Short version of the story: The shock cord broke, so no record.
Long version of the story:
The launch was delayed first by my own technical difficulties and then by some serious wind. Just when we were getting close to wrapping up for the day, the wind calmed down to about 10 mph and I got the F record attempt bird ("Shuttlecock in a Haystack") all set up. The liftoff was straight and smooth, and with the F10's characteristic lack of noise and smoke. Not too much drama except that it just kept accelerating all the way through the point where it winked out of the sky.
I had a strong signal from the BeeLine, though I think I could hear a little modulation from the spin rate. The signal stayed good well past a minute, so no ballistic re-entry. But then the signal stopped about 2.5 minutes after liftoff, instead of about 4. Hmmm. I walked up to the top of the SW ridge, still no signal. I walked all around on the ridge, sweeping with the antenna all around, still no signal. I give up the walking and head back down for some advice as to where to go to keep trying. Joe points out that the wind probably carried it pretty far East of where it looked like it was heading (SW), so I hop in the car and take a left onto the main road, and stop on the ridge. Hop the fence, and then Eureka, the much-anticipated signal. I was pleasantly surprised to find that the audible locating beep function of the Parrot alt picks up about where the BeeLine signal becomes non-directional from being too close. I walk toward the sound, but where's the parachute? Uh oh. There's a 29mm nosecone laying in the dirt. The beeline and the Parrot seem to be happy in there, though.
On the off chance I might get lucky, I walked about a mile from there, straight downwind, but never saw the body tube or the chute.
Here's my first cut at the flight data. The match between the accel and baro data is so good only because I tweaked the accel data for the purpose of getting a more accurate coefficient of drag. More on that later, if anyone is interested. The baro data is from the "factory" calibration. It looks like there is some trans-sonic and/or acceleration effect in the pressure signal. And my accelerometer-based mach inhibit worked.
A quick look a the data also explains why my shock cord broke. I was worried that I didn't have the right amount of powder, though I only used a teeny bit. See below.
The cord actually survived the first bounce (in addition to a test flight), and the force works out to be only about 2 lbs. But notice that the bounce that caused the separation also had a large lateral force. I checked where the cord broke, and sure enough, the stretched-out end of it lines up with the edge of my fiberglass-reinforced charge holder/piston, so I think it just got cut. This is round nylon/elastic, by the way, a short section of non-Kevlar to spread out the impulse over time. The deployment happened at about 90 mph, mostly due to weather-cocking despite a small downwind tilt of the pad. 9000 feet might be reachable on a calm day in the summer.
I assume you're talking about a 29mm bird flying on an Apogee F10? If so, I have to admit some skepticism about that altitude.
While the F10 is nearly a 100% F motor, that altitude is well beyond any current F altitude record. The current Tripoli F record and John Wilke's NCR club record were all set with F32 or F79 motors; 24mm near 100% F motors flown in 24mm minimum diameter birds. Optimum flight weight for a bird flown on the F10 is well below that for F32 or F79 powered birds due to the low initial thrust and longer burn, yet you flew with a GPS and tracking transmitter.
The additional drag caused by the larger diameter airframe would pretty much make your reported altitude an impossibility even with absolutely ideal conditions not to mention the additional weight of the GPS and transmitter. Not that I disbelieve your data, but I just can't see how you could get such an altitude from an F10.
I'd really like to see how your altimeter compares to a stock Missileworks or Perfectflight altimeter flown at the same time in the same bird. I'll certainly volunteer to fly a Parrot in one of my larger birds with a Missileworks mini-RRC2x and a Perfectflight MAWD. While both of them are far less sophisticated altimeters than your unit, they are the standard against which all others are judged. Maybe they read consistently low?
Did you ever add the temperature compensation code to your unit? It's such a sophisticated unit and the feature set just makes me drool. You should turn the design into a full flight computer with multiple pyro outputs and add timing and pull pin capability.
Warren
I assume you're talking about a 29mm bird flying on an Apogee F10? If so, I have to admit some skepticism about that altitude.
While the F10 is nearly a 100% F motor, that altitude is well beyond any current F altitude record. The current Tripoli F record and John Wilke's NCR club record were all set with F32 or F79 motors; 24mm near 100% F motors flown in 24mm minimum diameter birds. Optimum flight weight for a bird flown on the F10 is well below that for F32 or F79 powered birds due to the low initial thrust and longer burn, yet you flew with a GPS and tracking transmitter.
It didn't have GPS, just the BeeLine tracker and the Parrot. The rocket mass, minus the motor, was right about 40 grams. With the motor, it was about 120 grams, which is pretty much the optimal mass for an F10. Rocksim predicts 8000-9000 feet for this design, depending on the crosswind and surface finish. I've also got calibration verification data for this particular altimeter, if you're interested.
The additional drag caused by the larger diameter airframe would pretty much make your reported altitude an impossibility even with absolutely ideal conditions not to mention the additional weight of the GPS and transmitter. Not that I disbelieve your data, but I just can't see how you could get such an altitude from an F10.
Warren, you're making my day; thanks! 😀 Doing something hard isn't nearly as fun as it is when someone says it's impossible. Go ahead and sim a 29mm F10 rocket with a Cd of 0.25 and a non-motor mass of 40 grams and let us know what Rocksim tells you. (ROTFL)
I'd really like to see how your altimeter compares to a stock Missileworks or Perfectflight altimeter flown at the same time in the same bird. I'll certainly volunteer to fly a Parrot in one of my larger birds with a Missileworks mini-RRC2x and a Perfectflight MAWD. While both of them are far less sophisticated altimeters than your unit, they are the standard against which all others are judged. Maybe they read consistently low?
I'd like to see that, too. If the RRC and the Perfectlflite are accurate, they will read the same as the Parrot.
Did you ever add the temperature compensation code to your unit?
Yes. The Parrot reads the temperature and updates the scale and offset coefficients for each of the other measurements at 50 Hz during the flight. The coefficients used for the temperature compensation were calculated based on doing the full pressure and accel calibrations at two different temperatures. For the Parrot, the scale factor changes by around 1.6% for the accel and 3% for the baro unit for a 40 degree F delta temperature. I'd also love to see someone do an experiment with one of the other altimeters to see what their temperature sensitivity is.
It's such a sophisticated unit and the feature set just makes me drool.
Wow, thanks. I'm blushing.
You should turn the design into a full flight computer with multiple pyro outputs and add timing and pull pin capability.
Hey, I thought you said you would think about buying one once I put apogee deployment in. 🙄 😀 That's o.k, the record book is still wide open for you to win one. I had about $1000 of prototyping costs in developing the Parrot and much more than that in lost work time, so I'm not in a hurry to make another version.
It has sufficient current to reliably fire e-matches? If so, I've got an F32 and a bird to fly it in come the December launch. Will it fit 24mm airframes/noses? I seem to remember it will.
Warren
Yes, the on-board switch is rated for 4A continuous. But it also depends on what battery you use for the deployment; deployments require a separate battery from the Parrot's on-board battery. I have been using the same-size (65 mAhr) Li-poly cell I use in the Parrot, which only weighs about 1.2 grams, and I also use it to power the BeeLine transmitter. It fires Estes ignitors reliably, which require more current than any ematch I'm aware of. A 9V would work too. I'm not sure about an A23 cell; it would probably depend on e-match.
I have gotten a Parrot and a Beeline side-by-side into a 24mm nosecone. The Parrot fits by itself into an 18mm nosecone. See the Parrot Challenge thread for dimensions.
I'm curious how the F32 will go. I had agreed with my wife to make the November launch the last one for a little while, but I'm really itching to give it another shot, and it would be fun to be there for your record attempt, also.
By the way, I have re-started the email dialog with Tom Rouse about certification, so hopefully I'll hear something more definite soon.
Conrats, Adrian - It is a monster boost by any measure!. Sorry (truly!) that it didn't hang together. Go with Kevlar, ground testing to keep bp to a minimum, and a wee wrap of duct tape where the rub occurs and you will keep things together.
One question I did have -- how do you calculate your Cd at 0.25? For a reference, my J record rocket was w/in 1% of simmed altitude, and Rocksim predicts a Cd of .66465. My H rocket, which was within 0.10% of simmed altitude had a projected Cd of over .72. Bottom line is that all my rockets, which are notorious for tiny fins, (and are all weighed to the gram before plugging into Rocksim) have a Cd of 0.6000++. In some cases the Cd is significantly higher, often 0.7000+. These rockets are often wet-sanded on site, etc. I can't even imagine a Cd of 0.25 as you report? I'm curious if that is a plugged value or if Rocksim is kicking it out?
Warren is right on - if you have accuracy on your numbers and all other things being equal, the 24mm full F will destroy the 29mm F. The 29mm is just too big of a hole to punch vs. 24mm. Pi x R x R for the two diameters means a 29mm motor must cut a hole that is 46% larger vs. 24mm motor. That is an extraordinary amount of baggage to overcome... Minimum diameter is key to any altitude attempt.
FWIW, on my F32 boost, I was right at optimal mass. I got 6K (sims for that flight were way higher, though only about 7,000')
My first estimate of 0.25 came from the accelerometer measurements during the coast. I went to Rocksim to compare, and Rocksim predicts 0.28 at 500 mph. Then I realized that I assumed a 5400 foot ASL density for the whole flight. I'll update my post-flight analysis to calculate the density as I go along, but for now I stuck in a value for the density at 11,000 feet above sea level and got a measured value of about 0.29 at 500 mph.
Check out Rocksim's Cd analysis. It's pretty informative. The base drag, which is the only part directly tied to the cross-sectional area, is 0.09. The rest of the drag is from the nosecone, body, and fins, which are all pretty much proportional to the wetted area, given subsonic flight. So going to a 29mm diameter would increase the base drag by 46%, but would only increase the overall drag by about 15%, if the total wetted surface area were the same. And the "Shuttlecock in a Haystack" had really low total surface area, because the body tube was only 4.5" long. Add in the fact that the bird stayed in the lowest-drag, subsonic part of the Cd curve the whole time, and you get a flight that beat the old TRA record by 25%.
I agree that you need to stay subsonic, or get (and stay) supersonic for a long, long time. The getting-and-staying supersonic is tough. Not many motors fit that bill.
Getting your altimeters sanctioned by TRA would be a terrific thing. Adding the ability to fire an ematch was a major piece of the puzzle as well -- Nice job!
Side thread... Many flyers that I look up to constantly warn about Mach 2. I've seen a lot of shreds with rockets at what is probably about M2. I wonder what the big deal is? I wonder if Max Q is the real culprit? Hard to figure. I do know that M2 is really hard to reach from a logistical stanpoint. I'm not sure what my fastest boost ever was, but I'm pretty sure it was well below M2.
JW
Well, as far as Cd goes, my Cirrus (which was optimized for super/transsonic speeds) had a supersonic cd of 0.63 or so, and a transsonic cd of 0.70. Subsonic however was in the .4 range. For an optimized subsonic rocket, I could believe 0.3 or less.
I put in the density calculation into my post-flight analysis, and got the following plot of Cd as a function of velocity:
This is the same basic shape of the curve as predicted by Rocksim. Tonight I may try to see if I can plot the Rocksim Cd on the same chart.
This Cd was calculated by measuring the acceleration after burnout, and using the burnout mass along with F=MA to derive the drag force. I also calculated what the drag force would be for a Cd of 1 using 1/2 density*V^2*Area, and the Cd is then the ratio of those two values. In my spreadsheet now, the density is looked up as a function of altitude above sea level. The reference area I used was a 29mm diameter circle.
The Cd is higher now (0.3) since I adjusted the density for both altitude and temperature. (The 65 deg air temperature at launch is about 25 degrees F higher than the standard atmosphere value)
Very interesting 🙂
I notice that you got better data at low speed than I did with my Cirrus - likely due to the lighter rocket and lower altitude. Very nice cd though - 0.3 is quite impressive. I hope you can make it next month with the G25.
I filtered the accel data with a simple moving-window average before using it to calculate the Cd. I think you're right that the lighter rocket and lower altitude would help also.
In the analysis I'm also propagating the vertical and horizontal velocities independently, in order to estimate of the horizontal velocity at apogee. In this last F flight, for example, I had about 90 mph of horizontal velocity at apogee, because of the long burn into the wind. In the analysis I put in an angle from vertical right after it leaves the launch rail, representing the post-weathercocking angle. Then gravity turns the trajectory over from there. I adjust the assumed initial flight angle so that the rocket is going horizontally by the time the barometric apogee occurs. It would be interesting to do the same thing with your Cirrus data. Note that I've been finding that the accelerometer has some drift in the offset and scale factor besides just the temperature effect, so in your data reduction, don't be afraid to tweak those somewhat to better match the baro data near apogee.
I'm sure your Cirrus Cd would be lower if the rocket were shorter. Long, slender rockets will have a higher Cd because of the higher ratio of the wetted area to the frontal area. A fat boy probably has a really low Cd, but of course its drag is still high because Cd*A is high.
Oh, and by the way, after averaging the correct pre-flight altitude, I'm now getting 8422 feet AGL. So that's my final answer for the launch log.
I'm going to put that down in the launch logs... and I'd like to put it down as a new club record and encourage you to submit it to Tripoli... That beat John's F record hands down and then some. However, I just can't do that until your altimeter is certified and I know TRA won't honor it unless it's an accepted altimeter.
PLEASE, please, push TRA on certifying the altimeter for competition use. I want one....and I have some questions...
1 - Can the data be streamed out the serial or USB port in real time?
2 - Can the board be queried in real time via the serial or USB port?
Warren
Thanks for the launch log update, Warren. But another reason for not putting it down as a record is that I didn't safely recover the rocket, only the nosecone. 🙁
I actually printed out and brought a Tripoli record form with me on Saturday, but that probably jinxed it.
I re-started the email thread with Tom Rouse on Saturday morning, and he replied a few hours later saying he was impressed with the altimeter, in a way that suggested maybe he didn't actually look very closely at the data I sent him in July. He had a couple of questions about connecting the Parrot via Hyperterminal, which I answered on Saturday night, and I haven't heard back since. I pinged him again tonight. If he doesn't get back to me tomorrow or so, a polite question from someone else about how the Parrot certification is coming probably wouldn't hurt. John has volunteered.
1. The data for all the sensors streams out at 1 Hz via the USB port whenever it's in WAIT mode. So if you're capturing text in Hyperterminal, you could have it recording for hours, days, or longer. That's how I captured the following calibration verification data on my way to visiting my folks in Summit County:
2. If I understand your question, the board can be commanded to download the data from any of the 5 flights via single-character commands through Hyperterminal. A keyboard command (E) is also the only way to erase a flight.
Well, I still haven't heard back from Tom Rouse recently. Anyone have a suggestion?
Latest Post: Hello World Our newest member: ollieclutter8 Recent Posts Unread Posts
