Real World Pilots, please state your feedback about the flight model

Need to find some more time for detailed testing, but some quick tests in the G36 really pushed that thread home. Was extremely docile in a power off wing level stall, was able to just hold it in a stall without the nose dropping. I’ll admit it’s been a while, but I remember the nose on the Bonanza would drop rather sharply.

It is interesting that the FM starts to fall apart in scenarios like that, but at the same time accurately depicts how resistant a 172 is to spinning unless you use an aggressive power on entry.

REGARDING STALL BEHAVIOR

The JF Arrow developer (@GrimPhoenix9349) commented on how the config files can affect stall behavior such as nose drop, wing drop, and other things. See the link I posted below. He is frequently posting in that thread.

And by the way, sorry for the decrease in amount of recent posts. Unlike CantankerousCiabatta, I actually have a real job, and it just got white hot the last few days. I should be able to take tomorrow off, but I’m getting the JF Arrow III, so I’ll be flying that the next few days. I’ll try to report on it now and then.

Stall Config

1 Like

MSFS 2020 stall behavior depends on max elevator angle. Normally the difference between no stall and stall&spin is 5°. Change these parameter (add 5):
elevator_up_limit = 20 ; Elevator max deflection up angle (DEGREES)
elevator_down_limit = 15 ; Elevator max deflection down angle (absolute value) (DEGREES)

I was flying a C152 once and was seeing 1,200 fpm climb rate when flying under a cloud that had a concave bottom. Thermals and wind sheer would make things more realistic, but not necessarily more enjoyable.

I agree. I like the “easy” mode, too. But on “realistic” in a sim(!), I like to get “realistic”, not “fantasy”.

I have flown every sim that is out there except DCS. I love the way the Mooney handles in MSFS 2020. A friend of mine owns a Mooney tried it on my PC, and said he thought it was a good simulation. I myself have about 600 hours in SEL aircraft.

Crash Course in Propeller Aerodynamics

Concerning the last Developer Q&A, I’m not entirely sure the developers understand the problem of “missing propeller drag” in the sim. Seb was talking about drag from internal engine components instead of aerodynamic propeller drag. In all fairness, the drag of internal engine components had an effect on propeller windmilling speed and therefore aerodynamic drag produced by the propeller. I’m not a developer to be clear but I doubt this is the real problem. Also it would not explain the lack of drag on turboprops in MSFS as the PT-6 for example is a free turbine engine and therefore not influenced by drag of the engine itself, sure it has some drag from the reduction gearbox and power turbine, but I don’t think this is the real problem.

Edit: after doing some testing it seems like there is propeller drag modelled on certain aircraft. The single engine piston aircraft I have tested seem fairly accurate, the Baron isn’t accurate at all, the landing gear on the Baron does not produce any noticeable drag & able to fly way below Vmca without departure from controlled flight. All three turbo-props have no propeller drag modelled at all. None of the MSFS aircraft have the ability to feather the propellers although on the Kingair and TBM the RPM does drop significantly when selecting feather (but with no change in aerodynamic drag). In short: multi-engine piston aircraft and all three turboprops are significantly flawed.

Definitions

First it is important to understand the following definitions when talking about propellers:

  • Blade / Pitch angle - This is the angle between the propeller plane of rotation and the chord line of the blade. On a fixed pitch propeller this angle is fixed, on a Constant Speed Propeller (CSP), this angle is adjusted to balance engine power and propeller power absorbed, controlling the propeller RPM.
  • Washout or twist - The blade angle on a propeller blade is not the same along the whole range of the propeller, as the propeller rotates, the root of the propeller has a lower tangential velocity when compared to the tip of the propeller (compare this with a baseball bat, the tip of the baseball bat has a higher velocity compared to the root). In order to have the same angle of attack on the whole propeller blade the blade angle on the tip is lower than on the root (twist or washout).
  • Windmilling - Is the condition describing the propeller rotating by the forward speed of the aircraft rather than by absorbing power by the engine. When a propeller is said to be windmilling, the propeller is driving the engine instead of the engine driving the propeller. Windmilling is caused by a negative angle of attack on the propeller blade, essentially producing negative thrust (drag). You can look at it as a thrust reverse without having a negative blade angle. The windmilling RPM of a propeller is depending on the drag from the engine (in that regard the devs. are right) and aircraft forward airspeed (TAS). Lower RPM = lower drag, higher RPM = higher drag.
  • Forces - There are a lot of forces acting on a propeller, we will be looking at Lift (thrust) and Drag, the rest of the forces are of importance when trying to understand constant speed propellers. Centrifugal Twisting Moment (CTM), Aerodynamic Twisting Moment (ATM), centrifugal (feathering) weights added to the proppeller, springs in the propeller hub etc. we will skip those for now.
  • Fine & Coarse Pitch - On a constant speed propeller the blade angle is variable, a small blade angle is called “fine pitch” while a large blade angle is called “coarse” pitch. The finest pitch possible is the low pitch stop in flight and reverse pitch on ground (if installed), the most coarse pitch is the feather position (if installed).

Basics Aerodynamics

A propeller is an aero foil and essentially behaves like a aircraft wing orientated in the vertical plane. The working principle of a propeller is the same as for a wing, the only difference is that the resultant force is acting mostly vertical on a wing and can be divided into a vertical component (lift) and a horizontal component opposing the direction of movement (Induced Drag), on a propeller this resultant is acting mostly horizontal and can be divided into a horizontal component (Thrust) and a vertical component (Drag).

Very important, this is not the drag we are talking about in regard to the missing aerodynamic drag from (windmilling) propellers!

The engine power has to overcome the drag produced by the propeller, if the two are in balance the propeller RPM is constant, if the engine is producing more power than the propeller is absorbing, the RPM increases, increasing the thrust (and drag) until the two forces are in balance again and vice versa when reducing power.

Regarding constant speed propellers, this becomes a little more complex as there is an extra variable introduced. The propeller blade angle is variable, a constant speed propeller is able to MAINTAIN a certain RPM with an increase or decrease in engine power by modifying the blade angle, therefore changing the power absorbed by the propeller (higher blade angle is more drag and vice versa). For example if engine power is increased, the propeller tends to speed up but the blade angle is then increased producing more thrust (and drag), the increase of power is absorbed by the propeller with no increase in RPM! There is a limit to this, when power is reduced far enough the propeller will eventually reach the minimum blade angle, if power is reduced further the RPM will drop, if power is reduced further then at some point the propeller will start to windmill and drive the engine instead.

Propeller Angle of Attack

The angle of attack on a propeller is based on two components, the tangential velocity of the propeller and the True Airspeed. The tangential velocity of the propeller is depending on the propeller RPM, higher RPM is higher velocity and vice versa. We can’t call it “RPM” because the RPM is the same for the entire propeller, whereas the tangential velocity is not, it is depending on the location of the propeller, the tip of the propeller having the highest tangential velocity and the root having the lowest (zero velocity). To keep it simple we will call this component RPM, but technically this isn’t correct. I hope the picture below also illustrates clearly how we are looking at the propeller cross section in further examples.

In the first example the aircraft is stationary (TAS = 0), the propeller has a high angle of attack (AOA = blade angle) creating high drag on the engine. If we now release the brakes and start rolling the TAS increases which reduces the angle of attack, the reduced angle of attack causes a reduction in drag. The engine is now producing more power than the propeller is absorbing and therefore the RPM increases. This increases the angle of attack again to bring it back in balance, this is the constant force balance in play during flight. It is also the reason why on a fixed pitch propeller (e.g. Cessna 172) the RPM increases with increasing airspeed for a given engine power setting and vice versa.

Windmilling

Now the core of the issue. What happens if the we reduce power all the way to idle in flight? If the engine is not producing any power why doesn’t the propeller (and therefore engine) stop rotating? As soon as the engine fails or power is reduced to idle, the RPM drops significantly as the propeller absorbs more power than the engine is producing, eventually the angle of attack on the propeller becomes negative and the forces acting on the propeller will basically flip. Positive thrust becomes negative thrust (THE DRAG WE ARE MISSING IN MSFS) and drag becomes the driving force, keeping the propeller and engine rotating. This drag is very significant in real life and cuts the glide ratio short by a significant amount, the deceleration in real life is therefore also significantly higher than in MSFS and it is possible to make steeper approaches without picking up speed.

Note that a windmilling propeller basically acts like a thrust reverser, but at a positive blade angle.

Single Engine versus Multi-engine

Without diving deep into the working principle of a Constant Speed Propeller, Propeller Governors etc. There is a noticeable difference between single engine and multi-engine aircraft. The principles are all the same independent if we are talking about single-engine, multi-engine, fixed pitch or constant speed propellers. Also on a constant speed propeller the engine will windmill, the reason is simple, a constant speed propeller (it is in the name) wants to maintain a constant propeller RPM. If an engine fails or power is reduced to idle the constant speed propeller will drive the propeller to the fine pitch stop and from there it will essentially behave like a fixed pitch propeller.

Belief it or not, windmilling on a single engine aircraft is actually beneficial. From an aerodynamic and glide range standpoint, windmilling obviously isn’t a great benefit. But only having one engine, the main priority should be to restart the engine, windmilling of the propeller keeps the engine turning and increases the possibility of restarting the engine in flight. This is the principle difference between single and multi-engine aircraft. On a multi-engine aircraft an engine failure will not only cause asymmetric thrust, on top of this the windmilling propeller will cause a load of drag. On a multi-engine aircraft the priority is therefore to reduce the drag, secure the engine and continue flight on the remaining engine.

Reducing the drag of the propeller and stopping the engine is done by “feathering” the propeller. The blade angle is increased to a near 90 degrees which results in zero resulting force on the propeller blade, the propeller stops rotating and drag from a windmilling propeller is canceled. Feathering can be accomplished manually or automatically by the aircraft (i.e. auto-feather). Nice detail, the propeller blade angle when feathered is not exactly 90 degrees (few degrees lower in fact), if the propeller blade angle would be increased to 90 degrees the propeller blade would start to act like a wing and will produce a slight lift force (and therefore drag), the angle of attack is therefore kept slightly negative.

This is exactly the next problem in MSFS, last time I checked propellers are not able to feather, neither does the propeller produce drag as we know so the propeller performance in MSFS is basically a feathered prop anyway. When propeller drag becomes implemented feathering would become very important. At this moment, accurate simulation of asymmetric thrust and drag on multi-engine aircraft is unavailable.

In all cases below, flight model set to MODERN, assistance set to HARD, community folder completely emptied, ISA conditions, no wind.

TBM 930

Maximum weight, 120 kts glide speed, flown with FLC on autopilot, prop first feathered and then windmilling. I cut the engine (by closing fuel selector) overhead an airport at 3000 ft, ISA, no wind, glided down to 0 feet, paused the sim, took the drone camera to check my location, looked it up on Bing maps and measured the distance.

Prop. Feathered:

Propeller speed = 260 RPM
Glide range: 17.5 km = 57400 ft
Glide ratio = 57400 / 3000 = 1:19,13

Prop. windmilling:

Propeller speed = 960 RPM
Glide range: 17.1 km = 56088 ft
Glide ratio = 56088 / 3000 = 1:18,70

My methods are a little bit mickey mouse and completely accurate, the autopilot started to level off slightly when doing the glide ratio test with feathered propeller because I put 0 ft in the altitude preselect so thats were the extra 400 m likely comes from. It is safe to assume that there is absolutely no drag modeled on the TBM…

Next test, FLC at 120kt in descend, measuring vertical speed:

Flight idle: -700 ft/min
Windmilling: -650 ft/min
Feathered: -700 ft/min
Full reverse: -900 ft/min

Conclusion, no windmilling drag at all. In all situations the vertical rate is fluctuating between 650 and 700 ft/min.

Beechcraft G36 Bonanza

Speed 110 kts at FLC, max. weight:

Power idle: -1100 ft/min
Engine off: -1300 ft/min (mixture cut-off)
Low RPM: -1100 ft/min (mixture cut-off and prop full back)

Below, same test as TBM, power cut at 3000 ft overhead an aerodrome. Mixture cut-off, prop in low RPM, 110 kts at full weight according to the POH it should result in a glide ratio of 1.7 nm per 1000 ft.

Glide range: 4.9 nm
Glide ratio: 4.9 / 3 = 1,64 nm per 1000 ft

Conclusion, the G36 is pretty accurate! Maybe the problem only concerns turboprops?

Beechcraft Baron

Speed 110 kts at FLC, max. weight:

Power idle: -1000 ft/min
Engines off: -1450 ft/min (mixtures cut-off)
Props feathered: -1450 ft/min

Conclusion, not sure if the increase in rate if descent is because of losing idle thrust or if it is because of propeller drag. The props don’t feather, absolutely no response to propeller lever (RPM stays the same when feathering). Interestingly I accidentally left the gear down during this test so I rerun the test with gear-up with same results! The landing gear on the Baron does not seem to create a noticeable increase in drag.

Beechcraft Baron - Vmca demonstration

Beechcraft Baron
One engine full power, the critical engine windmilling
Take-off flaps, gear-up
ISA conditions
Max. TOW, aft CG

Seems the aircraft stalls before losing control due to asymmetric thrust and drag. I haven’t flown the Baron myself but I have heard stories about how it is handling single engine go-arounds and I even know someone who died practising single-engine go-around on a Baron. Can anybody with experience on the Baron confirm whether this behaviour is correct? Is it still climbing this well on one engine with flaps at take-off at speeds way below Vyse?

By the way the rudder axis on my joystick is seriously borked so I’m lacking precise control over the rudder…

Edit: I noticed I shutdown the wrong engine, I had the right engine in my head as being critical, forgot that the aircraft I used to fly has counter rotating props, oops…

Second attempt: now with the critical engine failed, 50% payload, gear down for some extra drag, first one flaps take-off, second one flaps landing:

Beechcraft Kingair

The Beechcraft Kingair is an absolute joke. I have flown this aircraft in real life, I can see it is seriously flawed.

Max. take-off weight, speed 140 kts on autopilot using FLC:

Power idle: -300 ft/min
Full reverse: -1150 ft/min
Engines shutdown, props full forward: -850 ft/min
Props feathered: -850 ft/min

I didn’t note the RPMs, the RPM does drop significantly when selecting feather, but with no effect on the aircraft drag. The difference between flight idle and engines shutdown is likely not an increase in drag but rather a way too high flight idle power setting…

When shutting down the critical engine (windmilling) and adding full power on the life engine, the aircraft stalls way before running out of rudder to counter the asymmetric thrust and drag. Second try with flaps in landing position to lower the stall speed, able to fly just 77 kts with the joke full back, rudder full in just keeping the slip indicator centered, again no departure from controlled flight below Vmca.

Cessna Caravan

Max. take-off weight, 110 kts on FLC (no idea what best glide speed is, I just picked a speed).

Power idle: -650 ft/min
Full reverse: -1050 ft/min
Engine shutdown, prop full forward: -900 ft/min
Prop feathered: -900 ft/min

On the Cessna Caravan the prop RPM does not respond when selecting feather. RPM remains the same. Conclusion, the three turbo props seem to suffer the most from incorrect (completely missing) propeller drag.

9 Likes

Got a chance to test this out on the G36 again finally, I actually just used the new menu option to increase max deflection up by 1 degree, and went from not dropping the nose at all to aggressively breaking into a left wing drop. So it will break at the stall now, but is exhibiting the same issues @Nijntje91 has documented of always wanting to drop a wing, as well as being able to flail the ailerons around up until the break without a care in the world or any influence on the stall behaviour.

1 Like

Yes, this is my experience, too. You have some influence on the stall behavior. You can change it from “parachuting” to “spin”. But still a lot is hard-coded.
For me, other missing features in the MSFS 2020 flight model are more important. Maybe Asobo will stay with this “elevator limit defines which stall animation you see” forever. Hopefully they make rudder more realistic. Or cross wind effects. Or so many other things!

1 Like

I created some new bug reports recently, you could find them here:

2 Likes

You are a busy bee !

Besides flying every other day and an office job on the side, yes :joy:

Anyone ideas?

It took a whole afternoon, but I found at least of the problems with the MSFS flight model:

Edit: I think I found atleast one of the issues, when looking at the elevator deflection it is somehow depending on the airspeed, so full back sidestick at Vmo only means a few degrees of elevator deflection. I think this is due to another “improvement” they’ve incorporated to make make planes behave more “realistic”. I believe this was explained in some developer Q&A, but I can’t find it.

They’ve explained something like, its harder in real life to pull the stick full back due to the feedback from the flight control system and G-forces so they’ve limited something. Just tested on the Cessna 172, exactly the same thing going on, no full elevator deflection at high speed. This would explain the weird stall dynamics and inability to pull the aircraft in an accelerated stall, it does however not explain or solve the sensitivity issue.

The implementation is done very poorly, to the point that you can’t even stall an aircraft anymore, can’t maintain altitude during a 45 or 60 degree steep turn etc. From the video below, you can see the elevator deflection above my mouse pointer. When inside the cockpit you have no hind that this is even happening as the yoke position mimics your joystick position and is unrelated to the actual elevator position!!! Not how it works in real life at all of course, totally unrealistic.

Asobo, please, please remove this or make it an option!!!

4 Likes

(2) Finds a thunderstorm and goes into it with a surprise ending - YouTube

It can def simulate walls of rain…

Something else interesting I found is that the elevator trim on for example the Cessna, does not change elevator position or yoke position in the cockpit. The aircraft does climb or descent when trimming up or down, my question is how? When the elevator itself is not moved the trim tab should work OPPOSITE from the elevator and trimming up will move the nose down and vice versa. How can they simulate the pitch trim without moving the actual elevator?

How does this work on other sims? It seems it doesn’t really matter what you dive into, with FS2020 you will always end up at something that doesn’t work…

Nijntje91 the reason is only you. You dig and dig until you find a bone. And because it is software, you call that bone a bug.

This nice picture explains everything (more or less). Your airplane is just a number of bits in the computer. Some of these bits are used for calculation AND for graphical representation. That is if you pull your stick, you get an updated calculation and an updated graphical representation. Other bits are only used for calculation. The elevator trim does not show in the graphical representation, but you get an updated calculation.
This is not completly correct. In non-cockpit view I have graphical representation of elevator trim.

See file:///C:/MSFS%20SDK/Documentation/03-Content_Configuration/SimObjects/00-Aircraft/Flight_Model.html#what-is-a-flight-dynamics-model

Maybe the same way it works in FSX/P3D. (There are a few different theories how this works)
Fact is that if you trim fully nose up/down you will notice a difference in pitch authority, something which shouldn’t be the case without a moveable stabilizer.
One of the theories is that FSX slightly shifts the CG to simulate trimming.

As long as there will be self-centering devices (joysticks, yokes, xbox thingies, …) there won’t be a one-for-all solution to this as it is simply not how planes work in the real life

All of the trimming (and sensibility curves, and “twitchyness”) issues and related workarounds would be solved with a proper FFB implementation
THAT is where a lot of effort should be spent. Having experienced FFB it is a definite game-changer (surprisingly even more so for rudder pedals where it is a MUST)

Edit: of course that doesn’t apply to FBW or possibly hydraulic systems

I agree, a FFB controller is essential for a realistic experience. I read about this one in a facebook group which works with MSFS: Brunner - CLS-E NG FORCE FEEDBACK YOKE