LR customers can now get the first public beta of X-Plane 11.40 by running the installer, checking the box to get betas, and completing the update. Steam will most likely be available tomorrow, barring any unforeseen crises with this build.

Release notes here.

This youtube video has a thorough explanation of the new flight modeling stuff.

PLUS new stuff for “official” beta 1:

Stall is now TRIPPED and you really have to lower the nose to recover the non-separated flow, as per reality. This really drops a wing more in the stall!

Rotor bounce back in ground effect now modulated by forward speed. Which is cool and causes a little bit of settle as you start to get going before ETL.

Thermals for gliders are found per element now, so the glider raises a wing in the edge of a thermal.

New wind PROFILE from wind sock to ground. This is from data on wind above ground carefully compiled for wind turbines!!! Har!
So while the metar or real weather or other weather setting sets the wind at 10 meters (the wind sock height), the profile from that down to the ground (by definition 0 for the atoms touching the pavement, and a curve in between) follows the wind turbine wind profile model

Now we have EXHAUST THRUST as well for the turbines and recips!
Plane-Maker > Default menu > Engines screen> Engines 1 tab > Top sort of close to the right.
Enter how many pounds of thrust the exhaust puts out per horsepower output of the engine. We find it to be about 0.10 for the PT-6 turbines, and is almost certainly very close to 0 for any reciprocating engine.

Wake turbulence, better to track other planes! Command-m to see wake turbulence from your plane and others when in flight. Fly into the wake if you like! 3-D per-plane wake turbulence following wind and settling profile! Some of this wake turbulence tech comes from Xavion, where I developed it to avoid (another) wake-turbulence hit in the real airplane.

Better propwash swirl based on 4 different tech reports! Now we have thew swirl angles at all radial distributions dialed in! The insight here was to NOT track propwash swirl in meters per second, but instead track it in degrees of swirl from straight aft! In reality, the SWIRL ANGLE does NOT change with propwash speed or longitudinal location along the propwash stream tube, so tracking this swirl ANGLE in X-Plane instead of swirl SPEED gives us the most accurate propwash swirl in all locations, no matter how far behind the prop.

Now, X-Plane uses the efficiency and drag of the prop to estimate the swirl angle using conservation of energy, and then applying that swirl angle at all prop radial locations according experimental data, and holding that swirl as you move aft in the prop stream tube as per experimental data. So this is a very nice start with analytical conservation of energy to DETERMINE the swirl, then using empirical data to apply that swirl at the right places in the prop stream tube. So X-Plane now tracks experimental data very very very well in propwash swirl! This is good for tail-dragger and other handling where swirl angles over the tail matter.

So now, we can finally say that we have good propwash speed AND SWIRL over the tails of those tail-draggers! (And all other planes with engines in the front blowing propwash on the tail!)

Also, propwash now spirals in FRONT of the prop when in reverse! Good find for the person that reported this bug!

Also, propwash now tuned a bit more to follow the orientation of the prop, as we have found to be the case both with tail-draggers doing take-offs and the BackFly e-VTOL, which runs its’ wing in propwash without stalling, as long as there is propwash over them, proving that the flow is aligned to the props as long as power is put into the flow. So this should align propwash with the wings and fuselage more, which should be more accurate in the cases that I am seeing.

Delayed propwash over the tail! It is well known that the wings cast downwash over the tail based on the coefficient of lift of the wings, with the various flaps and ailerons on them. But here is where it gets tricky: Did it ever occur to you that it takes TIME for that downwash to make it back from the wing to the horizontal stabilizer?

This DELAY helps STABILIZE real airplanes in pitch, because as the nose is raised for example, the horizontal stabilizer is initially pushing down into UNDISTURBED air: The increased downwash from the wings might take 0.05 to 0.10 seconds or so reach the tail and push it down farther, de-stabilizing the airplane!

The only way I could think of to simulate this in X-Plane was to keep track of the downwash at all points in space and move it backward from wing to tail over time…this is completely NON-feasible for memory and CPU reasons, so I have never simulated this delayed downwash effect before in X-Plane! As a result, the simulator has always had just a little less pitch-damping than the real airplane, making it touchier in pitch than the real airplane.

Now, I have a solution, and here is what it is: Let’s say the angle of attack of a given airplane is increasing at 20 degrees per second as the nose is suddenly pulled up. And let’s say that that we know that each degree of angle of attack causes 0.5 degree of downwash. Then I am sure we can agree that the downwash is increasing at 10 degrees per second, right? It’s un-deniable!

So we KNOW that the downwash is increasing at a rate of 10 degrees per second, and let’s say that it takes exactly 0.1 seconds for the wash to make it from the wing to the tail (this is typical and easy to find: Simply the distance between the wing and the tail divided by speed of the airplane). So we know that the downwash is increasing at EXACTLY 10 degrees per second, and we know it takes EXACTLY 0.1 seconds for the wash to make it from the wing to the tail (at the speed and size of our imaginary airplane for this example). Then how much less wash will the tail have during this maneuver due to downwash delay? The answer is EXACTLY 0.1 seconds times 10 degrees per second equals one degree! COOL! This answer is exact, no approximation! (NOTE that we assume a constant angle-of-attack change-rate for our math to be perfect, so this is still an approximation since the angle-of-attack rate changes over time in the real airplane, but this is pretty darn close to perfect propwash tracking, at a cost of NO memory and basically NO CPU, since the downwash delay is found with some VERY simple algebra, instead of a ridiculously-complex and memory-hogging operation of trying to track an entire flow-field over time.

So, now, we have more accurate pitch-damping due to delayed downwash with simple algebra… and no CPU or RAM cost to get it! Cool!
This makes the airplanes in X-Plane less squirrely in pitch than they were before, and probably closer to reality!

Varied propwash across the prop disc!

For helicopters or VTOLs with running the prop discs edge-wise into the wind, the propwash through the prop is now less at the leading edge of the disc, and more at the trailing edge of the disc, to simulate the propwash building over time and distance as air flows aft over the advancing prop disc. This results in more rotor-flapping and rotor or rigid-disc pitch-up, with performance matching our eVTOL AVA, shown below. So we used the aircraft below as a test-vehicle to measure nose-up pitching moment to tune the rotor downwash model in X-Plane!

Control effectiveness improvement:
Looking now at chord ratio and deflection and a few different tech reports, we have the control effectiveness accuracy improved for ailerons, elevators, and rudders. Like most all of the stuff above, you need to have the experimental flight model ON to get this!

Better supersonic flight dynamics and transition TO supersonic:
If the experimental flight model is on, we now transition smoothly from subsonic to supersonic model on the wings, bodies, and downwash.

As well, in Plane-Maker, go to the fuselage (and any other body) screens where you set the coefficient of drag. You can now set area-rule factor for supersonic airplanes. This defaults to 1.0 (a multiplier of 1 means use standard compression shocks and expansion fans), and can normally be left there, but for planes like the T-38, which have very necked-in bodies in the middle to do an excellent job of area-ruling, which keeps a constant cross-section area along the length of the airplane by narrowing the body where the wings add equal area, you can change this number to a multiplier of something like 0.75 or so to reduce the supersonic drag that X-Plane computes, to account for the air moving around the wing to neatly fill the space in the necked-in body and inlets, as the T-38, for example, has.

Better wing/body separation:
Vert and horizontal stabs for T-tail no longer consider themselves to be one long wing for efficiency purposes, because the vert stab does not act as an endplate for the horizontal stab, because it is in the center, where there is no lateral flow. Vertical stabs and wings individually still connect into one effective wing just fine though.

The thermals are now a bit more tightly defined, to form stronger gradients for the thermals so you can feel a bit of pitch or wing-rock as part of your plane enters a thermal. Turn on thermals in weather as always and cycle through the visual flight model with control-m as always while in flight at altitude and below the tops of the thermals to see the vertical speed profile of the thermal pattern nearby… kind of cool. And the wings and tail will all grab the edges of thermals as well as you enter or traverse them, giving the tell-tale wing rock away from the thermals, nose rise when entering the thermals, etc.

Also for gliders, the tow-plane will stay on the centerline a bit better to get you going even if there is some crosswind.

Wave shape changed from sin wave to wave shape closer to what we actually see on the water as the wind picks up. This should help make seaplanes a little more accurate. Seaplane float dynamics also improved to be more accurate with better damping in the water.

If you re-save in Plane-Maker, then the old: “Engage burner at 50% above this throttle setting” will become the new: “Engage burner at low level at this throttle setting, smoothly moving to max burner at full throttle” so the burner will smoothly increase as the throttle goes to maximum!

As well, if you re-save the aircraft in Plane-Maker, then: The throttle will run the N1 and N2 up to 100% by the time you reach the critical throttle entered in that slot in Plane-Maker. Then, ABOVE that throttle, since N1 and N2 will be at maximum dry thrust, the burner will add on top of that.

So, in summary, if you enter a throttle to engage the burner above: The N1 and N2 will come up to 100% as the throttle reaches that critical throttle position you entered in Plane-Maker, and THEN engage the burner smoothly above that throttle position as the throttle goes to maximum!

As well, jet engines with afterburners often do not show the afterburner fuel flow in the engine fuel flow indicator. So, we now have:

FLT_ARR_EXPR1("sim/flightmodel/engine/ENGN_FF_dry", sim_flightmodel_engine_ENGN_FF_dry, flt[p0].enpr[N].act_fm_FF_kgs__dry,flt[p0].enpr[N].act_fm_FF_kgs_dry,8)
FLT_ARR_EXPR1("sim/flightmodel/engine/ENGN_FF_", sim_flightmodel_engine_ENGN_FF_, flt[p0].enpr[N].act_fm_FF_kgs_tot,flt[p0].enpr[N].act_fm_FF_kgs_tot,8)

Where the dry datarefs are fuel flow without the afterburner, and the the regular fuel flow is the total of engine and burner together.

For Xavion:
Full next-gen ADS-B weather from X-Plane to Xavion (which you can get at the App Store) which makes for some pretty nice weather flying when setting convective weather in X-Plane! This includes icing, turbulence, and cloud-top, wind, temperature, freezing level, and airport METAR weather display in Xavion from X-Plane! A pretty cool weather setup!

Airfoil Maker foil-load really does load the foil now when you load a foil from the menu, even if that same foil is already open.

Winds come down near the surface now, as per wind-profile chart created for wind-turbine designers, of all things!

Stowing the prop for motor-gliders now does a complete job of hiding the prop from the airstream.

Various panel brightness rheostats now saved for situations and replays.

Tail-rotor trim option now in Plane-Maker for any helos that might have that. (X-Plane used to use full-deflection with full tail rotor trim but now you control the amount of trim available.)

New individual flight control SURFACE failures for individual flight control surfaces on the wings. Used in engineering-level e-VTOL simulation now, but can be applied to any airplane!

Electric motors can spin BACKWARDS for reverse with fixed-pitch props! Just equip them with a reverse checkbox in the Engines screen in Plane-Maker like any other engine that goes into reverse, and go into reverse with a joystick button configured to toggle reverse, like for any other airplane!

New command line: —lock_fr=xxx, which limits the frame-rate to this speed, and also takes you to just 1 flight model per frame to maximize the per-frame speed and maximize your chances of hitting that target frame-rate.

See my latest work and other projects at

31 comments on “Experimental Flight Model Changes in X-Plane 11.40

  1. I definitely love these flight model improvements, to me this is really what X-Plane is about. Thanks!

    Thinking about the command “—lock_fr=xxx”: I like this ability to limit frame-rate but some aircraft require more than 1 flight model per frame in order to work correctly. Since flight model calculations per frame already exist as a parameter, why force this value to 1 when locking the frame-rate? To me these settings would be better separated.

      1. Is the option supposed to slow time when the minimal frame rate is not met, like the usual 20fps floor? The command does nothing for me, nor in Linux, nor in Windows.

        1. My bad, this is an upper limit, not a floor. Not useful for VR !
          But I found that it can be used to make the vsync option work even when frame rate is dependant on the CPU.
          When doing some testing, I also noticed the 15 or 20 fps limit where X-Plane starts to slow down time seems to have vanished. Is it since XP 11?

    1. New command line: —lock_fr=45 : it just occurred to me this is huge for VR !!!
      Very happy with Austin’s new flight model breakthroughs. Cheers!

  2. When are more cities, new planes and eye candy coming? Weather enhancements and ATC still lack.

    1. Some of the features above normally come in major x-plane revisions. Laminar themselves normally state whether those features are due in the current iterative run of X-Plane (of which, almost none are announced.) But in general, developers blog is not really the best place for comments like this.

  3. I believe with 11.40 update a lot of aircraft developer will eager to try out the new flight model and may event require customer to turn on “Experimental Flight Model” option even though Ben specifically said not to do this in payware addons (e.g. Magknight 787). So question: will the 11.40 flight model be the standard one after 11.40 becomes final? What do you suggest developers do with the new “Experimental Flight Model”?

    1. We aren’t sure yet but probably we will _not_ make the experimental one official at the end of the 11.40 beta – the whole point of the experimental one is to give it more than a beta worth of test time. But I think this may be the last _update_ – that is, when we do make an experimental FM official, it’ll be the one from 11.40 RC.

  4. Hello,

    The X-Plane 11.30 experimental flight model changes are still available and affecting flight model only if you check the “use experimental flight model” checkbox at general settings or are they integrated into main code and affecting the airplanes flight model even with the “use experimental flight model” unchecked?

    Thanks in advance!

    1. The 11.30 experimental flight model is _replaced_ by the 11.40 experimental flight model. So you have access to two FMs in 11.40 – the newest experimental 11.40 (it’s like 11.40 with more accuracy, bug fixes, tuning, etc.) and the non-experimental one.

      1. So experimental flight model from 11.30 will become normal (default) flight model in 11.40 or not? It is confusing. Don’t know if it is confusing also for developers or not. As a regular user, I don’t care very much as I don’t check any checkbox with any “experimental” in it. However, if aircraft developers are sure their aircraft work better with experimental flight model, why you don’t give them the option to override experimental flight model setting in aircraft configuration? At least the user would not need to keep track for which aircraft he/she needs to turn the setting on or off. Although you asked developers to not ship aircraft with requirement for experimental flight model, some are apparently ignoring it.

        1. Experiment FM from 11.30 will _not_ be normal or experimental in 11.40. Basically 11.40 has “the experimental flight model rev 2” – it’s been further modified to be more accurate.

          We did not give authors the option to _require_ the experimental FM because it is _still changing_. Imagine if aircraft authors had had this option in 11.30. All of the new changes in 11.40 would have screwed up their aircraft. We’ll give them the option to use it when it becomes the _official_ flight model and not just experimental.

  5. Sounds very interesting, I have to test the stall behavior at first 🙂

    10% thrust from the exhaust output was a surprise for me, reading about it the first time in an article of Mr. Clements.

    Is it OK to tweak this value in an existing airplane?
    Will the exhaust thrust be added to the given horse power or will it be included?

    I guess added, so Total Power equals to SHP + Exh Thrust (?)

  6. Will we ever get a fix for ancient bugs such as the reflection bouncing and massive performance cost bug or the AA not working correctly since 11.3 (Even 4×4 is horrible with taxi lines and other thin lines flickering like crazy, only maxed out calms it down enough not to be a total distraction and also the flickering skybands bug most noticeable with the Cesna dawn or dusk which is horrible.
    Or are many fixes on hold awaiting Vulkan?

  7. Guess you wanted to mean “(it’s like 11.30 with more accuracy, bug fixes, tuning, etc.)” – 11.30 instead of 11.40, right?

  8. About three weeks ago someone reported that X-Plane and Apple Catalina Beta are a no go.
    I always upgrade, or at least try apple’s new betas.
    Do they seem to work together now?
    No, I do not have any plug ins!

  9. Sorry, you lost me.
    If I install 11.40b1 and do NOT choose “use exp flightmodel” what is different in view of 11.36?

  10. If I’ve understood right, a flaw of Microsoft Flight Simulator is a consider of counterclockwise spiral wash.

  11. Big dif with stalls, in fact, as close to reality as ya can get 😉 Thanks for your work people and thanks to Austin’s math spinning brain ;))
    Incidentally, I put on experimental flight model in settings and flew the Milvitz and the LR stock 172

    1. The stall behavior may be better, but still not as it should.
      Near a stall using aileron e.g. to the right the plane turns to the right(?)
      It should turn (stall) to the left.
      Looks like the change of the angle of attack by using aileron is not part of the flight model?

      My test was done flying the C90

  12. Flying on IVAO, the other planes have no wake turbulence. Is it possible to add wake turbulence to obj8 planes? That would be great. Otherwise I will only get wake turbulence from myself, which is normally not a factor…


    1. The plugin would have to manipulate the physics of the user’s plane directly — X-plane can’t “see” OBJ-drawn or plugin drawn “aircraft”.

  13. Great flight model update! Flying characteristics of all airplanes I tried are so much better. Thanks Austin!
    Unfortunately, the modifications did not solve the issue with taildraggers, which are still not usable on unpaved runways in cross-wind. But I am an X-Plane old-timer (iirc since version 2 or 3), and I know that eventually Austin will come through and will solve the taildragger issues brilliantly. For now I will go back to X-Plane 10 because my favorite plane is a Cessna 185.

  14. Dear friend, i have a problem when i am flying any jet airline (b732 , b733 or b727 etc) , Can`t reach de optimum flight level , the aircraft only can climb until 3000 feet below from optimum flight level. ( all this situation is with experimental flight model in ON ). Thanks .

  15. Hi LR,
    Not so much time right now, but taxying the C172 and stalling it feels definitely much more accurate. I know there is so much more that you have tweaked and modelled, but it is a very good foretaste. Congratuations for all this work, much needed!
    And thank you for all these high quality explanations!

Comments are closed.