Thursday, August 27, 2009

Thunder & Lightning mark II

Several years ago it took me 5 years to build and fly my first PML Thunder & Lightning. I was psyched on the paint job at the time and my amazing and beautiful ex took this photo for me:
Unfortunately I set the upper stage timer for three seconds rather than 0-1 seconds and that beautiful rocket powered into the hard desert dirt at about mach 1 on its maiden flight. That flight also killed my prized AltAcc2A altimeter. Undaunted I lucked out and replaced that altimeter with the 2C version and purchased another T&L kit. While I had this kit lying around for about a year and a half I finally started it and finished the sustainer build and paint in under two weeks. It'll look very much the same as my first build above with one exception: car paint!:
Now that's some gloss!! This was my first and only paint job in the paint booth below before I realized it wasn't going to work and abandoned it. My final option to master car paint will require adequate ventilation in my garage and I'll park my car outside while I paint. The 500th time's the charm right?!
Finally I should note that, while I'm leveraging the original booster, I still have parts to build a second booster for a 3-stage version:
I'm a bit surprised that it remains totally stable without additional modification. That should be a fun flight if/when I attempt it.

Saturday, August 22, 2009

Spray booth as dust buster?

It's seriously frustrating to build something for weeks or months then only to fail on the last step, the paint job, due to dust. I bit the bullet and built a semi-permanent paint booth in a spare area of my house. Fisheye lenses make indoor photography a snap and here's a side view of the booth:

Here's the entrance view:

And here's the lovely view I'll enjoy while spraying fumes:
That fan is very powerful and pulls the 4-mil plastic walls all the way inward with it's high setting. I think it'll remove most of the fumes and overspray particles on its low setting. I plan to try some clear coat as a test this weekend and this setup should nail the problem.

Wednesday, August 19, 2009

M750 flight APPROVED... Drag race style!

I completed my altitude prediction report for EM-SEM-FITY on an M750 and our president and range safety officer approved the flight for Plaster Blaster! (click the thumbnail below for a legible JPEG):

Yay. My best prediction is the unweighted average of all the simulations and historic data: 21,333' (4.04 miles!). I tried to post this as a PDF as well as the RockSim file but I'm having way too many FTP issues right now so I give up.
I also shared this packet with Frank Hermes who's also attempting the President's 20K challenge. He had the excellent idea to launch our attempts simultaneously as an M750 drag race. Our club president, Paul Snow, approved this drag race as well! This should prove spectacular and Frank has promised to be a good sport when he loses. :)

4" V2 painted but not quite perfect

So several entries below you saw below that the white paint applied perfectly, then I masked (and re-masked 3x), then painted the black. Again dust was my enemy outside and the black ended up severely pock-marked. I patched the diffuse color with a black paint pen and, while not perfect, it looks pretty good from this distance:
Next I plan to set up a paint room inside my house (with robust ventilation of course) and I'll apply a clear coat to remedy the specular gloss. It'll probably still be dimpled but the clear coat should dampen the craters and obscure the bronzing of the paint pen patches. I vow to get this right some day!

Sunday, August 9, 2009

12,744' FTW!

[that's "For The Win" not "F*#k The World"]
Yesterday at Lucerne Dry Lake I flew EM-SEM-FITY on a Cesaroni L610-Classic motor to 12,744' and that's a new personal record!:
Here are some flame details (hot day, distant shot means serious convective optical warping):

That motor burns for 8 seconds and the flight was very cool with two exceptions:
  • I've optimized this rocket for motors of full-M weight so the CG/CP relationship yields a static stability margin of ~1.6 fully loaded. Since the L610 motor is half that weight the static margin shoots up to ~2.1 making the rocket slightly overstable. This caused mild coning on the way up and I lost some altitude as a result.
  • The drogue ejection charges popped exactly at apogee but the nose cone popped off as well thus dragging out the 7' main 'chute. This was supposed to be a dual-deploy flight so popping the main at over two miles can mean a big walk. And walk I did. I could see the parachute the whole time and found it about 1 mile out across the road. Luckily a fellow flyer named Steve kindly drove out to pick me up otherwise it would have been a long walk back.
Oh and I encountered a highly improbably coincidence yesterday. It took about 2 hours to prep the flight and before I carried it out to the pad I wanted to verify that my radio tracker was working. I turned on the receiver and got two out-of-sync signals. I walked around for a bit and located a flyer named Julian whose transmitter was running on the same frequency! There are roughly 50 transmitter frequencies available so how insane is this? I pulled the battery from my transmitter, Julian successfully recovered his rocket, and kindly pulled his battery shortly thereafter so I could reassemble and resume.
Now for some data... the initial fully loaded weight was 25.3 pounds and length is 94". Here are the flight summary and graph from the ARTS2 altimeter:
And the same for the Parrot 2 altimeter:
The average of the two barometric altitudes is 12,744'. I assumed a CD of 0.45 in RockSim and this predicted an altitude of 14,120' so this represents a large error. I attribute this both to the overstable coning I mentioned above as well as a guessed CD. If I iterate on CD in RockSim to match the average altitude I get 0.57. If I then re-run the M750 simulation with this CD I get 20,147' which is well below the San Diego club's 25,000' FAA waiver. More on this in my next blog entry.

Friday, August 7, 2009

Car paint + dust + wind = FAIL!

OK so you can't really tell in this photo but the booster has a large number of crater repellencies that I've narrowed down to a serious dust concentration in the air:
This was my second attempt to get it right and, while it's better than the first, there will be a third to nail it. Anyway the finish is very close to final from a coefficient of drag perspective so it will yield good data tomorrow.

P.S. I'm normally a much more rigorous photographer and post-production guy but I need to start packing...

Tuesday, August 4, 2009

Electronics bay, sled, and layout

I'm really very happy with the final electronics bay. Again Jack Garabaldi provided these fine components and the results exceeded my expectations. Here are the top and bottom of the sled sandwiched between the two plywood/composite caps for illustration purposes:
On the top-most image at top right is a Featherweight Parrot 2 altimeter. This is an unbelievably capable device and this will be the first time I've used it. On the lower left is my trusty ARTS2 altimeter and there are three total 9V battery mounts as well.
I should also note that I had a design epiphany during a meeting this morning... I always struggle with the best way to mount external switches to arm the electronics in a vertical position. Every hole you drill or slot you cut has the potential to weaken the structure under potentially tremendous flight forces.
Both altimeters use barometric sensors to determine altitude over the course of the flight and descent. Therefore you must drill a static port (hole) to allow access to external pressure. I calculated that I needed a single 1/4" diameter hole to ventilate the bay. I then carefully mounted the screw switch for the ARTS2 at top center of the sled aligned with that static port. I also positioned the side-mounted arming button for the Parrot 2 right next to the screw switch (mostly obscured in the photo). Now all I need to do is place a skinny screwdriver into the static port, screw the switch down to turn on the ARTS2, and then track right to push the button on the Parrot2 and arm that device. Nice!! One hole serves three purposes. Heh heh. And I have no external switches or extra weak spots!! Ist fantastisch!
Finally I wanted to strengthen the thin plywood caps so I simply laid out two circles of carbon fiber/Kevlar fabric on polyethylene film, saturated them with 15 minute West Systems epoxy, placed the caps on the fabric discs, overlaid another layer of polyethylene, and placed a 10 pound weight on top. Once cured I end up with a super smooth, glossy composite reinforcement on each cap:

Build/paint progress and simulation refinements

I've made quite a bit of progress on EM-SEM-FITY in the last few days. I applied car primer to the booster section and found a few more rough areas. I allowed this to dry then sanded down the remaining asperities. Tonight I applied the final coat of primer and that's drying now. Tomorrow I'll spray the booster and part of the nose cone glossy white and allow them to dry/cure overnight. Then on Thursday I'll apply the vinyl decal and at least one layer of clear coat to seal the decal and maximize gloss overall.
Now rewinding back to last night I fully assembled the rocket including an M1939 reload to determine an optimal amount of nose weight to yield a static stability margin of 1.65. This turned out to be 25 oz of led and epoxy in the nose tip. I butted the nose cone against the wall as the center of pressure and center of gravity are measured and reported from the nose tip. The wall makes it easier to use a tape measure:
Here's the technique I've always used to determine the amount of lead shot required for good stability. This allows easy addition of shot to the bag and the position provides and accurate static moment:
In addition to ensuring an optimal amount of nose weight I also recorded center of gravity and overall mass measurements that allowed me to further refine my RockSim design and simulations. That program does a pretty good job of calculating starting points for component mass and center of gravity but you really need to override both with measured values to generate the best simulations.
Here I've built a bulkhead for the nose cone, fit it into the shoulder, and applied 15 minute West Systems epoxy thickened with colloidal silica. This increases viscosity and prevents the epoxy from running down gaps around the disc while curing:
Tomorrow I'll add a single wrap of carbon fiber/epoxy to the inside of the shoulder to strengthen it and render the bulkhead impossible to extract.

Coefficient-Of-Drag-Tard

Ya learn something new every single day... Jack Garabaldi of What's Up Hobbies informed me of a feature of my ARTS2 altimeter that I didn't even know about. After a flight you download the accelerometer data to your computer and enter pertinent flight info (initial weight, size, etc). The program will then provide a motor analysis (letter class and average thrust)(I knew about this feature) but also calculates a coefficient of drag (CD). This coefficient ranks among the most influential unknowns in flight simulations. The good news is that it remains roughly constant as long as the structure and finish of the rocket stay fixed. In summary a single flight in the regime of target speeds will yield a CD that you can use to vastly improve future flight simulations in RockSim 9. I was floored when I performed the following post-analysis of my successful level 3 flight:
  • The original, optimistic RockSim 8 simulation was roughly 6,100' on an M2030 motor.
  • The actual, average barometric altitude of the two altimeters was 5,257' (ARTS2 = 5113, AltACC2C = 5400). This is 14% lower than the simulation.
  • I re-loaded the ARTS2 data into the Data Analyzer v1.6 program and ran the CD analysis. The data is noisy but the visual high average CD is roughly 0.44.
  • I took this actual CD and plugged it into the original RockSim design and the revised simulation is 5,215'!! That reduces the error to 0.7%. Damn. That's some predictive power.
Here's the calculated CD as a function of mach number (speed):
This, in combination with predominantly constructive feedback from The Rocketry Forum, has convinced me to pre-fly EM-SEM-FITY on either an L610 or L952 this next weekend at Lucerne. After this flight I can then calculate an actual CD and provide an excellent prediction of altitude for the actual M750 flight at Plaster Blaster.
This will be required as the range safety officer (RSO) is concerned that I'll exceed the club's standing 25,000' FAA waiver with an M750. If I preliminarily assume a CD of 0.45 for EM-SEM-FITY RockSim predicts a peak altitude of 22,087'. My research turned up only one flight with an M750 that attained 26,000 feet at Black Rock (Jack also provided this data point). I've found only four other historic altitudes all below 23,000'.
If my justification still cuts it too close to the waiver then I'll take a step down in motor size. Right now I'm looking at the Cesaroni M1060 but that sims to only 19,443' so that's not likely to exceed 20,000' which is the reason I'm doing all this in the first place. Fingers crossed...