Balanced Moments and Roll and Coning

[wclaybaugh2]

SolarYellow presented a good explanation of coning. The rockets fins are the strong corrective force. I'm looking at base drag is a possible resonant damping coefficient. The low-pressure region behind the vehicle dampens the overcorrection of the fins. Even when the Input frequency/Natural frequency =1. It's a theory that needs further work.

It started as a project to study launch exhaust flow patterns of finless cone and pyramid rockets when launched close to a flat plate flame deflector producing a low-pressure venturi effect region. It morphed into a base drag pressure measurement study of large diameter single engine rockets.

The limited data I've collected is trending to greater dampening after engine burnout.

I took a flight dynamics class long ago and I assume the rules haven’t changed…although I expect the experience base has grown some.

Which is to say I can see how the low pressure area at the base of a rocket in flight might dampen the pendulum but not effect spin (to use SolarYellow’s excellent top level description).

But what are the aerodynamics that are absorbing that energy? If the vehicle swings a few degrees to one side, does not the low pressure area at the base follow? Is there some time constant to that following that serves to dampen the swing? Or?

Bill

 

[Spacedog49Krell]

I took a flight dynamics class long ago and I assume the rules haven’t changed…although I expect the experience base has grown some.

Which is to say I can see how the low pressure area at the base of a rocket in flight might dampen the pendulum but not effect spin (to use SolarYellow’s excellent top level description).

But what are the aerodynamics that are absorbing that energy? If the vehicle swings a few degrees to one side, does not the low pressure area at the base follow? Is there some time constant to that following that serves to dampen the swing? Or?

Bill

55 years ago, my Fluid Dynamics professor wouldn't discuss Knudsen flow in class. His quote in class was "nothing interesting happens there." I was the one who asked the question. I spent 26 years of my engineering career working on Knudsen flow issues in the semiconductor industry. Our knowledge base has changed over the years.

The vehicle swings change the shape and intensity of the low pressure trailing envelop, in theory, dampening the pendulum motion. It does not affect the roll action. Decoupling pitch and yaw from the roll results in a simpler problem. Proving the theory is the next step. I wish I had access to a high speed wind tunnel. Testing would be simplified, but eventually I would need to fly a test article and induce coning. To answer your other questions, I need data.

I'm currently testing arrays of MPX4115, BMP280, and BMP390 sensors. The MPX4115 I can sample at 1000+Hz. The BMP sensors are limited to 200Hz.

 

[SolarYellow]

55 years ago, my Fluid Dynamics professor wouldn't discuss Knudsen flow in class. His quote in class was "nothing interesting happens there." I was the one who asked the question. I spent 26 years of my engineering career working on Knudsen flow issues in the semiconductor industry. Our knowledge base has changed over the years.

The vehicle swings change the shape and intensity of the low pressure trailing envelop, in theory, dampening the pendulum motion. It does not affect the roll action. Decoupling pitch and yaw from the roll results in a simpler problem. Proving the theory is the next step. I wish I had access to a high speed wind tunnel. Testing would be simplified, but eventually I would need to fly a test article and induce coning. To answer your other questions, I need data.

I'm currently testing arrays of MPX4115, BMP280, and BMP390 sensors. The MPX4115 I can sample at 1000+Hz. The BMP sensors are limited to 200Hz.

It's stuff like this that makes me wish our members at Texas A&M would get involved and work on basic questions with their wind tunnel.

https://lswt.tamu.edu/technical-specifications/

 

[shockie]

Gordon Mandrell's Fundamentals of Dynamic Stability Parts 1-4 are in the 10/68,11/68,1/69,2/69 issues of Model Rocketry magazine. Available to NAR members in the Magazine Archives.

If TRA:

https://www.ninfinger.org/rockets/ModelRocketry/ModelRocketry.html
some interesting info on Radial Moment of Inertia

These eventually became Chapter 1 in "Topics in Advanced Model Rocketry" in 1973

 

[wclaybaugh2]

55 years ago, my Fluid Dynamics professor wouldn't discuss Knudsen flow in class. His quote in class was "nothing interesting happens there." I was the one who asked the question. I spent 26 years of my engineering career working on Knudsen flow issues in the semiconductor industry. Our knowledge base has changed over the years.

The vehicle swings change the shape and intensity of the low pressure trailing envelop, in theory, dampening the pendulum motion. It does not affect the roll action. Decoupling pitch and yaw from the roll results in a simpler problem. Proving the theory is the next step. I wish I had access to a high speed wind tunnel. Testing would be simplified, but eventually I would need to fly a test article and induce coning. To answer your other questions, I need data.

I'm currently testing arrays of MPX4115, BMP280, and BMP390 sensors. The MPX4115 I can sample at 1000+Hz. The BMP sensors are limited to 200Hz.

Thx. That is what I had suspected.

Gordon Mandrell's Fundamentals of Dynamic Stability Parts 1-4 are in the 10/68,11/68,1/69,2/69 issues of Model Rocketry magazine. Available to NAR members in the Magazine Archives.

If TRA:

https://www.ninfinger.org/rockets/ModelRocketry/ModelRocketry.html
some interesting info on Radial Moment of Inertia

These eventually became Chapter 1 in "Topics in Advanced Model Rocketry" in 1973

This makes a good point: all of the math behind the Datcom software is publicly available in “Topics”.

I’ve used Excel to automate some of those relationships (notably the fin cant required to induce some specific roll rate) and there is no reason the whole design process could not be so automated.

Bill

 

[Spacedog49Krell]

Gordon Mandrell's Fundamentals of Dynamic Stability Parts 1-4 are in the 10/68,11/68,1/69,2/69 issues of Model Rocketry magazine. Available to NAR members in the Magazine Archives.

If TRA:

https://www.ninfinger.org/rockets/ModelRocketry/ModelRocketry.html
some interesting info on Radial Moment of Inertia

These eventually became Chapter 1 in "Topics in Advanced Model Rocketry" in 1973

For the past 50+ years the model rocket community has been flying finless, dynamic stabilized rockets of various designs. I find equations that validate the stability of these designs, but very little numerical data that supports a real understanding of the problems or solutions for dynamic vortex stability.

"When you can measure what you are speaking about, and express it in numbers, you know something about it."
Lord Kelvin

We missed a giant opportunity for data with Bryce Chane's Sasquatch rocket. I hope he flies it again.

 

[kjhambrick]

True, data is not magic. It requires a lot of work to interpret what it means. If the AV-bay location does not give you a clear picture of what the rocket is doing, then move the sensors to where they do give a better picture. A 240 frame/second video gives you a snapshot of what the rocket flight looks like every ~4 ms. You still need to interpret the image. An Arduino based flight computer can record 10 channels of data every 4 ms. I have less interpretation with the digital data. I still need to understand what the data means, but I can apply equations to the data for understanding.

The trend in flight computers is to higher data recording speeds. SparkyVT has been flying 1000Hz systems for 5-6 years, I've been flying 500Hz systems for 4 years. Blue Raven does 500Hz and there are other systems at 400Hz. I use video as confirmation of my interpretation of the data. On-board video is an important tool. I'm looking at the drone barebones 4K video boards for my next avionics bay.

@Spacedog49Krell --

Some of the best high-speed data and analysis I've ever seen is in @Adrian A's thread: 2-stage J total impulse controlled by Blue Ravens.

There are a lot of flight dynamics to ponder in Adrian's data and I believe he has captured pitch-roll coupling in multiple flights with his 500 Hz Blue Ravens.

Especially his two-stage 29K flight on May 17'th 2023 starting at Post #36

I am pretty thankful that Adrian is so generous with his data -- there is always a lot to learn in the raw data even more from his analysis.

What I meant by 'But it's not magic -- all we can measure from within an AV-Bay are the effects of real-world causes of any flight anomolies that we might experience.' is not that we need faster, better data.

What I meant is that everything that our sensors can measure from within the AV-Bay are 'only' effects.

The Accelerations we measure are the vector sums of unknown forces acting on unknown masses.

Similarly, when our gyro sensors measure rotation rates, we can't really directly see the cause of those rotations.

That's where the magic lives: it's all in the interpretation of the events that we CAN measure.

-- kjh

 

[kjhambrick]

Gordon Mandrell's Fundamentals of Dynamic Stability Parts 1-4 are in the 10/68,11/68,1/69,2/69 issues of Model Rocketry magazine. Available to NAR members in the Magazine Archives.

If TRA:

https://www.ninfinger.org/rockets/ModelRocketry/ModelRocketry.html
some interesting info on Radial Moment of Inertia

These eventually became Chapter 1 in "Topics in Advanced Model Rocketry" in 1973

Thanks for the link @shockie !

I read part 1 years ago but I lost the xerox copy I had.

I never found parts 2 and 3 and I didn't know they were in 'The Topics Book'.

So now I need to buy the book.

Thanks again !

-- kjh

EDIT: Yikes ! $500 !! Maybe I had better wait for Santa to bring me one: AMAZON: Topics in Advanced Model Rocketry

 

[Adrian A]

In case anyone wasn't already aware of this, a simple method for observing a lateral mass imbalance, at least for rockets launched from towers, is to stick the fins off the edge of a table or countertop and roll the rocket. When I was doing my I and J altitude rockets 10 years ago, the nosecone I had had a small misalignment or mass imbalance, and so did the rest of the rocket, so I used this method to rotate the nosecone to cancel the offsets as best as I could.

 

[kjhambrick]

In case anyone wasn't already aware of this, a simple method for observing a lateral mass imbalance, at least for rockets launched from towers, is to stick the fins off the edge of a table or countertop and roll the rocket. When I was doing my I and J altitude rockets 10 years ago, the nosecone I had had a small misalignment or mass imbalance, and so did the rest of the rocket, so I used this method to rotate the nosecone to cancel the offsets as best as I could.

So now you tell me, after I've already epoxied those rail button warts to my airframe

But I'll definitely try this on my upcoming 38mm rocket before I ruin the roll.

Thanks for the idea !

-- kjh

 

[SolarYellow]

Towers FTW.

 

[Spacedog49Krell]

@Spacedog49Krell --

Some of the best high-speed data and analysis I've ever seen is in @Adrian A's thread: 2-stage J total impulse controlled by Blue Ravens.

There are a lot of flight dynamics to ponder in Adrian's data and I believe he has captured pitch-roll coupling in multiple flights with his 500 Hz Blue Ravens.

Especially his two-stage 29K flight on May 17'th 2023 starting at Post #36

I am pretty thankful that Adrian is so generous with his data -- there is always a lot to learn in the raw data even more from his analysis.

What I meant by 'But it's not magic -- all we can measure from within an AV-Bay are the effects of real-world causes of any flight anomolies that we might experience.' is not that we need faster, better data.

What I meant is that everything that our sensors can measure from within the AV-Bay are 'only' effects.

The Accelerations we measure are the vector sums of unknown forces acting on unknown masses.

Similarly, when our gyro sensors measure rotation rates, we can't really directly see the cause of those rotations.

That's where the magic lives: it's all in the interpretation of the events that we CAN measure.

-- kjh

Adrian's Blue Raven flight computer is the direction our hobby must take to make significant steps in understanding and measurement of fight dynamic issues. Utilizing advanced measurements, we can make design and product improvements and advance our knowledge base with better safety for future participants.

To better understand flight pitch-roll coupling issues we may need microsecond data. Real-world anomalies are not predictable in simulators unless we understand mathematically the anomaly. The more data points you have to study, the greater the chance of understanding and solving the problem. If the sensors in the AV-Bay do not give you a picture of the issue, then move the sensors to where you do get a picture of the issue. If you suspect fin flutter, put strain gauges on the fins. Gyros in the AV-Bay may not detect fin flutter anomalies.

Something I can measure but not explain is why a rocket coasting with a 1.6 rps CW rotation suddenly stops and rotates CCW at 2 rps.

 

[Sooner Boomer]

If you are running redundant altimeters and you have both on one side of the sled and both batteries on the other side, that can be a problem if the batteries are significantly heavier than the altimeters (which they often are nowadays). Ditto for the hardware... U-bolts, charge wells, etc. Putting them on opposite sides will help balance your rocket axially.

I agree that the mass should be distributed as equally about the axis (in the direction of flight) as possible. My experience with rockets that are imbalanced is not that they cone (the aft describes a circle), but that if the rocket has spin, the rocket body nutates. "Coning" is a form of precession, nutation is a "wobble" in the precession. A rocket that is well balanced about the axis can cone if it has a high CG/CP ratio (the nose is very heavy); ie. the rocket is overstable. An example of overstable coning is here:

 

[kjhambrick]

<<snip>>

Something I can measure but not explain is why a rocket coasting with a 1.6 rps CW rotation suddenly stops and rotates CCW at 2 rps.

@Spacedog49Krell --

Wow !

Do you have an example of 'roll reversal' that you could share ?

That's sounds very interesting !

Thanks !

-- kjh

 

[kjhambrick]

I agree that the mass should be distributed as equally about the axis (in the direction of flight) as possible. My experience with rockets that are imbalanced is not that they cone (the aft describes a circle), but that if the rocket has spin, the rocket body nutates. "Coning" is a form of precession, nutation is a "wobble" in the precession. A rocket that is well balanced about the axis can cone if it has a high CG/CP ratio (the nose is very heavy); ie. the rocket is overstable. An example of overstable coning is here:

Yes, @Sooner Boomer !

What is an "aluminum plate glider" ?

I didn't see it ...

Does your 7xC7 cluster Fat Boy always fly that way ?

I did see rod whip as the rocket left the rod but I can't see a lot of roll in the video.

But then, my video manipulation skills leave a lot to be desired

I believe I have also seen similar wobble without any noticable roll -- where the nose and the tail describe a flat arc that decays during the coast phase.

Maybe related: I've also seen a lot of posts here on TRF where a short, wide diameter, overstable leaves the rail at a random angle.

Sometimes they wobble and sometimes they take off at the original random angle without any wobble or coning.

There are sure a lot of flight modes that I don't understand !

-- kjh

 

[kjhambrick]

55 years ago, my Fluid Dynamics professor wouldn't discuss Knudsen flow in class. His quote in class was "nothing interesting happens there." I was the one who asked the question. I spent 26 years of my engineering career working on Knudsen flow issues in the semiconductor industry. Our knowledge base has changed over the years.

The vehicle swings change the shape and intensity of the low pressure trailing envelop, in theory, dampening the pendulum motion. It does not affect the roll action. Decoupling pitch and yaw from the roll results in a simpler problem. Proving the theory is the next step. I wish I had access to a high speed wind tunnel. Testing would be simplified, but eventually I would need to fly a test article and induce coning. To answer your other questions, I need data.

I'm currently testing arrays of MPX4115, BMP280, and BMP390 sensors. The MPX4115 I can sample at 1000+Hz. The BMP sensors are limited to 200Hz.

@Spacedog49Krell --

Trying to read between the lines ...

I only understand Knudsen flow at the conceptual level.

Is there something I could read to understand how Knudsen flow applies to a model rocket flying in the lower atmosphere ?

What are you measuring with the pressure transducers at high sample rates ?

Are you mounting them outside the airframe ?

If so, how do you set them up so they don't interfere with the airflow about the airframe ?

Thanks !

-- kjh

 

[Sooner Boomer]

Yes, @Sooner Boomer !

What is an "aluminum plate glider" ?

I didn't see it ...

Does your 7xC7 cluster Fat Boy always fly that way ?

I did see rod whip as the rocket left the rod but I can't see a lot of roll in the video.

But then, my video manipulation skills leave a lot to be desired

I believe I have also seen similar wobble without any noticable roll -- where the nose and the tail describe a flat arc that decays during the coast phase.

Maybe related: I've also seen a lot of posts here on TRF where a short, wide diameter, overstable leaves the rail at a random angle.

Sometimes they wobble and sometimes they take off at the original random angle without any wobble or coning.

There are sure a lot of flight modes that I don't understand !

-- kjh

"aluminum plate lighter" - flash pan (thingie...).

No, it doesn't always fly that way. I've had a problem until recently in getting all 7 motors to light. I've also made a nosecone that's a lot lighter - by about half. It flies better (straighter, higher) on the lighter nosecone.

Another flight:

The rocket doesn't necessarily need to roll to cone. The aft end moves in a circle in the direction of travel.

I built it with a *bunch* of nose weight - 6 or 7 ounces. I didn't use the "tail cone hack", so that much weight, a lot more than it needed, made it overstable.

 

[kjhambrick]

<<snip>>

The rocket doesn't necessarily need to roll to cone. The aft end moves in a circle in the direction of travel.

<<snip>>

Thanks @Sooner Boomer

I understand what you're saying now !

I think I need to start with a glossary of terms of the common modes of dynamic stability ( or instability ) so I can discuss the phenomena better

-- kjh

p.s. Dooh ! "Aluminum Plate Lighter" -- I need a hearing aid

 

[Spacedog49Krell]

@Spacedog49Krell --

Trying to read between the lines ...

I only understand Knudsen flow at the conceptual level.

Is there something I could read to understand how Knudsen flow applies to a model rocket flying in the lower atmosphere ?

What are you measuring with the pressure transducers at high sample rates ?

Are you mounting them outside the airframe ?

If so, how do you set them up so they don't interfere with the airflow about the airframe ?

Thanks !

-- kjh

Knudsen flow does not apply to model rockets in the lower atmosphere. If you fly above 40km, then you will experience Knudsen flow. The Falcon 9 stages between 40-60km, a prime pressure region for Knudsen flow. The beautiful exhaust gas patterns seen just before MECO on the west coast at sunset, are the interactions of the exhaust gas molecules in Knudsen flow. The second stage exhaust gas patterns appear as straight lines indicating Molecular flow.

I'm measuring the changes in air pressure as the air transitions from flow in the boundary layer (over the body) to turbulent, low pressure, flow behind a rocket with extreme high base drag. To answer the question; "Does base drag dampen resonant frequency pitch-roll coupled events."

They will be mounted both inside and outside the airframe. Those mounted inside will have short connecting tubes to the specific areas to be measured. Those mounted outside will be located in the turbulent flow at the base of the rocket. The mere fact of measuring anything changes what you are measuring, the "Uncertainty Principle".

Krell