I could use just a little guidance

Your data matches what I saw from the flight line, the flight looked great. Do you have a plot of derived velocity that you could super impose on your angle plot. From what I see, I do believe you could critically damp your system.

Random Flying Object said:

Your data matches what I saw from the flight line, the flight looked great. Do you have a plot of derived velocity that you could super impose on your angle plot. From what I see, I do believe you could critically damp your system.

Click to expand...

Here you go. Velocity from a Raven.

Jim

Heck you may just want to turn the gain down a notch and let it ride.

The onboard did a great job showing the vertical! The video from flight line was a little deceiving but maybe the angle, great flight Jim, looks like you are making headway.

Nick@JET said:

The onboard did a great job showing the vertical! The video from flight line was a little deceiving but maybe the angle, great flight Jim, looks like you are making headway.

Click to expand...

The flight line video isn't all that helpful because the rocket goes out of view just at the point where stabilization starts. By the time it comes back into view, the deed is done.

One other piece of data from the flight was the gps file. It wasn't helpful with respect to the path of the rocket, but it did give the position at apogee. Turns out the rocket was located about 200 feet to the north of the apparent position over the road. That is, it was at an angle of about 3 degrees to the north (or to the left from the perspective of the video). That works out to about a 3° tilt, which is pretty consistent with the tilt data. Assuming a south wind, I would have expected the rocket to be angled to the south and not the north. It's possible that the assumption of a south wind is not correct. The forecasted winds at altitude were a little confusing that day, and it didn't really look like there was much wind at all until closer to the ground, where there clearly was a south wind. Just another piece of data.

At this point, I think I can use my data point (the rate of turn shortly after engaging stabilization) to estimate what might happen in my Balls flight. I have several data points that I can use. It sure would be nice though to be able to actually calculate the rate of turn. In essence, if you have two opposing fins that are at a slight angle (to induce a pitch rather than a roll), how fast does the rocket turn? I'll figure it out eventually.

Jim

JimJarvis50 said:

...
It sure would be nice though to be able to actually calculate the rate of turn. In essence, if you have two opposing fins that are at a slight angle (to induce a pitch rather than a roll), how fast does the rocket turn? I'll figure it out eventually.

Jim

Click to expand...

That's easy, you just take your lift from your fins and divide by the moment of inertia... and adjust for velocity and changing air density... and wave hands at the impact of the main fins... and mumble mumble added mass and damping... [Throws hands up in the air] Or just flight test it. Looking forward to hearing about the big flight at BALLS!

I wasn't really planning to do another test of the stabilization system, but it's so darn much fun to fly that I just can't resist. So, here's the video record from the 11th test flight.

https://youtu.be/cQT6fxZ0kKk

This flight was just supposed to be straight up with no particular events (just somewhat lower gains this time than in previous flights). Mother nature had different ideas and decided to weathercock the rocket off the pad, so there was the opportunity to watch the rocket go back vertical after getting knocked off course.

There are more oscillations in this flight than in previous flights. It could be because less gain was used. However, I think the steady-state configuration just has a little roll in it. Maybe one of the canards wasn't exactly straight or the cameras cause some turn, or whatever. The most direct evidence of this is the slow roll closer to apogee. I think the rocket was just fighting the control system a little. This is the third good flight in a row, so I think the hardware and control changes that have been made are working pretty well.

My wife didn't want me to fly the rocket because it was pretty windy on the ground. I told her not to worry, that the rocket would go up over the pad, drift that direction for about 2,000 feet, and then land right over there. I missed the target by about 100 yards - not bad for an 8,000-foot flight. But the system (for sure this time) goes back to the shelf until Balls.

Jim

JimJarvis50 said:

I wasn't really planning to do another test of the stabilization system, but it's so darn much fun to fly that I just can't resist. So, here's the video record from the 11th test flight.

https://youtu.be/cQT6fxZ0kKk

This flight was just supposed to be straight up with no particular events (just somewhat lower gains this time than in previous flights). Mother nature had different ideas and decided to weathercock the rocket off the pad, so there was the opportunity to watch the rocket go back vertical after getting knocked off course.

There are more oscillations in this flight than in previous flights. It could be because less gain was used. However, I think the steady-state configuration just has a little roll in it. Maybe one of the canards wasn't exactly straight or the cameras cause some turn, or whatever. The most direct evidence of this is the slow roll closer to apogee. I think the rocket was just fighting the control system a little. This is the third good flight in a row, so I think the hardware and control changes that have been made are working pretty well.

My wife didn't want me to fly the rocket because it was pretty windy on the ground. I told her not to worry, that the rocket would go up over the pad, drift that direction for about 2,000 feet, and then land right over there. I missed the target by about 100 yards - not bad for an 8,000-foot flight. But the system (for sure this time) goes back to the shelf until Balls.

Jim

Click to expand...

Nice flight! I'm a little behind on reading this thread so I don't totally get the purpose of the rotating fin can but it all looks really good.

ArchitectOfSeven said:

Nice flight! I'm a little behind on reading this thread so I don't totally get the purpose of the rotating fin can but it all looks really good.

Click to expand...

The purpose of the spin can is to avoid "control reversal". The canards generate vorticies that interact with the fins below, and can actually cause the rocket to turn in the wrong direction. The cartoon below attempts to illustrate the effect. I have several flights, with one flight in particular, that shows this effect. With the spin can, the vorticies hit the fins but produce no torque on the air frame. The spin can is not all that difficult to make.

Jim

Nice work, again! I was planning on putting an upward facing camera on my next 2 stage launch but after seeing your video, I think maybe not.

JimJarvis50 said:

The purpose of the spin can is to avoid "control reversal". The canards generate vorticies that interact with the fins below, and can actually cause the rocket to turn in the wrong direction. The cartoon below attempts to illustrate the effect. I have several flights, with one flight in particular, that shows this effect. With the spin can, the vorticies hit the fins but produce no torque on the air frame. The spin can is not all that difficult to make.

Jim

Click to expand...

Oh so the fin can just spins freely? I thought that it's rotation was dictated by the flight computer as some sort of way to counter vehicle spin. That is a really cool idea. And yeah I understand the idea of control reversal. 10/10 for elegance in solution.

timbucktoo said:

Nice work, again! I was planning on putting an upward facing camera on my next 2 stage launch but after seeing your video, I think maybe not.

Click to expand...

Oh, by all means use the upward facing camera. Just don't do it in Texas around noon near the summer solstice with a vertical-seeking rocket. I have seen things with upward facing video that I never would have figured out otherwise.

Jim

ArchitectOfSeven said:

Oh so the fin can just spins freely? I thought that it's rotation was dictated by the flight computer as some sort of way to counter vehicle spin. That is a really cool idea. And yeah I understand the idea of control reversal. 10/10 for elegance in solution.

Click to expand...

Thanks. None of it was my idea, but I will claim the cost-effectiveness prize. The spin can is a few tubes and about $20 in ball bearings.

Jim

JimJarvis50 said:

Oh, by all means use the upward facing camera. Just don't do it in Texas around noon near the summer solstice with a vertical-seeking rocket. I have seen things with upward facing video that I never would have figured out otherwise.

Jim

Click to expand...

Took your advice and mounted camera looking up. Does that look like clean separation to you?

https://youtu.be/ajazlVDFEGA

timbucktoo said:

Took your advice and mounted camera looking up. Does that look like clean separation to you?

https://youtu.be/ajazlVDFEGA

Click to expand...

Dang Tim, That is way cool! Neat video. Kurt Savegnago

timbucktoo said:

Took your advice and mounted camera looking up. Does that look like clean separation to you?

https://youtu.be/ajazlVDFEGA

Click to expand...

Yep, you get to see some things you might otherwise miss (or might never figure out at all). You got the separation, the motor ignition and deployment in one vid! Nice flight.

Jim

So, I am keen to see what comes next!

Great project Jim!!

MattSR said:

So, I am keen to see what comes next!

Great project Jim!!

Click to expand...

Matt, I didn't see that you asked a question. Sorry.

I'm not quite sure what comes next. One thing that I know is that if I try more high altitude flights (I hope to), I want to be able to use this system. At Balls, my flight was at the maximum 6 miles from the launch point. A friendly jet stream moved it back to under three miles (i.e., it goes vertical, weathercocks into the jet stream, and then rides the jet stream back towards the pad). However, had the rocket not stabilized, as depicted in the picture, the landing point would have been 20 miles or more further away. There are not many directions where you can go 20+ miles and get your rocket back.

Another thing I know is that flying the system is a lot of fun. I had four flights with it between Balls last year and this year, with three of them being successful, and they were some of the most enjoyable flights I've had in the hobby. I'm going to fly it some more.

Another thing I know is that this particular system is not going to be widely adopted. It is too difficult and expensive for most people to try, even though it is a lot of fun. I have been fortunate to have someone with existing hardware and code, who was capable of modifying the program for rockets and willing to work with me to do it. Even with that, I was not able to get more than a few people to take a serious shot at designing a stabilization system. As people here have pointed out, the RC-type equipment used in the system has some limitations and can't possibly work (even though it does). Given that, and given that the system won't be widely adopted, I don't really see the benefit of asking my Guru to spend a lot of time on further development. I have what I need for the objective I set out to accomplish (just a little guidance), and maybe that's the best I can do. I hope others can come up with a system that is better and can be made more widely available.

So, for now, I'm just going to fly the system and have fun with it. I'll post the flight results here from time to time.

Jim

JimJarvis50 said:

As people here have pointed out, the RC-type equipment used in the system has some limitations and can't possibly work

Click to expand...

Says who?! Have we got news for them .

Hi Jim,
Obviously a tremendous accomplishment and a hard-won success. I was wondering what your approach to 'parameter estimation' has been through this? Were you able to perturb your control system to get initial values for your gains? I am expecting to need to implement something equivalent in my project once I get the booster finished.

br/

Tony

TonyL said:

Hi Jim,
Obviously a tremendous accomplishment and a hard-won success. I was wondering what your approach to 'parameter estimation' has been through this? Were you able to perturb your control system to get initial values for your gains? I am expecting to need to implement something equivalent in my project once I get the booster finished.

br/

Tony

Click to expand...

Thank you. I was very pleased that this system worked basically as intended on the Balls flight, and on the flights leading up to Balls.

Determining the correct gains for this is not easy. I can share a few thoughts, but I don't know how helpful it will be to you. With respect to yaw/pitch control, I was able to get perhaps 4 flights that showed the rate of response when a change was required. These were done by launching the rocket at an angle, either with the control system active or activated after the launch, and in one case, an initial angle caused by weathercocking. The Balls flight was an example of the system coming on during flight (after the booster separated) where there was a flight angle to correct. What I came to realize is that the rate of correction was much slower than I would have expected, taking on the order of 5 seconds perhaps for the rocket to change direction by 5 degrees or so. This change will depend on a lot of variables, and while I got some insight into how to calculate these things, I have never been able to come up with something that would quantitatively relate the control action to the rate of change in direction. I gave it a good try at one point, but just couldn't get there. In my case, a slow change in direction was just what I wanted, but I never approached flight conditions where rapid changes would have occurred.

For roll, the calculations on roll rate (relating control action to rate of change) are much easier. The opportunites to employ control theory are much greater, not that I did that. From a practical perspective, the problem I was trying to solve was to optimize the control response as best as I could. If the response was too great, then things could go unstable given the type of equipment I have access to, or cause changes in the roll rate in excess of the gyro range. On the other hand, too slow might result in more roll than you want or uncontrolled roll that was also outside of the range of the gyros. I think I have probably used higher gains for roll than I should for most flights, but adjusting these to be in balance with yaw/pitch is very difficult when using one set of canards, I think an approach of using large canards for yaw/pitch and small canards for roll would probably work much better.

The other obvious problem with both yaw/pitch and roll control is the change in velocity during the flight. I have dealth with this by limiting the portion of the flight where control was active so that the velocity range was manageable. I don't think it would be that hard to allow the gains to vary with time of flight, although I have not done this.

If you think I can help answer any questions you have, feel free to ask.

Jim

I flew my VTS last weekend for the second time. Since the first flight I have turned the gains down a little, but the last flight will be of higher velocity (much bigger motor) and so the fins will have more authority (I don't have any gain scheduling). A quick look at the data showed it seemed to actually control to a nice vector, and there were no power supply failures this time . Roll, pitch and yaw were all greatly reduced, but unfortunately the vector was not vertical. I suspect the commercial autopilot I am using has some sort of drift correction algorithm that skews things whilst sitting on the pad. I need to check that next.

As for the roll gains, I think it is easy to have too much of that. That was apparent on my first flight. If your results are non-oscillatory then it wasn't too high (for that flight speed) .

I am also considering migrating the hardware to a different platform and adding a pitot static airspeed sensor to possibly help with gain scheduling for the control loops.

This is a fun area to play in but not without its foibles.

I need some more guidance...

The project for this year is shown below. It might go pretty high. Due to the size of the sustainer tubing, I won't be able to use my stabilization system. Therefore, I am considering spinning the rocket. There are a couple of scenarios here. The first would be to use an electronics package in the sustainer that is built around the EasyMega for tilt inhibition. I am concerned about the spinning rocket affecting the tilt estimate prior to sustainer ignition, so in this case, only the sustainer would be spun, starting at stage separation. The purpose of doing this would be to avoid the tumbing that I experienced in my Balls flight this year (), and maybe the sustainer would go straighter.

The second approach would be to use an electronics package from a certain "secret santa" that would be imune to spin for tilt inhibition . Then, the idea would be to spin the stack from the start of the flight. Maybe that would be better for reducing dispersion.

I don't know much about how to spin rockets. I believe I have a way to determine the fin angles that I would need (target 6 Hz), but I don't know much about how to balance the rocket and what the accuracy of this would need to be to avoid problems.

At present, I am not planning to de-spin the rocket (a project for another day), so it'll be a crappy video. I'm open to any thoughts, experience, pitfalls, or other guidance on this topic.

Jim

Hi Jim,

So I imagine you already know that the Black Brant XI uses spin motors that light right after launch to get the angular momentum right away rather than wait for canted fins to get it going. I believe they are embedded in the transition. Canted fins are pretty standard too, either should reduce dispersion. The effect on you inertial measurement will be important to understand as you are already aware.

Balancing the rocket will be easiest if you enforce internal physical symmetry as much as possible [see below].

One thing to keep in mind is that your rocket will be expected to tumble like last time, but probably faster if you are adding more angular velocity. Physics says it will eventually spin about its largest inertia axis [and won't be terribly 'happy' spinning about an axis not aligned with a principle axis], so a good de-spin strategy would be the thing to address the eventual tumbling.

Some scale test flights may give you some ideas about how its working [or not]

br/

Tony

TonyL said:

Hi Jim,

So I imagine you already know that the Black Brant XI uses spin motors that light right after launch to get the angular momentum right away rather than wait for canted fins to get it going. I believe they are embedded in the transition. Canted fins are pretty standard too, either should reduce dispersion. The effect on you inertial measurement will be important to understand as you are already aware.

Balancing the rocket will be easiest if you enforce internal physical symmetry as much as possible [see below].

One thing to keep in mind is that your rocket will be expected to tumble like last time, but probably faster if you are adding more angular velocity. Physics says it will eventually spin about its largest inertia axis [and won't be terribly 'happy' spinning about an axis not aligned with a principle axis], so a good de-spin strategy would be the thing to address the eventual tumbling.

Some scale test flights may give you some ideas about how its working [or not]

br/

Tony

Click to expand...

I find this whole topic to be quite confusing. I have been told by a really smart guy that what I should do is to either spin the rocket fast (6 Hz) or slow (1 Hz) but not at a frequency that is close to the pitch frequency. It seems intuitive to me that there should be some gyro benefit related to spin. What am I missing?

Jim

There is something called 'roll-coupling' [may go by other names too] that is the interaction between the roll speed [spin] and the natural frequency of the rocket in pitch. It is often seen as a corkscrew flight path [or shred, or both]. Nothing is as simple as one would like in aerodynamics, as you have already encountered in your stabilization system. Wallops Flight Facility spin balances their rockets on a large machine, so getting that right one would think is important. Regarding the 'gyro' benefit, one has to spin pretty fast for it to work well [slow spinning tops wobble a lot more than fats spinning tops]. Spinning slowly may avoid roll-coupling, but it also won't do very much. I am not sure if I still have specs, but 20 rev/s sounds familiar for the BBXI. I believe Gofast used yo-yo de-spin on theirs. WFF was happy when they got spin down to < 1 rev/s by that method. What that would mean for you is that you would need to know what your final spin rate was, then you could design the yo-yo to remove that much angular momentum and be happy.
In a nutshell:
1. make rocket perfectly symmetrical left to right from all side views [aligns principle axes with the rocket axes]
2. characterize final spin rate from whatever method used.
3. design yo-yo weights and cord length to remove expected angular momentum

the yo-yo could be replaced by a momentum wheel or de-spin thrusters. A wheel or thrusters could be used for closed loop control, but the yo-yo is probably the simplest.

hope this helps,

Tony

So I did a little looking and found references to the BBXI spinning at rates of 2Hz and 5Hz, so not sure what I was remembering. My other payload flew on a Terrier Orion, I have that as spinning up to 6.66 rev/s, with roll frequency crossing the vehicle's pitch frequency very rapidly at 1st stage separation [a big drop in inertia perhaps], and then the spin rate stays above the pitch frequency for the rest of the flight. Fins were canted 55 arcminutes. You might want to peruse MIL-HDBK-762 DESIGN OF AERODYNAMICALLY STABILIZED FREE ROCKETS (17 JUL 1990), I recall it seemed pretty thorough.

br/

Tony

Last Airfest, the evening I gave you the "pixie pellets", We went across to trailer and talked to [can't remember name] guy who had the trailer set up for tracking multiple rocket transmitters/gps [up tp 17 of them at once].

Topic was spin stabilizing, i think you were lazer focused on your high altitude motor ignition issue and may have forgot.
Anyhow I discussed at length how I was going the route of spin stabilizing my full scale Sparrow-Arcus vs your active method.
I may be the only person so far on forum to have done this so far. I too reached out ..to no avail...and spent 2 yrs gathering PDF files and data research on th subject. Now 2 yrs into the project, this is my plan to go over 150,000.
First picked Sounding rocket closest to my goal.No point in re-inventing the wheel.
2nd do 1/3 scale....1/2 scale versions to work out spinning issues and electronics needed.
3rd full scale version & flight.

Issue at hand was high altitude staging, which I believe I have figured out[for me] & you validated by using one of my BKNO3 pellets. I have been using them on every motor been lighting to sort things out and for practice in sizing correctly.
May even try some for separation as I have found some sounding rockets do so.



Other issue I wanted to conquer wad the interstage, after 1/3 scale success with spinning, wanted to do I//s like the real deal..requiring a pointed I/S that fits into nozzle of sustainer, like the real sounding rocket. Kurt did same. [but we did not realize that till after his flight was made public]
Thinking it may solve any binding problems spinning may induce...once again, no point in re-inventing the wheel, since this is how the full scale version does it.
The problems and solving of them saved for another time, your issue at hand spinning.








So far at 1/2 scale version and awaiting flight.

Telemega tilt will not function after 5.5 revolutions , sensor is only good for 2000 degrees [2000/360=5.56 revolutions] at that point it looks like an accelerometer after reaching G limit.

MY first stage does NOT spin, only upon separation does the sustainer begin spinning...make things a bit simpler for us.

Here is except from my thread on my journey so far that may help;

Success!!

[Remember readers, nothing spins till sustainer separation, full stack is non-spinning during boost phase]


Finally after 10 months of work. Validation, under boost the full stack had a slight positive roll of 1.4 rps, as soon as separation charge fires at 2.3 seconds the sustainer attains full spin of negative 12 rps. Fins are small enough not to have any effect on the whole stack. Due to fin cant the sustainer always shows negative roll rate.
Booster positive role attributed to angle/wind or slight misalignment of fin mounting.

360 degrees is one full revolution...sensor pegs at -2000 [5.5 revolutions] [this is where we must extrapolate data will Excel to determine actual from raw data points, since data is pretty much linear.]

Anyhow you can see the slight rise under boost to approximate 500 degrees [rotation], then at 2.3 seconds [separation] full spin pegs the sensor to 2000 negative.
At around 11.5 seconds , sustainer spin slows enough to register again slowing to 825 degrees [820/360=2.2 rps] this was due to high arc at apogee, the flight was not vertical.[due to small booster motor to keep on our field] First 2 flights sustainer slowed down to 1.2-1.5 Rps, well with in design expectations.
As a matter of fact if this trend follows up-scaling, there is no need to de-spin [yo-yo] for apogee deploy.



After all the extrapolating in Excel...I determined to spin rate could easily be determined using the magnetometer feature of the Telamega...that's further in the thread. Just counting the tips of sine waves for any one second period.
I also am going the route of designing for single use and bringing the small payload [11 inches] and nosecone down separately from sustainer to further eliminate issues with spinning and fouling recovery.


Thread covering this for those interested is here:

This link takes you first post after successful spin flight.
https://www.rocketryforum.com/showt...2-stage-sounding-rocket&p=1592729#post1592729

Things like needed fixture to align fins, how angle was determined [1 degree] etc. are covered starting at the beginning would be advised.I linked to 2nd page for those just interested in spinning results.

Also added weight in NC and careful mounting of gear on center axis balanced to aid in spinning..not coning, is needed. Not that difficult at all. See pics of that later in link. Kinda hard to answer needed issues in one post, after taking several years to figure out......lol

The other JIM

PS after all this work, I now see Kurt has gone well over 200,000 and as far as I could tell....did not spin..so there ya go...lol

"pixie pellets" - It sounds like CJ may be dealing more than just motor starters! LOL

Seriously, Jim and Jim,....thanks for posting this cutting-edge R&D for our hobby. It's informative and inspirational for the rest of us.

Here is part of post after full up launch of 1/3 scale which went 17,000 something.
This was a 4 grain 54mm to a 6 grain 29mm. largest motors that would fit in 1/3 scale. Pic for size reference.






Here's what did it....the "spin" factor.
Just a quick look shows the spin rate jumping up at motor burn out. Yes the booster prevents sustainer from spinning till separation. No fin cant on Booster.

First seconds of flight show a moderate roll rate of 1.5-2 revolutions-per-second.
At 1.6 seconds booster burns out [at 1.500 ft], stages separate & sustainer immediately spins up to full speed under coast....26-29 rps.
In 5 sec. of coast, sustainer reaches 5,500ft where ignition occurs sending spin rate up to 36-37 rps.
[You count the peaks or valley's between any 1 second time slot to determine spin rate. magnetometer]



Hope this may start your path to "spinning". I think 5-6 eps is too low for you and the harmonics of your large rocket. One reason the first Go-Fast went wonkers was too low of spin rate, [5-6]they increased it by[don't remember] and it then flew fine. First one tumbled into oblivion.
My entrance into this was via Jeroen after much discussion on how to & why. I miss him dearly.

With some serious tinkering, you may be able to use your active stabilization on lower stage and just spin the top, getting best of both.

My spin rate is determined by the full scale which used 18 or more, depending on motor used in sustainer. There were several based upon weight of payload flown.