Adverse yaw? - Page 4 - RC Groups
 Jan 30, 2013, 11:14 AM Registered User Many people's intuitions about turning are driven by their experiences twisting the handlebars of a bicycle or twisting the steering wheel of a car. It seems that a net "twisting" torque or yaw torque must be applied to keep the turn going. When we turn by banking a wing, we see that that isn't so. When we turn by stepping on a rudder pedal, again there seems to be a requirement for a steady net yaw torque, but that's only because we are fighting the "directional stability" or "weathervane effect" of the vertical fin. If we are neutrally stable in yaw, we can generate a steady sideforce without a steady rudder force. Once we estabilish the desired yaw-slip angle, it will stay there even with the rudder centered. If we are actaully yaw-unstable we may need to apply pressure to the outside rudder pedal not the inside rudder pedal, to hold the yaw/slip angle constant as we fly a constant-rate unbanked "flat turn", even though the yaw string is still streaming toward the inside of the turn just as it would in a flat turn in a more "normal" aircraft with positive directional stability. We would never want to fly such a yaw-unstable plane (or a plane with neutral yaw stability) but the situation could arise in theory, or in a computer-controlled plane. Likewise, bicycles and cars can be designed in such a way that they don't require a steady "twisting" force on the controls to keep the turn going, even in the case of an unbanked flat turn. It has to do with the rake of the connection to the axle of the wheel, etc. The heart of a turn is a centripetal force, not a steady yaw torque. To create a condition where a sideforce tends to act in a constant direction rather than a changing, centripetal, center-pointing direction, you need to expose additional constraints to keep the flight path from curving. That is what you are doing Vespa when you say it is impossible for a sideforce to cause a turn. Steve
Jan 30, 2013, 11:29 AM
Registered User
Quote:
 Originally Posted by vespa Any airplane can fly flat circles, even a flying wing with no fuselage or vertical tail whatsoever. All you need is something to create a yaw rate. For example a split aileron/drag rudder. Lateral area greatly improves the effect but it's the yaw rate that makes it possible.
No, all you need is for something to create a sideforce. Particularly in the case of a glider where we have no directed thrust issues, we can yaw the nose sideways to the flight path all day long but if we aren't generating sideforce, the flight path is NOT curving and the heading is not changing apart from that initial heading change involved in yawing the nose sideways.

You are really missing something crucial Vespa. Here's a real, actual example. I mounted a rudder to hang glider. Side area of the hang glider was quite low. Steady left rudder definitely, positively swung the nose to point some degrees further left. A yaw string on the nose streamed to the left-- we are now flying sideways, with the rudder deflected. But flying sideways is NOT the same as establishing a yaw rotation. In a conventional aircraft in this situation, IF we forced the wings to stay exactly level, the airflow impacting the right side of the fuselage would generate a sideforce to the left. The flight path would curve left and the aircraft's "directional stability" would prevent the yaw/slip angle from getting any larger, which means that the "directional stability" dynamics would generate a brief net left yaw torque to overcome yaw rotational inertia and initiate a steady left yaw rotation.

HOWEVER in the case of the hang glider, side area was so low that the dominant sideforce was actually the way that the left-deflected rudder was shoving air to the left, creating a sideforce to the right. Yes the nose swung left a few degrees and the yaw string on the nose streamed to the left, but there was no net sideforce to the left. Rather, there was a very weak net sideforce to the right. When I kept the wings exactly level, and the rudder was deflected to the left, the flight path tended to curve right. Again, the aircraft's "directional stability" tended to prevent the yaw/slip angle from getting any larger, which now means that the "directional stability" dynamics generated a brief net RIGHT yaw torque to overcome yaw rotational inertia and initiate a steady RIGHT yaw rotation.

To prevent the flight path from curving right, I had to lower the left wing slightly.

A key point is that in this particular aircraft, the "weathervane effect" or "directional stability" was provided by the swept shape of the wing, not by a fixed vertical fin. If an aircraft's "directional stability" is provided not by wing sweep but rather by surface area located aft of the CG but ahead of the rudder, then it will always be the case that a left-deflected rudder provides a net left side force, except in cases where exotic effects are dominating like asymmetric thrust, asymmetric wing drag, etc. Because when the pilot applies heavy left rudder, the yaw/slip angle will only come to equilibrium when the nose has yawed left relative to the flow until it has reached the angle where the fin is making a right yaw torque equal to the rudder's left yaw torque. Since the fin is ahead of the rudder, it must make more force to make the same torque, so the net sideforce must be to the left. This is how most aircraft behave.

If you can't make heads or tails of what I'm talking about, you might perhaps find a way to expand your paradigm... rather than defend it!

Steve
Last edited by aeronaut999; Jan 30, 2013 at 12:12 PM.
 Jan 30, 2013, 11:50 AM Registered User I saw this and it seems to fit the turing problems of my plane! Too stable! https://www.rcgroups.com/forums/showthread.php?t=280238 The Lacey is known for it's remarkable stability, even though it has no dihedral. This is a good thing for freeflight versions which go very well with scale flat wing. The reasons are I believe twofold:.. First due to the short low aspect wing the speed differential between the inner wing and the outer when the plane turns is low. This means the tendency for the plane to ‘fall into’ the turn is much less than a longer wing model. Second and perhaps most importantly the big slab sided fuselage acts like a huge wing fence. When in a turn and banked any aircraft 'slips' inward producing a sideways component to the airflow... (this is how dihedral works in keeping the model stable). On the Lacey the fuselage 'masks' the outside wing from the sideways airflow component, as the wing is so short this masking effect impacts a large proportion of the wing area. This masking produces a roll effect opposite to the turn causing the model to 'recover' to straight and level flight. I can’t think of much that could be done to get around it, other than adding anhedral (which would look awful).. It's just a 'feature' of the design. As you have discovered moving the vertical CG upward may help a bit but probably wont be a complete cure.
Jan 30, 2013, 12:15 PM
Registered User
Quote:
 Originally Posted by vespa Any airplane can fly flat circles, even a flying wing with no fuselage or vertical tail whatsoever. All you need is something to create a yaw rate. For example a split aileron/drag rudder. Lateral area greatly improves the effect but it's the yaw rate that makes it possible.
I'd argue that yaw rate is just the accidental byproduct of the rate of curvature of the flight path plus the rate of change of the slip/skid angle. Centripetal force is what makes the turn possible.
Jan 30, 2013, 12:21 PM
Registered User
Quote:
 Originally Posted by grant31781 I saw this and it seems to fit the turing problems of my plane! Too stable! https://www.rcgroups.com/forums/showthread.php?t=280238 The Lacey is known for it's remarkable stability, even though it has no dihedral. This is a good thing for freeflight versions which go very well with scale flat wing. The reasons are I believe twofold:.. First due to the short low aspect wing the speed differential between the inner wing and the outer when the plane turns is low. This means the tendency for the plane to ‘fall into’ the turn is much less than a longer wing model. Second and perhaps most importantly the big slab sided fuselage acts like a huge wing fence. When in a turn and banked any aircraft 'slips' inward producing a sideways component to the airflow... (this is how dihedral works in keeping the model stable). On the Lacey the fuselage 'masks' the outside wing from the sideways airflow component, as the wing is so short this masking effect impacts a large proportion of the wing area. This masking produces a roll effect opposite to the turn causing the model to 'recover' to straight and level flight. I can’t think of much that could be done to get around it, other than adding anhedral (which would look awful).. It's just a 'feature' of the design. As you have discovered moving the vertical CG upward may help a bit but probably wont be a complete cure.
Grant that "masking" effect you speak of is indeed a well-known part of the reason that a wing mounted on top of a slab-sided fuselage acts like a wing with dihedral, even if it actually has none. You get the opposite effect with a wing mounted on the bottom of a slab-sided fuselage-- if the wing is flat, it will act like it has anhedral (negative dihedral). For example see figures 9.3-9.6 here http://www.av8n.com/how/htm/roll.html

That is why low-winged aircraft are generally given more dihedral than high-winged aircraft of similar performance and handling characteristics.

Steve
 Jan 30, 2013, 12:21 PM REMOVE TRUMP NOW! Sorry aero, a slip (sans rudder) occurs purely due to the bank angle and resulting lateral lift component. Therefore the slip is always in the direction of the bank. The yaw force produced by the outboard wing flying faster has nothing to do with slip or bank -- it is the result of, and in opposition to, the yaw rate. Control line planes are a great example of the required yaw rate. They work because a string attached to the wingtip is pulled aft in an effort to spin the plane around. Your hang glider example of yawning left with a rudder and turning right is plausible, but not for the reasons you think. What's happening here is that you have a tremendous amount of dihedral (low c.g., high sweep) so a tremendous effort (weight shift, ailerons) is needed to hold the wings level during a slip. With effort comes drag, so just like the split aileron/drag rudder example, it's conceivably possible to produce a yaw rate. But what really makes this work is that a hang glider has almost no fuselage, so if the rudder is effectively the majority of the aerodynamically useful fuselage, deflecting it is akin to rotating the fuse relative to the wing in a scissor fashion like some of NASA's oblique wings. In other words the rudder only shifts the fuselage alignment, it doesn't create much continuous yaw moment needed for a turn, so it could be overpowered by asymmetric wing drag.
Jan 30, 2013, 12:31 PM
Registered User

# sideforce and Zagi

Quote:
 Originally Posted by vespa Any airplane can fly flat circles, even a flying wing with no fuselage or vertical tail whatsoever. All you need is something to create a yaw rate. For example a split aileron/drag rudder. Lateral area greatly improves the effect but it's the yaw rate that makes it possible.
Continuing along the same lines as post #93-- see this Zagi modified to have a tail boom and rudder. The tail boom is very thin (about 1/8 ") as seen from the side. There is only a very narrow (about 1/2" wide) vertical fin in front of the rudder, for hinging. Very little side area apart from the rudder.

Deflect the rudder left, and the nose yaws visibly some degrees (10?) but then just stays there. As long as you keep the wings level, there is NO visible turning tendency.

For the reasons outlined in post #47, I suggest that left rudder actually creates a slight right turning tendency with the wings level, and that a slight left bank is needed to hold the heading constant. I intend to confirm this by flying with an on-board video camera.

I'm not sure whether it will be better to hand-fly the plane, or to connect the ailerons to a heading-hold gyro, for the smoothest video that best confirms that a slight left bank is indeed required to hold the heading constant, when the rudder is deflected left.

Then I'll repeat with tip fins or with a sideforce generator ( vertical surface near the CG), and show that with the increased side area, left rudder now drives a left turn, and slight right bank is needed to hold the flight path straight.

All videos will include a view of a yaw string to show that left rudder does indeed always yaw the nose significantly to the left, relative to the flight path and airflow.

It's all about side area and sideforce!

Steve

### Images

Jan 30, 2013, 12:59 PM
Registered User

# sideforce

Quote:
 Originally Posted by vespa Sorry aero, a slip (sans rudder) occurs purely due to the bank angle and resulting lateral lift component. Therefore the slip is always in the direction of the bank. The yaw force produced by the outboard wing flying faster has nothing to do with slip or bank -- it is the result of, and in opposition to, the yaw rate.
I couldn't disagree more. The wing's lift force acts in the "upward" direction in the aircraft's own reference frame and thus has no more tendency to drag a left-banked airplane sideways to the left in the earth's reference frame (creating a slip), than it has tendency to to drag the same aircraft straight up relative to the ground (creating a skid), (Imagine a steep bank if you are having trouble seeing how this would create a skid.)

The slip we normally see in a constant-banked turn without rudder, is due mainly to the faster airspeed and higher drag from the outboard wingtip. This creates a yaw torque that can only be balanced by allowing the nose to be yawed slightly outboard, exposing the outboard side of the vertical fin to the flow (this is a sideslip.)

Quote:
 Originally Posted by vespa Control line planes are a great example of the required yaw rate. They work because a string attached to the wingtip is pulled aft in an effort to spin the plane around.
Couldn't disagree more... net yaw torque is zero in steady flight of a control-line plane.

Quote:
 Originally Posted by vespa Your hang glider example of yawning left with a rudder and turning right is plausible, but not for the reasons you think. What's happening here is that you have a tremendous amount of dihedral (low c.g., high sweep) so a tremendous effort (weight shift, ailerons) is needed to hold the wings level during a slip.
Vespa, I have to tell you that the hang glider has a mild anhedral effect not a dihedral effect. The leading edges have very significant downdroop in relation to the mean chord line. From a side view we can easily find a viewing angle where we see the top surface of the wing nearest our eyeballs and the bottom surface of the wing furthest from our eyeballs-- this is an anhedral geometry. To hold the wings level (or at any constant bank angle) with the left rudder deflected, I had to command left roll torque, more so at high airspeed, but a little bit even at low airspeed. I didn't follow your argument through but if it is based on a dihedral effect then it cannot be correct.

A key point is that a pilot hanging from a flexible strap connected to the CG of the wing, flying with a loose grip on the controls, acts like his weight is at the CG of the wing, not below the CG of the wing. If we are analyzing what happens when the pilot tightly holds himself in some fixed position without regard to how much muscle force is required to do so, then that is very different, and we have to analyze the pilot's mass as acting well below the wing. If we are interested in control forces not control positions, we have to analyze as if the pilot's weight is at the CG of the wing. But we're getting a little off track. Both in terms of control position AND in terms of control force, I had to command a roll input toward the left, to hold the bank angle constant when flying with left rudder.

Quote:
 Originally Posted by vespa But what really makes this work is that a hang glider has almost no fuselage, so if the rudder is effectively the majority of the aerodynamically useful fuselage, deflecting it is akin to rotating the fuse relative to the wing in a scissor fashion like some of NASA's oblique wings. In other words the rudder only shifts the fuselage alignment, it doesn't create much continuous yaw moment needed for a turn, so it could be overpowered by asymmetric wing drag.
No, even though the fuse area is small, the swept wings are creating a strong "weathervane" yaw torque. The rudder must fight this torque and so the fuse can not simply swing to the position that would streamline the rudder with the flow. The flow is still pushing hard on the left side of the left-deflected rudder. I know because the control cable was hand-operated. The rudder is generating PLENTY of yaw torque, toward the left. However, it is ALSO generating sideforce toward the right, and unlike a more normal aircraft, there is no fuselage to generate a stronger sideforce to the left. So the net sideforce ends up being to the right and the flight path curves to the right.

It's all about side area and sideforce!

With the Zagi, we could if we wanted, try to replicate the effect while setting the wing at the anhedral angle that requires NO aileron input to keep the wings level (total "effective dihedral" is zero). But this might introduce enough side area, compared to a totally flat wing, to make the effect vanish. I'm not sure. Maybe I'll have a chance to find out!

Steve
 Jan 30, 2013, 01:03 PM Registered User Ok enough for me to today. Maybe I've given some people something to think about.... Steve
Jan 30, 2013, 01:03 PM
Registered User
Quote:
 Originally Posted by ShoeDLG Most planes can be made to do a level "steady heading sideslip" (hold a constant bank angle with no turn). In most cases, all that is required is to apply sufficient rudder opposite the bank angle. An airplane with lots of side area, strong apparent dihedral effect, strong adverse yaw due to aileron deflection, and weak directional stability can even do a steady heading sideslip with no rudder deflection. In extreme cases, an aircraft like this could do a level turn to the right in a bank to the left, with no rudder deflection. This unlikely be a desirable flying quality.
I wasn't referring to adverse yaw
These models do nice flat very tight turns in the direction the rudder is deflected -
fuselage area and it's distribution is the key - no dihedral in the wings
Years back, we fought with layouts which had some couple effects which were hard to counter
A common one was a design which, given hard left rudder -would produce a diving right roll.
in trying to get good neutral aerobatic setups we have seen some really odd roll couple /yaw etc., results .
Jan 30, 2013, 03:15 PM
Registered User
Quote:
 Originally Posted by ShoeDLG Most planes can be made to do a level "steady heading sideslip" (hold a constant bank angle with no turn). In most cases, all that is required is to apply sufficient rudder opposite the bank angle. An airplane with lots of side area, strong apparent dihedral effect, strong adverse yaw due to aileron deflection, and weak directional stability can even do a steady heading sideslip with no rudder deflection. In extreme cases, an aircraft like this could do a level turn to the right in a bank to the left, with no rudder deflection. This unlikely be a desirable flying quality.
I didn't read Shoe's post in detail before. I still say that what he described is exactly what I observed in the full-scale Ka-6 sailplane, when I let the rudder float freely. I'll try it again the next time I get to fly one. Also I still say that sideforce from the fuselage and other surfaces is absolutely key to understanding how this could happen.
Jan 30, 2013, 03:30 PM
Registered User

# turns, slip, dihedral

PS thinking some more about how dihedral generates a roll torque--

Immediately after an increase in bank angle, or while the bank angle is increasing, yes there is some tendency for the plane to be dragged sideways through the air at a non-zero (and increasing) slip angle. The net force acting on the aircraft, including both lift and gravity, is indeed a horizontal force. This horizontal force could serve as the centripetal force that drives a non-slipping turn, but the aircraft has not yet had a chance to initiate the required rate of yaw rotation to allows there to be no slip (nose pointing directly into airflow), or allows there to be non-zero but non-increasing slip angle (yaw string deflected at a fixed angle). The flight path is starting to curve, but the aircraft heading isn't yet yawing around at the rate required to keep pace with the developing turn, and so the yaw string swings to the side at an increasing angle. The more the yaw rotational inertia, and the smaller the vertical fin or other equivalent surfaces, the longer it will take to establish the required yaw rotation rate to hold the slip angle constant. In other words it takes some time (perhaps a second or two?) for the yaw rotation to "ramp up" to the required value to keep pace with the turn rate, and during this time interval, sideslip is non-zero and increasing. This sideslip will interact with dihedral to create a roll torque toward the high wingtip-- this is a stabilizing effect that tends to stop the bank angle from increasing.

This process doesn't operate when the turn rate is constant and the yaw rotation rate is constant. Now yaw rotational inertia is irrelevant. To understand why we tend to see some slip even when the turn rate and yaw rotation rate is constant (in a turn with no rudder input or insufficient rudder input), we need to note that the outboard wingtip is moving faster, and creating more drag, than the inboard wingtip. If we are applying little or no rudder input, the high drag of the outboard wingtip tends to yaw the nose outboard, until sufficient yaw torque is generated by the sideways airflow against the vertical fin that the net yaw torque can be zero. At this point the slip angle will be constant, but not zero. At the same time, the slip interacts with dihedral to create a rolling-out torque, toward he high wingtip.

This process also operates when the bank angle is increasing and the yaw rotation rate is increasing, not constant. In other words it can operate hand-in-hand with the inertial effect described at the top of this post. So it is a more general effect than the inertial effect-- the inertial effect is limited to when the bank angle is increasing, but the differential-tip-speed effect is not.

Note that a slip driven by the inertial effect alone could never bring a plane with dihedral back to wings-level after a large disturbance. After all, if the plane was upset into a right bank and turn, then as soon as the bank angle begin decreasing and the turn rate started decreasing, yaw rotational inertia would tend to keep the nose swinging to the right at a rate more appropriate for the earlier, steeper bank angle. This would tend to swing the nose to point to the right of the flight path and airflow (relative wind), exposing the left side of the fuselage to the airflow, and creating a left-to-right airflow over the aircraft that would interact with dihedral to make a right roll torque, tending to increase the bank angle.

In other words the inertial effect tends to interact with dihedral to resist changes in bank angle, or slow down the rate of change in the bank angle, but this is not the same as returning an aircraft to wings-level after a large disturbance. This must involve a slip driven by the difference in airspeed, and drag, between the two wingtips in turning flight. Since dihedral is generally observed to produce a rolling-out torque (i.e. a roll torque toward wings-level), rather than a general resistance to changes in bank angle, this suggests that sideslip driven by the difference in airspeed between the two wingtips in turning flight tends to dominate over sideslip or skid driven by yaw rotational inertia.

In an aircraft with a short wingspan but a long fuselage, another effect tends to create a sideslip during a turn-- since the flight path and relative wind (airflow) are curved, if the vertical fin is streamlined to the flow, then there must be a sideways flow component across the wings. See http://www.av8n.com/how/htm/yaw.html#sec-long-tail-slip The curvature in flow between the wings and the tail, and the difference in airspeed and drag between the inboard and outboard wingtips, are really just two different manifestations of the fact that the flight path and the relative wind / airflow are all curved in turning flight.

Now, really logging off for the rest of the day!

Steve
Last edited by aeronaut999; Jan 31, 2013 at 08:33 AM.
Jan 31, 2013, 12:00 PM
Registered User
Quote:
 Originally Posted by aeronaut999 To understand why we tend to see some slip even when the turn rate and yaw rotation rate is constant (in a turn with no rudder input or insufficient rudder input), we need to note that the outboard wingtip is moving faster, and creating more drag, than the inboard wingtip. If we are applying little or no rudder input, the high drag of the outboard wingtip tends to yaw the nose outboard, until sufficient yaw torque is generated by the sideways airflow against the vertical fin that the net yaw torque can be zero.
With the outboard wingtip moving faster than the inboard wingtip in a steady turn, I think you can accurately say that the drag on the inboard wing will generally be different than the drag on the outboard wing. I don't think it is generally accurate to say that the the outboard wingtip will create more drag than the inboard wingtip.

For example, a glider flying near min sink airspeed will have more induced than profile drag. By slowing down the inboard wingtip in a steady turn, you are likely increasing the net drag on the inboard wing (and reducing the net drag on the outboard wing).

By increasing speed you might see a crossover to where you eventually see more drag on the outboard wingtip due to the turn. However the significance of this is probably limited by the fact that as you go faster at a given bank angle, the difference between the speed of the wingtips goes down in a level turn:

tip-tip_delta_speed = g*span*sin(bank_angle)/true_air_speed

so as you go faster at fixed bank angle, the difference in speed between the tips gets smaller.

At higher airspeeds (where profile drag is much bigger than induced drag), I think the yawing moment coefficient due to the difference in tip speed is proportional to:

span*g*sin(bank_angle)/true_air_speed^2

So although this effect probably changes sign above a given airspeed, it also becomes very small at higher speeds.
 Jan 31, 2013, 12:27 PM Registered User Shoe, re the above, I know of no case whatsoever where a circling aircraft, pilot giving no rudder input, tends to show a skid (yaw string blowing to inside) rather than a slip (yaw string to blowing to outside). Do you know of any such case? My hang glider always shows a slip in a constant-banked turn. At any airspeed. To the best of my recollection, sailplanes, flying with rudder floating free or rudder held centered, always show some slip. Of course, P-factor etc could create a skidding tendency but that's really not what we are talking about here... An aircraft that tended to skid rather than slip in a constant-banked turn would experience some very "interesting" effects-- anhedral would create a rolling-out torque, and dihedral would create a rolling-in torque! Steve
 Jan 31, 2013, 12:37 PM Registered User PS Shoe re your equations above, the difference in tip speed may get smaller as speed increases, but since lift scales according to airspeed squared, does the difference in lift get smaller as airspeed increases? I have always guessed the answer was "yes" but have never been completely sure... Steve