Low G and Rrpm decay rate

[Vance]

Vance, I imagine that a forward pulse on the stick actually could set off a PPO sequence in a vulnerable gyro -- one with a high thrustline and inadequate H-stab. A gyro with this kind of layout is a loaded mousetrap, waiting to spring.

A low-mounted H-stab, out of the propwash and with no incidence, will slow or even stop the PPO process -- even if the gyro has a small amount of HTL.

However, IMHO, it's even better to set the H-stab and/or prop thrustline so that the nose rises (stick fixed) when rotor thrust goes to zero. By being immersed partially or completely in the propwash, the H-stab will do this work even at zero airspeed -- at least it will when power is up, which is when it matters.

In terms of Mike G's curves, such a gyro would show a recovery of RRPM and G-load with the stick locked.

I find clients often stab at the cyclic when they encounter rough air.

One of my demonstrations to show them the futility of this is to take the controls and move the cyclic rapidly in a four inch circle.

It appears to me there is no effect other than the rotor slowing down a little.

I typically stop after two revolutions of the cyclic.

As I watch the video it appears to me that the accident pilot’s stabbing at the cyclic had little if any effect.

I defer to your knowledge and judgment.

I feel the best response in a gyroplane with a high thrust line in relation to the center of gravity when the response is not what I had hoped for is to pull the power to idle.

The high thrust line gyroplanes I have flown all pitched up when power was removed.

Most have a significant yaw when thrust is removed.

If there is no thrust I feel the thrust line in relation to the center of gravity becomes less important.

 

[Abid]

In my opinion based on my gyroplane flying experience a quick jab of the cyclic won't initiate anything.

In my opinion if he had three hundred pounds of thrust a foot above the center of gravity there are three hundred foot pounds of torque trying to rotate the aircraft around its center of gravity.

When the opposing rotor thrust is removed the gyroplane rotates forward without pilot input.

I feel this is an important distinction.

At one time I imagined if I kept the stick back a forward tumble was impossible.

I discovered this was not the case at the Cable Air Show flying a Cavalon performing my typical routine that I fly in the near centerline thrust Predator.

Near the top of a zoom climb I discovered I had no control of the rotor with the cyclic.

I was not able initiate a turn.

I closed the throttle and as she began to sink rotor control was restored.

I found this experience disquieting and suspect the delay of the rotation from the horizontal stabilizer is what gave me enough time to explore my options.

Upon reflection I realized the horizontal stabilizer is not very effective at low indicated air speed.

Hi Vance:
A quick down jab on the stick is definitely "not" good and should not be attempted specially if accompanied with high power setting. It does change the AoA of the rotor blades and although you may feel nothing happened at first, it has happened and it can create a blade stall beyond a limit. We do not want anyone to be testing those limits. In general, quick jabs in a gyroplane or rotorcraft are dangerous, in an airplane or trike are sloppy pilot technique and I am being "kind" in saying that. Smooth control inputs are always the way of the pilot.

Your deceleration on top of the zoom climb can also create low G for sure. You were likely unloading the rotor also with engine power starting to carry part of the burden of lift hanging the machine on the prop. Closing the throttle is exactly the right thing to do in this situation and great that you did. And yes the horizontal stabilizer did dampen and give you time to correct. Horizontal stabilizer is effective whenever "it" sees airspeed/airflow. That could be at zero indicated airspeed for you in the cockpit but 100 mph for the horizontal stabilizer due to local airflow from the pusher prop. Without that stabilizer you would have been toast.

Edit Add: In high thrust line gyroplanes it is possible to set things up so they don't behave as you state. AR-1, pitches slightly up when power is applied and pitches down when is reduced. It is still slightly high thrust line (4.5 inches).

 

[All_In]

Thank you!
U-ROCK, Mike!!!

 

[Doug Riley]

Fara: Yes, you're correct that a mildly HTL gyro can be made to have stable responses to power and G changes. The H-stab must be set up so that it catches some propwash and provides a download on the tail. The download supplies the nose-up response to low G, and (if the H-stab catches propwash) the H-stab's download will decrease when the throttle is closed, providing a net nose-down moment that helps preserve airspeed when you throttle down.

A HTL gyro that lacks this setup will flare when you close the throttle, as Vance reports. Annoying at best. Bensens did that.

In contrast, my tandem Dominator, with both an immersed H-stab and LTL, would OVER-compensate for throttle closure. It dropped its nose and sped up 10 mph or more. You needed to crank in some back-stick trim when entering a power-off glide to avoid breaking the sound barrier.

Compensating for HTL with tail download becomes impractical if the HTL is more than a few inches. You end up needing so much download that the tail boom must be half a mile long and/or the H-stab must be the size of Nevada. The download looks to the rotor exactly like actual weight, so it will cut into performance if you add much of it. E.g. a stock RAF needs something around 100 lb. of continuous HS download to achieve stable throttle and G responses. Who wants to haul 100 lb. of phantom weight around the sky?

The new wisdom about momentary "jabs" on the stick is interesting. Igor Bensen harped on this constantly, claiming that you MUST use short control jabs. He compared the "correct" use of the controls to a boxer's fist jabs. What's more, he wrote, you need to follow your initial jab-and-return with a quick OPPOSITE jab. He specifically warned against "deflect and hold."

In my very green youth, I absorbed Dr. B's advice. I therefore would use very rapid jabs when taking what little dual instruction I got, from Ron Menzie in his SxS gyroglider. He said "stop that." When taking a checkride with Ed Alderfer (for my BFI) decades later, I reverted to the same technique and got the same scolding. So what's the deal with control jabs?

Jabs are a standard technique for maintaining control of an unstable system. Professional test pilots use just this technique when testing a new aircraft that just MIGHT get away from them. Dr. B's gyros were not especially pitch-stable, what with slightly HTL, a fast, light rotor and no effective H-stab. So, out of caution, Dr. B. tried to teach us beginner gyro pilots a test-pilot technique.

Today's gyros, with heavier, slower rotors, tandem configurations and H-stabs, don't react much to Bensen-style jabs. We relax and use more leisurely inputs. In fact. my usual advice to gyro students has been "be lazy." But the pilot-victim in the PPO video was flying something pretty close to a Bensen, probably with the fast, light rotor that Bensen favored. I imagine that this pilot got a reaction to his forward jab, but that this "tripped" a PPO sequence that back stick then couldn't stop.

 

[Vance]

I can tell when my clients become comfortable flying as their control inputs become smooth and progressive.

Perhaps I am confused because I was not taught the jab technique.

With all of the gyroplanes I have flown smooth control inputs worked best for me.

Smooth control inputs seemed even more important in a light single place gyroplane.

 

[Tyger]

Magnis do certainly nose up when power is reduced, as Vance has reported. It's one reason to cut power to idle a little early during landing.
My understanding is that Magni have designed their H stabilizer, which is lower than most, to be as much out of the propwash as possible...

 

[Doug Riley]

Right, Tyger. If the Magni's H-stab had negative incidence, I suspect that the nose-rise at power reduction would be less or non-existent.

For a long time, I though all gyros experienced a mini-flare when power was cut (or when the engine quit!), because of "rotor drag." We all were to be very wary of this so as not get too slow upon engine failure.

Duh. The real cause of the nose rising is the location of the rotor thrustline ahead of the CG, necessary to counteract the HTL. If you counteract the HTL with something else, such as a downloaded stab, the rotor thrustline is free to settle into a line that points through, or even behind, the CG.

 

[Mike G]

Doug
could you be more specific when you say:

"In terms of Mike G's curves, such a gyro would show a recovery of RRPM and G-load with the stick locked."

Stick locked in which position? I'm not promising to try it but if it's feasible I might.

Tyger
You are not misreading the graph, you are simply not reading my first post which explains the lag between g and Rrpm decay rate and hence Rrpm😁.

Vance & Abid, all control movements are as smooth as possible when I'm doing this.

Mike G

 

[Tyger]

I understood what you said about the lag from change in G to decay rate, but I'd have thought the decay rate and rrpm would be synchronized on the new graph. Bad assumption, I guess.

 

[Doug Riley]

Mike G: I assumed (usually a bad idea) that you obtained these curves using the following control inputs: (1) induce >1G using back stick pressure, then (2) push forward to induce <1G, and then (3) remove the forward pressure to restore 1 G. We'd be interested in a description of your actual stick motions or pressures.

By "stick locked" I meant locking the stick mechanically at the #2 point. This can be quite dangerous if the aircraft is not designed specifically to nose up on its own when rotor thrust is reduced.

Also interesting is the fact that, stick free (hands off), a gyro with a Bensen-type rotorhead will, on its own, tend to move the stick back during low G. By "Bensen type," I mean an offset gimbal head, with excess* offset that is counteracted by one or more trim springs pulling down on the torque bar. This wholesome tendency is the reason for the "stick float" technique that pilots of HTL gyros are often taught.

The problem with "float the stick" is that it requires a reasonably pitch-stable airframe. The lower end of the trim spring is normally attached to the frame. If the frame has enough HTL instability, it will pitch over so much and so fast in low G that the trim spring will go slack. The spring will then fail to pull the rotor aft (hands off) as we'd like. Duane Hunn's "stabilator" worked around this problem by detaching the lower end of the spring from the airframe, and attaching it instead to a floating vane that lined up with the airstream.

In any event, a locked stick defeats the gimbal head, for better or worse.
____________________________________
*"Excess" means more than necessary to allow hands-off flight at cruise. Our usual 1" aft offset of the pitch pivot from the spindle axis is "excess" for a Bensen-class gyro. Without a trim spring, a 1" offset head will have a pretty strong forward stick pressure.

 

[Mike G]

Tyger
You have to remember that these recordings are from a device that is designed simply to give the pilot advanced warnings of a potential blade flap, behind the curve or bunting scenario.

The recordings are strictly for post accident (black box) analysis where this lag probably wouldn’t be of much interest and could be compensated for if really necessary. We are limited by the power of the computer chip doing the math, if we wanted to make a laboratory recorder for analysis (which Jean Claude would love) it would all be much more expensive.
I could fiddle the numbers to make a better looking graph but what you are seeing is what we get.

At the moment the GWS is proving to be remarkably accurate in it’s predictions, allowing me to make flapping take-offs that are really on the edge knowing that even though I’m within 1 or 2° of my 8° flapping limit the GWS is predicting that I won’t hit it. I am currently making the flapping alarm a bit more conservative, I think we’ve pushed it too close to the limit.

The bunting is a similar situation. I’m pretty confident that I’m safe but see no reason to go any further. The bunting alarm is probably still a bit too conservative compared to what it could be but pushing it nearer to the limit only gives the pilot taken by surprise less time to react.

The downside of making the alarms come on earlier to give the pilot more time is the risk of more spurious alarms. This has been one of the most difficult areas to address because when things like flapping and bunting happen they happen very quickly and the predictive software is limited by the computing power and the accuracy of the sensors.

Doug,
let me think about what you've written. I'll see if the video I took during these trials is any good, there was a lot of sunlight and the image wasn't that good.
Mike G

 

[Mike G]

Doug
As you can see it's not very aggressive but I was trying to maintain 0.6 g with a relatively fixed stick. You end up having to pull back because you're heading for the ground.

 

[Abid]

Fara: Yes, you're correct that a mildly HTL gyro can be made to have stable responses to power and G changes. The H-stab must be set up so that it catches some propwash and provides a download on the tail. The download supplies the nose-up response to low G, and (if the H-stab catches propwash) the H-stab's download will decrease when the throttle is closed, providing a net nose-down moment that helps preserve airspeed when you throttle down.

A HTL gyro that lacks this setup will flare when you close the throttle, as Vance reports. Annoying at best. Bensens did that.

In contrast, my tandem Dominator, with both an immersed H-stab and LTL, would OVER-compensate for throttle closure. It dropped its nose and sped up 10 mph or more. You needed to crank in some back-stick trim when entering a power-off glide to avoid breaking the sound barrier.

Compensating for HTL with tail download becomes impractical if the HTL is more than a few inches. You end up needing so much download that the tail boom must be half a mile long and/or the H-stab must be the size of Nevada. The download looks to the rotor exactly like actual weight, so it will cut into performance if you add much of it. E.g. a stock RAF needs something around 100 lb. of continuous HS download to achieve stable throttle and G responses. Who wants to haul 100 lb. of phantom weight around the sky?

The new wisdom about momentary "jabs" on the stick is interesting. Igor Bensen harped on this constantly, claiming that you MUST use short control jabs. He compared the "correct" use of the controls to a boxer's fist jabs. What's more, he wrote, you need to follow your initial jab-and-return with a quick OPPOSITE jab. He specifically warned against "deflect and hold."

In my very green youth, I absorbed Dr. B's advice. I therefore would use very rapid jabs when taking what little dual instruction I got, from Ron Menzie in his SxS gyroglider. He said "stop that." When taking a checkride with Ed Alderfer (for my BFI) decades later, I reverted to the same technique and got the same scolding. So what's the deal with control jabs?

Jabs are a standard technique for maintaining control of an unstable system. Professional test pilots use just this technique when testing a new aircraft that just MIGHT get away from them. Dr. B's gyros were not especially pitch-stable, what with slightly HTL, a fast, light rotor and no effective H-stab. So, out of caution, Dr. B. tried to teach us beginner gyro pilots a test-pilot technique.

Today's gyros, with heavier, slower rotors, tandem configurations and H-stabs, don't react much to Bensen-style jabs. We relax and use more leisurely inputs. In fact. my usual advice to gyro students has been "be lazy." But the pilot-victim in the PPO video was flying something pretty close to a Bensen, probably with the fast, light rotor that Bensen favored. I imagine that this pilot got a reaction to his forward jab, but that this "tripped" a PPO sequence that back stick then couldn't stop.

Hi Doug:
In airplanes the tail is always producing down force and the whole time they are flying the airplane is basically carrying ~105% of its weight because of the tail. Certain airfoils selected by the designer for specific AoA range where normal cruise is expected can make this better with very low pitching moments but in general airfoils with low or no pitching moments also tend to be less efficient so its all an art of compromise by the designer. Though 100 pounds of constant load for a gyro whose gross is only a 1000 pounds does seem like a bit much.

I honestly know nothing about Bensen gyroplanes or designs nor his instructional manual nor his thoughts nor really anything about him. Zilch.
I would never think that self learning in a flying machine like a gyroplane is a good idea at all and if I did make a manual for that, certainly quick control pulses would not be part a lesson plan. I don't know what Mr. Bensen or any of those guys were doing. Small smooth movements are all that is needed to properly test response of aircraft and its control interface. Slowly building up.

 

[Doug Riley]

Fara: The more we know about the Bensen era, the better IMHO. We are better able to progress if we know where we've been (and what we're trying to overcome!). We gyronauts all suffer from the awful reputation established by Bensen gyros decades ago. I find it helpful to able to explain exactly what was wrong back then, and how we've fixed it.

As for FW design: yes, exactly as you say. A bit of tail download, with CG ahead of the wing's center of pressure, is the normal FW layout. This arrangement results in the plane's dropping its nose when the main wing stalls -- a wholesome thing.

A traditional term for this arrangement is "horizontal decalege." The extra load on the wing from the downloaded HS does result in trim drag. Fixed wings being far more efficient than autorotating rotors in their lift-over drag performance, though, trim drag isn't a very big deal in FW planes. OTOH, gyro rotors have about a 5:1 (or worse) L/D, so a big HS download creates trim drag that really hurts. Much better to lay out your frame so that very little HS download is needed for G stability.

We gyronuts face the opposite problem from FW designers -- our danger zone is not high disk AOA, but sudden low disk AOA, resulting in low G and loss of RRPM. We'd like our aircraft to show some self-correcting traits, just as the FW self-corrects by dropping it nose and un-stalling the main wing.

It's nice that the solution for gyros is the same as for FW planes: put the CG ahead of the main lifting surface's center of pressure (what we call the rotor's thrust line) during steady flight. Horizontal decalege rides again!

In a HTL gyro, the rotor's thrust line will have to be AHEAD of the CG to hold the nose up, unless a downloaded HS does that job. Using rotor thrust to hold the nose up is undesirable and potentially unsafe, since loss of rotor thrust leads to a nose drop and the beginning of a PPO sequence.

If OTOH the rotor's thrustline is behind the gyro's CG, then loss of rotor thrust will result in a nose rise (from the tail download). This will tend to increase disk AOA and restore 1 G.

With HTL and a HS lacking download (such as the Magni), the rotor's thrustline will necessarily be ahead of the CG. Upon loss of rotor thrust, the nose will drop (which is what Magni pilots report in such situations). However, once the gyro assumes this nose-down stance, the HS starts to "bite" and to create a download. With enough airspeed and HS volume, the HS can arrest the developing PPO before it proceeds very far. We know from the Magni's lack of PPO crashes that this arrangement is a huge improvement over the no-HS, HTL setup.

Still, my own preference would be zero nose drop, and in fact some nose rise, stick fixed. Dominators do just that in low-G events: when transitioning to a Dom from older gyro models, I had to un-learn the old "float the stick" technique. Stick-fixed works well as long as you have a truly stable airframe.

 

[Abid]

Fara: The more we know about the Bensen era, the better IMHO. We are better able to progress if we know where we've been (and what we're trying to overcome!). We gyronauts all suffer from the awful reputation established by Bensen gyros decades ago. I find it helpful to able to explain exactly what was wrong back then, and how we've fixed it.

As for FW design: yes, exactly as you say. A bit of tail download, with CG ahead of the wing's center of pressure, is the normal FW layout. This arrangement results in the plane's dropping its nose when the main wing stalls -- a wholesome thing.

A traditional term for this arrangement is "horizontal decalege." The extra load on the wing from the downloaded HS does result in trim drag. Fixed wings being far more efficient than autorotating rotors in their lift-over drag performance, though, trim drag isn't a very big deal in FW planes. OTOH, gyro rotors have about a 5:1 (or worse) L/D, so a big HS download creates trim drag that really hurts. Much better to lay out your frame so that very little HS download is needed for G stability.

We gyronuts face the opposite problem from FW designers -- our danger zone is not high disk AOA, but sudden low disk AOA, resulting in low G and loss of RRPM. We'd like our aircraft to show some self-correcting traits, just as the FW self-corrects by dropping it nose and un-stalling the main wing.

It's nice that the solution for gyros is the same as for FW planes: put the CG ahead of the main lifting surface's center of pressure (what we call the rotor's thrust line) during steady flight. Horizontal decalege rides again!

In a HTL gyro, the rotor's thrust line will have to be AHEAD of the CG to hold the nose up, unless a downloaded HS does that job. Using rotor thrust to hold the nose up is undesirable and potentially unsafe, since loss of rotor thrust leads to a nose drop and the beginning of a PPO sequence.

If OTOH the rotor's thrustline is behind the gyro's CG, then loss of rotor thrust will result in a nose rise (from the tail download). This will tend to increase disk AOA and restore 1 G.

With HTL and a HS lacking download (such as the Magni), the rotor's thrustline will necessarily be ahead of the CG. Upon loss of rotor thrust, the nose will drop (which is what Magni pilots report in such situations). However, once the gyro assumes this nose-down stance, the HS starts to "bite" and to create a download. With enough airspeed and HS volume, the HS can arrest the developing PPO before it proceeds very far. We know from the Magni's lack of PPO crashes that this arrangement is a huge improvement over the no-HS, HTL setup.

Still, my own preference would be zero nose drop, and in fact some nose rise, stick fixed. Dominators do just that in low-G events: when transitioning to a Dom from older gyro models, I had to un-learn the old "float the stick" technique. Stick-fixed works well as long as you have a truly stable airframe.

Correct.

I guess for L/D of 5:1 is for the gyroplane not the gyro rotors. Because a 8H12 rotor by itself would be much better L/D. Its the combination of the fuselage and rotor that increases the drag.

I do not believe in complete disregard of tail and over emphasis of center line thrust in well intentioned but most likely questionable studies like the Glascow one for instance. I do agree that excessive HTL should be reduced for reasons you have mentioned. I still prefer a HS much father back than where say configurations like a Dominator put them because I believe there are advantages to tails farther back as well.

 

[WaspAir]

I guess for L/D of 5:1 is for the gyroplane not the gyro rotors. Because a 8H12 rotor by itself would be much better L/D. Its the combination of the fuselage and rotor that increases the drag.

Airplanes are no different in that regard. If I could fly my sailplane wing without the rest of the aircraft, the L/D would be nothing less than astonishing. Until somebody can figure out how to make the equivalent of a Northrop Flying Wing for a gyroplane (a rotor with no fuselage), it ultimately doesn't matter. The net gyroplane L/D sucks, and the theoretical rotor-alone L/D is still lousy compared to the corresponding theoretical fixed-wing-alone L/D, because it drains so much energy to spin the thing around.

There is some academic discussion of rotor vs. whole airframe to be found here for those who are curious:

https://www.icas.org/ICAS_ARCHIVE/ICAS2012/PAPERS/434.PDF

 

[C. Beaty]

The miserable efficiency of a gyroplane is the result of its rotor going 500 mph while the rest of it goes 50 mph.
Compounding the inefficiency of a gyroplane is the method of powering the rotor.
Rather than by a gear train with an efficiency of 90%+ as is the case of a helicopter, the gyroplane rotor is powered by a windmill whose wind is generated by a propeller with an efficiency of no better than 70%.

 

[Abid]

Airplanes are no different in that regard. If I could fly my sailplane wing without the rest of the aircraft, the L/D would be nothing less than astonishing. Until somebody can figure out how to make the equivalent of a Northrop Flying Wing for a gyroplane (a rotor with no fuselage), it ultimately doesn't matter. The net gyroplane L/D sucks, and the theoretical rotor-alone L/D is still lousy compared to the corresponding theoretical fixed-wing-alone L/D, because it drains so much energy to spin the thing around.

There is some academic discussion of rotor vs. whole airframe to be found here for those who are curious:

https://www.icas.org/ICAS_ARCHIVE/ICAS2012/PAPERS/434.PDF

Yes I just wanted to be a little precise in the wording for clarity. You can may be do a flying rotor like a flying wing but you would need the pilot to be very very hardy

 

[All_In]

Airplanes are no different in that regard. If I could fly my sailplane wing without the rest of the aircraft, the L/D would be nothing less than astonishing. Until somebody can figure out how to make the equivalent of a Northrop Flying Wing for a gyroplane (a rotor with no fuselage), it ultimately doesn't matter. The net gyroplane L/D sucks, and the theoretical rotor-alone L/D is still lousy compared to the corresponding theoretical fixed-wing-alone L/D, because it drains so much energy to spin the thing around.

There is some academic discussion of rotor vs. whole airframe to be found here for those who are curious:

https://www.icas.org/ICAS_ARCHIVE/ICAS2012/PAPERS/434.PDF

Thank you for sharing. I save it to the PRA library. The Section= Science, Subject= Flight Performance, and I'll need to read it to add the tags.

 

[Doug Riley]

I get dizzy easily. I'd barf if I rode around on a rotor-only gyro -- presumably sitting on a giant hub bar in the middle. Yecch.

There's no free lunch when it comes to trim drag. If you want a stable airfoil, you need to apply that download to the trailing edge. You can use an HS boomed out back (the normal approach) or you can make the download using trailing-edge reflex. In the latter case you're simply attaching your downloaded H-stab to the aft end of the wing (either fixed or rotorblade; the same rules apply; a rotorblade is simply a small flying wing).

Some designers try to cheat on this rule. My closest gyro-flyin' buddy died in the crash of a sailplane that was intentionally designed with a non-downloaded HS. The wing had drooped trailing edges to add even more nose-down moment. To keep the nose up, the designer put the CG way aft at 40% or more of wing chord. But this arrangement only works in a narrow range of airspeeds; either faster or slower, the wing's pitching moment varies as the square of airspeed, while the weight at the CG obviously doesn't vary at all. Bill might have been OK if anyone had told him about this aggressive setup -- but no one did. Sad.

We in Gyro-land had a rotor blade "design" foisted upon us some years ago with the same sort of drooped trailing edge airfoil -- an unworkable design.

Less radically, some experimental gyro rotor blades have been laid out with little or no reflex, in the hope that the blade's mechanical stiffness alone would keep the blade from "tucking." Yes, it would be nice to get rid of the trim drag caused by reflex. Rotorblades are pretty limber, though, and such a design amounts to yanking on the dragon's tail. Bensen put a little extra reflex into his metal blades at the tip, probably as a precaution against this sort of instability.

Fara, I agree that a farther-aft placement of the HS has real benefits. With a longer lever arm, the download in pounds (hence trim drag) can be less for a given number of foot-pounds of nose-up torque. Even better, the HS's role in dynamic stability (damping) increases as the square of the tailboom length.

OTOH, I disagree vehemently with Magni's decisions to (a) use no HS download and (b) place the HS so low that it catches no propwash. We need that propwash on the HS in these slow aircraft; HS lift varies as the square of airspeed. Immersion also makes the HS respond in proportion to throttle setting -- a desirable arrangement if you have HTL.

Finally, an HS centered in the propwash helps counteract torque roll. You can have a far-back, wash-centered HS as long as you have the rotor clearance, but the structure gets a bit unwieldy on a pusher.