Sometimes the model flies in a bubble and sometimes it doesn't

[Tom Harper]

http://www.rcgroups.com/forums/showthread.php?t=522513&highlight=ground+speed+effects+airspeed

A problem with the discussions above is the absence of instrumentation.

Much of the discussion challenges the ability of the modeler to properly interpret what is being observed. Other discussions concentrate on classical theory. Of course the theory is correct in it's limited context, but may not apply generally.

The recent introduction of "AttoPilot" increases the scope of the available information. AttoPilot has GPS and barometric systems to provide altitude and path control. All flight information is recorded at a .2 sec rate. Airspeed is recorded for both GPS (ground speed) and barometric (airspeed) measurements.

Many modelers believe that a model will turn into the wind and hang there (windmill). This is frequently observed with free-flights, and is also attributed to RC models. Such a thing could not happen if the model were tightly coupled to the moving air mass by 'equilibrium'.

The three images below are plots of 3 consecutive automated flights. The model is flying at a controlled altitude of 200' and a controlled airspeed of 48 KPH (+10%,-6%). The altitude variation was less than 8'.

These flights were made with increasing amounts of system gain as part of the tuning process:

1. In the first image the model has insufficient gain to follow the programmed path. It follows the programmed circle, but drifts to the West with the wind. It is flying in equilibrium with the moving air mass.

3. In the third photo the gain is high enough to maintain circular flight. Equilibrium is evident by the distortion of the circle and by the recorded groundspeed. The airspeed is constant but the ground speed varies with heading. The model is flying in equilibrium with the moving air mass.

2. In the second photograph the model is not in equilibrium during the early portion of the flight. It exhibits the characteristic described above. The flight is detailed on the image.

This does not violate any laws of physics. It simply illustrates that the model may or may not operate in equilibrium with the moving air mass. It all depends on the conditions.

Equilibrium is a special case.

[JetPlaneFlyer]

Tom,
What do you mean by 'not in equilibrium'?..Are you claiming that there is some other external influence on a plane in flight other than the air that surrounds it (and gravity)?

What's your point?

At least the fact that the recorded airspeed and altitude is constant throughout de-bunks the old 'downwind turn' falacy! (though I'm sure some will still argue the point)

Steve

PS.. I mainly fly free flight and I've never noticed a tendency for a free flight model to consistently turn into wind and 'hang there'... Occasionally it may happen but no more often than the law of averages would dictate, equally often they disappear directly downwind. You want proof then look at the direction almost all free flight modellers run after their models.. (i.e. downwind). If models really did ‘weathercock’ into wind then most would fly away upwind.

[bjr_93tz]

PS.. I mainly fly free flight and I've never noticed a tendency for a free flight model to consistently turn into wind and 'hang there'... Occasionally it may happen but no more often than the law of averages would dictate, equally often they disappear directly downwind. You want proof then look at the direction almost all free flight modellers run after their models.. (i.e. downwind). If models really did ‘weathercock’ into wind then most would fly away upwind.

Too right with the above statement, free flighters trimmed to fly perfect circles on calm days will ALWAYS turn down wind and fly perfectly straight when the wind picks up :D :D

IF you're really lucky they'll build up just the right amount of kinetic energy before dropping a wing into the ground and shearing it off.....

[Tom Harper]

JPF,

You make my point. When a FF runs with the wind it is not circling in equilibrium.

A trimmed model that is in equilibrium with a moving air mass will follow the pattern shown in photo 1. That is like a FF in a light morning breeze.

The same airplane shown in photo 2 does not follow that pattern. There are two cases where it weathercocks into the wind. I believe this is because the wind velocity is not constant. When the model is adjusting to changes in velocity, of the air mass, it is not in equilibrium. It is then acted upon by the change in velocity of the moving air mass.

The plots are of a model under automatic control. They are not being directed by the pilot on the ground.

A downwind turn in an air mass that is changing velocity will result in disaster.

The point is that equilibrium with the air mass is one mode of flight. It is not the only one. Sometimes the model flies in a bubble of moving air and sometimes it doesn't.

[Montag DP]

A downwind turn in an air mass that is changing velocity will result in disaster.Disagreed. It depends on how the velocity is changing. If the wind velocity is decreasing the plane may actually gain airspeed by turning downwind.

The point is that equilibrium with the air mass is one mode of flight. It is not the only one. Sometimes the model flies in a bubble of moving air and sometimes it doesn't.By equilibrium you just mean that the wind is steady?

[mlbco]

Tom,

I'll be the first to admit I don't understand what you're getting at, but the results you show have everything to do with the autopilot's performance and little to do with airmass "equilibrium". I haven't seen the navigation logic for the autopilot you're using, but as the ground speed approaches zero the navigation control might become unstable. The failed attempts to re-capture the waypoint look like the system is unstable or confused about what to do when the groundspeed approaches zero. I don't think this has anything to do with the airmass argument you're suggesting.

One other thing to check regarding your autopilot's navigation logic is whether the system has been demonstrated to be stable when navigating upwind vs. downwind. Some systems command turn rate (i.e. body axis heading rate) to close the course line guidance control loop and this will often go unstable when traveling upwind in high wind conditions. The reason for this is that the course over ground rate (i.e. GPS "turn" rate) is what should be commanded in order to navigate relative to the earth, not the aircraft's heading rate.

Steve

[JetPlaneFlyer]

JPF,

You make my point. When a FF runs with the wind it is not circling in equilibrium.



I'm still not sure what your point is much less how I'm making it... :confused:

The models that I refer to that sometimes fly straight downwind or sometimes hold an upwind heading are the ones that are not trimmed consistantly to circle. When a model is trimmed to fly straight then it may sometimes fly straight into wind, it may head straight down wind, or it may head in any other direction. This has nothing to do with wind direction it's just down to normal flight variables, launch direction, turbulence etc.

'Gusts' obviously can effect a model but this would not consistantly make a model weathervane into wind because if the average wind speed is steady then every gust is acompanied by a similar 'lull'. The increasing speed gust may sometimes tend to cause a plane to turn into wind but the decreasing speed lulls have the opposite effect.

If your only point is that gusts, turbulence and other 'non-equilibrium' wind conditions have an effect on a plane then i dont think anyone will argue and you dont need flight plots to prove it.

Steve

[JetPlaneFlyer]

The plots are of a model under automatic control. They are not being directed by the pilot on the ground.



Agreed that the control is 'automatic' but i assume it's GPS directed so the automatic control is attempting to hold the model in a steady rate of turn relative to a 'geostatic' (ground based) reference..
In a wind the 'real' rate of turn (relative to the air) and therefore the bank angle of the model will be much higher in the downwind leg which potentially could still cause problems.
Human control or by GPS the basic issue is still the same; attempting to fly the plane by ground based reference.

Steve

[eflightray]

Pity it would be too tricky to launch a free flight balloon and free flight glider at the same point, at say 200ft. The glider would be set to fly a circle.

Interesting to see the 'pattern' they fly in the hope it may lay a few ghosts to rest.

But then, should the glider be set to fly a right hand or left hand circle in the Northern Hemisphere ? :)

[Tom Harper]

mlbco et al,

What I am curious about is the belief, by many, that a model will 'weathervane' into the wind - that it is effected by the wind. This gives rise to 'downwind turn' and related stories. There seems to be two schools of belief.

One holds that the model flies in equilibrium with the air mass. It moves in a bubble of air. I can confirm this in data logs of flights where the wind is steady or the gusts build and ebb slowly.

The second belief is that the wind constantly effects the performance of the model. I listen to the comments of RC pilots about the model turning, hanging, gaining or losing altitude in response to the wind. Often this looks like a problem of reference, but sometimes it looks like a wind issue.

On the second flight shown above (image2) the ground comments were that the model was weathervaning. The flight was being logged, so I had a record. Examining the log, I'd say that is exactly what it was doing.

Steve (mlbco) - Attopilot is good down to ground speeds of 15kph or less. The 5Hz GPS speed readings were 23, 16, 10, 6, 22, 23, 30. I doubt that the 10 and 6 values ever got past the smoothing algorithm. Attopilot gave full throttle and pitched down about 5 degrees to counter the gust.

As I said above, these flights were part of the tuning process. I increased the max bank angle to 30 degrees and airspeed to 62 KPH. That allowed the model to deal with the level of gusts present that day (image 3).

The point? Sometimes the model flies in equilibrium with the air mass and sometimes it doesn't.

[slipstick]

Sorry I must particularly thick tonight. Can you explain "in equilibrium with the air mass", perhaps in some different words because those don't mean anything to me.

I can't think of any case where "equilibrium" means anything in this context.

Steve

[JetPlaneFlyer]

Tom,
It's really impossible to say from the flight plots what behavior was due to the 'autopilot' and what was due to the wind/gusts. A much more meaningfull test of the effects of the wind would be to turn off the autopilot and set the model into a steady turn and watch what happens. That way you could at least be reasonably sure that any observed behavior was due to 'wind effects' and not autopilot 'interference'.

By the way.. i dont see anything in plot 2 that looks like 'weathervaning' to me. The flight tracks at points B and C look to me like what you would expect from a circling model being blown downwind in a wind that almost matches flying speed. If the model was 'weathervaning' then I'd expect the model to adopt an upwind heading and hold it (that's what weathervaning is after all).. That's not what occured at points B and C, if anything it appears that the autopilot had not kicked in and the model was following an almost 'freeflight' steady turn rate circling patern.

I'm still not really sure what this 'equilibrium thing' is meant to mean in this context...

Steve

[Montag DP]

Well of course, if you have gusts there will be an effect on the flight. If you don't have gusts then they don't affect the flight. A steady wind will have a constant effect on the plane's velocity vector, relative to the ground. I too am still not really sure what the point is. :confused:

[BMatthews]

Let's kill a couple of myths. I've had the opportunity to witness a LOT of free flight flying in my time so I feel I'm fairly competent to comment on some of this stuff.

A free flight model in circling flight in wind with no significant thermal content and where the wind speed is steady will alter ground speed throughout the turn but does not lift up or dive down in different parts of the circle. This supports the stable within the air mass claim. However it does not take much thermal or turbulence to kick a model around and make it seem like it is responding to airspeed velocity changes.

There's at most a very small percentage of free flight models that "weathervane". And when they do it can be into the wind or off the wind at an angle or even downwind :D The direction the nose points when this occurs is rarely consistent from model to model or even flight to flight with the same model. Out of the times I've seen this the "obvious" one of nose directly into the wind is no more common than any other direction. In all cases it seems to be more linked to a model with not enough turn trimmed into it and what happens is that the model suddenly finds an area of light lift that upsets the glide speed. When that happens it noses up or down and with the change the cross linking of the rudder tab and wing warping counter each other and it points cross country. I've even seen models where they circle faithfully but in a very open diameter in light breezes but when sent up in dead calm late evening conditions they decide to go for a cross country jaunt. The owner never did figure THIS one out but once the model was set for a smaller diameter turn the glide circle stayed consistent with varying wind conditions. We all just shrugged and decided that there's still SOME mysteries in the field of free flight and moved on.... So from all these observations I would suggest that models do not "weathervane" in the sense that folks say the wind is making them point.

A case in point was a towline glider I flew many years ago. In light lift calm air it had a combination of wing warping and rudder trim that produced a nice 60'ish foot diameter turn. In the right conditions it would circle around pickup lift tighten to about 50 feet in the thermal. At other times it would "weathervane" on a straight flight and even turn the other way at rare times and then suddenly lift the nose a bit and hook into a thermal. When I first observed this I was dumbfounded and thought I had a real dud on my hands. Then I realized it wasn't coming down but instead was milking the air and staying up and even climbing. Took me a while to study and understand this back then but the light finally dawned. Sadly it suffered from a stuck timer and few away after only about 4 contests but it always had this odd ability to wander. But at no time was it actually weathervaning in the proper sense of the word.

I have to say though that it would be nice to send up an old timer that would have a sensitive airspeed sensor built in along with a directional sensor so that it would be easy to take the data from flying in a trimmed hands off circle and plot the heading to airspeed for a few circles. If there's no significant and repeating change in airspeed it would sure lay to rest the whole downwind turn issue. But to do this I say it should be an old timer or regular polyhedral RC glider so that the trims can be set up for stable flight and the Tx be held thumbs off for a series of glide circles. Why? Because it's pretty safe to say that 99% of the claims to all these behaviour issues are more accurately caused by the thumbs on the sticks than the aerodynamics of the models.

[Tom Harper]

Bruce,

That's the point.

The autopilot provides hands off circular flight at a constant altitude.

In image 1 the model flies much like a free flight drifting with a constant breeze.

In image 2 the model is too lightly controlled to overcome gusts, so it weathervanes. The data log shows it hanging directly into the wind. You can see that at points B and C.

In image 3 the model has the velocity and control gain to overcome gusts.

An old timer would act the same. I do have a Buccaneer 'B' kit that I could put together and power with an electric. Wouldn't be difficult to install AttoPilot as a data logger. That would provide the FF test, but I'm too old to chase it.

Tom

[vintage1]

Tom. The autopilot is not referenced to the air, its referenced to the ground.
Its trying to keep the model in one place ground wise.

That makes complete nonsense of all your assertions.

[Tom Harper]

Perhaps, but how so?

[mlbco]

Tom,

Here is some data from my UAV flying autonomously at 40 mph between 2 closely spaced waypoints in steady winds of 30 mph. The funny shape of the trajectory is solely due to the guidance laws and the bank angle limit that they obey. The UAV drifts downwind at max bank angle and then intercepts the course at nearly 90 degree approach angle. Altitude and airspeed are nearly constant during the flight and bank angle is commanded by the guidance algorithm to capture the courseline and waypoints. Simulation results of a point mass model of the UAV yield almost identical trajectory results. The point mass model assumes zero sideslip, constant speed, and constant altitude. Do you see any "equilibrium" effects in this data?

Steve

[JetPlaneFlyer]

In image 2 the model is too lightly controlled to overcome gusts, so it weathervanes. The data log shows it hanging directly into the wind. You can see that at points B and C.


Tom,
What's your definition of 'Weathervaning'?

In track two local to points B and C I see a plane flying what appears to be pretty much constant radius (relative to the air) right hand circles and getting blown downwind by a wind who's velocity is close to the model's own airspeed. Obviously as the plane is circling part of the circle will be into wind and because the wind speed almost matches flying speed when the model does face into the wind it will be observed to 'hover' relative to the ground.

If your definition of weathervaning is any model that while circling at some point flies directly into wind then 'yes' it is weathervaning... But by this definition all models that fly in a circle in a wind 'weathervane'.

If by weathervaning you mean a model departs from its normal trimmed flight path, points directly into wind and holds that heading for an extended time period (which is what i would recognise as weathervaning, if it existed) then sorry but i just dont see it in plot 2, not even a trace of it.

And I still dont see how one flight is 'in equilibrium' (whatever that is in this context) and the other is not :confused:

Steve

[MCarlton]

Perhaps, but how so?

Surely, what your plots show is the path of the model in reference to a fixed point, ie the ground, so I would assume that the variations in flightpath are due to the autopilot attempting to maintain a ground track in turbulent air, thus each gust/lull creates a response in the model to try to maintain that track.

I'm no expert, but it seems to me that in order to evaluate the "downwind turn" hypothesis, the ground is entirely irrelevant, it is the course which the model flies in relation to the air that is at issue.

Thus, the only way to measure this, would be to take measurements of the model as it circles around a point which is free to move with the air mass. So for example, if you could have an autopilot which was set up to follow a circular flight path centred on a dragless balloon, and then measure (by what means I don't know) the distance between the model and the balloon.

"weathervaning" is usually something I would attribute to a model which is trimmed to fly straight, and is something that old FF slope gliders (pre- magnet steered) used to maintain heading. Usually by employing a lot of fin area.

If the model is flying with the controls/trim deflected, then that deflection/trim will have the same effect on the model, regardless of the wind, the only difference will be a visual difference which we perceive due to our fixed position relative to the air in which the model moves.

[eflightray]

Change the 'bubble' to a cube.

Imagine a plane on a calm day flying round in a circle.

Now using the latest Star Fleet technology, cut a great big cube of air out including the plane, (all protected by a force field naturally). ;)

Now move the cube across the sky to an area of sky that has a wind moving at 20mph, now let the cube move with the 20mph air so it match speed.

Does the plane know it's now moving in a wind at 20mph ?

Will the shape of the circle change relative to the cube ?

Will the shape of the circle change relative to the ground ?

Oh!, don't ask me, I'm only the local idiot who keeps getting thing wrong :o

[Tom Harper]

mlbco,

Interesting plot. The effect of equilibrium would show up as a significant increase in groundspeed on Easterly headings and a significant decrease in groundspeed on Westerly headings.

Here is a normalized chart for one of the circles in image 3. Airspeed is constant while groundspeed correlates with heading.

Tom

[Tom Harper]

JPF,

Based on the comments of local RC pilots, the term 'weathervaning' seems to mean that the model tends to turn into the wind.

The three plots shown were the first of about two dozen tuning flights. These first three flights had minimum system gain (control). The amount of control in image 1 is about what you would expect for a free flight model. In image 2 the autopilot is trimming the model but not enough to overcome the impact of gusts. In image 3, system gain has been increased to the point that the model is under full control of the autopilot.

Tom

[Tom Harper]

MCarlton,

The path shown in image 1 is close to the scenario you propose. The model is drifting downwind as it circles. The amount of control imposed by the autopilot is just sufficient to produce a circle, but not enough to make progress toward the way point.

In image 2 the system gain is high enough to allow progress toward the way point, but not enough to overcome the impact of gusts. In each case the model turns iinto the wind, pauses and then falls off to run with the gust.

Tom

[MCarlton]

In image 2 the system gain is high enough to allow progress toward the way point,

Agreed, but you are talking about a waypoint which is on the ground, which has nothing to do with the path the model follows in the air.

All the paths you are looking at are in reference to fixed points on the ground, which is the whole point of the downwind turn business, that what appears to be an effect on flightpath and airspeed is ONLY an effect when viewed from a fixed point. If the reference point is allowed to move, the effect is different.

Here's a couple of analogies.

Imagine a boat, going round in circles on a lake, tethered to a free floating buoy. Now put the same set up on a flowing stream. Because the tether point is able to move at the same speed as the current, the boat will still go around the buoy in circles, but those circles will move in relation to a fixed point on the bank.

Or imagine a control line model in a high wind. If the operator does not move, the model will appear to blow towards the operator on the upwind side of the circle. But, if the operator walks downwind at the same speed as the wind, it will appear to the operator as if the model is flying in still air.

So what the autopilot in your setup is doing, is trying to maintain a track over the ground, not maintain a constant circle, because if it were flying a constant circle, those circles would move downwind and the track over the ground would be skewed.

[Brandano]

Can you set up the AP to fly for a length of time at a constant bank angle, coordinated, and record the rate of turn and airspeed through that time?

[Tom Harper]

Brandano,

Here is a 24 second clip from the flight shown in image 1. The target Roll Angle is 15 degrees.

Tom

[Brandano]

noticed how the airspeed keeps pretty constant through 60 degrees of heading change?

[Tom Harper]

Yes - you can also see that in #22.

[BMatthews]

Tom, the chart in post 22 confirms what we've been suggesting. The airplane sees a constant airspeed while the ground speed alters throughout the turn.

So what's left? The last suggestion that "weathervaning" is active would be to look at the rate of heading change. If the rate of heading change or deltaHeading rate is constant through a circle or two again it says there's no such thing as weathervaning. But if the deltaH should alter and reduce for part of the circle in a consistent manner for a few turns then it says there's something at work. Interacting with thermals will do this so it would need to be check over a variety of flights to ensure you find some truly calm air other than steady wind speed or see a steady trend thru a number of flights.

I noticed that the heading curve varies all over which seems to support the idea of weathervaning being an issue. But without knowing what the autopilot was doing it's hard to validate this data. What you really need is to fly the model on trim and set it for circling flight without the autopilot being able to alter the servo position but so that it still reads the flight information. Then let it glide downwind through 3 or 4 circles totally hands off and see what the data says.

The whole point of my earlier posts is that much of the reported woes are actually due to the pilot's thumbs not understanding how the plain responds to the moving air as seem from the fixed observation point. So the key is to remove the pilot's thumbs from the equation. And that includes the virtual thumbs of the autopilot as well. The plane needs to be left to alter speed and direction on it's own with nary an input to mess it up other than that needed to bring it home or avoid crashing.

[Tom Harper]

Bruce,

Even a free flight carries the virtual thumbs of the builder's errors and adjustments.

I believe it would be interesting to examine the data around the alleged weathervaning'. I'll post some graphs and data tomorrow.

Tom

[slipstick]

Even a free flight carries the virtual thumbs of the builder's errors and adjustments.

But during any single flight these are fixed...and that's the whole point. If either a ground based pilot or a ground referenced autopilot can fiddle with the controls during the flight they can mess up the plane's flight pattern. So, as so often in the "downwind turn" debates, we're seeing artifacts of the control methods and not genuine aerodynamic effects.

Steve

[Tom Harper]

Slipstick,

Are you moving us toward a conclusion that aircraft with moving control surfaces are not effected by wind gusts or shear?

I do agree that some ground observations regarding wind effects are incorrect because of the frame of reference of the observer. However, gusts do impact the model. An automated model provides us with an instrumented probe that could enhance our understanding of gusts.

[kcaldwel]

Tom,

Wind gusts have been studied to death, mostly because of the wind power industry. Gusts follow a logarithmic law distribution, and are direction invariant. This means that the model will have no idea what direction the wind is, unless it is being guided by a ground based pilot, or a ground based GPS coordinate system, in which case it can be commanded to do strange things.

http://www.bom.gov.au/amm/docs/1965/deacon.pdf

http://twist.physik.uni-oldenburg.de/unicms%20Hydro/PDF/BLM2003.pdf

http://www.springerlink.com/content/?k=wind+gust+power+law

http://findarticles.com/p/articles/mi_m0QRG/is_27/ai_n11836129/

Kevin

[Tom Harper]

Kevin,

Thanks for the links. The log model is interesting.

However, these articles simply describe gusts and the degree to which they can be predicted or at least 'nowcast'. That doesn't relate to flight dynamics.

Tom

[Tom Harper]

I believe the issue was well stated by Vintage1:

Tom - ".....years ago, you convinced me that an airplane flies at equilibrium, oblivious of the motion of the supporting air mass."


Vintage1 - "Oblivious of the steady state speed of the air yes, bot not oblivious to changes in that."

http://www.rcgroups.com/forums/showpost.php?p=12570277&postcount=11


Tom

[HX3D014]

I believe the issue was well stated by Vintage1:

Tom - ".....years ago, you convinced me that an airplane flies at equilibrium, oblivious of the motion of the supporting air mass."


Vintage1 - "Oblivious of the steady state speed of the air yes, bot not oblivious to changes in that."

http://www.rcgroups.com/forums/showpost.php?p=12570277&postcount=11


Tom
That sounds right.

Steady wind is one thing. but changes to the wind speed will kick the plane around.

When that change happens (at what part of the flight, what heading to the gust) will cause different effects. Just as the Ratio of several important factors will cause different results.
Ratio of influencing factors. (Of the top of my head, got any others)
Aircraft Wing loading : Wind speed : Aircrafts Speed : and Gust/Lull Values

I think aircraft wing loading is not the best to try to capture a value that suggests the aircrafts inertia over aerodynamic drag. and even the aerodynamic drag value will change from direction to direction of gust intercept.

Bryce.

PS do you have any other photos of the field? preferably one that may show any terrain detail.

[eflightray]

What's the difference between say a wind gust and a thermal/downdraft ? or are they the same thing ?, (you can't actually see either).

I occasionally fly thermal gliders, and one of the important part of watching the plane is for any 'disturbance', and then trying to decide where the thermal is. We watch the wing to see if it tilts or the model is pushed left or right out of a thermals path, rarely have I flown head on into one. A head on sometime is seen by the model slowing.
Like wise, your also on the look out for the down drafts and trying to avoid them.

I was just thinking that whatever any 2D tracks may show, you don't really know what the air is doing at that instant in time. Do you also carry altitude monitoring gear on board so that a 3D plot can be shown ? It would give a clearer idea of whether the model hit rising air, a downdraft, or stayed level during a deviation.

Sorry, I do tend to ramble a bit. :)

[Tom Harper]

eflight,

The log shows that the model lost about a foot of altitude.

The normalized plot of the gust event is shown in the attached image. As the model turned into the gust the autopilot pitched it down about 3 degrees. In the plot you can see an airspeed loss of about 10%.

If the model was in equilibrium with the air mass there would have been no change in airspeed. If the autopilot had caused the change in airspeed it would have increased due to pitching down, not decreased 10%.

[Tom Harper]

Bryce,

The field is flat desert covered with creosote. The model altitude was 360 feet.

Tom

[BMatthews]

As slipstick said already the virtual thumbs are locked in place for any single flight.

And yes when in turbulence a free flying model will be subject to gust and thermal effects. But the problem is that I don't know of anyone that flies free flight where the model is completely symetrical other than by accident. Because free flights are trimmed to fly in one direction of circling only during the glide (power patterns can vary in direction) they are set up with some twist in the wing as well as rudder tab offset. So no, most free flight models will not react "normally" to turbulence. That story of my towline glider with "wanderlust" is an extreme example but one that is common and typical to lesser degree with any free flight contest or sport flying model.

Now an RC model with symetrical washout or washin in the wing trimmed for a circling hands off flight pattern is a different story. Because it's not assymetrically trimmed it should not respond to vertical gusts in the same way my towline glider did, Instead it should maintain the steady circling flight with no signs of this badly named "weathervaning" issue.

Now turbulence is another story. Obviously if the model passes from one area to another where the wind speed and/or direction is shifted then there will be a fluctuation in the flight while it accelerates in whatever direction in response to the change. But once back to equilibrium the flight circle trim will force it to continue with the turn as before.

Let's not discount how the model handles either. I've got an RC glider that is notoriously short coupled in the fuselage department. If I maintain a decent speed it will rail up and pivot in circles standing on a 45 or more angle in tight lift. But if I screw up and let it slow down below a comparitively high cruise speed the model will level the wings and head away from the turn despite massive control inputs. To recover I need to push forward momentarily to regain a semi fast cruise. At that time full rudder authourity is restored. Later on I made a model with a longer tail moment boom for electric power. The longer version due to the tail boom as well as the greater weight of the electrics on board all worked to ensure a faster speed in flight so this tendency is greatly reduced. But it's still there. The point to this being that if it does something like this the pilot could construe it as "weathervaning" when it's something entirely different.