• Which the release of FS2020 we see an explosition of activity on the forun and of course we are very happy to see this. But having all questions about FS2020 in one forum becomes a bit messy. So therefore we would like to ask you all to use the following guidelines when posting your questions:

    • Tag FS2020 specific questions with the MSFS2020 tag.
    • Questions about making 3D assets can be posted in the 3D asset design forum. Either post them in the subforum of the modelling tool you use or in the general forum if they are general.
    • Questions about aircraft design can be posted in the Aircraft design forum
    • Questions about airport design can be posted in the FS2020 airport design forum. Once airport development tools have been updated for FS2020 you can post tool speciifc questions in the subforums of those tools as well of course.
    • Questions about terrain design can be posted in the FS2020 terrain design forum.
    • Questions about SimConnect can be posted in the SimConnect forum.

    Any other question that is not specific to an aspect of development or tool can be posted in the General chat forum.

    By following these guidelines we make sure that the forums remain easy to read for everybody and also that the right people can find your post to answer it.

External Flight Model

Hi Folks,

Just something I've been curious about - not planning on making one... Could anyone explain a high level overview of just what an "external flight model" is ? I'm somewhat familiar with the tabular lookups in a normal flight model... Is the external flight model a completely separate program executable that launches with the aircraft and all the sim information is just passed back and forth between this external program ? Would it still do lookups in tables - or is it computing the dynamics in a completely different way ? Again - not looking for a "how to" - just the overall concepts to satisfy my curiosity...




Resource contributor
External flight models can be done any way you like them. They aren't a separate process, but code within a gauge or module that controls the 6DOF velocities of the aircraft. You can use Blade Element theory, lookup tables, CFD, anything you like to generate the forces, integrate into velocities and using SimConnect to send them to the sim.

Roy Holmes

Resource contributor
For what it is worth, I do not see a big case for using an external flight model for fixed wing sim planes. Sure there are a few things the sim does not do too well, like stalls with flaps, LE flap effects, TSFC and supersonic dry thrust, but it is generally possible to closely match any flight manual performance.

It does a rather poor simulation of helicopters probably because the team was shut down before the EH-101 model was finished. Engine control is bad if you want to simulate a multi-engine start up sequence and rotor lift is not realistic. This is where a lot of assistance outside the normal implementation is needed.



Resource contributor
Actually there is quite a case for one, namely when it comes to the simulation of slats, elevator behavior in the transonic region, mach drag, and external stores effects. Nevermind variable geometry wings.

Slats: They don't increase lift, they shift the critical AoA. Using Leading Edge flaps in FSX doesn't really do this. I got it right in the Milviz F-4E by hooking and dynamically controlling table 1545 although that has now been replaced in the ADV version with the external flight model.

Elevator behavior in the transonic region: Table 420's mach resolution is far too low. Not to mention that the center of lift movement with the change to supersonic flight isn't realistically simulated. Eg, for the Milviz F-4E, I had to hook the sim to locate table 420 in memory and overwrite it in order to get the very rapid change in elevator effectiveness as you move through Mach .95-97 that will cause you to suddenly overstress the airframe if you have the stick too far back while slowing down in that region. The same behavior is present in our F-4E ADV with its external flight model.

Mach drag: I don't know much about it but I do recall all our flight modelling guys at Milviz saying something about it being impossible to get it right without an external flight model.

External stores: Even with hooking the airfile to control Cd0 for dynamic drag, asymmetric drag is not possible without an external flight model.

External flight models also allow for resolving FSX's flawed inertia model so more realistic handling can be achieved.

Edit: After reading Roy's post more carefully, yes performance can be done fairly closely, but when it comes to handling there is simply no substitute for an external flight model.

Roy Holmes

Resource contributor
Actually there is quite a case for one, namely when it comes to the simulation of slats, elevator behavior in the transonic region, mach drag, and external stores effects. Nevermind variable geometry wings.
I agree with much of that but the big question is "Will the average Joe notice it"? Just because a test pilot did and wrote it up does not mean it is that obvious and hence without it you have a deficient model.
Yes to slats. They delay the separation of airflow at high AOA and so increase the AOA where it occurs. That can not be modeled without some measure outside the norm.
There are not many supersonic airplanes that have elevators per se but I guess you mean longitudinal pitch control. The F-4 pitch control in the transonic region was very good. Mach 0.95 to 0.97 is a tiny range in the Mach envelope. I flew 1200 hours in the F-4 and was never disturbed by that region. I never spent much time there in day to day flying but it was probably way down on my list of pitch controls issues.

To put this in perspective I lost control badly twice when flying the F-4, Both occasions involved high AOA and low speed. The first was in my early days on the airplane, I was #3 of a stream take-off and it was customary for that guy to climb more steeply than the lead on climb out to avoid his slipstream. Customary to me from my Hunter flying, but not a good idea in the F-4. I raised the nose and checked forward but it kept pitching up. Full forward stick did not stop it. For a while it was looking really grim but 44,000 lbs of reheat thrust saved the day, the airplane accelerated and stab control improved. You do not find much mention of that in the manual. The second occasion was on finals a bit too close to the guy in front. I flew into his slipstream and the airplane banked 90° right then 90° left and was uncontrollable. Again I put on power and accelerated, it sort of shrugged out of the wake and I went around. That behaviour at slow speed and high AOA was a much greater cause for concern than anything to do with Mach.

Mach drag can be accurately modeled. The intervals in 154a are up to you, not fixed at 0.2 Mach intervals like most of the Mach tables.
External store drag can be accurately modeled and will give commensurate fuel flow rates. Granted assymetric drag is not possible within the normal measures.
Variable geometry can be modeled. I did it in the SSW Tornado. The performance differences are modeled and so is the handling. Verified by current real world pilot.

Inertial effects are an issue as is assymetric thrust effects which are grossly exaggerated, though not too bad in the close coupled F-4.
You can simulate center of lift change effects. The effect is that center of lift diverges in its relationship to center of gravity. So move the center of gravity which gives the same effect. Yes you cant do that using normal measures but it is easy to do.

I guess the issue is what is meant by an external model. I agree you can do a lot by just writing to existing variables or invoking tables outside the norm. I would hesitate to call that an external model, it is just a way to fine tune some aspects.
I do not see the need for a totally separate model.
I also do not agree
when it comes to handling there is simply no substitute for an external flight model.
You can make a very good simulation of handling, and meet the Mil requirements. Do not forget that airplanes with dangerous handling do not sell because they will not pass handling requirements..
Hi Folks,

Thanks - all definitely interesting - appreciate the responses one and all... Yeah - from what I've gathered from others the turbo-prop model isn't the best either...

Roy - you've answered my questions a bunch over the years in great detail - plus countless other posts - do you have any work out there for public consumption ?

Last edited:

Roy Holmes

Resource contributor
I have no models published that are all my own work.

My published contributions are FDE for SimSkunkWorks:
AV8B (except STOVL/VTOL aspects)
Mirage F-1 (not available outside VAAFSE)

FDE for IndiaFoxEcho:
Eurofighter Typhoon

Hi Roy,

Sorry - forgot to check back here...

Thanks - I'll take a look - haven't messed with too many warbirds in FS - last was the old Cloud 9 Phantom package in FS9...

I've been tempted lately by the MilViz Tallon...

I usually take my warbird desires over to DCS where it's setup for weapons... Over there the F86 is a favorite - actually flew RW (SEL) with an 80+ year old F86 jock - very interesting guy and had some great stories to tell - they don't make them like that anymore... The flight dynamics might even be better than P3D ? They feel pretty good...

I'll definitely take a look at some of these now - appreciate the help and pointers...


Sent from my iPhone using Tapatalk

Ellover Dee

I am in complete agreement with Mr. Holmes here.

The native FDE in FSX/P3D is outstanding in creating fidelity in all important aspects of flight. The problem is that it takes a great deal of time and dedication to determine just how to do it.

THAT is what the MSFS team built into the "game" or simulation: the ability to work things out for yourself. I think Mr. Holmes would agree that when we have an idea, and try it, and "Voila!" it turns out to be right on the money...and aerodynamically correct...it's a great "aha!" moment.

How many civilian pilots flying subsonic aircraft spend a lot of time exploring post-stall behaviour, flat spins, compressibility, and the like? None. That stuff is for test pilots, and for subsonic flight, it's all been done a very long time ago.

If we focus on the expected performance of aircraft commonly owned and flown privately, FSX/P3D gets it as right as we could ever hope. If, that is, we know how to do it.

The military and manufacturers have huge budgets to develop their own simulations and use their test pilots and staff to work things out. They are not using FSX or P3D for that kind of testing. For basic familiarity, these PC-based programs are quite good, but not for advanced testing and pilot training.

Several big advantages of using the native FDE are:

-uses far less processor resources that aircraft laden with hundreds of thousands of lines of code
-external hardware usually works seamlessly
-what you see is what you actually get in terms of what the flight model is doing

After 22 years and thousands of hours of time doing flight model development, I am still making discoveries and find this simulation, now P3D V5, greatly rewarding. I purchase quite a bit of payware and am always gratified when the flight model is not laden with all kinds code that ruins my attempt to create a personal (and far more accurate) flight model for my own use and enjoyment.

I've deleted hundreds of dollars worth of payware because the flight dynamics are not accurate and no fixes are possible. Beautiful visuals and a constant stream of "realistic" failures that take half a day to sort out before one can take off are certainly not for me. And then there are the constant patches, fixes, and updates required to repair products that weren't really ready to be delivered.;)
How many civilian pilots flying subsonic aircraft spend a lot of time exploring post-stall behaviour, flat spins, compressibility, and the like? None. That stuff is for test pilots, and for subsonic flight, it's all been done a very long time ago.
Not only that flat spins are of course possible in FSX/P3D and aerobatics are generally fun with good FDEs, IMO stall, post stall behavior and spins are an essential part of the FDE since many RW accidents are stall/spin accidents.