So, maybe I'll try a three-stager

Charles_McG said:

If I understand the analysis of one of the other high altitude flights (maybe even one of Jim’s) keeping the spin all the way up is an issue because angular momentum naturally transfers to the longest axis. So spin becomes tumble.

The other thing is that there’s no air for the fins to work on. If I understood the post about two up, the rocket went 90 degrees sideways at Mach 2. Not possible in substantial air.

If anything, I’m waiting to see them try a de-spin mechanism.

Click to expand...

You might be remembering the three stage flight from last year that tumbled.



I think spin will eventually become tumbling, but coning will for sure. I think the idea, if it is not possible to spin the rocket, is to delay the onset of tumbling as long as possible. In the three stage flight, tumbling started at 140K and cost approximately 10K in altitude. This flight has significant angle of attack at 100K. Between the coning and the subsequent tumbling, we lost at least 90K in altitude. So, either spin or don't spin, but don't spin around 2 hz. I wonder if rollerons are an option?

Jim

RCS

Charles_McG said:

RCS

Click to expand...

I have a buddy who is supposed to looking into that (you know who you are). Maybe at some point.

Jim

Jim,
one obvious way to reduce spin is to make sure your fins are straight and symmetrical. I'm sure you are already working towards that, but have you measured it? Some budgeted fin misalignment would be something to add to the flight sim. I put a lot of effort into perfectly straight fins [as have some other folks] and it seems to pay off.

Also your thrust curves don't actually look anything like what was used in the flight sim. It would be interesting to see the projected altitude with the as-flown thrust curves compared to the initial prediction.

If you do try 'spin stabilized', yo-yo despin is pretty straightforward in terms of the physics and the mechanism. Spin balancing is another matter, and precession is then really becomes your adversary.

br/

Tony

TonyL said:

Jim,
one obvious way to reduce spin is to make sure your fins are straight and symmetrical. I'm sure you are already working towards that, but have you measured it? Some budgeted fin misalignment would be something to add to the flight sim. I put a lot of effort into perfectly straight fins [as have some other folks] and it seems to pay off.

Also your thrust curves don't actually look anything like what was used in the flight sim. It would be interesting to see the projected altitude with the as-flown thrust curves compared to the initial prediction.

If you do try 'spin stabilized', yo-yo despin is pretty straightforward in terms of the physics and the mechanism. Spin balancing is another matter, and precession is then really becomes your adversary.

br/

Tony

Click to expand...

Fins being straight and symmetrical is not easy to measure on a rocket with a fin can. I did try to do this using multiple approaches, since I didn't build the fin can, but it's hard to get a good frame of reference to verify what straight really is. Not sure I'm explaining this well, but hypothetically, what would you do to verify this if you had to?

I don't believe you can use the thrustcurve to represent the motor if there is drag that isn't quantified. What I did do is to manipulate the booster motor file impulse to match the actual performance. What I learned is that the total impulse of the booster motor was as predicted, just a different curve shape. Can't really do that for the sustainer, but it was the same propellant made in the same way.

I would like to try yo yo despin at some point. As you say, it's pretty straightforward. I just can't visualize doing spin balancing properly.

Jim

JimJarvis50 said:

Fins being straight and symmetrical is not easy to measure on a rocket with a fin can. I did try to do this using multiple approaches, since I didn't build the fin can, but it's hard to get a good frame of reference to verify what straight really is. Not sure I'm explaining this well, but hypothetically, what would you do to verify this if you had to?

Jim

Click to expand...

I've never done this so take it with a huge block of salt...

I've seen machinists mount a shaft on a lathe and then use a dial indicator to adjust a 4-jaw chuck so that the shaft is perfectly (within 0.001") centered. They would then run the indicator down the shaft to measure run-out. He could get less than 0.002" run-out over a 6' long shaft. Could you mount the airframe into something like a lathe and use a dial indicator to first ensure the airframe is mounted parallel to the indicator and then run it down the fin to see if the fin is mounted 100% parallel to the airframe (aka straight). It would then also be trivial to see if the fins were also symmetrical around the airframe as most lathes have indexes at regular intervals (ex. every 10 degrees)

Like I said, I've never done it and I don't have access to large lathes so I don't know how difficult it would be for you to do either.

(edit: or like Tony says as his method sounds alot cheaper and probably easier)

Hi Jim,
estimating the thrust curve is an iterative process. The accel plot is not it, but the sim provides drag vs velocity. One makes an estimate, run the sim, compare to flight and adjust until the two converge 'close enough'. Something is always unknown, but you can back out a thrust curve to a knowable quality from the flight data.

For fin straightness you need to locate the center line of the rocket and then measure the distance of the fin surface from it. I use a very straight 6 ft level [that I take good care of] for my straight edge and calipers for finding the center line. I shim the level to parallel with the center line using gauge blocks [not that expensive on ebay] but one could use shim stock, sheet metal or paper if needed. I use a dial indicator on a base to get the relative distances from the straight edge. If they are equal then the fin is straight.

a diagram would probably help but that will be a while.

br/

Tony

heada said:

I've never done this so take it with a huge block of salt...

I've seen machinists mount a shaft on a lathe and then use a dial indicator to adjust a 4-jaw chuck so that the shaft is perfectly (within 0.001") centered. They would then run the indicator down the shaft to measure run-out. He could get less than 0.002" run-out over a 6' long shaft. Could you mount the airframe into something like a lathe and use a dial indicator to first ensure the airframe is mounted parallel to the indicator and then run it down the fin to see if the fin is mounted 100% parallel to the airframe (aka straight). It would then also be trivial to see if the fins were also symmetrical around the airframe as most lathes have indexes at regular intervals (ex. every 10 degrees)

Like I said, I've never done it and I don't have access to large lathes so I don't know how difficult it would be for you to do either.

(edit: or like Tony says as his method sounds alot cheaper and probably easier)

Click to expand...

Well, take my answer with a small grain of sand....

The fin can was made by a very good machinist (Jerry McKinlay) using a really big lathe. There were tabs cut for the fins and I suspect that they were perfectly aligned. At one point, we were considering spinning the rocket, and Jerry asked me what fin angle I wanted, to the nearest 10th of a degree. What I am less sure about are the materials themselves. Aluminum tube is not perfect. I don't know the details of the machining of it, and I would only be speculating as I'm not a machinist, but I can imagine difficulties in trying to ensure that everything is exactly square with the centerline down the tube. So, I am just wondering if there is a way to check that other than the obvious suspects. I've tried lasers, gravity, angle iron, etc., but i find it difficult to determine mechanically exactly where straight is.

Jim

So 1 degree is 17.5 milliradians, and 0.1 degree is 1.75 milliradians, or nearly 2 thousandths of offset for every inch of root length. A 10" long fin root would be offset 0.0175" at 0.1 degrees, that seems like a lot. I'm sure the machinist did better than that on the can, but as I think you mentioned, tolerances add up. If you need fins straight, I think it is worth investing a big in being able to tell if they are.

Just as a reference, I aligned my fins to better than 100 microradians [and a confidence of about +/-100 microradians]. For the part of my flight that is on video, there was no visible rotation of the rocket and no obvious rotation in the plume trail. My point is with careful measurements using basic machinist tools, one can do very accurate work.

My suggested shopping list:

1) 6" calipers [dial or digital, Mitutoyo or comparable brand, NOT from HF or other discount stores], get 8" ones if you have 'regular size' motors, but you will also need jaw extensions [brake disk adapters].
2) a dial indicator with a 2" face and ~1" travel preferred, 0.001" resolution. Fancier ones can work, but it is a long way from the far side of the straight edge to the fin so they might not reach without a longer bar.
3) a magnetic base for the indicator.
4) a couple sets of 1-2-3 blocks, cheap as you can find are usually fine. Very handy for holding things up and spacing things.
5) a very straight level, I got mine from Lowe's, I went to every Lowe's in town and compared their levels to each other and bought the straightest one. Hold two lightly against each other face to face and back to back. If there is no gap both ways then they are both pretty straight.
6) a set of precision gauge blocks from eBay.
7) a granite countertop. Makes a very serviceable surface plate, but a very flat floor/countertop/plywood is better than nothing.

br/

Tony

Oh, and to answer your earlier question: carefully shim a motor case [with something like kapton tape] until it is a close sliding fit and stick it in the body tube so that some sticks out. That will give you a cylindrical reference for the back end, the body should work for the front end. Measure the roundness of the motor case to be sure you are on a high or low point on the diameter [a high point is better]. If any part of the body is not uniform enough to make a good reference for finding the center line, then you have revealed a potential source alignment issues.

JimJarvis50 said:

I also added some redundant electronics just in case Kate has a brain freeze (not going to happen though).

Jim

Click to expand...

Jim, what electronics did you end up using as a back up to Kate?

ChuckH said:

Jim, what electronics did you end up using as a back up to Kate?

Click to expand...

I just used a simple timer for Hail Mary deployment at apogee and a PFCF for redundant main. You can hear the timer "fire" well after the actual apogee (it is what causes the siren sound).

Jim

JimJarvis50 said:

I just used a simple timer for Hail Mary deployment at apogee and a PFCF for redundant main. You can hear the timer "fire" well after the actual apogee (it is what causes the siren sound).

Jim

Click to expand...

Interesting. Did you go with a timer because current altimeters (other than Kate) cannot reliably detect apogee at the expected altitude?

ChuckH said:

Interesting. Did you go with a timer because current altimeters (other than Kate) cannot reliably detect apogee at the expected altitude?

Click to expand...

I went with a timer to have something that would absolutely fire and wouldn't depend on "detecting" anything. If we had not used Kate, I would have used timers anyway, along with barometric apogee if apogee was detected at less than 90K feet. I might have used integrated altitude instead, except that I expected tumbling, and integrated altitude would become inaccurate if the tumbling rate was high enough (with the electronics in the cone, tumbling would look like motor acceleration and deployment would be late).

Jim

TonyL said:

Hi Jim,
estimating the thrust curve is an iterative process. The accel plot is not it, but the sim provides drag vs velocity. One makes an estimate, run the sim, compare to flight and adjust until the two converge 'close enough'. Something is always unknown, but you can back out a thrust curve to a knowable quality from the flight data.

For fin straightness you need to locate the center line of the rocket and then measure the distance of the fin surface from it. I use a very straight 6 ft level [that I take good care of] for my straight edge and calipers for finding the center line. I shim the level to parallel with the center line using gauge blocks [not that expensive on ebay] but one could use shim stock, sheet metal or paper if needed. I use a dial indicator on a base to get the relative distances from the straight edge. If they are equal then the fin is straight.

a diagram would probably help but that will be a while.

br/

Tony

Click to expand...

yes, a diagram and a maybe a whole thread in the techniques forum would be super informative

Every fillet build i've seen involves at least some hand finishing, wouldn't non-symmetrical fillets mess up even perfectly straight fins? Do you have a method of producing fillets as symmetrical as your fins are straight?

TonyL said:

So 1 degree is 17.5 milliradians, and 0.1 degree is 1.75 milliradians, or nearly 2 thousandths of offset for every inch of root length. A 10" long fin root would be offset 0.0175" at 0.1 degrees, that seems like a lot. I'm sure the machinist did better than that on the can, but as I think you mentioned, tolerances add up. If you need fins straight, I think it is worth investing a big in being able to tell if they are.

Just as a reference, I aligned my fins to better than 100 microradians [and a confidence of about +/-100 microradians]. For the part of my flight that is on video, there was no visible rotation of the rocket and no obvious rotation in the plume trail. My point is with careful measurements using basic machinist tools, one can do very accurate work.

My suggested shopping list:

1) 6" calipers [dial or digital, Mitutoyo or comparable brand, NOT from HF or other discount stores], get 8" ones if you have 'regular size' motors, but you will also need jaw extensions [brake disk adapters].
2) a dial indicator with a 2" face and ~1" travel preferred, 0.001" resolution. Fancier ones can work, but it is a long way from the far side of the straight edge to the fin so they might not reach without a longer bar.
3) a magnetic base for the indicator.
4) a couple sets of 1-2-3 blocks, cheap as you can find are usually fine. Very handy for holding things up and spacing things.
5) a very straight level, I got mine from Lowe's, I went to every Lowe's in town and compared their levels to each other and bought the straightest one. Hold two lightly against each other face to face and back to back. If there is no gap both ways then they are both pretty straight.
6) a set of precision gauge blocks from eBay.
7) a granite countertop. Makes a very serviceable surface plate, but a very flat floor/countertop/plywood is better than nothing.

br/

Tony

Click to expand...

JimJarvis50 said:

I went with a timer to have something that would absolutely fire and wouldn't depend on "detecting" anything. If we had not used Kate, I would have used timers anyway, along with barometric apogee if apogee was detected at less than 90K feet. I might have used integrated altitude instead, except that I expected tumbling, and integrated altitude would become inaccurate if the tumbling rate was high enough (with the electronics in the cone, tumbling would look like motor acceleration and deployment would be late).

Jim

Click to expand...

If you ever need a really good straight edge, the 36" Veritas straight edge is an excellent tool. Specs say to 0.0015" over 36". I even got it out for a rocketry project recently.

Granite countertops are not really that flat, though you might like to think they are. Some may be, and a few will have bad spots in them. It all depends on what it took to get them polished smooth. Definitely would not trust one for any critical measuring. A reference granite surface plate is a great tool. That, and a machinists square will tell you when your airfame is plumb to the plate.

Chris_H,
Granite countertops are typically mass produced on large automated systems that crank them out from blocks about half the size of an 18 wheeler, sliced like bread, and run through very large polishers and then stuffed into sea containers for export. I have been to a factory overseas that does this. This approach promotes very uniform results, and so while it does not guarantee any particular flatness, the results are generally flat. The cheapest granite countertop I could find indeed had flaws and still was very flat, as expected.

To be clear, one does not need to be 'plumb' or square to the "surface plate". One needs to identify the centerline of the airframe, and those things are not the same. One can assume that the reference surface is not flat, that the straight edge is not straight [or rigid] and that the airframe is not round, or cylindrical for that matter, nor is the fin flat or straight either. One does need to estimate how flat, straight or round things are in order to use that information to improve the quality of one's measurements. One also can estimate the errors inherent in the measurement technique such as flex of the straight edge and misalignment of the dial gauge. One can even account for the flatness of the surface depending on the characteristics of its variation.

Chad,
You are correct regarding hand finishing, some variation is always possible. Also velocity gradients in the air currents will produce a net force so some spin is always going to be expected. My fin brackets are machined aluminum for structural reasons, so they are pretty consistent. Folks doing fillets can look at developing a very uniform technique and/or a way of using a reference gauge to get as symmetrical as possible.

br/

Tony

TonyL said:

Chris_H,
Granite countertops are typically mass produced on large automated systems that crank them out from blocks about half the size of an 18 wheeler, sliced like bread, and run through very large polishers and then stuffed into sea containers for export. I have been to a factory overseas that does this. This approach promotes very uniform results, and so while it does not guarantee any particular flatness, the results are generally flat. The cheapest granite countertop I could find indeed had flaws and still was very flat, as expected.

To be clear, one does not need to be 'plumb' or square to the "surface plate". One needs to identify the centerline of the airframe, and those things are not the same. One can assume that the reference surface is not flat, that the straight edge is not straight [or rigid] and that the airframe is not round, or cylindrical for that matter, nor is the fin flat or straight either. One does need to estimate how flat, straight or round things are in order to use that information to improve the quality of one's measurements. One also can estimate the errors inherent in the measurement technique such as flex of the straight edge and misalignment of the dial gauge. One can even account for the flatness of the surface depending on the characteristics of its variation.

Chad,
You are correct regarding hand finishing, some variation is always possible. Also velocity gradients in the air currents will produce a net force so some spin is always going to be expected. My fin brackets are machined aluminum for structural reasons, so they are pretty consistent. Folks doing fillets can look at developing a very uniform technique and/or a way of using a reference gauge to get as symmetrical as possible.

br/

Tony

Click to expand...

My comment on granite countertop stock not being a guaranteed flat is based on time I have spent in a friends shop, surfacing slate for making turntables. Yes, the slabs are fed though a big sander /polisher, but a lot of granite has flaws, and the equipment is designed to be able to fix this easily by focusing on individual spots. Yes, granite countertop material can be used as a flat surface. I have used it for this, and still do sometimes. I mainly use my granite measuring plates, unfortunately the biggest one is 18" x 24".

Also, at least some of the machinery used for surfacing these slabs uses a platen, which is a wear part.

And in commenting on the flatness of countertop granite as a measuring reference, I was imagining up a jig for measuring fin alignment, possibly down to 0.001". I can see a few ways to set this up, and almost commented on that. I would not want to go through the effort to make jigs and setup for measuring fins, and have a countertop surface that could cause a problem. The straight edge that I linked to could be used to check the flatness of a countertop.

Yes, I understand about aligning to the axis of the airframe. That is not difficult in concept. Knowing that an airframe end is cut square is also useful, and that was what I was commenting on. 'Plumb' as describing the relationship between an airframe and a plane is best measured from the axis of the airframe, so maybe I used the word in a way that confused you. However it is spelled out, there is some sort of an axis, and a perpindicular plane to the axis is a valuable reference. In that contect, the axis is 'plumb' to the measuring plane. Maybe I used that word too loosely? If so, sorry.

I work with precision measuring tools every day, and spend a good bit of time in front of lathes and mills and such. I specialize in precision bearing work, and commonly work in 'tenths'.

So yeah, if I am going to the extent of 'blueprinting' fin alignment, a countertop might be useful for a worksurface, but the unreliability of it as a flat surface would need to be checked and compensated for. If it needs to be flat, my straght edge told me that countertop granite is not reliably flat. I have a sink cutout that I picked up as a flat surface for my vacuum bag, it has wavyness that approaches 0.010".

A machinists level combined with a good straight edge would be a useful tool for this task.

Chris_H,
nice to know we both have cool jobs However what we 'could do' at work is different from what I actually do in my garage because I have to make do with what is available. Your sink cutout seems like a good choice, except you apparently got a crappy one. Something that has been spot polished is definitely undesirable because it disrupts the uniformity of the surface error.

The counter in my kitchen is flat to ~.001 over a running 12" or better, as is my bargain granite slab [so about 170 urad of worst case tilt]. One can take advantage of consistency and not need to eat the 170 urad of tilt if the material is generally 'dished' or arched, and so the induced error at one end is about the same as the induced error at the other. For my setup the worst case theoretical induced error would be about 7 tenths, so I am still pretty happy. Sure my surface 'could' have been worse, but if it is known, it still would be effective.

br/

Tony

Yep. Lot's of ways to skin a cat, so to speak.

I like to put together videos of the Balls flights each year. Since the project this year got put on hold, I decided to put together a second video of the flight from two years ago (the three-stager to 175,000 feet). I have lots of extra video and pics. This second video has more footage from the flight line and more emphasis on the stabilization system, that performed well on this flight. Enjoy!

Jim

JimJarvis50 said:

I like to put together videos of the Balls flights each year. Since the project this year got put on hold, I decided to put together a second video of the flight from two years ago (the three-stager to 175,000 feet). I have lots of extra video and pics. This second video has more footage from the flight line and more emphasis on the stabilization system, that performed well on this flight. Enjoy!

Jim

Click to expand...

I like to watch them thanks Jim!

Awesome video, and amazing to watch. The amount of work that you put into these systems and rockets is truly amazing.

Balls is coming up soon! Woohoo! Sometimes I try a three stager and sometimes I go with two. The 2018 three-stager was a pretty good flight, but in the 2019 two-stager, we left some meat on the bone. The sustainer coned, and we only got 142K of what might have been a much higher flight. Since there is room for improvement, we're going to try the two stager again.

In 2019, it was the SVJ2 (for Stu, Vern, Jerrry and Jim). This year, it will be the SVJ3, unless someone on the team can come up with a better name. Stu made the motors (P to O) in 2019. They worked quite well, but it was a tough build. So this year, we have a new guy on the propellant (Q to O this time). He doesn't really want to be involved though. When I ask him why, he just says it's none of my bismuth, but I'm confident we're going to have great motors.

Some of the preparation of the rocket was done last year, before Balls was cancelled. This year has been simulations and paperwork (now submitted to FAA). But the preparation is now going full speed, including a handful of things that will hopefully stave off coning, and I was reminded this week of just how complex this project is. A year off isn't helping either, but we'll get there.

Jim

Sounds like your motor man knows his bismuth.
Good luck

FredA said:

Sounds like your motor man knows his bismuth.
Good luck

Click to expand...

Yep. I think that was demonstrated over the weekend.

Jim

I'd love to hear about your anti-coning secret sauce(s)

Finicky said:

I'd love to hear about your anti-coning secret sauce(s)

Click to expand...

It is my belief that the rocket in 2019 exhibited pitch-roll coupling. The roll frequency was 2-3 revolutions per second (if I recall). Generally, rockets should spin at 6+ rps or less than 1.5 rps to avoid the range where coupling can occur. So, one strategy is to try and avoid spin. Previously, we discussed trying to measure the angle of the fins to see if that could be an issue. They are on very straight; however, it is possible to measure the small errors that are still there (given that the motor tube itself is not even perfectly straight). I have done the measurements and found the rotation of the spin can that minimizes the total error in the fin angles. I don't know if this will be an improvement since I don't know where the fins were set of the 2019 flight, but it is as good as it can be.

The 2019 flight also used some really crummy ablative. Odds are this contributed to some spin. The ablative has been removed. I get lots of opinions on the need for ablative on aluminum fins. These are fairly thick at 3/16", and we'll try to light the sustainer as high as reasonbly possible and extend the burn of the sustainer motor as long as feasible.

I've also gone through and tried to balance the rocket as best I can. We added some back up electronics in the cone prior to the 2019 flight (to back up Kate) but didn't rebalance the cone. Also, there is a bulkhead in the normal altimeter bay location that contains the deployment charges and the eyebolts for both the drogue and main harnesses. The eyebolts are not centered. So, I have tried to separately balance the cone and the upper air frame. About 100 grams of weight has been added to roughly offset the weight and weight radius of the imbalances. I'm not going to be able to balance the motor case (I don't have it), but I can probably try to balance the fin can. Just haven't tried that yet.

So, that's my list at present. I don't think there is much I can do with the chutes except to pack the main so that it isn't loose in the air frame. Any other suggestions would be welcome.

Jim

I vaguely recall reading that having more than 4 fins can help, but I'm pretty sure this is not a direction you want to go for a rocket going this fast and high !

Finicky said:

I vaguely recall reading that having more than 4 fins can help, but I'm pretty sure this is not a direction you want to go for a rocket going this fast and high !

Click to expand...

I believe you're right, but more than 4 is not going to happen on this flight. At least it's 4 and not 3. Pretty hefty altitude penalty to go to 6.

Anyway, the guy that's making the motors, who isn't really involved, has an interesting way of assembling the motors. Basically no casting tubes. I asked him how it works and he offered an explanation, to which I said "huh?". Second explanation... "huh?". Finally, I said just draw me a picture, and keep it simple. He did that, and I get it!

Jim