Gyro stability Test Criteria and Methods - Introduction

[gyrogreg]

This thread is to discuss and seek peer review on flight test methods to determine the quantitative static and dynamic stability of a gyro. This thread presents, for peer review and education, the currently proposed stability standards in the ASTM gyroplane standard, and the testing methods currently envisioned.

There are lots of forum discussions about theories of gyroplane stability. Some of these get very intricate. I have been one who suggests that these theories, with all the complicated interactions in a gyro, may not be fully defined or proven. I also espouse that there are no “cookbook” solutions to gyro stability. I suggest that such theories and “cookbook” may be valuable concepts for initial designs, but that the final determination if the final configuration is actually stable, must be determined by flight testing.

Let me also mention that “flight testing” does not mean going out to see how if “feels” or handles wind, or how easy it is to fly. Such “subjective” determinations, especially when provided by pilots experienced in flying that machine, are not true indications of how that machine might fly in the hands of other, less experienced pilots, in other realms of the flight envelope (speed, power, loading, wind turbulence, etc.) Often, it is the experienced pilot who actually stabilizes the gyro – it is very difficult for an experienced pilot to avoid commanding stabilizing control inputs because they may be so unapparent to even that pilot. Subjective pronouncements of “stable” can be very misleading and often, even unconsciously, biased.

The only true way to rate a gyro’s stability performance is to perform objective flight tests that basically isolate pilot flight inputs from the equation. This is what professional test pilots do, and what aircraft standards specify as criteria. Such criteria should be readily comparable to similar testing data from other machines so that objective comparisons and decisions can be made.

The object of the stability criteria and the testing methods is to assure, as well as practicable, that a particular gyro would not exhibit flight and stability characteristics that would lead to accidents. We all know that there are certain pitch stability accident types (mostly PIO and buntover) that have been the historic risk of flying gyros. The intent of these stability criteria is to address the stability issues that may lead to such accidents, and to help identify what those issues might be in any particular gyroplane. The intent of these criteria, and associated testing methods are to help gyro pilots evaluate and understand the potential stability-related risks in any gyro. These criteria may not be so confining or strict as some older aircraft stability standards, but we do want them to effectively address issues that might prevent likelihood of these traditional accidents.

These stability criteria, attempt to establish objective “flight performance” criteria. The criteria should be based on “results” of flight testing. The criteria and test methods should isolate, as much as practical, any pilot intervention or bias on the results. It is often tempting to base criteria on a “prescriptive solution”, rather than objective and analytical testing results. These criteria are intended to avoid, as much as possible, “prescribing” what a design should look like. These criteria should not care what a design looks like, they should only care about how it works – “Flight results testing”.

If possible, these criteria and test methods should be phrased in terminology readily understandable to a normal pilot with normal training – not confusing terminology or intricate theoretical descriptions.

These test methods are intended to be relatively simple and inexpensive to accomplish – require no exotic sensor or recording instrumentation. This is to both allow most gyro pilots to actually conduct some testing themselves, and to minimize costs of manufacturers who wish to verify compliance with the standard.

Below are individual posts addressing the four stability areas, or criteria, that the ASTM gyroplane standards subcommittee has proposed. These are divided into basically the following four areas:

Static Airspeed Stability
Static G-Load (or maneuvering) Stability
Static Power Stability
Dynamic Stability

The following posts initially address these four areas. For this initial thread, I hope we attract constructive peer review and comment. We are also hoping to refine some testing methods for each of these stability criteria. More information is included in these first several posts below. If extensive discussions on any particular effort evolve, to avoid confusion from too many subjects, I will break that area off into a new thread for more specific discussion.

[gyrogreg]

Standards Criteria:

Static Longitudinal Airspeed Stability:

The longitudinal control must be such that, with constant engine power, an aft force and movement of the cyclic control is necessary to achieve an airspeed less than any available trim airspeed;

The longitudinal control must be such that, with constant engine power, a forward force and movement of the control is necessary to achieve an airspeed greater than any available trim airspeed; and,

The control force gradient must not reverse during any progressive application of control movement at airspeeds greater than VMIN up to VNE.

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.

Explanation: This is the classic “test pilot” criteria for static airspeed stability – the slope of the control vs. airspeed curve must be positive. This means that a steady higher airspeed must not require an aft stick or pressure, and steady lower airspeeds must not require a forward stick or pressure. This criteria results in the tendency of the aircraft to self-restore itself to “trimmed” airspeed upon a disturbance to airspeed. A tendency for airspeed to diverge (to higher or lower airspeeds), without active pilot intervention, could result in severe aircraft divergent higher or lower airspeed. Aircraft with statically unstable airspeed static characteristics, requires constant pilot attention to maintain “trimmed” airspeed upon a disturbance to airspeed.

Suggested testing method:

Perform tests at different trimmed airspeeds, and loadings, in calm air

Increase airspeed above trimmed airspeed – note that forward stick pressure and position is required to maintain the increased airspeed.

Decrease airspeed below trimmed airspeed – note that an aft stick pressure and position is required to maintain the decreased airspeed.

Repeat test for different trimmed airspeeds from Minimum Power Required Airspeed (MPRS) up to Vne.

Repeat test for various allowed aircraft loading – gross wt. and CG limits

Note 1: Stick position can be determined or measured by a retractable tape measure secured to a solid point on the instrument panel and extended to the cyclic stick. Any other reasonable method to determine stick position deviation can also be used.

Note 2: Stick pressure or stick position is not required to be identified quantitatively other than the direction and that some force and stick movement is required.

[gyrogreg]

Standards Criteria:

Static Longitudinal Maneuvering (G-Load) Stability:

The pitch control forces during turns or load factor maneuvers greater than 1.0g must be such that an increase in load factor is associated with an increase in aft pilot control force, and a decrease in load factor is associated with a decrease in aft pilot control force.

Conditions--Static longitudinal maneuvering 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.

Explanation: This is the classic “test pilot” criteria for maneuvering (G-Load) stability. This criteria results in the tendency of the aircraft to self-restore itself to 1g rotor loading upon a disturbance to rotor G-loading. A tendency for G-Loading to diverge (to higher or lower G-Loadings), without active pilot intervention, could result in severe rapid aircraft divergent pitch up or pitch down. Aircraft with statically unstable maneuvering static characteristics, requires constant pilot attention to maintain 1g loading upon a disturbance. Such G-Load instability upon rapidly decreasing and divergent G-Loading is thought to be a critical element in gyro buntover incidents – especially in the hands of an inexperienced pilot.

Suggested testing method:

Perform tests at different trimmed airspeeds, and loadings, in calm air
Steady the airspeed in straight and level flight.

Maintaining the same airspeed and power setting, bank the aircraft into an approximate 30 degree spiraling bank (to increase rotor G-Load) – steady the airspeed at the original trimmed airspeed.

Note that aft stick forces are required to maintain the original trimmed airspeed in the spiraling descent.

Repeat test for different trimmed airspeeds from Minimum Power Required Airspeed (MPRS) up to Vne.

Repeat test for various allowed aircraft loading – gross wt. and CG limits

Note 1: Stick pressure is not required to be identified quantitatively other than the direction and that some aft force is required.

[gyrogreg]

Standards Criteria:

Static Longitudinal Power Stability:

At the cyclic stick position established in level flight at MPRS at MPRS power, a change in power from MPRS power to full power shall not result in a change in a steady state trimmed airspeed of more than 10 % from MPRS.

At the cyclic stick position established in level flight at MPRS at MPRS power, a change in power from MPRS power to idle power shall not result in a change in a steady state trimmed airspeed of more than 10 % from MPRS.

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

Explanation: This criteria establishes that the “Sum of Static Moments”, or balancing of the HS to other static moments such as prop thrustline offset and drag line offset, reasonably prevents rapid aircraft pitching upon sudden power changes. This criteria also avoids severe commanded stick inputs upon power changes, such as during power reduction for flare or power application to “save” a botched landing. This criteria addresses the concerns for “unbalanced” high or low propeller thrustlines that might result in a buntover upon sudden rotor thrust or G-load decrease. This criteria results in a tendency for the aircraft AOA attitude and airspeed to remain relatively constant upon power changes – a characteristic the FAA suggests to avoid pilot over-reaction upon power changes. This criteria also initially assures that the CG/RTV relationship will not severely change under different power applications. Because this CG/RTV relationship is a critical factor in the first two static stability criteria above, this test Power Stability testing is recommended to be conducted, and any deficiencies corrected, prior to conducting the other stability tests.

Suggested testing method:

Perform tests at different loadings in calm air

Steady the airspeed in straight and level flight at MPRS.

Maintaining the same fixed stick position, increase power to full power.
Note that airspeed does not increase or decrease more than 10% from the original MPRS.

Repeat test for a power reduction from MPRS to idle power, noting that the airspeed does not deviate more than 10% from the original MPRS.

Repeat test for a power reduction from MPRS to idle power and then shutting the engine down. Note that, with power off, the airspeed does not deviate more than 20% from the original MPRS.

Repeat test for various allowed aircraft loading – gross wt. and CG limits

NOTE: It is suggested that an acceptable means to maintain the “fixed stick” condition for this test, would be to force the stick against a bar or spacer between a solid point on the instrument panel or pilot’s seat.

[gyrogreg]

First, I must note, that this DYNAMIC testing should be done only by professional test pilots who are experienced in the specific gyro type. Dynamic testing on a truly unstable gyro can cause rapid dynamic divergence.

CAUTION: This testing requires experienced testing methodology and technique. This DYNAMIC testing should not be conducted by an inexperienced pilot. This DYNAMIC testing should not, in any case, be attempted prior to meeting the STATIC stability test criteria above!

Standards Criteria:

Dynamic Longitudinal Stability

The gyroplane under smooth air conditions must exhibit no dangerous behavior at any speed between MPRS and VNE with:
(1) Primary cyclic controls fixed and
(2) With primary cyclic controls free.

No longitudinal oscillations with periods less than 5 s shall be exhibited with:
(1) Primary cyclic controls fixed and
(2) With primary cyclic controls free.

Any excitable longitudinal oscillations with periods longer than 5 s do not diverge with:
(3) Primary cyclic controls fixed and
(4) With primary cyclic controls free.

Conditions--Dynamic 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.


Explanation: This criterion establishes two dynamic requirements:

1. That there be no rapid natural oscillation tendencies with periods less than 5-seconds. This is because more rapid oscillations or response rates to disturbances may be too quick for the pilot to properly apply commanded corrective control inputs. Rapid oscillation or response tendencies can lead to pilot reactive over-control or Pilot Induced Oscillations. This minimum of 5-second period criteria essentially means that any tendency to oscillate so quickly should be at least “critically damped” (no over-shoot) so as to not require or excite pilot corrective response.

The 5-second period limit is not a magic number concluded by extensive testing or historical evidence. This limit was presumed reasonable in that it is slow enough that normal pilot skills are assumed to be adequate to respond correctly to such rates. Some airplane criteria go much further than this, requiring the Phugoid oscillation period to be as much as 10-40 seconds. This gyro criterion does not distinguish between Phugoid oscillations or other short-period oscillations. This 5-second criterion allows that high maneuvering rates are a desirable characteristic of gyros and that a 5-second period seems reasonable for a normal pilot to correct.

2. That ANY natural oscillatory tendency be “damped” – this means induced oscillations would not inherently continue or diverge to higher amplitudes of oscillation. This assumes that the pilot should be able to apply proper corrective control responses to “pilot-dampen” or correct any oscillations or disturbances slower than those defined by the 5-second period requirement above. Such oscillations might be initiated by external uncommanded inputs (wind gusts), or by pilot over-reactive commanded control inputs. Many airplane criteria go further than this requirement and often require damping rates to 10% within 3 cycles. This gyro criteria presumes that normal pilot skills can stop or correct these longer period oscillations – as long as they are at least inherently damped somewhat.

Suggested testing method:

Suggest that indicators of oscillations can be airspeed indicator and/or rotor RPM indicator (G-load indicator). Testing of gyros suggests that these two parameters dynamically vary in proportion to aircraft AOA, altitude, climb rate,g-load, etc. Only one indicator would be required to establish damping and/or oscillation tendencies.

The magnetic pulse-wheel type rotor RPM indicators are handy for recording the RRPM deviations on a standard audio tape recorder. The recorded “hum” frequency variations are readily perceived and timed and can be measured electronically afterwards if necessary.

All testing must be done with for both “fixed-stick” and “free-stick” modes. Fixed-stick can be accomplished with the use of a spacer or bar to hold the stick against after the input excitation.

Positive damping (and measurement of damping rates) for longer period oscillations can be determined by using a singlet control input (step displacement of the cyclic stick from one position to another). This induces the longer-period Phugoid oscillations to be observed to be damped. This should not be attempted by any pilot that is not thoroughly accomplished and experienced in THAT gyro! This testing should not be attempted unless the above static criteria are fully met! A gyro with poor dynamic stability might be induced into a transient reaction that a less skilled pilot may not be able to stop!

Investigation for SHORT-period oscillations requires careful “doublet” control inputs - slowly increasing the “doublet” rate, and watching carefully for onset of oscillations. (A “doublet” is one complete cycle of control input.) This is the really critical testing safety concern that requires an experienced test pilot to accomplish properly and safely. Short period natural oscillating tendencies may only be induced when the exciting “doublet” input is close to the same frequency as a natural short-period oscillation frequency of the gyro (within 20%?). The test pilot starts with longer period “double” inputs, slowly increasing the frequency of the inputs while watching for oscillation tendencies. This is called a frequency “sweep” test. This is similar to how test pilots slowly approach “flutter” tendencies on airplane testing. The objective is to slowly approach the critical frequency, without actually matching that frequency where severe aircraft response might be possible. A full and careful “sweep” of the higher frequencies is needed to assure that there are no natural oscillation tendencies in that short-period range.

I have proposed that this DYNAMIC testing is probably not essential for gyro owners who want to verify their gyro is stable safe! I believe, that the HS volume and power required to meet the above static stability criteria will be more than adequate to provide adequate dynamic damping to meet these dynamic criteria. Much more testing data is required to make that assurance completely though. However, manufacturers wishing to comply with the ASTM gyroplane LSA standard WILL need to do the dynamic testing in order to determine compliance with the LSA requirements and to sell LSA gyroplanes in the U.S.

[Dean_Dolph]

Greg, first of all I want to make it clear to everyone that I don't have the credentials to be addressing any issue in this thread. But I do have a comment and question or two and maybe three!

1. What the heck is MRPS in the post dealing with Static Longitudinal Airspeed Stability? The acronyms you have used are another example of why a list of acronyms needs to be created to support the gyro glossary. Not all of us are knowledgeable enough to follow the presentation without knowing what they stand for. (Ha! Just found the definition further down but my comment still stands.)

2. I would think that the flight conditions would need to be recorded. You mention conducting tests in calm air but what does that mean? Zero, 5 mph, less than 10 mph or…….

3. What about specified trimmed air speeds? Is there some reason not too?

4. Why not specify how much above trimmed airspeed the airspeed needs to be increased? Also why not specify specific speed, loading Cg limits and etc. for all the tests?

5. Why not specify a required method of measurements? I'm specifically referring to the stick movement measurement but there may be other measurements taken that I won't comment on.

6. What about including in a final stability report the measured Cg, pilot weight (part of MOI?) and all the other design factors that influence stability?

What I'm driving at is that left on their own, the testers may properly conduct the tests as you have outlined and determined that the machine passes the standards. But without specific test conditions there will be no way that a comparison can be done between makes, models or types by anyone interested. This includes the manf. when changes are made or a new machine is developed. Sure, they can subjectively say some thing was improved but without presenting test results that were conducted to a specific standard they are leaving themselves open to question. I recognize that without data gathering equipment that a lot of the testing will be subjective anyhow but I would like to see personal opinion/judgement reduced to a minumum.

I can appreciate keeping the testing as simple as possible for the individual owner but as a consumer, I want to see standardized data from manufacturers of the light gyro. There might not be any value for the consumer in being able to compare data but I suspect there is. I would think that a data table would indicate not only that a gyro passed/failed the stability test but also the degree of stability/instability in all three axis.

[MGman]

Greg - I think this is probably the most important post I have seen on this forum since I have been reading it ( a couple of months now). It goes to the heart of gyro flying, in my opinion - To attempt to define and document the criteria for making gyros safer to operate and overcome the bad publicity the sport seems to have attained.

I am not a pilot (yet) so I don't have enough experience to comment on most of the aerodynamic concepts and criteria you have presented. However, I am a mechanical engineer and do understand the engineering concepts involved, and they definitely make sense to me. Part of my background has been involved in specification writing and applications, and also in testing procedures writing and documentation, and interpretation of the test results.

I hope that others on this forum that do have extensive experience will jump in and offer their comments and thoughts. This may be the only chance to affect the outcome of this very significant proposal before it is enacted.

The only significant comment I think I can offer is to suggest that the criteria provide for specific documented "success" criteria for each testing category. The specific success criteria for passing (or failing) each specific test must be clear and recognizable as the specific machine passing or failing the criteria. It should not be left to the individual builders to provide their interpretation of the test results. I guess my comments are really very much the same as Dean's, in that a data table could be produced to allow prospective purchasers of any specific machine to be able to easily see and understand the performance of a given machine in order to decide if it has accepatable operating characteristics for their conmfort level.

Please keep up the good work - This is a great piece of work.

Dave Bohler

[Dean_Dolph]

Greg, Dave clarified what I should have said! Spec writing? Test layout? Result interpretation? Yeah, been there, done that and it was always understood that if clear specifics were not built into the test procedure then the it was likely that Murphy would follow whoever was responsible for the testing and run the test exactly opposite of the way you intended. Not Good!

There is always danger in leaving anything open for interpretation in a test procedure. It is always black and white with absolutely no gray areas. This is one area where I believe being prescriptive is not only warranted but is an absolute necessity.

Now it is time to get out of the way while the knowledgeable people get some serious work done here.

[gyrogreg]

Dean and Dave,
1) Thank you Dave for pointing out that MPRS should probably be added to the “Glossary of Gyro Terms”. I haven’t forgotten your request to including acronyms in the “Glossary”, and I will do that when I get a bit more time. I do think, however, that the “Glossary” has been helpful to some people, despite its deficiencies.

2-6) Specifying numbers in the criteria: Your experience with writing procedures is really coming through. This is exactly the kind of feedback I am hoping to get from this forum thread! There would be no problem with quantifying and documenting the conditions and data from this testing. I think I draw a distinction between the standard’s criteria (minimum requirements) and a Test Method to verify compliance to those criteria. I think you are suggesting the Test Method should include more data and precision. The “Test Methods” is not developed, and may not have to be, for the LSA rules. But, if we develop formal “Test Methods” for the gyroplane LSA standard, we may indeed SUGGEST more precise specifics. The standards criteria specifies the minimum requirement, and a manufacturer might, and probably would choose to document these specifics, and more, for their confidence and files – but the LSA rules only require that a manufacturer state compliance with the standard, the intention is to leave the method of verifying compliance (test methods) up to the manufacturer if possible. I do agree that a manufacturer providing a more complete data set – above and beyond the standards requirement – could be an advantage for a sale. But, the manufacturer has that choice for competitive reasons – whatever they feel indicates their compliance to the standard the best. The standard does not make it a requirement for a more complete data set, but that may be an advantage for a manufacturer and some assistance for the more technically astute customer – manufacturer’s choice – not an FAA requirement.

The FAA also intends that we make verification of these criteria as simple and inexpensive as possible, while still assuring safety. And, to encourage, or at least not to discourage, the average gyro flier to understand and assess their own machine, I am inclined to keep things as simple as possible – trying to think outside the usual FAR box in language and procedures a layman can understand and accept!

I’m not sure we know how to more specifically quantify the conditions and criteria more precisely at this point anyway. I have conducted these tests numbers of times and have found no real technical need for more precision. For instance, if the stick is further forward and the pilot is having to also push the stick forward, positive static airspeed stability (positive slope of the curve) is indicated – current thinking is that positive slope is all that is required to make the gyro stable-safe from a static airspeed criteria. Also, for instance, for the “calm wind” criteria, if the wind is too turbulent, it is simply difficult for the test pilot to make the assessment – they will have to try again in calmer air – I have readily assessed most of the static criteria in winds with up to 10 mph gust factors – but, in some cases, it may be easier to make the assessment in calmer winds! As far as specifying exact conditions, my testing has found that three test points within the allowable or critical range (airspeed and power and load) – especially if the point near the worst extreme condition is tested - is more than adequate to eliminate the possibility of different results at some finite intermediate test condition. In other words, my testing hasn’t found it necessary to get over-precise here, and nobody has yet provided safety arguments for otherwise.

For better understanding and encouragement to use these criteria by less technically professional people, I am inclined to keep the guidance to the general public as friendly and simple as possible. But, we will likely be providing some guidance to manufacturers as well. And that guidance might be that manufacturers might be much more comfortable if they developed and recorded and filed a very complete testing data set – as part of their design development, and for their own, and their customer’s confidence of compliance with the standard. We will also be encouraging persons shopping for a gyro to ask for convincing data that shows a machine complies – a very complete data set might be more likely to make that sale – even if it doesn’t really make any better assurances. Although, in my testing, I do not find it necessary to require more specific data points and measurements than the standards requirements currently state, a manufacturer might more surely make a sale if they can provide a full data set with more data points. I think competition might also raise the standards of the test reporting as well – but I’m not sure that is really necessary for safety and should be required by the standard.

Our ASTM LSA standard is only intended to verify minimum criteria adequate to assure the gyro is stable safe. This is not intended for any more comparison between models other than that. IMHO, the degree of stable-safe beyond the simple requirements of these criteria would not really improve the safety of the gyro – but might be more a matter of “harmony” of control “feel” or personal preferences. For instance, a tuned “harmony” between airspeed stability and g-load stability might present a different “feel” between one machine or another – but not really influence the risk of an instability accident. Such variations – i.e.: stronger airspeed or g-load stability, might be more of a marketing choice, rather than a safety issue requiring more precise specification. By not specifying or quantifying parameters beyond what is needed for safety, we allow more choices for manufactures and customers.

But, at this point, we are not totally sure (at least I’m not) just what degree of stability really minimizes the risk of stability related accidents, without compromising desirable attributes of a gyro. That is one reason for my creating this thread – I am seeking experienced judgments or comments on whether the criteria are good enough to ensure stable-safe! I will certainly keep your “Testing Methods” suggestions in mind for if/when we provide such guidance to manufacturers. And, if you would have some good safety arguments for more tightly specifying conditions or results, we need to consider those. But, to this point, these are the degree of requirements that the LSA consensus says does assure stable-safe gyros. I’m looking for good arguments to support this or to change the requirements - many of these criteria are the best opinions of experienced people but may not have a lot of analytical or accident statistical support. The FAA has encouraged us to “start somewhere”. If we find from experience or technical argument that some criteria need to be tightened up (or loosened), that is what the ASTM review and revision process is for. If this thread uncovers improvements in these criteria, our ASTM gyroplane subcommittee will consider such suggested changes – the reason I have started this thread.

Thanks, Greg Gremminger

[Mike Jackson]

Originally Posted by gyrogreg

Standards Criteria:

Static Longitudinal Maneuvering (G-Load) Stability:

Explanation: This is the classic “test pilot” criteria for maneuvering (G-Load) stability. This criteria results in the tendency of the aircraft to self-restore itself to 1g rotor loading upon a disturbance to rotor G-loading. A tendency for G-Loading to diverge (to higher or lower G-Loadings), without active pilot intervention, could result in severe rapid aircraft divergent pitch up or pitch down. Aircraft with statically unstable maneuvering static characteristics, requires constant pilot attention to maintain 1g loading upon a disturbance. Such G-Load instability upon rapidly decreasing and divergent G-Loading is thought to be a critical element in gyro buntover incidents – especially in the hands of an inexperienced pilot.

.

In fixed wing testing , one usually looks for a positively increasing slope with no objectionable "nonlinearities" up to about 3 Gs. The slowly varying G method at constant airspeed as you described is one of the better methods to use. If the curve "flattens" , the stick force per G lightens, at higher G we might begin to find objectionable maneuver flight characteristics which can aggravate or possibly initiate a PIO situation.




Mike

[Mike Jackson]

Originally Posted by gyrogreg

1. That there be no rapid natural oscillation tendencies with periods less than 5-seconds. This is because more rapid oscillations or response rates to disturbances may be too quick for the pilot to properly apply commanded corrective control inputs. Rapid oscillation or response tendencies can lead to pilot reactive over-control or Pilot Induced Oscillations. This minimum of 5-second period criteria essentially means that any tendency to oscillate so quickly should be at least “critically damped” (no over-shoot) so as to not require or excite pilot corrective response.

The 5-second period limit is not a magic number concluded by extensive testing or historical evidence. This limit was presumed reasonable in that it is slow enough that normal pilot skills are assumed to be adequate to respond correctly to such rates. Some airplane criteria go much further than this, requiring the Phugoid oscillation period to be as much as 10-40 seconds. This gyro criterion does not distinguish between Phugoid oscillations or other short-period oscillations. This 5-second criterion allows that high maneuvering rates are a desirable characteristic of gyros and that a 5-second period seems reasonable for a normal pilot to correct.

2. That ANY natural oscillatory tendency be “damped” – this means induced oscillations would not inherently continue or diverge to higher amplitudes of oscillation. This assumes that the pilot should be able to apply proper corrective control responses to “pilot-dampen” or correct any oscillations or disturbances slower than those defined by the 5-second period requirement above. Such oscillations might be initiated by external uncommanded inputs (wind gusts), or by pilot over-reactive commanded control inputs. Many airplane criteria go further than this requirement and often require damping rates to 10% within 3 cycles. This gyro criteria presumes that normal pilot skills can stop or correct these longer period oscillations – as long as they are at least inherently damped somewhat.

Greg, I have found that most FW systems have a short period between 2 - 5 seconds. Most systems will NOT be critically damped and will have some degree of overshoot. We want to make sure these overshoots are damped and have a short "time constant" - time to 63% of your stable initial value of whatever you are using to measure your short period. This could be fuselage ref line, rotor rpm, disc AOA - but not airspeed by assumption. We assume the doublet you described is sufficiently quick to leave airspeed constant. This time constant (Greek letter TAU) will be on the order of a second or less for higher perf FW aircraft and probably close to that for a rotor system. I need to look at Houston's data. Point is, it will be quick and none of us will adequately be able to respond quickly enough if we wanted to. That's where sufficient HS size and placement comes in. Let the design take care of it.

The Phugoid is one of those things that is more annoying than anything else - especially to autopilot designers. Most are damped, maybe neutral with a long period 20+ seconds and I believe fall into the WGAS category where gyro flying is concerned. To excite, slowly pull back on the nose (FRL) a finite amount and watch the AS go down, yawn a little then watch the AS build and the nose come up. You can do it stick free or fixed.

Here's a little beauty that will allow you to get a feel for both maneuvering flight (stick force/G), and the short period mode (gust stability). NASA pilots and engineers refers to this as the "Concave Downward" requirement emphasizing the short-time response. Might be helpful where the FAA LSA requirement descriptions are concerned. It requires that the time history of normal acceleration (G ) in response to a cyclic control step be concave downward within 2 seconds. It is intended by this feature, to provide a que of the coming acceleration (G response) , and to guarantee a finite (damped) response to the control deflection. The requirement provides for reasonable maneuvering characteristics in forward flight - possibly the most compelling single handling quality.

How do we do it in practice? Pick an airspeed in the heart of your gyro's envelope and trim hands off at a constant power setting. Quickly pull to a specified G level. Don't get too carried away on the pull - try about 2 to start. When you've reached your target G, quit pulling. You can hold the stick fixed or release it (stick free). What do you feel? The G may increase for up to 2 seconds but you should then feel the G load decay. This is your indication of a finite damped response.

You may have doubts about G continuing to increase after you've stopped pulling but there seems to be lag in any pitch response. One of the better films I've seen of this is Ernie Boyette's Dominator video where he is pumping his stick fore and aft flying straight and level. Watch the stick position and the slight lag in fuselage (G) response as he does this.

You can now change airspeeds and CG locations if you wish but it is a quick test technique to get a feel for dynamic stability and maneuver flight. You might also say, all we've really done is excite the short period with a single pulse. This is the case, but because the airspeed will decay here, you might see the phugoid also excited .

Mike

[MGman]

Greg - Thanks for your reply - You are extremely patient with people that don't have direct flying experiience, and I appreciate it. I feel a little reticent doing any posting because of my lack of experience, but what the heck - I guess everyone who is interested has something to offer :-).

I need to give Dean the credit for the acronym comments BTW. My comments were toward seeing what your thinking was in providing more hard numbers for the pass/fail criteria. From your response you have obviously given these same ideas a lot of though and I can see where you are coming from. You can't make them hard and fast for a lot of flying situations where it is mostly a pilot feel sort of thing.

It looks like you are starting to get the experienced flyers making comments now, so I'll sort of drop back and stay out of the way (unless I read someting I don't agree with (or don't understand), at which point you'll hear from me again :-).

Thanks again for your efforts - they are helping me learn - Dave Bohler

[raghu]

Thanks Greg for putting out the proposed tests for peer review. In this post I will give my comments on the dynamic stability test and on a later one I will post comments on the static tests.

I have two major comments 1) the manner in which oscillations are measures; 2) the criterion that no oscillation must have a time period less than five seconds I see Mike J. has brought this up as well;

Originally Posted by gyrogreg

No longitudinal oscillations with periods less than 5 s shall be exhibited with:
(1) Primary cyclic controls fixed and
(2) With primary cyclic controls free.
.....
.....

Suggest that indicators of oscillations can be airspeed indicator and/or rotor RPM indicator (G-load indicator). Testing of gyros suggests that these two parameters dynamically vary in proportion to aircraft AOA, altitude, climb rate,g-load, etc. Only one indicator would be required to establish damping and/or oscillation tendencies.

1. THE NATURE of DYNAMIC MODES
-----------------------------------------------------
In order to measure the oscillations one must first be aware of the different dynamic modes and what parameter ( airspeed, rotorspeed, pitch varry). It turns out in a gyro that passes the G load static stability test ( this is really AOA static stability...more on that later) there are essentially three oscillation modes
a. short period mode: varies in time period between ~2-10 seconds and is alwys damped though almost never critically. The damping for example in the magni VPM m16 is ~ 1 second to half in amplitude. The oscillation is primarily one of pitch attitude but it may involve (particularly at high speeds) some rotor speed variation. In my opinion the way to measure it would be by looking at pitch variations, given that at low speeds you may see no rotor speed changes in this mode) . A video camera should do the job. As Greg has said the way to capture this mode is by a stick doublet.

b.Long period or Phugoid mode: In the order of 10+ seconds and at best weekly damped and can be mildly unstable as well. The oscillation involves airspeed and also rotor speed. The rotor speed slightly lags airspeed due to rotor inertia. For the magni vpm 16 the time period is 12-15 seconds and time to damp to half amplitude is between 30 and 40 seconds.

c. rotor mode. This by the nature of autorotation is critically damped ( deadbeat response) and can be ignored ( see *)

2. MEASURING PERIOD and DAMPING
Given this my view is that the long period should be measured by air speed, though rotor speed may also suffice. But for the short period measuring pitch variation may be the best option.

3. FIVE SECOND CRITERION
If the criterion is just for the long period mode then its fine. However the short period mode will be in the 5 second or below range. For example the magni VPM 16 at 30 mph has a short period mode of 4 seconds! Similarly I would guess the littlewings would be low as well particularly the elevator model that relies on pitch changes to make cyclic input. Both these aircraft from all accounts are very pilotable and must pass the stability test.

Things are a little different with gyros that have a lack of g load (what is really AOA stability, more on this later) stability, the short period merges with the rotor mode and a new second phugoid mode of a smaller time period ( ~6 seconds) develops and things get very hairy....... though the mode may be marginally stable and could induce PIO. For this reason, this mode must be taken out by insisting on G load ( AOA ) static stability.

Hope this helps and I will post other comments as and when I get a chance to write.

[gyrogreg]

Mike and Raghu, thank you very much for your thoughts. It is very helpful to have experienced and qualified technical input. I think these are very important and helpful issues to explore. I’m not sure I am technically savvy enough to understand all that you have presented, but let me see if I can summarize – and then I have a few specific questions to help clarify this:

Both Mike and Raghu suggest that there WILL be natural oscillation tendencies below 5 second periods – AND, these WILL NOT be “critically damped” – they will have some overshoot. BUT, Raghu suggests that these short-period oscillations will be quickly damped (1 second or more?) IF the aircraft has G-Load static stability (Raghu says this is really called “AOA stability”).

QUESTION 1: Are we sure of these short-period assumptions for Gyroplanes, or are they unproven transferences from FW fundamentals – can we trust these assumptions?

QUESTION 2: Our ASTM standards criteria says “No longitudinal oscillations with periods less than 5 seconds”. In my explanation, I had assumed this would be met if those oscillations were at least “critically damped”. Since they are not “critically damped”, but may be quickly damped (1 second +), how would you recommend we define “no longitudinal oscillations”? Should some point of “quickly damped” be considered acceptable? - Or is “no oscillations” an improper simplification (remember, keep it simple!). What could we use as a measurable dividing point between “oscillations” or "no oscillations"?

QUESTION 3: If a gyro meets the Static G-Load Stability requirement (AOA stability?), would this technically suffice for adequate damping of short-period oscillations – therefore making the “no oscillations shorter than 5 seconds” criteria unnecessary?

Raghu, is suggesting that the short-period oscillations would not be observable via RRPM. I would agree that a video camera recording nose attitude would certainly be a simple means to observe and record those oscillations – if this is a necessary requirement (Question 3 above).

QUESTION 4: Mike, you seem to agree that the requirement that longer-period be “damped” is adequate – not a safety issue! - rather than requiring a specific damping rate. Or, are you suggesting that even undamped longer-period oscillations are not really required?

QUESTION 5: Raghu, you are pointing out that the VPM16 (as tested by Dr. Houston) has a short period of 4 seconds, but is damped to ½ amplitude in 1 second. The short period of less than 4 seconds would arguably not meet the standard’s 5 second minimum. You suggest similar results are likely for the LW. But, since you suggest that both these aircraft are “very pilotable”, are you suggesting that the criteria of a 5 second minimum oscillation period is too stringent – for safety? Would you define a 4 second period with a 1 second damping rate as being an “oscillation” per the criteria in the ASTM standard?

QUESTION 6: I have tried hard to initiate a “short-period” oscillation in the Magni M-16 (“sweep testing” with “doublet” excitation!). I would expect this to repeat what Dr. Houston reports. What am I missing? Is the damping rate so rapid that they cannot be readily detected? Was Dr. Houston’s determination from actual flight testing, or from the computer model? Are there better ways to verify there are “no oscillations” - short period? In light of QUESTION 3 above, is this even necessary?

QUESTION 7: What is the difference, or is there any difference, between “Maneuvering (G-Load) Static Stability” and “AOA Static Stability”? Are these just terminology differences, or are there real differences that influence real safety determinations we should be considering?

I ask the above questions for clarification and advancing our understanding of what the stability requirements to assure a stable-safe gyro should be. I invite anyone with thoughts on all this to please post.

Thanks, Greg Gremminger

[raghu]

Greg, Here are the answers. feel free to ask for any clarifications.

QUESTION 1: Are we sure of these short-period assumptions for Gyroplanes, or are they unproven transferences from FW fundamentals – can we trust these assumptions?

Ans: I was giving data from actual flight test of a gyro. So, yes the short period does exist. Now, do all gyros have such a mode? Well the pure short period mode will always be there as long as you have AOA stability. This mode will often be resonably damped ( I will explain why in a later post if you want an explanation) but there could be exceptions.

On the other hand if you do not have AOA stability ( say neutrally stable) then the pure short mode will ironically become critically damped ( yes this is true) but as you increase the instability it will rapidly merge with the rotor mode and form a new mode that is very much like the phugoid in terms of damping ( weekly damped) but is of a lesser time period ( though not as low as the pure short period mode). Based on this high frequency and low damping, I argue that this may well be the mode that is suseptible to PIO and shoul be taken out of the equation by requiring sufficient AOA stability.

Nb. in theory the transition from short period mode to critically damped mode will occur at zero AOA stability, but in practice this transition may occur at some point just above or below AOA stability as other factors may come into the equation, though AOA is the dominent factor and so the transition is very nearly at neutral AOA stability.


QUESTION 2: Our ASTM standards criteria says “No longitudinal oscillations with periods less than 5 seconds”. In my explanation, I had assumed this would be met if those oscillations were at least “critically damped”. Since they are not “critically damped”, but may be quickly damped (1 second +), how would you recommend we define “no longitudinal oscillations”? Should some point of “quickly damped” be considered acceptable? - Or is “no oscillations” an improper simplification (remember, keep it simple!). What could we use as a measurable dividing point between “oscillations” or "no oscillations"?

Ans: I think we are going into handling qualities issues here and you may need more input from test pilots, but I would say make no limitation to the maximum frequency as long as the damping is in the range of damp to half in 1 second. These oscillations will not be noticable to the pilot possibly just as you do not spot the magni short period mode



QUESTION 3: If a gyro meets the Static G-Load Stability requirement (AOA stability?), would this technically suffice for adequate damping of short-period oscillations – therefore making the “no oscillations shorter than 5 seconds” criteria unnecessary?

Ans: It is likely in practice given current configurations you will be OK ( well damped SP mode), but you cannot guarentee it ( at least until we have more data). For example, it is concevable that you have a scenario where a very short lever arm is used with a sufficient HS to provide static AOA stability. In this case damping may be low and things could get squirly ( though there will almost always be positive damping). One would thus need to test.


Raghu, is suggesting that the short-period oscillations would not be observable via RRPM. I would agree that a video camera recording nose attitude would certainly be a simple means to observe and record those oscillations – if this is a necessary requirement

Ans: The sp mode may have some RRPM digressions ( particularly at high speed) but that is not the primary component and so pitch variations is a good attribute to measure the sp mode.


(Question 3 above).

QUESTION 4: Mike, you seem to agree that the requirement that longer-period be “damped” is adequate – not a safety issue! - rather than requiring a specific damping rate. Or, are you suggesting that even undamped longer-period oscillations are not really required?

QUESTION 5: Raghu, you are pointing out that the VPM16 (as tested by Dr. Houston) has a short period of 4 seconds, but is damped to ½ amplitude in 1 second. The short period of less than 4 seconds would arguably not meet the standard’s 5 second minimum. You suggest similar results are likely for the LW. But, since you suggest that both these aircraft are “very pilotable”, are you suggesting that the criteria of a 5 second minimum oscillation period is too stringent – for safety? Would you define a 4 second period with a 1 second damping rate as being an “oscillation” per the criteria in the ASTM standard?

Ans: see my answer to question 2.

QUESTION 6: I have tried hard to initiate a “short-period” oscillation in the Magni M-16 (“sweep testing” with “doublet” excitation!). I would expect this to repeat what Dr. Houston reports. What am I missing? Is the damping rate so rapid that they cannot be readily detected? Was Dr. Houston’s determination from actual flight testing, or from the computer model? Are there better ways to verify there are “no oscillations” - short period? In light of QUESTION 3 above, is this even necessary?

Ans: The figure I quoted where from actual flight testing. It is tricky to catch and will require you pumping the controls at the right frequency, but you should spot it, though you cannot measure it as easily as the phugoid. you may need to video and verify after the flight. Perhaps even that may not be accurate enough as freezing the stick during such a short duration is hard and you are going to get inadvertend cyclic noise. In general as long as you see enough damping this mode not a problem

QUESTION 7: What is the difference, or is there any difference, between “Maneuvering (G-Load) Static Stability” and “AOA Static Stability”? Are these just terminology differences, or are there real differences that influence real safety determinations we should be considering?

Ans: This is something that needs some clarification. I hope to devote a separate post ( when I get a chance) to articulating how engineers think about dynamic stability and what tools they use to better get an insight. The current approach in many posts is a mish mash and very unstructured approach with direct and indirect effects all combined, often resulting in circular arguments. In the meantime more to the point...

1. Static AOA stability: Gives the relationship between AOA and pitching moment. If an increase in AOA results in a nose down moment then the gyro is AOA stable. The turn test actually measures this.

2. static Rotor G load stability: This is a relationship between RRPM and pitching moment. If an increase in RRPM produces a down pitching moment then the gyro is rotor G load stable. A test for this may need more thinking. Out of the top of my head photos at trimmed flight could help determine CG versus RTV relationship. Or perhaps a more radical one, dropping a sack of potatoes in flight and noteing the effec on pitch!

Usually AOA stable gyros are also rotor G load stable because AOA satbility is often derived by placeing RTV behind the CG. However this is far from guarenteed. Consider a gyro that has RTV in front of a CG ( at some air speed) and also has a large HS with zero incidence to the relative wind. Now this gyro may very well be AOA stable if the HS is large enough as the slope of the HS is much steaper than the rotor. However, the gyro would be G-load unstable.

why bother with this distinction? Well it turns out that if you are G load unstable the long period mode losses a bit of damping ( by interacting with the rotor speed). This may still be perfectly fine, but the trend is adverse. Thus the RTV BEHIND THE CG is a very good criterion as it kills the proverbial two birds with one stone- both AOA and G load are stable.