Question to HTC: Does propwash affect controls?

[dtango]

On atmosphere
Badboy has already spoken on the topic.  To add we see evidence of it in AH such as the existence of true air speed, variation of parasite drag with altitude, variation of mach number.  All these depend on an atmosphere being in place.

On modeling of propwash:
I absolutely understand what Krusty is saying.  It's however totally misleading to say that propwash is not modelled in Aces High.

Here's a definition of simulate (verb): the act of imitating the behavior of some situation or some process by means of something suitably analogous

The only indication that airflow exists around an airplane is the resulting effect it has on the aircraft.  If we model the effects of the propeller airflow on the aircraft, things like:

• asymmetric wing stall,
• differences in power-on and power-off stalls,
• yawing moment from propeller slipstream, etc.

then it's a suitable analogy for the flow field.  There is no difference between modeling the flow field versus its effect because its only the effects of the flow field we care about.  If the aircraft demonstrates effects of propeller sliptstream on the aircraft then for all intents and purposes the flow field virtually exists with respect to the aircraft.  The presence of the effects mean the presence of the flow field exists because without the flow field there would be no effects from them.  As Kweassa points out, there's really no difference as it relates to our virtual aircraft.

On approaches to modeling aerodynamics:
Of course HTC could have chosen to implement a CFD approach with nasty Navier-Stokes (NS) based equations (derived in 1822) of flow fields.  The only difference that would make is greater fidelity in some things (for NS based CFD approaches have their own limitations as well) which would price most of us out of flying Aces High because we would all need linux clusters to run the game .

Infact "The history of theoretical aerodynamics to date can be largely be described as the quest for solvable simplifications of the NS equations." (Raymer).  Prandtl's lifting line theory of which so much of the basis for calculating lift and induced drag of finite wings comes from is an example of one such simplification.  The implication here is that appropriate simplification produces results that are (like the case of lifting line theory) remarkably good at an economy of effort or computation.  

In other words the presumption that CFD is always better than other approaches is simply not true because for many things you would end up with virtually the same results.  And please don't ask me to scan in various images I have access to comparing lifting-line, panel code, and NS CFD solutions on their prediction compared to actual data!

A brief note on X-Plane
I'm sure X-Plane is a great simulation.  Let's be accurate though.  From what I can gather X-Plane uses some form of panel method for it's basis for deriving it's key aero coefficients.  This is technically not considered true CFD because it's not using a Navier-Stokes method for it's solutions.  Of course their secret sauce is a secret .

What about propwash effect on rudders and elevators?
I purposely left this for last.  Has HTC somehow forgotten the effect of propwash on the tail because there isn't a virtual flow of air molecules modeled in AH?  

There's already a huge hole in this line of thinking to begin with.  Firstly how do we know that HTC hasn't modeled some virtual stream-tube of air molecules to simulate propwash?  Krusty makes the assumption that they don't.  There's no way we would know without knowledge of the code because as stated previously the only reason we know that propwash exists is the effects that it has on the airplane.

Regardless of how HTC has chosen to model propwash the answer to the question "have they somehow forgotten propwash effects on the tail?" is - no they haven't forgotten it.  As I alluded to earlier in this thread, presence of yawing moment at takeoff is indication that it exists.  And because it's almost Christmas I've done some testing to demonstrate more clearly that it does .  But alas, I'll leave everyone hanging because I need a little bit of time to organize the theory and observations.  Stay tuned!

Tango, XO
412th FS Braunco Mustangs

[moot]

Pardon me if this is a stupid question, but isn't lifting your tail off the runway while on full throttle and gear brakes due to propwash?

[clerick]

Originally posted by Kweassa
though, I fail to grasp why that is anything important, since if the equation already takes into account how the atmospheric conditions would interact with the FM equation/calculation, then by all means, that's practically no different from having a plane flying through an environment where the "atmosphere" is really modelled in.

The only difference that i could see would be that the "air" would be treated similarly to another terrain object, but far more dynamic.  If the air is/were present and not just its expected affects, then you would be able to extrapolate a number of affects, such as the affect of one aircraft's prop wash on another and other turbulence.  The first plane would actually affect the "air" it is moving through and any disturbances would then be remembered by the game regardless of whether or not another aircraft moves through the same spot.

As i understand it now, the game says "if there was real air, this is what would happen with these engine settings, speed, AOA et.c." instead of actually using some data derived from the atmosphere at a given point in time and space within the arena.  I would imagine that modeling air to that level of complexity would be very hard.

[mtnman]

Originally posted by clerick
The only difference that i could see would be that the "air" would be treated similarly to another terrain object, but far more dynamic.  If the air is/were present and not just its expected affects, then you would be able to extrapolate a number of affects, such as the affect of one aircraft's prop wash on another and other turbulence.  The first plane would actually affect the "air" it is moving through and any disturbances would then be remembered by the game regardless of whether or not another aircraft moves through the same spot.
 

Wow, could you imagine the complaints??  I think it would be cool, of course, but the whines would be heard 'round the galaxy!  With lag the way it is, and the confusion with the collision model, it would be insane.  

We'd effectively be colliding with affects on the atmosphere caused by airplanes passing through it.  What would happen if you flew through smooth air that should have been roiled?  Or through a propwash that was 50ft too far to the left/right/up or down?  (Due to lag)

Don't get me wrong, I'd love to see it, or any other atmospheric effects such as wind, rain, fog, massive smoke from burning objects, contrails, etc...

MtnMan

[dtango]

Moot:  that’s absolutely correct!  Details below.

As promised, details around propwash over tail testing.

Elevator & Rudder Effectivness
One of the variables that define elevator and rudder effectiveness is velocity.  Here are equations for:

Elevator effectiveness:
(1)

Rudder effectivness:
(2)
From these equations we see that elevator and rudder effectiveness is directly proportional to velocity on the tail h-stab (Vh) and v-stab (Vv).

Propeller Induced Velocity
Propellers accelerate the air through their disk which results in thrust as per Newton’s laws.  From propeller momentum theory, total velocity at the start of the propeller slipstream is:

V + Vi  (3)

where V is the aircraft forward velocity and Vi is the velocity induced by the propeller.   Momentum theory tells us that prop induced velocity can be found as:

Vi = sqrt[ (V^2/4) + Thrust/(2*Area(prop)*air-dens)] – V^2/2  (4)

Vi is not constant but varies with freestream velocity and thrust (which also varies with  freestream velocity) as demonstrated by the following graph:



The bottom line is that prop slipstream velocity is at it’s maximum at static thrust (0 forward velocity) and then reduces as forward velocity increases.

If we take the simple assumption that the tail of the aircraft is not too far distant from the propeller we can approximate the total velocity acting on the tail as V+Vi.  So from equation (1) and (2) we can conclude propeller induced velocity affects elevator and rudder effectiveness by the addition of Vi.  The magnitude of this effect is most prominent at low airspeeds and least at higher airspeeds.

Testing Aces High
So how can we test to see if propwash over the tail is modeled in Aces High?  One way is to see if elevator and rudder effectiveness varies with thrust while keeping forward velocity constant.  In other words if we keep velocity constant and vary thrust by changing engine horsepower, if elevator or rudder effectiveness changes then we can conclude that the change in effectiveness is caused by prop induced velocity.

Of course complications exist in performing this.  Among them are 1) how to keep a constant forward velocity while varying engine horsepower and 2) how do you perform the test at low enough airspeeds where prop induced velocity effects would show up prominently.

Elevator Test
Thanks to moot pointing it out, there’s actually a pretty simple way to test elevator effectiveness.  The procedure for the tests conducted were:

• start with the aircraft at rest,
• stand on the brakes,
• use whatever rudder necessary to keep the nose straight,
• push and hold the stick forward to apply down elevator,
• modulate the throttle to see if the elevators would rotate the plane and lift the tail off the ground with throttle applied.

It was no surprise to see tail lift off the ground with throttle input while the aircraft strained against the brakes on the ground with virtually no forward velocity.  The pitch rate of the rotation also fluctuated with the amount of throttle input given as well.  This test is best viewed from an external side view to see the effect.  This shows conclusively that propwash over the elevators is modeled.  Moot, thanks for the idea for the procedure!

Rudder Test
What about the rudder?  Trying to come up with a test procedure was a bit trickier that satisfies the near constant velocity and low airspeed constraints.

The propeller’s helical slipstream produces a yawing moment on the plane.



The propwash impinges the vertical stabilizer producing a sideforce.  Also since v-stab is an airfoil, the slipstream creates an airflow at some angle of attack with respect to the v-stab.  This produces a lift force in the direction the horizontal axis.  The combination of the side force and lift force from the slipstream produces a yawing moment on the airplane.

Tests were conducted on a P-51D at 50% fuel with fuel burn set to infinite (no burn).  Here are the procedures I used to test if propwash effect on the rudder existed:

• Enable combat trim
• Climb to about 2k AGL
• Set speed command to 170 mph
• Alt-x into auto climb to try and fix forward velocity
• modulate throttle at 5” man increments
• do this slowly and watching E6B to keep TAS as constant as possible
• Watch what happens to rudder trim to see if it varies

The tests were done for manifold settings from 20”-61” both directions.  With the TAS holding nearly constant around 200 mph, combat trim automatically adjusted the rudder trim to keep the nose pointed forward as I modulated throttle.  This test shows again that propwash must be modeled on the v-stab because the rudder trim is being adjusted to deal with the changing moments while forward velocity was near constant and throttle was modulated.

The change in rudder trim was much more pronounced conducting the tests at an alt-x speed of 150 mph, and TAS nearly constant around 160 mph.  Of course p-factor could also play a part to the yawing moment besides prop induced velocity.  It’s difficult to eliminate it with my procedure but I tried to minimize the effect by picking airspeeds in a constant speed climb to try and minimize large aoa’s which would produce large amounts of p-factor.

So from the testing I think it’s safe to conclude that propwash on the tail in Aces High is indeed modeled.  It’s clear that elevator effectiveness and rudder effectiveness varies with thrust while holding velocity near constant.  The whole topic can be quite a bit more complex but for the sake of understandability I’ll keep it simplified at the level I’ve presented.

Tango, XO
412th FS Braunco Mustangs

[TimRas]

Originally posted by dtango
It was no surprise to see tail lift off the ground with throttle input while the aircraft strained against the brakes on the ground with virtually no forward velocity.  The pitch rate of the rotation also fluctuated with the amount of throttle input given as well. .This shows conclusively that propwash over the elevators is modeled.

While I don't doubt that the propwash is modeled, this is not conclusive proof. The thrust from the propeller tends to tip the taildragger plane over its nose even without propwash.

[dtango]

Originally posted by TimRas
While I don't doubt that the propwash is modeled, this is not conclusive proof. The thrust from the propeller tends to tip the taildragger plane over its nose even without propwash.

Agreed.  I assumed thrust not being offset from CG which is not valid.  

In the tests however, the nose rotates sooner & easier with elevator deflected vs. not which would indicate propwash effects however.

Tango, XO
412th FS Braunco Mustangs

[moot]

I think it's one of the 47s or the N1K that yaws even as you hold the tail up with the brakes on.

[BlauK]

Is it so that each plane is kinda flying alone in clean air? They only affect their "own air" and are not disturbed by other planes' air flows or turbulences?

[dtango]

I'm not aware if they do or not.  As others have mentioned it would be challenging to model.  

We should note that some simpler forms of that do exist in AH.  Wind layers can be set up in AH with stratified at user definable altitude bands, directions, and speeds.  There is also a "rise" variable which can be set from negative to positive.  I have no idea what this does, maybe some way of similating gusts?

Another point, prop wash wake or turbulence is probably not the issue if they were going to model aircraft to aircraft wake impact.  I don't have the theory but turbulence created by propulsion systems tend to be short-lived.

What we would be more worried about vs. propwash from other airplanes is this instead...



...wake turbulence due to wing tip vortices.  They are magnitudes stronger and much longer lasting than propwash disturbances.

Some of the possible issues that could occur are depicted in the diagram below:



Tango, XO
412th FS Braunco Mustangs