliffs/2/3 airframes will be sold later this year as separate modules seemingly identical to FC3 versions but with AFM added, it would be great if we could get an explanation of what this meant.

 

What is AFM and how will it affect my experience as an end user?

 

Thanks.

This should cover it mate. :)

Aircraft dynamics are calculated on the basis of the same physics equations describing translational and rotational motion of a solid body under the influence of external forces and moments, disregarding the nature of their origin.

Trajectory and angle movements look more natural due to correct modeling of the aircraft’s inertial properties.

Transitions between the flight modes in a smooth manner without abrupt changes of angle rotational speeds and attitude (for example: after a tail-slide or when landing with an angle of roll on one landing wheel).

Gyroscopic effect with the aircraft’s rotation taken into account.

The asymmetric effect of external forces is taken into account, along with the effect of external forces not going through the center-of-gravity (for example: engine thrust, drag chute forces). These forces are correctly modeled at any flight mode and cause an adequate rotary moment.

The center-of-gravity can change its location within the speed axis system.

The modeling of lateral and longitudinal center of mass has been introduced. This can change depending on fuel load and weapon loads.

The asymmetrical loading of weapon and fuel pylons, which influence the characteristics of lateral control (depending on flight speed, regular overload, etc), is also modeled.

When calculating aerodynamic characteristics, the aircraft is represented as a combination of airframe components (fuselage, outer wing panel, stabilizer, etc). Separate calculations for the aerodynamic performance of each of these components are performed. This is done over the entire range of local angles of attack and slip (including supercritical), local dynamic pressure and Mach number. This takes into consideration the change and level of destruction of control surfaces and various airframe components.

Aerodynamics are accurately modeled in the entire range of angles of attack and glide.

The efficiency of lateral control, and degree of lateral and static lateral stability, now depend on the angle of attack, longitudinal and lateral center-of-gravity.

The wing autorotation effect when performing a rolling rotation at high angles of attack is modeled.

Kinematic, aerodynamic and inertial interaction of longitudinal, dihedral and lateral channels (yaw movement when performing a rolling turn, rolling motion at rudder pedal forward, etc).

Angle of glide availability is determined by the pilot’s efforts and the plane’s position.

Possibly the most important distinction between SFM and AFM is this......"When calculating aerodynamic characteristics, the aircraft is represented as a combination of airframe components (fuselage, outer wing panel, stabilizer, etc). Separate calculations for the aerodynamic performance of each of these components are performed."

When I win the Euromillions lottery (it would probably help if I actually bought a ticket lol) I'm going to bribe An.Petrovich and his flight model posse with a ridiculous amount of money to work on an AFM for Razbams Harrier. If he nobly decides that he cannot be bought then I will have no alternative but to hire a squad of elite ex-Spetsnaz mercenaries to kidnap him and force him to work on a Harrier AFM whether he wants to or not. He will not be harmed in any way. Chocolate digestive Biscuits and vodka will be provided, I'll even fit out his temporary dungeon with a surround sound cinema complex to make his stay as a prisoner a little happier. This way progress will be achieved. This way the Harrier will excel, and be all that it can be. This will come to pass, and eventually you will all realise that "Biggles was a wise man" ; although his methods may have seemed a bit extreme, it was all for the best.....and the greater Utilitarian good.

*Falls off chair laughing*

:D

lol

Seems legit, keep on proceeding this. For the sake of more hi fi modules we have to use uncommon means.

AFM was introduced with the Su-25T in the first Flaming Cliffs expansion. It basically makes the aircraft more challenging to fly and allows for a more dynamic flying experience. Less train along a rail kind of flying.

Can't find the original movie, because youtube is bubbling everything. But theAFM was pretty exciting news when it first came :)

Compare this with this at 01:30 minutes in.

And you will see the A-10 "sticking to the rail" at touchdown.

I think AFM is using blade element method for calculating forces acting on a wing. In short you divide the object into smaller sections and calculate the effects of flow on each of them and then combine the forces together to get the net effect. It seems ED is also using some kind of additional model for simulating the flow not directly attached to the aircraft body. This is especially evident in Ka-50 flight model where the induced airflow has profound effects on the forces acting on the blades.

The aerodynamic forces in BEM (blade element method) are calculated by finding the velocity vector of the flow in relation to the element being examined and then using tables or formulas derived from experiments to determine the resulting lift, drag and torque. You can find these tables for different airfoils from the internet or from the manufacturers of the aircraft if you have the required connections. You can also make the experiments yourself if you have access to the exact airfoil shape and wind tunnel or CFD program (both methods are expensive and time consuming though).

BEM doesn't model the flow around the object like wake effects. For that you need some additional modelling. Doing it accurately requires using Navier-Stokes equations (this is what CFD or computational fluid dynamics programs do) which is complicated and requires lots of computing power and is currently not possible in real time with home computers. We are actually on the verge of having CDF flight models being reality as there's already some simple programs that can do this in real time but not with enough accuracy to model an airplane or with enough processing power left for any other calculations besides it. Having 5-10x the processing power of current high end gaming machines should do the job. Anyway for wake modelling (like propwash or induced flow in helicopter rotor) you need some kind of model that approximates the flow in a reasonably accurate manner but requires only simple calculations. I think this is the secret sauce that ED has that makes them the top dog in game industry flight models. BEM and CFD are both well known methods for aerodynamics modelling and used in engineering and science but there's no method that I know of for approximating the wake effects in conjunction with BEM.

X-Plane is yet another simulation platform long using BET.

Try to fly with Su-25T and then shot a heavy missil, the airplane discompesates for change of center of gravity. When you are landing, the air plane does not appear to levitate, I don't know how to explain it in words, but it's shows

Ok thanks.

So it is an increase in the complexity of the modelling of the forces acting upon the 3d model.

Hopefully ED will spell out how what the changes are. Rather than just saying 'it has a AFM.'

That PDF is a great insight into what goes behind the scenes in DCS. Thanks for posting CptSmiley :)

But despite all that math that has to be done in real time, it is not the main load in the sim engine. It is the graphics work that kill the FPS.

For ED's idea of what the advanced flight model is, its described in the Su25T pdf manual.

That PDF is a great insight into what goes behind the scenes in DCS. Thanks for posting CptSmiley :)

 

But despite all that math that has to be done in real time, it is not the main load in the sim engine. It is the graphics work that kill the FPS.

I don't think it is.

The engine ED uses has always been CPU limited, not GPU.

I don't think it is.

The engine ED uses has always been CPU limited, not GPU.

And because the way the graphic engine in the DCS works, it's held back by the CPU. Number of objects in the 3d model is in some way handled by the CPU and not GPU. And more objects that are in one scene the lesser FPS you get.

Tried, tested and documented.

Ok thanks.

 

So it is an increase in the complexity of the modelling of the forces acting upon the 3d model.

 

Hopefully ED will spell out how what the changes are. Rather than just saying 'it has a AFM.'

What you are thinking would be based on blade element theory. However I don't think AFM in any of the DCS products rely it. Thus the 3d model has no bearing on the flight characteristics. If that were case it wouldn't take forever to develop the AFM per plane.

Actually none of you are correct. AFM is whatever you want it to be.

 

Let me first define SFM. SFM is a way of defining a flight model utilizing a flew dozen parameters. While this can do a good job, SFM has no way of modeling the advanced flight control laws of modern fighter jets, which is why you get odd dynamic and control responses for the FC3 aircraft.

 

The AFM is essentially a blank slate API where the coder decides how they want to define the forces and moments acting on the aircraft. You can utilize whatever algorithms/methodology you wish, you can simulate a falling brick, you can use blade element theory, you can use JBSim configs, you can utilize CFD look up tables, you can use AVL/DATCOM or something like that. It's all up the the coder.

 

A while back I made a quick and dirty F-16 "AFM" using NASA and some other sources for the flight control system data: http://forums.eagle.ru/showthread.php?t=95985 Feel free to take a look at the code as an example of just one way of doing it, how ED is doing it, I have no clue, but they AFM models so far have been pretty awesome.

 

What I am trying to get at is that you can not assume that AFM is high fidelity, it is all up to how it gets coded up.

Hello CptSmiley. :) I have to disagree vehemently that the term-acronym AFM as coined by the ED-DCS developers themselves means 'whatever you want it to be', on the grounds that a baseline definition of its core constituent elements (and precisely how it differs from 'SFM') has been provided by the developers themselves in my previous post. :) Being a flight modeller yourself, you'll know that the laws of physics are not arbitrary, debatable, or "whatever you want them to be" either. :D

I'd like to point out that no 'opinion' or 'impression' of mine personally was offered regarding what the term advanced flight model pertaining to DCS actually means, but rather a concise explanation of it quoted verbatim from the developers themselves. :) I'm therefore slightly bemused as to how you can suggest that the definition of such as provided by the creators of said 'DCS-AFM' themselves can somehow be 'incorrect' lol.....maybe take up the issue with them by PM (jk). :)

It may well be that your vision of what constitutes an 'advanced flight model' may differ, but with the greatest of respect this does not matter and is irrelevant for the purpose of this discussion. The OP asked for a definition of what the term-acronym 'AFM' means in DCS and how it is distinct from their own 'SFM' and this was provided. :)

It is of great importance that the term AFM is given a clear and at least baseline definition for the simple reason that in the absence of such, it becomes as you wrote an arbitrary, nebulous term and "whatever you want it to be". The practical consequence of which would be that any third party developing for DCS could claim an aircraft implements DCS 'AFM', but if any of the aforementioned AFM criterion were not satisfied or missing, this would not strictly be true. This would be a misrepresentation, and misleading and unfair to the paying customer.

I take your point on board that individuals may go their own way regarding coding etc of said AFM, and I am not arguing that DCS AFM is the 'Alpha and Omega' or Holy Grail of all FM philosophies, but I'd strenuously argue that these baseline criterion as defined by DCS themselves in the previous post should be satisfied before anyone can claim an aircraft implements AFM as defined previously (I understand they are farming out the tools to do so to select 3rd parties). This is not to say that it is 'set in stone' and that evolution of AFM cannot (and probably will-already has) take place over time by either DCS themselves or some Genius 3rd party, but rather that it establishes at least a baseline standard; giving the consumer a clear reference regarding definition, and consequently peace of mind. :)

BTW, I should point out that what I know about flight modelling and physics could probably be written on the back of a postage stamp :D, I am most definitely no expert on these matters, but I thought I would quote those who matter in terms of defining the DCS term-acronym AFM, (the developers) and who are.

Cheers. :)

What CptSmiley described is what was originally referred to as the AFM API provided by ED to select 3rd parties (and then later included in the base install folder).

There have been many discussions on these forums about this topic, so I'll keep it short.

ED uses the term AFM to describe the fidelity of their flight models (and the process they use to create them, whatever that may be), but it has also been used by third parties to describe the fidelity of their own work.

Some people prefer using the term, EFM (External Flight Model), for third parties, as we are tasked with creating a physics model that plugs into ED's code through the use of external functions.

Personally, I use the terms interchangeably, as I consider the work I have done on modelling the L-39 to be more advanced than SFM.

While some may prefer to allow ED to define what an AFM really is, I consider it to be the blank slate that Smiley mentioned. The developer is free to build and define all forces and moments acting on the aircraft, without being restricted to defining a pre-set list of coefficients or characteristics. This is what gives these aircraft the truly dynamic responses when flying the edge of the envelope that are not possible with the on-rails standard flight models that require scripting to handle these regions.

Yesterday while testing the settings for viewpoint system in Su25A I decided to take off for a quick flight and after climbing to 6000m I made a dive on Mozdok airfield to see if I can make the air go boom! after pushing the Mach gauge pass 1.0... And I did... but ground was also coming real fast and the elevator was pulling my stick forward more and more so I deployed the brakes and dropped flaps... as I was pulling away from mother earth another bang shuttered my plane... I was under Mach 1.0 for some time for now so that noise was meaning something bad happened...

The plane was out of the dive by now, engines were still roaring just some lamps on right side of the dashboard were lit... but they were in russian so I ignored them. The plane was not flying right... it was rolling slightly by itself and instinctively after receiving no feedback to the stick movements I was pushing the rudder pedals to make it roll.

I manage to guide my stricken plane to be aligned with the runway and call the base for inbound... they were not responding. That was strange but I proceed anyway and while on final I push the airbrakes button once more... no sound, no vibration... I looked then on my windows leaning left and right as much as I could... I couldn't see the the wings... Oh sh...

So I remember a discussion between Wash and Mal while reentering with Serenity in a planet's atmosphere:

" This landing is gonna get pretty interesting."

"Define "interesting"."

"Oh God, oh God, we're all going to die?"

I dropped the gear but no 3 greens just a bleeping central red one... I touched at around 300kph plus something and after gear collapsing and all the noise and shuttering I push the shute release and abruptly the plane came to a stop. I've cut the engines open canopy and said to myself... "there... that's AFM!".

What CptSmiley described is what was originally referred to as the AFM API provided by ED to select 3rd parties (and then later included in the base install folder).

 

There have been many discussions on these forums about this topic, so I'll keep it short.

 

ED uses the term AFM to describe the fidelity of their flight models (and the process they use to create them, whatever that may be), but it has also been used by third parties to describe the fidelity of their own work.

 

Some people prefer using the term, EFM (External Flight Model), for third parties, as we are tasked with creating a physics model that plugs into ED's code through the use of external functions.

 

Personally, I use the terms interchangeably, as I consider the work I have done on modelling the L-39 to be more advanced than SFM.

 

While some may prefer to allow ED to define what an AFM really is, I consider it to be the blank slate that Smiley mentioned. The developer is free to build and define all forces and moments acting on the aircraft, without being restricted to defining a pre-set list of coefficients or characteristics. This is what gives these aircraft the truly dynamic responses when flying the edge of the envelope that are not possible with the on-rails standard flight models that require scripting to handle these regions.

Hiya Eagle, valid points, rgr and understood. :) Obviously speaking for myself only, I would absolutely prefer as a customer to have the ED-DCS definition of their 'AFM' as a basic standard, and for a clear distinction to be made otherwise; the term 'EFM' would do just fine. :) Semantics really but important in my opinion for the reasons outlined above and in terms of 'quality control'.

For my part I feel that having a baseline standard using a set definition for the description 'AFM' is profoundly important.....and to think of it as being 'restrictive' is perhaps a matter of perspective, I would take the viewpoint that these conditions and criteria should be satisfied as a bare minimum in order to be described* as such.....and still feel that any FM that does not include these core constituent elements and is described as 'AFM' (as I dare say most would expect and understand as being in DCS) would IMO have the potential to be misleading. Just an opinion like any other, but there you go. :) Any improvements upon this baseline and additions by 3rd parties such as yourself are of course most welcome! :thumbup:

I wish you the best of luck in your development endeavours too Eagle, keep at it and its great from a sim fan viewpoint to see people creating new stuff. Its all good. :thumbup:

AFM was introduced with the Su-25T in the first Flaming Cliffs expansion. It basically makes the aircraft more challenging to fly and allows for a more dynamic flying experience. Less train along a rail kind of flying.

 

 

Can't find the original movie, because youtube is bubbling everything. But theAFM was pretty exciting news when it first came :)

 

Compare this with this at 01:30 minutes in.

 

 

And you will see the A-10 "sticking to the rail" at touchdown.

I'm a bit confused with this post if someone can help me out. What does this guy mean by the first Flaming Cliffs Expansion? Is he meaning that Su-25T reacts like that once you upgrade to FC3, or is he talking about Flaming Cliffs2 or Lockon? Would seem silly not to have included this model in FC3.

I'm a bit confused with this post if someone can help me out. What does this guy mean by the first Flaming Cliffs Expansion? Is he meaning that Su-25T reacts like that once you upgrade to FC3, or is he talking about Flaming Cliffs2 or Lockon? Would seem silly not to have included this model in FC3.

The AFM first saw the light of day with the Su-25T in the first FC expansion to Lock On. Naturally, it followed into FC2 aswell as DCS World :)

I'm a bit confused with this post if someone can help me out. What does this guy mean by the first Flaming Cliffs Expansion? Is he meaning that Su-25T reacts like that once you upgrade to FC3, or is he talking about Flaming Cliffs2 or Lockon? Would seem silly not to have included this model in FC3.

Flaming Cliffs was the first expansion to LOMAC, ie LockOn Modern Air Combat. With the release of FC it became LockOn Flaming Cliffs, introducing the SU-25T and the AFM.

The AFM first saw the light of day with the Su-25T in the first FC expansion to Lock On. Naturally, it followed into FC2 aswell as DCS World :)

So just to confirm then that the Su-25T is a AFM in FC3 or Not? jeez wish I knew that if it is lol. I would have been learning to fly that first instead of the F15.

So just to confirm then that the Su-25T is a AFM in FC3 or Not? jeez wish I knew that if it is lol. I would have been learning to fly that first instead of the F15.

Yes it is Dudester (the first AFM overall) and its great fun to 'fly'. :)

I dropped the gear but no 3 greens just a bleeping central red one... I touched at around 300kph plus something and after gear collapsing and all the noise and shuttering I push the shute release and abruptly the plane came to a stop. I've cut the engines open canopy and said to myself... "there... that's AFM!".

I'm not with you, you call a successful landing without wings, AFM?