Gyro stability Test Criteria and Methods - Introduction

Spring centered sticks provide pseudo stick forces. Although the stick forces are not from the rotor directly the direction of the forces are in the right direction. From a pilots perspetive, for the most part, it does not matter whether the forces are created artificially by the spring or actually from the rotor. Fly by wire aircraft use spring centred stick forces to provide feedback to the pilot quite effectively.

Chuck, Stick free neutral stability may be OK to fly as your experience with a non stick centered gyro shows but I am more concerned that stickfree instability and I think it would make for a very uncomfartable piloting experience. How was the original bensen with spindle control to fly?
 
Last edited:
I’ve never flown a spindle head gyro, Raghu.

In case you’re not familiar with its geometry, the spindle head used a spherical roller bearing as the main rotorhead bearing with the rotor teetered at the top of the spindle and cyclic control applied to the bottom. In forward flight with cyclic flapping, the rotor thrust vector passed forward of the pitch control axis.

In a story related to me by Dave Prater, he and Bill Parsons teamed up and built gyros together; Dave using the spindle head and Bill using the gimbel head.

Dave said that when flying together, he was all over the sky while Bill was rock steady. He couldn’t understand why until he tried a gimbel head himself. I’ve heard similar stories from others that have flown spindle heads.

With appropriate offset and trim spring rate, a riding lawnmower could be made to display some degree of stick free stability, the reason for my de-emphasizing the importance of stick free stability testing. Trim spring rate has a more important influence on stick free stability than most people realize.

I agree that while it’s important to have some sort of stable stick force because the pilot is always more or less in the loop depending upon how the stick is grasped, only stick fixed stability testing demonstrates intrinsic stability of the machine.

Let’s have no more of the huckster’s pitch; “It’s so stable it flies hands off!” The most important indication of stability is the question; How stable it is when a low time pilot has a choke hold on the stick?
 
Proposed ASTM Dynamic stability criteria

Proposed ASTM Dynamic stability criteria

Chuck and all,

Just FYI, the ASTM standard for Dynamic stability does require that the dynamic criteria be met with BOTH fixed-stick and free-stick modes.

The less than 5 second period criteria is mainly the DYNAMIC criteria that Raghu takes issue with and that I would like to address specifically at this time with suggested and supported alternative DYNAMIC criteria and “FTT”.

I much appreciate the detailed posts, but I’m kind-of hoping for a simple answer to my static QUESTION 11: Do we all agree that the proposed STATIC criteria appear to be appropriate and encompassing?

Thanks for the valuable posts - Greg
 
A qualified yes to question 11, Greg.

Some rotorcraft controlled by swashplate means have a mechanism for disengaging the trim springs; if that is the case with the McC J-2, it would not meet the stick force requirement with the springs disengaged, or; if the springs are too soft to fully meet the stick force requirement, an out of trim forward bias could be cranked in. Ron Herron knows the answer to this one.
 
The three static stability tests

The three static stability tests

The three static stability tests.
------------------------------------

1) Test one: G-load test (turn test)

What does this prove: Stick fixed AOA (or g-load) stability.

Why do we need this:This is the most important factor in dynamic stability. It ensures that the gyro is always pointing into the relative wind. This is what gives the speed holding tendency to a stable aircraft. This point is often not appreceiated in the case of gyro due to the existence of velocity stability.


2) Test two: Speed test

What does this prove: Stick free aoa (or g-load) stability.

Why do we need this: Helps with pilot feel and also ensures that in case the pilot is temporarily incapacitated and the stick is free, the gyro will remain stable. The effect of the offset gimball is the dominating factor here.

3) Test three: Power test
What does this prove: It does not prove any standard stability criterion.

Why do we need it: Primarily for safety. In case the engine suddenly fails or full power is added rapidly the gyro must not make any abrrubt pitch changes. In my view this is the least critical aspect. Also, it is debateble what the tolerence should be. The exact speed difference is not as important as the instant pitch reaction.

** Do we need any other static stability tests?

Answer: A test to check weather the RTV is very close to or behind the CG at all speeds will be useful. This static stability is termed rotor G-load stability

Why?

Because this stability dereivative influences the damping of the long period mode. If it is unstable then the long period looses a bit of its damping. At an extreme this may mean that the long period mode is actually unstable.

Also, probably more importantly, rotor G-load stability acts aginst a PPO tendency.

How do we perform this test?

An obvious way is to photograph the gyro in flight and work out the RTV/ CG geometry. This method could work but I have no idea how accurate it would be.

Another method ( very impractical) is to carry a sack full of potatoes at the CG and drop it during flight. If the instant reaction is nose down then we have rotor G-load stability, if it is
nose up we have negaative G-load stability. The key is the instant reaction this may be hard to monitor. In both cases the gyro will nose up in the more medium to long term.

I guess you could ignore this test and just test for dynamic stick fixed stability and verify that the long period mode is sufficiently damped.
 
Question #11

Conditions--Static longitudinal airspeed stability must be met at the following power and airspeed conditions: Trimmed at:
(a) Steady altitude at MPRS,
(b) Full power at the lesser of VH or of VNE,
(c) Engine idle at MPRS, and
(d) Engine idle at 80 % VNE.

I suggest to add the following power speed combinations-
- Full power at 0.9 VMIN
- Full power at VX (Best climb airspeed)
- Engine idle at 0.9 VMIN

I would also add the above combinations to the G-load stability testing.

I would also expand the envelope for the static power stability tets. All the proposed tests Occure at MPRS. I would expand the range as follows:

- At the cyclic stick position established in level flight at VMIN at full power, a change in power from full power to engine off shall not result in a change in a steady state trimmed airspeed of more than 20 % from VMIN.

- At the cyclic stick position established in descent at VMIN at idle power, a change in power from idle power to full power shall not result in a change in a steady state trimmed airspeed of more than 20 % from VMIN.

-At the cyclic stick position established in flight at 0.9 VNE at full power, a change in power from full power to engine off shall not result in a change in a steady state trimmed airspeed of more than 20 % from 0.9 VNE.

Udi-
 
raghu said:
...Also, probably more importantly, rotor G-load stability acts aginst a PPO tendency...

Raghu - did you mean “against PIO tendency”? Aligning the engine thrust line with the CG is the only way I know of to reduce PPO tendency.

Also, for the benefit of having a clear terminology - when you say "rotor G-load stability", you are talking about the rotor specific derivative of the aircraft G-load (or AOA) stability, correct? I believe that although you can separate the two on paper, short of removing the HS from the machine, I don’t think you can test in flight rotor G-load stability. Every test, including the sack of potatos test, would include the airframe/stab effect on total G-load/AOA stability.

Udi-
 
Back at it. To summarize the Phugoid Mode - low freq (large period), lightly damped, alpha (AOA) assummed constant. Fn ~ 1/V , dp ~ Cd/Cl
Typical period ` 20+ sec, Houston's Magni ~ 18 - 19 sec (flt test). There is not much to "design" here. Easily controlled/trimmed in cruise fltt. No impact on Maneuver Flt.

SHORT PERIOD

This is the bread and butter of longitudinal stability. The same 2nd order eqns apply but the variables having strong influences on Fn and dsp (short pd damping ratio) are different. The static stability design factors have a large influence on short period characteristics.

Variables include AOA, deck angle, vertical g, and rotor rpm (rrpm). For simple testing, deck angle is the easiest to observe. Here's where rotor / airframe coupling discussions can enter. I will assume (?) for small pertubations about a trimmed condition, the rotor will follow the airframe with maybe some lag. For the math and flt test we define/set AS and Alt constant.

What are the variables?

Fnsp ~ Cma, 1/(Iyy)^0.5 , AS

dsp ~ Cmq, 1/(Cma)^0.5 , 1/(Iyy)^0.5

where

Fnsp - short pd. undamped natural freq.
Cma - restoring moment chg due to AOA chg, sign is neg by convention for statically stable AC. This is a strong term for both freq and damping and per static stability criteria affected by HS size and location as well as CG vs RLV
Iyy - here's your mom. of inertia about longit. (y) axis by convention.
Cmq - pitch rate damping - guys are getting creative naming this one. Moment change due to pitch rate or AOA rate chg (assuming rotor follows fuselage).

Changes in SP characteristics:

AS ^ - Fnsp v( v = down or lower)
Iyy ^ - Fnsp & dsp v Increassing mass about Y axis decreases damping &
freq
Cma ^ - Fnsp & dsp ^
Cmq ^ - Fnsp & dsp ^
Static Margin (cg vs RLV) v - Cma v and Fnsp v

Just like the Phugoid, thee characteristic eqn consists of a damping term
2 dps Fnsp and the frequency term Fnsp^2.

dsp - damping ratio can be described by the number of overshoots of deck angle, vertical g, AOA, or rrpm. For dsp between 0.1 and 0.7, dsp can be approximated by dsp ~ 0.1 (7 - #peaks). These are peak overshoots until totally damped.

Here's some common terminology wrt damping ratios:

Overdamped - d > 1
Critically damped - d = 1
Underdamped - 0<d<1 this is where we'll usually operate ( 0.1 - 0.7)
Undamped - d = 0
Negatively damped d < 0


2 dsp Fnsp - damping term which can be described as inversely proportional to the time to damp. We can measure this in flt test. Sometimes we refer
to what is called a time constant (Tau) = 1/dFn where one time constant = 63% of the final original value. Sometimes we use time to half amplitude -
T1/2 = .69/dFn.

I had an opportunity to fly a highly augmented Lear jet belonging to the Calspan Corp out of NY. We could "vary" many stability derivatives to see how flight characteristics changed. This was an outstanding teaching aid. Quoting from their manual:

It is important to remember iin stability and control work that neither frequency nor damping ratio, nor time to damp alone, will be sufficient to predict handling qualities. It is the combination of Fn and d and consequently 2*Fn* (time to damp) that is meaningful

It's difficult, if not impossible for specific cookbook numbers to apply to the LSA standards but we CAN talk about ranges of damping ratios and natural frequencies which are based on a large data base of pilot surveys while flying this augmented airplane. These same "goodness" ratings were bounded and put on a logarithmic chart with Fn vs d. The best tested boundary had d varying from 0.4 to 1.2 and Fn varying from 0.5 to 1.2 cycles per sec. More qualitative descriptions of these damping and freq ranges later.

I will try to get this graph thrown up on this forum. I contend the same favorable goodness, or pilot in the loop, ranges apply to sport gyros.
I'll later try to discuss simple FTTs for the short period including what some may be seeing as periods less than the phugoid but longer than a "typical" short period response.

I will also try to tie it all up in a Man. Flt discussion. Like Greg G. said all we may find is we need an adequate tail, reasonable cg wrt rlv and thrust line. I'm just trying to help put some #s (or reasonable ranges of numbers)on values we're not too used to defining. When all is said and done, it will be flt test and the inputs of many gyro pilots to come to a consensus. It will be as much qualitative as quantitative.
 
Mike, in the end, I'm waiting for you guys to boil it all down to where the back yard builder can apply the standards. Chances are, there are going to be a lot more backyard built machines than manufactured ones for sometime. In my mind, this whole exercise is about safety.

What the ASTM sub-committee is doing is establishing standards for the manufacturer and even there I think we are going to have to be careful that they are as easy to understand and implement as possible. Backyard builders have been known to become manufacturers, you know! This means minimum calcs and/or canned software to handle them.

And I think everyone agrees that flight testing has to be done in order to prove the paper design and the final construction.
 
Udi said:
Also, for the benefit of having a clear terminology - when you say "rotor G-load stability", you are talking about the rotor specific derivative of the aircraft G-load (or AOA) stability, correct? I believe that although you can separate the two on paper, short of removing the HS from the machine, I don’t think you can test in flight rotor G-load stability. Every test, including the sack of potatos test, would include the airframe/stab effect on total G-load/AOA stability.

Udi-

Not quite Udi! I have defined this earlier but here it is again:

Rotor G load stability: Keeping all other parameters fixed, if an increase in rotor RPM (RRPM) causes a nose down reaction of the gyro then this derivative is stable. Conversely, if the gyro pitches up due to an increase in RRPM the derivative is unstable. Note, what we are interested ( as in all static stability derivatives) is the instantaneous reaction.

Based on this it is obvious to see that the only way to make this derivative stable it to place RTV aft of CG. Consequently, the photo test will work. Also, the sack of potatoes will work because when you drop the sack there is an instantaneous increase in G -load- it does not matter whether the g-load increase is due to a drop in weight or an increase in RRPM. As you are looking for an instantaneous reaction the speed of the gyro does not change so none of the aerodynamic forces change and so HS or no HS is irrelevant. The catch though as I said in my earlier post is whether you can effectively spot this instantaneous reaction and not confuse it with the more medium trend.

The other test (particularly g-load of the entire gyro or AOA) are structured cleverly that they indirectly measure static stability qualities and thus avoid measuring the instantaneous reaction. The photo test I guess is indirect as well, but the sack of potatoes is a direct test and so its practicability may be an issue. You could do the same test by un banking from a turn, but again you should look for the instantaneus reaction and not other medium/longer term reactions. For example in both the unstable and stable case the medium term reaction is that the gyro will climb and so it will pitch up.
 
Last edited:
I’ve performed the potato drop test, Raghu and was unable to spot the instantaneous reaction.

Well, not exactly. I didn’t actually drop a sack of potatoes but the rotor rpm can be cranked up by holding a steep turn until rotor speed stabilizes and then abruptly rolling out and observing response.

All I’ve ever observed was the medium period response; the gyro simply climbs until excess rotor energy is consumed.

We’ve also had a skydiver jump from a gyro.

The young lady skydiver was an employee of mine at the time and was after unusual entries for her logbook.

The pilot to be, Lloyd Poston, was concerned as to what the reaction of the gyro might after she jumped. The spot on the gyro from which she jumped was very nearly on the CG, so I assured him (I didn’t know with certainty but OTH I wouldn’t be flying the thing) that all he need expect was a gentle climb until excess rotor rpm was dissipated.

As it turned out, that’s all that happened. Lloyd didn’t observe any significant pitch change.

The gyro was a Lycoming powered machine with fuselage pod made from a surplus drop tank. We bolted a plank to the side of it to serve as a running board of sorts and a handrail crosswise to the mast to prevent her from falling into the prop. She was instructed to climb out, sit on the plank and simply roll off from a seated position.

They rehearsed the exit routine on the ground and the actual jump went off without a hitch.
 
Dean_Dolph said:
Mike, in the end, I'm waiting for you guys to boil it all down to where the back yard builder can apply the standards. Chances are, there are going to be a lot more backyard built machines than manufactured ones for sometime. In my mind, this whole exercise is about safety.

What the ASTM sub-committee is doing is establishing standards for the manufacturer and even there I think we are going to have to be careful that they are as easy to understand and implement as possible. Backyard builders have been known to become manufacturers, you know! This means minimum calcs and/or canned software to handle them.

And I think everyone agrees that flight testing has to be done in order to prove the paper design and the final construction.

Hi Dean,

You're exactly correct in my view.

I guess what we're trying to find are design guidelines and associated "ranges of numbers" which is one way to describe a safe design. All of us want to keep it simple, but I think that a manufacturer may be interested in a range of damping ratio/frequency combinations which might emulate a Dominator. Or he may be more interested in a cruising XC machine, possibly like the Sparrow Hawk. Why are the damping and natural frequencies of these 2 machines different (if at all)? How do I modify the physical airframe/rotor to give me higher damping, higher Fs/g, lower frequency, etc. This is where we go if we want to dig a little below the surface.

The backyard builder, if he's smart, will jump on the shoulders of giants and could physically build a good design simply by inspection of successful designs, using the right materials. All the great design "gouge" over the years (possibly written in blood) might be collated and used on a machine that "looks about right". We've built a lot of great aircraft over the years before we analyzed them to death.
 
Getting closer - to consensus?

Getting closer - to consensus?

Good posts – keep them coming!

Chuck, we should not have to worry about swash plate gyros – the ASTM LSA standard at least is limited to fixed pitch rotors (or ground adjustable pitch rotors) – no collective controls! – Complexity eliminates in-flight pitch adjustments (for now).

Raghu, thank you for your comments on the three static criteria and tests. I think you are saying, that in lieu of the Rotor G-Load criteria (sack of potatoes test), the “turn” test criteria (what we have so far called static G-Load or Maneuvering stability) should suffice to cover this – even if it also covers static AOA stability as well. In my simple mind, I just feel that if we can show that the airframe nose reacts the correct direction to a change in G-load, we have covered this really important static stability issue.

Raghu, you also commented that the “Static Power Stability” criteria and test is not really a stability issue. But then, Udi is suggesting making this criteria more stringent. In my thinking, this IS actually a stability issue – that airspeed or AOA is stable wrt power changes – for the safety reasons you point out But also in my simple thinking, this is the issue that gets right at the heart of a possible PPO from the “popular” un-balanced high prop thrustline crowd. This criteria tells us that if the Sum of Static Moments is suddenly changed by the loss of RTV thrust, the machine will not be “pushed over” into a PPO. From a study of the static moments on a gyro, if an increase of power causes the nose to lower (increased trimmed airspeed), a loss of RTV thrust on that machine will also cause the nose to lower – possibly initiating the popularly classic PPO buntover. This criterion originally started out addressing only the nose lower with increased power issue, specifically to avoid the likelihood of a PPO if the rotor loses its thrust (or in some people’s more simple visualization – the rotor drag goes to zero). The criterion was expanded then, when the FAA Rotorcraft Directorate expressed to us that they expected gyroplanes to be “Power Stable” in both nose-down and nose-up directions! With no objections, I think this power stability criterion is appropriate, but I do admit, the tolerances are somewhat a guess that we would hope to refine with more data. My thought is that at 10% or a 20% trimmed airspeed change would not result in enough of an instantaneous attitude (cyclic) change that might stall a blade enough to vilently flap – precess stall. For now, the major gyro manufacturers have had no objection to the tolerances suggested at this time.

As for Udi’s suggestion to test this more vigorously at more speeds and power ranges, I can put that to the ASTM subcommittee for consideration. But, to me, this is not realy necessary to eliminate our initial concern for a PPO potential.

Mike and Dean, We really do intend that this criteria and the testing methods be simple and inexpensive enough to be accomplished by the individual gyro owner as well as designers/manufacturers. This is essential in order to achieve acceptance and understanding by gyro pilots in what is required for gyros to be stable/safe. But, also, we do not want to add undue costs to a manufacturer to verify compliance with the standard. It is my goal that we might be able to set the criteria and testing methods so that we can readily communicate those to the regular gyro pilot – and have them understand them, respect them and require them in what they buy or fly! The design details in what Mike and others are providing will be valuable for designers and manufacturers and more technically adept builders to help them meet, or adjust to meet, the criteria. And, some of the details will help us in determining – with time and data just what criteria and flight testing elements are necessary for a stable/safe gyro. For instance, we would want to watch the data to determine if just “damped” long-period dynamic responses are adequate for safety. Time and data may tell us we are overly-conservative in some areas, where our intuition had possibly set too liberal criteria in other areas. Our intention is to be biased on the safe side to start with right now!

When we have wrung this out as far as anyone cares to, I will present this material to the ASTM Gyroplane subcommittee for consideration of refinements to the existing standard. Your posts are great material for consideration.

DYNAMIC criteria: I think we still need to find the DYNAMIC criteria that would avoid PIO. I think Mike is agreeing that the criteria can be expressed in the shortest oscillation period allowed, and the damping rate for that oscillation. I think everyone is in good agreement that the long-period oscillations just need to be “damped. But, where is the breakpoint between the short-period criteria and long-period? And, how do we test for the criteria?

I like Rahgu's arguments that there might be some "residual intermediate" natural oscillations in the 7-10 second period range that might be THE concern. Is 10 second period the breakpoint, below which any oscillations should be damped to x percent within x cycles?

Thanks, and keep up the valuable input - Greg
 
gyrogreg said:
As for Udi’s suggestion to test this more vigorously at more speeds and power ranges, I can put that to the ASTM subcommittee for consideration. But, to me, this is not realy necessary to eliminate our initial concern for a PPO potential.

My intention was not to make the tests more stringent. I thought that the proposed tests do not cover the entire flight envelope. In my opinion, these tests should cover the entire range of normal operating conditions. Why leave entire regions of the envelope untested? I, as a customer, would like to know that my FSA gyroplane was tested before I fly it. This is no longer experimental aircraft!

One example in particular is engine out at airspeed VNE. Some gyros are at the end of the cyclic range at this airspeed and may not respond favorably in the event of engine out.

I say, cover the entire normal flight envelope. If you want to provide wider tolerances for airspeed at the edges, fine. But at least test it.

Thanks

Udi-
 
Power stability wider range

Power stability wider range

Udi,
I don’t really disagree with the wider range for Power stability testing. Your proposed speeds and power changes actually target some really critical issues – the slow speed at idle power then increasing to full power, seems to me to specifically target possible PPO tendencies at reduced g load at the push over the top of a zoom. The high speed with full power reduced to idle also aggressively targets the potential for bunt upon sudden power loss on a gyro that artificially holds the nose high under this condition.

Development of this criteria in the standards group was a bit dicey, as the whole issue of “unbalanced” prop thrustline was a bit touchy for some, and poorly understood by others. I’m still not sure how the group would feel about this – might be taken by some as a direct attack on CLT designs. Of course, a CLT should not have any problems with this issue, but some very low prop thrustlines are considered CLT by many. Your suggested speed and power change ranges would certainly target gyros with highly unbalanced prop thrustlines – high or low. And, that is probably a good idea – I just have not been comfortable pushing that issue so strongly because I have run into a lot of resistance when this issue might suggest some “perfect” designs aren’t so perfect. We’ll see how the suggestion goes, but I really feel the problem machines will be identified with the more moderate criteria.

Thanks, Greg
 
Would Udi's test not target unstable machines, not thrust line offsets?

I am not confident I am keeping up with the discussion, but I belive (think, guess?) udi's suggestion would be passable by a large thrust offset machine with enough coping mechinisms (tail)
 
That's right, eruttan. Some of the additional tests I have suggested are targeted at the edges of the envelope where the gyroplane must be well balanced for it to pass the test. Any gyro that is passing these tests may be considered pitch and power stable, even if it has a thrust line offset along with a correcting mechanism.

I know of at least two gyroplanes that would pass these tests right now, possibly three.

Udi-
 
Udi said:
My intention was not to make the tests more stringent. I thought that the proposed tests do not cover the entire flight envelope. In my opinion, these tests should cover the entire range of normal operating conditions. Why leave entire regions of the envelope untested? I, as a customer, would like to know that my FSA gyroplane was tested before I fly it. This is no longer experimental aircraft!

One example in particular is engine out at airspeed VNE. Some gyros are at the end of the cyclic range at this airspeed and may not respond favorably in the event of engine out.

I say, cover the entire normal flight envelope. If you want to provide wider tolerances for airspeed at the edges, fine. But at least test it.

Thanks

Udi-

Hi Udi,

We have tossed the term "flt envelope" around quite a bit. I don't know how we are defining "entire flt envelope". Most experimental gyros don't come with well defined published flt envelopes. I have not seen much discussion of what really should/might define a VNE limit. Just because we can reach a high calibrated airspeed, for example, doesn't mean we want to be there, maybe due to cyclic boundaries or pitch sensitivity.

IOW, we may want to define our envelope/limit to a point reasonably short of falling off the edge of the earth. Sound flt testing is conservative and works slowly from the heart of any proposed envelope. As we move closer to the boundaries, the testers say, "that's fast enough because...., here's the published limit - don't exceed it because..........."

Mike
 
Mike;
I humbly disagree.

If I sell a plane that can do 150MPH, but has undesirable traits that show at the speed. Not to mention break Sprot Pilot perhaps. I might put a Vne of 100 on it. It test it in this envelope and it is good. It is safe in this envelope. If a pilot goes beyong my tested and published envelope, thats pilot error. It is on the pilots head.

Conversly, If the standard gets into , "well this aircraft can go 150, so you have to test it there", we are beyong the scope of assureing stability and safety. We are forcing people to test where no one would want ANYONE to fly.

If a designer/builder/manufacturer says Vne = X, thats what it is.
Getting into the reasons for it is irrelivant and beyong the scope of the ASTM standard.

As long as flight testing is done for the entire published envelope, then the craft is tested.

I cant see anything other that the Published flight envelope working. But I will listen to the counter point.
 
I agree. I believe that we should test stability to the published flight envelope. If some U BEAUT designer can have his machine meet the standards with 20' of rope out the back with a brick tied to it, then it passes!!!!

Aussie Paul
 
Top