Honey Badger 2 Build Thread and Flight to 57k!

[Viperfixr]

Another awesome BALLS project...congrats! While the rocket is darned impressive, the motor is a total beast.

 

[butalane]

Being out at the away cell seeing this thing launch was seriously so sick. I was so pumped that I think I screamed louder than both of you guys haha. I don't think most people understand what an accomplishment this flight was, if you had flown any cti 6gxl load this thread would be 3 times longer with replies. It's great you both are sharing the details of the construction too, good ideas. The only thing that sucks is that you bested my 98mm minimum diameter rocket by ~700'! Lol once again great flight guys.

More replies but it would've doubled the cost of the whole rocket!

Thanks for all your help: GPS, fin shape, and last minute tower advice were all key to making it a smooth flight!

 

[jus_rockets]

Congrats on the awesome flight, I wish that someday I can fly a rocket of this caliber.

 

[CCotner]

Thanks James! Btw, we decided on the fin shape because of your flight last year, makes a big difference!
Our nose tip was threaded so we could retain the avionics bay via a piece of all thread. We were about to drill another hole in the nose shoulder for the arming wires but then realized we had a big hole up front if the tip was off So we loaded the rocket on the pad without the tip on (wires hanging out of the front part of the nose), armed the altimeters, then screwed the tip on.

Oh wow, that's actually pretty genius. So you had a screw switch or something on the end of the wires, powered everything on, then stuffed it inside and threaded the tip on? It's the little details like that that make this project outstanding. Genius. =)

 

[butalane]

Congrats on the awesome flight, I wish that someday I can fly a rocket of this caliber.

Thanks!!!

Oh wow, that's actually pretty genius. So you had a screw switch or something on the end of the wires, powered everything on, then stuffed it inside and threaded the tip on? It's the little details like that that make this project outstanding. Genius. =)

Is screw switch the technical term for "twisted the wires together?" Other than that, that is exactly what we did.

We had a pin that went through the NC shoulder and into the bulkplate of the altimeter bay to stop the shock cord from spinning the alt bay out of the NC tip. We also put some loctite on the threads of the NC tip to make sure it didnt spin free on its own.

 

[ClayD]

Thanks!!!



Is screw switch the technical term for "twisted the wires together?" Other than that, that is exactly what we did.

We had a pin that went through the NC shoulder and into the bulkplate of the altimeter bay to stop the shock cord from spinning the alt bay out of the NC tip. We also put some loctite on the threads of the NC tip to make sure it didnt spin free on its own.

I have a few rockets with that high tech method for arming altimeters. I like turning them off when its a no-go.

Congrats on your teams success!

 

[butalane]

Copy of the GPS data plot


View attachment 147171

I wanted to add some more details - this plot came from a Big Red Bee 70cm with the 100mW trasmitted and new NEO GPS chip. What a great experience! The GPS picked up track at ~40kft and relayed the rest of the flight to us. It was a little painful and expensive getting a HAM license, Radio, and finally the big red bee, but absolutely worth every bit of effort in the end!

 

[Reinhard]

Hey Alex,
Unfortunately our altimeter didn't record the data onto the sd card correctly (even though the sd card was installed correctly and the altimeter was potted for high g flights) so we don't have the flight data in a readable form, just a 2mb junk file :bang:. Luckily the altimeter fired the charges correctly but for our next flight, an improved altimeter that can handle the acceleration is a must. The only data we got was from the big red bee gps so we have altitude.

The altitude was within 2,000ft of the RasAero sim which predicted ~Mach 3 for top speed. That seems pretty reasonable to me.

Is this from a commercial electronics package or something self made (not many altimeters with SD cards out there)? If you understand the file format, you might be able to reconstruct the content with some low level tools. Two years ago a rocket that I was involved with "suffered an anomaly" and lost power in the process before the file on the SD card was closed. Regular file access didn't work, but low level access (I believe using the dd tool under Linux) allowed it to retrieve the data.

Reinhard

 

[Oberth]

Regular file access didn't work, but low level access (I believe using the dd tool under Linux) allowed it to retrieve the data.

Very interesting. Yes this was a commercially available altimeter. At the end of the onboard video you can hear the beep sequence which indicates no continuity on any of the channels...seems to me this is evidence the altimeter lost power or reset at some point of the flight.

I'll post the altimeter file, if anyone can make meaning of it that would be awesome! Here's what I know so far, the file won't open because it is missing a header. Someone suggested using a hex editor which we tried but didn't have any real success (definitely isn't my strong point).


View attachment FLT-001.zip

 

[THarrison]

Very interesting. Yes this was a commercially available altimeter. At the end of the onboard video you can hear the beep sequence which indicates no continuity on any of the channels...seems to me this is evidence the altimeter lost power or reset at some point of the flight.

I'll post the altimeter file, if anyone can make meaning of it that would be awesome! Here's what I know so far, the file won't open because it is missing a header. Someone suggested using a hex editor which we tried but didn't have any real success (definitely isn't my strong point).


View attachment 148738

$50 prize on me for anybody that can extract legit data from this file. I really want to see it.

Todd Harrison

 

[Reinhard]

Very interesting. Yes this was a commercially available altimeter. At the end of the onboard video you can hear the beep sequence which indicates no continuity on any of the channels...seems to me this is evidence the altimeter lost power or reset at some point of the flight.

I'll post the altimeter file, if anyone can make meaning of it that would be awesome! Here's what I know so far, the file won't open because it is missing a header. Someone suggested using a hex editor which we tried but didn't have any real success (definitely isn't my strong point).


View attachment 148738

This might be an interesting riddle to solve and/or a rather frustrating one too ;-).

I'm not familiar with the G-Wiz format, but it is probably rather "plain" (looks like lots of 16bit data in the hex editor). Do you have one or two a reference files for comparison purposes. It would be best if at least one was from the same unit with the same settings?

Reinhard

 

[butalane]

This might be an interesting riddle to solve and/or a rather frustrating one too ;-).

I'm not familiar with the G-Wiz format, but it is probably rather "plain" (looks like lots of 16bit data in the hex editor). Do you have one or two a reference files for comparison purposes. It would be best if at least one was from the same unit with the same settings?

Reinhard

Wow, thanks a bunch for taking a look!

Here is the data file from a pre-flight bench test we did of the unit.

I really hope this helps!

View attachment FLT-000.zip

 

[daveyfire]

Very interesting. Yes this was a commercially available altimeter. At the end of the onboard video you can hear the beep sequence which indicates no continuity on any of the channels...seems to me this is evidence the altimeter lost power or reset at some point of the flight.

I'll post the altimeter file, if anyone can make meaning of it that would be awesome! Here's what I know so far, the file won't open because it is missing a header. Someone suggested using a hex editor which we tried but didn't have any real success (definitely isn't my strong point).

I think I got it. Looks like you had some power losses on the way up—too much motor in too little of a rocket.




One hell of a flight! I miss the playa so much.

 

[daveyfire]

You know, I just realized something.

Those weird sags in the alt plot... those aren't power losses... they're awfully regular, ~0.6 second duration at 5 second intervals. That's gotta be your BRB GPS coupling into the board. They don't show up in the accel plot, so it must have an internal amp, while the baro side picks it up in the trace running from the sensor to the amp prior to heading into the ADC. Trippy!

 

[CarVac]

You know, I just realized something.

Those weird sags in the alt plot... those aren't power losses... they're awfully regular, ~0.6 second duration at 5 second intervals. That's gotta be your BRB GPS coupling into the board. They don't show up in the accel plot, so it must have an internal amp, while the baro side picks it up in the trace running from the sensor to the amp prior to heading into the ADC. Trippy!

View attachment 148816

Hmm, that's similar to the problem Adrian had with Beeline trackers tripping the Parrot main deploy soon after apogee.

 

[Marsman]

Great build! One question- where did you buy your phenolic for the nozzle and what grade did you use?

 

[Oberth]

We should have posted the file earlier

where did you buy your phenolic for the nozzle and what grade did you use

It was canvas phenolic from mcmaster.

 

[THarrison]

I think I got it. Looks like you had some power losses on the way up—too much motor in too little of a rocket.

View attachment 148814
View attachment 148815

One hell of a flight! I miss the playa so much.

Yes! YES!

 

[A5tr0 An0n]

Can you please provide some more information on the ablatives that you used?

 

[bobkrech]

Well, not sure if anyone is following the thread...but if you are, we also go some onboard video! The rocket is moving around a lot the whole flight, but there are some cool freeze frames and you can see the fast liftoff!


[YOUTUBE]7BELQLjgOrU[/YOUTUBE]

This was a great flight, and I've been following this thread. The last few comments made me go back and reread the entire thread again and I have a few questions.

Did you use spin fins or was the spinning unintentional? My interpretation of the video is that you had a pretty high spin rate near apogee and had a lot of coning which indicates a dynamic instability. (The apparent rotation reversals are an aliasing effect of the video frame rate and the rotation rate. This effect is often seen when you see videos of a spinning mag wheel on a race car or of a prop on an airplane.)

I'm wondering if the coning arose from a little bit of fin misalignment coupled to a little bit of asymmetric thrust possibly due to uneven ablation of the Garolite CE phenolic nozzle producing off center thrust, of if the CG/CP ratio was marginal under the flight dynamics?

Bob

 

[CarVac]

This was a great flight, and I've been following this thread. The last few comments made me go back and reread the entire thread again and I have a few questions.

Did you use spin fins or was the spinning unintentional? My interpretation of the video is that you had a pretty high spin rate near apogee and had a lot of coning which indicates a dynamic instability. (The apparent rotation reversals are an aliasing effect of the video frame rate and the rotation rate. This effect is often seen when you see videos of a spinning mag wheel on a race car or of a prop on an airplane.)

I'm wondering if the coning arose from a little bit of fin misalignment coupled to a little bit of asymmetric thrust possibly due to uneven ablation of the Garolite CE phenolic nozzle producing off center thrust, of if the CG/CP ratio was marginal under the flight dynamics?

Bob

Man, you're right about that spin speed! You can see the rolling shutter smearing the lines of clouds into arcs as it exposes the sensor line-by-line.

It doesn't look like coning to me, though: it looks to me like the rocket was spinning purely on its own axis. If it were coning I imagine it wouldn't have achieved nearly the altitude it did.

 

[CCotner]

By my count, it isn't spinning fast enough (15 Hz assuming that's 30fps video from a keychain camera) to fold or alias the signal. Also, how do you tell from onboard video whether a rocket is 'coning' (meaning it is traveling on average straight up but is actually undergoing some complex (multi-axis) rotation as well) or is just arcing over normally while spinning?

EDIT: I see rolling shutter artifacts also, but only two points where the rocket spin direction changes (transient moments where the vehicle does not appear to be spinning). These 'frozen' points, where no motion is apparent, should only happen at multiples of the framerate; since it's probably a 30fps keychain camera, that means those two reversals happen at 0Hz, 30Hz, 60Hz, etc.. By watching for features to cross the view repeatedly, I count a maximum omega before the first reversal to be about 1.5Hz, and after that the maximum omega before the second reversal is about 3.5Hz-not fast enough to reach the first signal folding point (where motion would appear reversed) at 30Hz.

After the second reversal it speeds up to maybe 5-6 Hz at most-pretty fast for a rocket not intended to be spinning, but not fast enough to cause signal folding or aliasing.

Edited again because I made a stupid mistake. Assuming 30 fps video, the motion will appear correct up to 15 Hz, then will appear to slow down as the true rpm increases from 15 to 30Hz (signal folding); at 30Hz it will appear frozen in place. As the true RPM increases from 30Hz to 45Hz, it will appear to begin rotating backward faster and faster (aliasing); from 45Hz to 60Hz it will appear to slow down (folding and aliasing). I'm pretty sure, at least.
Unless I learned nothing in my systems classes or are blind and can't see the video? I don't think rolling shutter artifacts indicate high enough speeds to cause signal folding.

 

[bobkrech]

Here's some interesting examples.

[YOUTUBE]LVwmtwZLG88[/YOUTUBE]
Sloughing blades and regrowing them?

[YOUTUBE]rVSh-au_9aM[/YOUTUBE]
Helicopter going through the air without the rotor spinning?

[YOUTUBE]ttvSzoqGQlY[/YOUTUBE]
What's going on?

[YOUTUBE]qwNT4rtdoDs[/YOUTUBE]
Really?

[YOUTUBE]J72rWG8W91w[/YOUTUBE]
Effects of rotating the scan angle of camera array from horizontal.

In all cases we know the directions of rotation doesn't change, or does it. :wink:

Lot's of optical illusions. This is not a frame transfer camera: It is an NTSC interline transfer array camera that is being processed on the fly. A Simple Fourier analysis doesn't account for the vertical and horizontal scan rates, scan angle, the exposure time blurring and the 60 Hz frame scan rate of the NTSC image. I'll leave the stroboscopic modeling to someone more mathematically inclined, but the video shows clearly that the rocket is coning and almost tumbling end over end at apogee.

How can we determine this? Due to an aerodynamic asymmetry, whether intentional or unintentional, we know the rocket is rotating because the FOV is rotating. We also know that unguided rockets undergo a gravity turn, and we know that the amplitude of the roll-yaw-pitch coupling is increasing with time as evidenced by the increasing vertical amplitude of the horizon position in the video. These effects illustrate the classical definition of inertial coupling aka coning. https://en.wikipedia.org/wiki/Inertia_coupling or mathematically describes in more detail here https://www.dtic.mil/dtic/tr/fulltext/u2/a349546.pdf

We can't determine if it is due to solely to asymmetric aerodynamics or if it enhanced by an asymmetric thrust vector from the propulsion system.

I've watched dozens of amateur high altitude unguided rocket flight videos and almost all exhibit some degree of coning which is difficult to prevent without active stabilization or a great deal of 6-DOF dynamic stability modeling. The rocket experienced minimal coning and reached apogee without wildly tumbling at high velocity and shredding like many do. This flight was a great success.

Bob

 

[CCotner]

Lot's of optical illusions. This is not a frame transfer camera: It is an NTSC interline transfer array camera that is being processed on the fly. A Simple Fourier analysis doesn't account for the vertical and horizontal scan rates, scan angle, the exposure time blurring and the 60 Hz frame scan rate of the NTSC image. I'll leave the stroboscopic modeling to someone more mathematically inclined, but the video shows clearly that the rocket is coning and almost tumbling end over end at apogee.

How can we determine this? Due to an aerodynamic asymmetry, whether intentional or unintentional, we know the rocket is rotating because the FOV is rotating. We also know that unguided rockets undergo a gravity turn, and we know that the amplitude of the roll-yaw-pitch coupling is increasing with time as evidenced by the increasing vertical amplitude of the horizon position in the video. These effects illustrate the classical definition of inertial coupling aka coning. https://en.wikipedia.org/wiki/Inertia_coupling or mathematically describes in more detail here https://www.dtic.mil/dtic/tr/fulltext/u2/a349546.pdf

We can't determine if it is due to solely to asymmetric aerodynamics or if it enhanced by an asymmetric thrust vector from the propulsion system.

I've watched dozens of amateur high altitude unguided rocket flight videos and almost all exhibit some degree of coning which is difficult to prevent without active stabilization or a great deal of 6-DOF dynamic stability modeling. The rocket experienced minimal coning and reached apogee without wildly tumbling at high velocity and shredding like many do. This flight was a great success.

Bob

I can visually see the vehicle omega smoothly go from 0Hz on the pad up to no more than 2Hz, then back down to zero; then in reverse up to maybe 4 Hz, then back down to zero, then up to maybe 7Hz in the original direction; Assuming that the horizontal scan rate is fast enough as to be instant (very reasonable assumption for line-scanning cameras), and the vertical scan rate is at the slowest possible rate for 30fps video (1/30 of a second to scan the entire frame line by line), I don't see how it is possible that the rocket magically reached a high omega of 15+Hz in those first two seconds of flight while we were able to see the lakebed clearly and smoothly pan past the FOV. Sorry, but I don't buy it.

You say the amplitude of coupled motion increases with time because the envelope of (apparent) motion of horizon line in the frame grows. However, if the rocket is experiencing no coupled motion whatsoever: if it is experiencing only a slow pitch over and a rapid axial spin (that remains axial as it pitches), wouldn't the video look exactly like that? the horizon would remain fixed while vertical, then the horizon point would execute sinusoidal motion with frequency equal to the spin rate as the rocket pitched through increasing angles, until it is flying nearly horizontal at apogee and the fov is either sky or earth; just like in the video.

I guess what I'm saying is I understand just enough of his stuff that I'm going to need more convincing that your complex explanation for the spin and motion is true and my simpler explanation is not, because generally the simplest explanation is correct. So, do some math and make me see! don't give me movies of a largely unrelated trait of line-scanning video cameras that I already know. Show me some numbers! And explain clearly why the video looks the way it does and not the ay I described.


EDIT: that's a fantastic document, and I look forward to digging deeper into it in the near future. I wish the DTIC search engines weren't so terrible, I looked for a paper just like this months ago and darned if I couldn't find a single bloody paper, and here you have the exact thing I wanted all along. =p

 

[CarVac]

Starting at launch, the rocket is rolling towards the right (clockwise from the top).

At 18 seconds, the rocket reverses direction towards the left.

At 23 seconds, the rocket reverses direction again, and starts spinning towards the right. It keeps this direction of spin for the remainder of the ascent.

However, at 45 seconds-ish, the ground appears to be moving rightward (as if the rocket started turning left again). This is because the rocket is spinning REALLY DARNED FAST. You can see that it's not actually stopped because the clouds have been distorted into arcs from the rolling shutter.

 

[bobkrech]

I thought Bob was going to tell us how the direction of rotation changed from left, right, left in the beginning of the video. I'm not seeing how stopping a propellar from moving with video artifacts equates to the same effect seen in the rotation reversals. I understand fourier analysis. But the rocket rotated slowly one direction (not even a full turn) then, a little faster in the other direction (maybe 2 turns) retracing the same objects in the field of view, before finally steadily increasing the rate of rotation in the original direction until burnout. Seemed perfectly normal to me. I don't think the rocket was actually rotating many times faster than was seen (from the start) to create that effect.

Angular momentum has to be conserved. If you can describe the physics (and aerodynamics) that first spins an ascending rocket up in one direction, then reverses the rotation into the opposite direction and then reverses the rotation back again in the original direction then you know some physics that I don't.

There isn't any rotational reversal. The rocket spins up then the spins down in the same direction. It's same video aliasing effect you see in the Harvard start video where it looks like the propeller reverses direction several times due to aliasing. Smaller diameter rockets spin at a much higher spin rate than larger diameter rockets and the spin rate is proportional to 1/D. You see this in any figure skating competition when skaters pull their arms in to increase their spin rate. The comparable diameter Super Arcas was designed to have a burnout spin rate of 25 Hz to prevent coning so a several Hz spin rate is not unexpected or unusual.

If you look at butalane's video in post 12, you can clearly see the periodic corkscrew in the exhaust plume from coning. To my eye it appears that the exhaust spiral is relatively evenly spaced. The rocket is accelerating up during the video and if the spacing doesn't increase, the spin rate has to be increasing (more complete revolutions per unit time) to account for the constant spacing.

Bob

 

[BLKKROW]

I can visually see the vehicle omega smoothly go from 0Hz on the pad up to no more than 2Hz, then back down to zero; then in reverse up to maybe 4 Hz, then back down to zero, then up to maybe 7Hz in the original direction; Assuming that the horizontal scan rate is fast enough as to be instant (very reasonable assumption for line-scanning cameras), and the vertical scan rate is at the slowest possible rate for 30fps video (1/30 of a second to scan the entire frame line by line), I don't see how it is possible that the rocket magically reached a high omega of 15+Hz in those first two seconds of flight while we were able to see the lakebed clearly and smoothly pan past the FOV. Sorry, but I don't buy it.

You say the amplitude of coupled motion increases with time because the envelope of (apparent) motion of horizon line in the frame grows. However, if the rocket is experiencing no coupled motion whatsoever: if it is experiencing only a slow pitch over and a rapid axial spin (that remains axial as it pitches), wouldn't the video look exactly like that? the horizon would remain fixed while vertical, then the horizon point would execute sinusoidal motion with frequency equal to the spin rate as the rocket pitched through increasing angles, until it is flying nearly horizontal at apogee and the fov is either sky or earth; just like in the video.

I guess what I'm saying is I understand just enough of his stuff that I'm going to need more convincing that your complex explanation for the spin and motion is true and my simpler explanation is not, because generally the simplest explanation is correct. So, do some math and make me see! don't give me movies of a largely unrelated trait of line-scanning video cameras that I already know. Show me some numbers! And explain clearly why the video looks the way it does and not the ay I described.

If I understand your concern correctly you think the rocket is changing its Omega in a sporadic manner? Instead of the supposed Omega decreasing until the rocket begins to cone, that is if we ignore the sign or the direction of rotation?

Sorry just trying to get a clarification on your concern.

 

[CCotner]

If I understand your concern correctly you think the rocket is changing its Omega in a sporadic manner? Instead of the supposed Omega decreasing until the rocket begins to cone, that is if we ignore the sign or the direction of rotation?

Sorry just trying to get a clarification on your concern.

My concern is two-fold. The video shows an apparently impossible change in angular momentum of the vehicle, but the explanation of aliasing and folding plainly does not jive with my year of classes in systems behavior.

By my (admittedly poor) understanding of aliasing and folding after only three classes in systems and controls analysis, for a 30 Hz video to show signal folding the vehicle RPM must exceed 15 Hz; and up to that speed, the video should (as one goes frame by frame) show clearly increasing angular speed as determined by the passage of terrain landmarks. When I step frame-by-frame through the first few seconds of vehicle flight (up until it appears to reverse direction), that is simply not what I see; I see a maximum omega of about 2 Hz. I am not a video analysis expert. This implies to me that it did change direction, contradicting apparent momentum conservation, which is unlikely.

Yes, conservation of momentum implies that the rocket should not change spin directions. My only high-performance rocket flight (that I have personally been a key part of) had a gyroscope on it sampled at 120 Hz (therefore nominally accurate performance up to about 60Hz of spin). The data from my flight clearly shows the vehicle begin spinning clockwise, then stop, then counterclockwise, all at no more than .5 Hz, right up until the motor coughed out and the flight went screwy. Can I explain the fundamental aerodynamic principle which caused it to reverse directions? no. Is it possible that my gyroscope was incorrect? it is extraordinarily unlikely but sure, it's possible. Does this mean that Butalene's rocket might have reversed spin direction? possibly, but he didn't fly a gyro, so we can't know.

In the end, what do we have? I can believe my rudimentary understanding of system behavior and fourier analysis, my video analysis and my eyes, and accept that an unknown aerodynamic phenomena that I have previous data of caused the vehicle to spin very slowly and to change direction, despite not knowing why exactly it did. Or, I can believe that my understanding of signal folding and aliasing is incorrect despite a year and a half of systems classes, my video analysis and feature recognition skills are wildly off, and that my previous data is either a fluke, in error, or irrelevant, and that the vehicle spun very, very fast during its ascent.

I don't feel any more comfortable ignoring my education and the evidence of my eyes than I do explaining away momentum conservation by some mysterious (though not unprecedented) aerodynamic force I know I don't understand.

Were it my project, I would be deeply concerned by my inability to solve the dilemma. It is not. Hence, I wash my hands of this weirdness.
https://i4.photobucket.com/albums/y...gifjacksparrowiwashmyhandsofthisweirdness.gif

 

[tuxxi]

In this video, there is clearly rolling in both directions. The roll rate is <10hz for both the sustainer and the booster, so I don't understand how the bi-directional rolling could be caused by aliasing of the cameras

sidenote- i'm tired and i skimmed the thread - this may or may not actually be what you gents are discussing here.

[video=youtube;krNFHr2YZBY]https://www.youtube.com/watch?v=krNFHr2YZBY[/video]

 

[XJChris]

Excellent point MarkH, the aerodynamics certainly changed during the M3+ ascension. We've already seen photos of how the fins eroded and changed shape. Also, there is evidence of tremendous heat generation which could have some effect on the fins or fin joints or main body. And at M3+ I'm sure the smallest change in any part of the rocket would have large effects on the overall aerodynamics.