Rocketry Planet

After a couple of partly-successful AeroTech I600 flights of my scratch built carbon fiber rocket, "Violent Agreement," earlier this flying season, I decided to try my hand at a multi-stage 38mm flight. In mid-August I had the male tooling plug for a nose cone, and a carbon fiber tube airframe that had been to 16,300+ feet on an I600, but had suffered a drogueless landing at high speed. With the damaged portion cut off, it was 14.5" long. I also had a 3" long, 38mm avionics bay with two Parrot altimeters and a Beeline transmitter. Everything else I would need to make from scratch, and XPRS (my first target of opportunity for this bird) was a little over a month away.

This project was a big step up for me in many areas. I have built around 10 minimum diameter rockets, all much lower powered than this one. My lone attempt at a multi-stage rocket was in 24mm, and it suffered a failure to ignite the sustainer due to a bad battery/igniter combination, but both stages recovered successfully. My highest-powered flight with a fully-successful recovery was my L2 certification flight on an AeroTech J285 and a Giant Leap Thunderbolt kit.

Theodore von Kármán (May 11, 1881 – May 7, 1963) was a Hungarian-American engineer and physicist who was active primarily in the fields of aeronautics and astronautics. He is responsible for many key advances in aerodynamics, notably his work on supersonic and hypersonic airflow characterization.

My other recent high-powered flights include a 125G shot on an AeroTech H999 with a apogee charge that was too large and jammed my main parachute piston; an AeroTech I600 flight that went Mach 2 and just under 16,000 feet but that apogee charge was also too large and deployed the main, resulting in a 5-mile drift; and another I600 shot on the same rocket with a good apogee charge but a main parachute charge that didn't get the main parachute out. I keep trying different dual-deployment configurations, trying to find a reliable approach that uses a single airframe break relatively far from the nose (to minimize drag and airframe flexibility) while packaging the altimeters and tracker efficiently. I think I found the design I've been seeking for the sustainer.

But first, some background on the other parts.

After some research into optimal nose cone shapes, I found that there are different types of theoretically optimal cone shapes that Theodore von Kármán invented. The shape most commonly referred to as a Von Karman nose cone minimizes the drag for a given diameter and length (length-diameter, or LD). This one is relatively low volume. There is a different shape for minimizing the drag for a given diameter and volume (VD). This one is a bit more bulbous. I also found out that the pressure drag (supersonic wave drag) for all of the types goes down with nose cone length. The longer the nose cone is, the more gradual is the increase in cross-section, and the lower the pressure drag.

Longer nose cones increase the surface area, however, which hurts the skin friction drag. For my rocket design, I decided to make the cylindrical part of the rocket only as long as it needed to be to contain the motor and the coupler, and put everything else into a Von Karman VD nose cone, making it as long as necessary to fit all the stuff inside.

I measured the volume required to pack a 35" parachute I wanted to use, the volume required for several hundred grams of tungsten-epoxy nose weight, and the volume required for my electronics. That, plus the diameter, determined the length of the nose cone using the VD shape. I subtracted off the thickness of the carbon layup I planned, and then a friend from work took those coordinates and turned the nose cone shape on a lathe in maple.

The carbon nose cone halfway out of the mold.

After making a couple of nose cones directly on this male tool earlier this year, I decided it was too hard to remove the part from a single-piece male plug, so I made a two-part female tool for this multi-stage attempt. It took two tries each for the mold and the nose cone, but finally I had my nose cone for this project. Before the nose weight, it weighed about 20 grams and is about 0.030" thick.

Sustainer airframe right out of the oven, before fin trimming.

Other than the fiberglass coupler glued into the front end of the airframe tube, the sustainer is constructed entirely of uni-directional plies of carbon fiber fabric, using high-temperature epoxy from Cotronics. I was concerned about aerodynamic heating from the expected Mach 3.3 speed, and the Cotronics resin I used was rated to 600 F. The fin stock has about 80% of the thickness with fibers oriented in one direction (perpendicular to the body tube), and after the fins were attached and filleted, another layer of unidirectional fiber was applied from tip to tip, again with the fibers oriented perpendicular to the body tube.

I accidentally ordered a model number of resin that does not cure at room temperature, so all of the sustainer layups were cured in my kitchen oven (we'll run the self-cleaning cycle before we use it again). This caused some problems, particularly since I couldn't use a vacuum bag for the tip-to-tip layup. The layup appeared to lay flat against the fillets, but it turned out that the tip-to-tip layer was lifted slightly off of the fillet surface.

Before final finishing, I filled the gap between the fillets and fins with Aeropoxy.

The sustainer tube layup consisted of a thin layer of fibers oriented in the circumferential direction, next to the mandrel, then a thick layer of axially-oriented fibers, and then at the front end of the tube, I applied another layer of circumferentially-oriented fibers.

The advantage of putting the circumferentially-oriented fibers on the inner and outer surfaces of the layup is that the two layers of circumferential fibers are separated by the thickness of the axial fibers, which maximizes the tube's resistance to crushing or deformation from couplers. The inter-stage couplers, in this case, are the second and third stage motors, which extend past the airframe three to four inches each.

Sustainer airframe tube, showing extra circumferentially-oriented fibers at the front end.

The sustainer uses a coaxial recovery system, in which the main parachute is contained in a sealed 24mm tube inside the nose cone, while the shock cord, tracking antenna, and apogee ejection charge are between the nose cone and the parachute holder. A fiberglass coupler is glued into the front of the sustainer body tube. The outside diameter of the coupler holds the nose cone on and the aft edge of the coupler provides a forward stop for the motor and avionics bay.

The avionics bay has a large bulkhead at one end, made of 0.093" fiberglass, with a 1.5" OD which rests against the coupler. The smaller end rests inside the coupler. The shock cord for the parachute is threaded from the front end, through two holes in the back bulkhead of the avionics bay, and back to the front of the avionics bay. One end is tied to the parachute, and the other end is tied to the nose cone. To seal the avionics bay from the apogee deployment charge, I wedged in a section of surgical rubber tube between the coupler and the parachute holder. This presses the antenna, apogee deployment wiring, and the two sections of shock cord against the parachute holder.

Sustainer showing parachute holder before nose cone installation.

The recovery system works by firing a charge located in front of the parachute holder's cap to blow the nose cone off at apogee. The shock cord that is arranged accordion-style around the parachute holder straightens out to about six feet. The shock cord for the nose cone is anchored a few inches away from the front of the main parachute tube so that at the main parachute deployment altitude, the main parachute has a clear path to exit when the charge inside the parachute holder fires. This system worked in ground tests and in its first flight at BALLS.

Its advantages are:

• It allows dual deployment with a single airframe break that's located only about an inch from the front of the motor.
• The nose cone shock cord can't pull out the main parachute (a problem with my first I600 flight).
• The main parachute's outward path is not blocked by the nose cone shock cord (a problem I've had before with dual deployment from one tube).
• No tether devices are required.

The second stage was originally also constructed of 2-3 layers of unidirectional fibers, similar to the sustainer. After the tube was complete, I decided that I wanted it to be a bit stronger, considering the large bending forces I was expecting for this stage, it being located in the middle of such a long, skinny rocket. So I applied an outer layer of heavy 0/90 carbon fiber fabric, to provide both axial fibers to resist bending and circumferential fibers to resist deformation from the motor couplers. At the time it seemed like overkill, because I could comfortably sit on or stand on the tube. This stage was constructed using Aeropoxy resin, since it was only expected to reach Mach 1.5, so heating was not a major concern.

The second stage also uses the parachute holder concept to do dual deployment from a single tube, but this parachute holder is a 38mm coupler that attaches to the avionics bay. The avionics bay/parachute holder is a piston that gets ejected at apogee to provide some drag. Then at the main parachute deployment altitude, the avionics bay fires a charge to eject the parachute. I was concerned about the avionics bay and parachute holder getting stuck in the tube when the charge fired, since the gas from the apogee ejection charge could escape through the arming and vent holes in the airframe once the avionics bay starts moving forward. So I sanded down the avionics bay and coupler so that it would slide freely in the tube, with the idea that the initial momentum would complete the ejection even if the ejection pressure was entirely lost through the arming and vent holes.

Second stage inner parts. From left: parachute holder, two Parrot avionics bay, spacer, anchor, spacer, insertion tool. Fine igniter wires are just visible against the insertion tool.

The avionics bay not only performs the dual deployment function, but also separates the first and second stages, ignites the second stage motor, and separates the second and third stages. That's a total of five outputs, in addition to using one output to turn on the transmitter at apogee.

The output functions of each altimeter are shown in the table below:

When the avionics bay is ejected at apogee, the other end of the shock cord is attached to an adjustable motor retention device that also serves as a motor forward stop. Since the avionics bay is moving relative to the motor, I used a simple 3-wire connector to connect the avionics bay to the second stage separation charge and the second stage ignitor. Those two functions shared a high-side wire to reduce the wires needed to run alongside the motor from 4 to 3. For both the sustainer ignition and the second stage separation and ignition, I used 30 gauge enamel-insulated wire sold at Radio Shack as magnet winding wire. The small gauge adds some resistance (about 0.3 Ohms per foot), but not enough to prevent the functions from working. I have experimented with copper foil tape in the past, but I found it too fragile and too difficult to connect at the ends.

Program settings in the Featherweight Interface Program, used for the 250G Parrot in the second stage avionics bay.

My original plan for the first stage booster was to use my short 38mm carbon fiber section with a small avionics bay and an AeroTech I1299 motor. The second stage and sustainer would use CTI J530 Imax motors. After some difficulty in procuring the I1299, I decided to use an AeroTech J570 that I already had on-hand, since it also has a strong initial kick to get the rocket moving quickly. But my existing carbon airframe was too short to use this configuration safely, from both an aerodynamic stability point of view, and a mechanical strength point of view. So, since I didn't have time to create a third carbon stage, I used a filament-wound fiberglass stage.

To reduce the overall length and mass, I decided to delete the avionics bay from the first stage, and just use the second stage to do the separation, plus use motor deployment for an apogee parachute deployment. I used a section of fiberglass coupler to make a combination piston/motor stop/parachute holder that would transmit the thrust of the first stage motor directly to the second stage motor, and protect the parachute from the first stage motor ejection charge. The back half of this piston is about 2" long, with the open end against the first stage motor. The front half is also about 2" long, and it constains the parachute.

To allow the parachute to come out when the holder is deployed, I split the front section in half, with one half still attached to the back half, and the other half meant to fall out of the way at deployment. Both halves push on the second stage motor. To protect the first stage parachute from the first/second stage separation charge, I made a parachute protector out of a piece of 0.030 fiberglass disk and a Nomex® parachute protector.

First stage parachute protector and thrust spacer, post-flight.

So altogether, I had a 38mm minimum diameter, three-stage rocket that weighs about 4.2 Kg loaded, is 7-1/2 feet tall.

The author on his way to the launch tower.

The three stages of Violent Agreement:

First stage (front) with an AeroTech J570, the second stage (middle) with a CTI J530 Imax, and the sustainer (back), also with a CTI J530 Imax.

The propellant mass fraction was 40%. Using Rocksim's conical nose cone for drag calculations, here is the rocket's projected performance:

• First stage max Gs: 28
• First stage burnout: 550 mph at 2 seconds
• First stage separation: 2.2 seconds
• Second stage ignition time: 6.3 seconds, 400 mph
• Second stage max Gs: 28
• Second stage burnout: 1150 mph (Mach 1.5) at 8 seconds, 6000 feet altitude
• Second stage separation: 8.2 seconds
• Third stage ignition altitude: 21 seconds, 17,000 feet, 400 mph
• Third stage max Gs: 54
• Third stage burnout: 2200 mph (Mach 3) at 23 seconds, 25,000 feet
• Third stage apogee: 56,000 feet at 62 seconds

The trip:

I finished the rocket at 4:30 PM on Thursday before BALLS. I took some pictures, and then at 4:40 PM, I broke off one of the first stage fins when the skinny box I was using fell over in my garage. After a quick Dremel session, superglue, and new Cotronics epoxy fillets, I was on the road about 5:10 PM. After a quick sleep period in Rock Springs, Wyoming, I got to Winnemucca, Nevada, by about 12:30 PM. After all the cautionary tales about taking the 85 mile Jungo Road "shortcut" to the Black Rock desert (rather than driving 220 miles via Reno and Gerlach), I was a bit nervous, but I soon found that the road was a piece of cake. It was recently graded, so it was smooth sailing at 50-80 mph the whole way. I think it only took about an hour and 45 minutes, including stopping for some pictures. It was fun, and I highly recommend it for those driving to the playa the East.

Typical condition of Jungo Road, here approaching the Sulfur mine. The playa is just in front of the furthest horizon on the right side of the picture.

The playa is immense, awe-inspiring, fun to drive on, and absolutely perfect for rocket recovery. The BALLS site is sort of at one skinny end of the playa, and it's still about 2.8 miles to the nearest plant life, let alone a hill. The dirt is firm enough to take tent stakes, but soft enough for safe 30-40 ft/second recoveries. It also gets absolutely everywhere, and attacks your skin. When the wind picks it up, it's a nightmare for the dirt-phobic.

The 2009 BALLS flight line, stretching as far as the eye can see.

The flight:

Violent Agreement on the playa.

After about 2 hours of setup, I was ready to go at 4 PM. I had been pushing to get the flight in on Friday, since the forecast called for deteriorating weather throughout the weekend. Friday was absolutely beautiful, with partly-cloudy skies, 0-7 mph winds, and temperatures in the upper 60s to low 70s.

Ready for the launch button.

The sustainer looks a little loose in the tower, but the first stage has to fit too, and has a 0.12" larger diameter.

The first stage burn went well, with only a slight wobble out of the tower. It was finished heading nice and straight.

Good flame-to-booster ratio.

And then we waited for the 2nd stage ignition. And waited. Then someone gave a report of a parachute out, a naked parachute. Darn. More waiting and scanning, another parachute reported, a few pops in the distance, and a report from an away pad of a rocket landing nearby. Binoculars showed a first stage lawn dart and another stage on the playa in the distance. Overall, it was pretty anti-climactic.

After a nice walk on the playa, I gathered up the first stage parachute and noted the first stage lawn dart location. Then I determined that the other stage I found on the playa is the sustainer, and it's putting out the transmitter signal I'm hearing (the second and thirrd stages are on the same frequency).

The sustainer all alone on the playa.

As I'm taking pictures and gathering up the sustainer, a helpful fellow in a pickup truck with my second stage in the back drives up and offers a ride back to the flight line, which I gratefully accept.

Post-mortem:

I got data from all three altimeters, and they tell a pretty interesting tale, along with the damage to the stages.

The first problem was that the first/second stage separation charge did a lot of damage. It charred the first stage parachute and somehow made it deploy at separation, well before the motor ejection charge went off. That snapped the first stage shock cord and a couple of shroud lines, leading to the first stage lawn dart. The separation also broke one or more of the wires to the second stage ignition, based on the Parrot's in-flight recording of the output continuity.

In the diagram above is the Featherweight Interface Program readout of the first 6 seconds of the flight, as recorded by one of two Parrot altimeters in the second stage. Note the brown continuity voltage indicates no continuity on the second stage ignitor shortly after the first stage separation.

Illustrated in the diagram above, the flight of the second stage up past apogee. Note the significant wobble in the coast as indicated by the lateral G sensor (purple line).

So now the second and third stages are traveling upward together. The sustainer ignition is waiting to see 17,000 feet, so holds its fire. The second/third stage separation event tries to go off at 5000 feet, but that battery doesn't put out enough current to fire my Estes/pyrodex charge, though it has done so with that type of charge several times in the past.

Looking back on it, that rechargeable cell had about a year's worth of hard flights, and it just wasn't up to the job any more. I had replaced the cell for the other altimeter in the avionics bay after it showed signs of damage, and I should have changed both of them out. At apogee, the other altimeter fired the apogee charge, simultaneously with the sustainer's apogee charge.

The diagram above illustrates the data from the other Parrot altimeter from the second stage.

The dip in the brown line indicates the output switch closed to fire the second stage/sustainer separation charge, but the battery was too weak to ignite the Estes ignitor used in for the charge. I should have used a Q2G2 and/or a deployment battery. Data ends when the apogee charge fired, due to the fireball entering the avionics bay through one of the arming holes.

Perhaps because of the nose cone ejection holding down the sustainer, the apogee charge blew out the side of my "overkill" tube in the process of ejecting the avionics bay. The fireball from the apogee charge also got into the avionics bay via the vent/arming holes and shorted out the altimeter that was supposed to do the second stage ejection. The data for that altimeter stops right there.

The second stage fell hard and broke one fin. The sustainer's Raven altimeter deployed the recovery gear cleanly at the expected times, and the rocket came in under parachute pretty much undamaged despite a 39 ft/second descent due to the extra unburned propellant. The only damage was to the parachute holder, which took some sort of hit that creased the tube, preventing future piston use.

I tried to get the sustainer ready for a re-flight on Saturday, so I worked into the evening. Before bed, the moonbow portended a change in the weather:

Moonbow over the playa and a well-decorated campsite.

But the front came through about an hour too soon:

Hiding behind my car moments after the gust front tore through the flight line.

Then it was time to go.

Another BALLS participant getting out while the getting was good.

The majestic black rock desert.

Lessons learned:

• Using the avionics bay as a piston is a bad idea because of the vent hole issue.
• Tiny Lipoly cells do degrade eventually. Test them before each flight after they've been around awhile.
• Use low current ematches or Q2G2 ignitors for all charges from now on.
• If you don't have time for complete ground testing, you don't have time to fly the rocket.
• Figure out a way to ground test stage separation without risking accidental motor ignition.
• Coaxial dual deployment works, but make the coaxial parachute holder stronger.
• Figure out a way to protect the ignition wires better from the deployment events.
• Put in a delay for one of the apogee events to reduce the chance they happen simultaneously.

Adrian Adamson is the owner of Featherweight Altimeters, manufacturer of the lightweight Parrot altimeter. Adamson, an engineer who worked with NASA's Jet Propulsion Laboratory on the electrical system for the Mars Rover project, more recently finds himself working on power system models for NASA's Constellation Program. For his contribution of this article, he will receive a free Rocketry Planet T-shirt. This sponsorship is made possible by our friends at Graphix & Stuff, producers of high quality hobby apparel and vinyl signage. Want your own free gifts? Read the program details page for complete information.

NOMEX® is a registered trademark of E.I. du Pont de Nemours and Company in the United States.

Reader comments:

#1 Re: Article: Violent Agreement: Adrian Adamson�s high altitude project

Very ambitious project. Sorry it didn't work out. Sounds like you learned a lot as did many of the flyers at BALLS this year. I look forward to seeing it fly perfectly next year.

Mike Fisher
Binder Design

Binder Design on 11-06-2009 11:42 PM

#2 Re: Article: Violent Agreement: Adrian Adamson�s high altitude project

Outstanding article and well worth the wait despite the anti-climactic outcome. An inspiring and educational build thread as well. Thanks Adrian.

watheyak on 11-07-2009 01:19 AM

#3 Re: Article: Violent Agreement: Adrian Adamson�s high altitude project

Thanks, Mike. I'm planning to come back to Black Rock sometime next year, hopefully at a launch where you're there too.

You're very kind, watheyak. I'm glad you enjoyed it.

And thanks to Darrell for editing the very long and complicated article to something readable.

Depending on the wind and how much I can finish up at the field, I may be flying the sustainer as a single stage tomorrow or the next day at the Northern Colorado Rocketry November club launch. Today was the first day since BALLS that I've tried working on my rockets, since I've been focused on shipping out the first production round of the Raven altimeters. Now that those are in the mail, I'm finding that one evening's worth of work isn't turning out to be enough time to get ready. But maybe between tomorrow at the field and tomorrow night the hotel I can get one flight ready for Sunday.

Adrian A on 11-07-2009 01:45 AM

#4 Re: Article: Violent Agreement: Adrian Adamson�s high altitude project

This is probably a stupid question, but, what motors were used in each stage? In the picture they are clearly CTIs.

dragonsp on 11-07-2009 08:22 AM

#5 Re: Article: Violent Agreement: Adrian Adamson�s high altitude project

Those are the new 6 grain Cesaroni xl J530s. Jeroen has been very helpful in providing flush tapered aft closures so I could use those brutes in the upper 2 stages. Now that I have flown them, I'm looking forward to igniting them!

Adrian A on 11-07-2009 11:26 AM

#6 Re: Article: Violent Agreement: Adrian Adamson�s high altitude project

Thanks for sharing the process-plus highlights, travails, issues etc.

Lots of USEFUL educational and applications info for us all.

I'm sure you'll be 100% successful next launch

Diosces on 11-09-2009 11:35 AM

#7 Re: Article: Violent Agreement: Adrian Adamson�s high altitude project

Quote:

Those are the new 6 grain Cesaroni xl J530s. Jeroen has been very helpful in providing flush tapered aft closures so I could use those brutes in the upper 2 stages. Now that I have flown them, I'm looking forward to igniting them!

I'll have to remember that ... "Now that I have flown them, I'm looking forward to igniting them". I know just how you feel there.

Well Adrian, we didn't have much luck at Balls did we. Maybe next year, right? I enjoyed your article and the build thread, but what I found to be amazing was how much you had to do right towards the end of the project. It was an accomplishment for you just to get there.

Jim

JimJarvis50 on 11-09-2009 02:51 PM

#8 Re: Article: Violent Agreement: Adrian Adamson�s high altitude project

Thanks, John. I'm glad you found it interesting.

Quote:

I'll have to remember that ... "Now that I have flown them, I'm looking forward to igniting them". I know just how you feel there.

Well Adrian, we didn't have much luck at Balls did we. Maybe next year, right? I enjoyed your article and the build thread, but what I found to be amazing was how much you had to do right towards the end of the project. It was an accomplishment for you just to get there.

Jim

Thanks, Jim. I have to credit my poor planning and time management skills for that! And a tolerant family. Also, I actually missed my deadline by about 50% because I was originally planning to fly at XPRS. But now that I have quite a few months to the next high-altitude opportunity, I'll have the chance to do it right. Still not Jarvis-style build quality, but hopefully with a better chance of success than the last time. There are a number of us from the Northern Colorado Rocketry club who are in the early planning stages of a group pilgrimage to Black Rock. I'm not sure yet which Black Rock launch I'll be shooting for, but I think I'll be there for one of them.

But since I'll still have some 12kft launch opportunities before May, I'm getting tempted to take a little break from 38mm and see what I can do with CTI's new 29mm full G motor.

Adrian A on 11-09-2009 03:48 PM

#9 Re: Article: Violent Agreement: Adrian Adamson�s high altitude project

I think what's most admirable and amazing about this project and Featherweight avionics is that Adrian is always pushing the envelope of what's possible. In the case of his altimeters, setting a new standard for the industry--making amazingly capable altimeters in such a small package. For his rockets, I am sure he'll get this project to full success and set new records with innovative ideas such as he tried here. In fact, I think Adrian's already set records in 24mm altitude, officially or not (could not believe he was penalized for using his own altimeter).

Anyway, thank you Adrian for this project's write-up--you're inspiring and assisting us all.

Stealthfixr on 11-30-2009 04:20 PM

#10 Re: Article: Violent Agreement: Adrian Adamson�s high altitude project

Quote:

I think what's most admirable and amazing about this project and Featherweight avionics is that Adrian is always pushing the envelope of what's possible. In the case of his altimeters, setting a new standard for the industry--making amazingly capable altimeters in such a small package. For his rockets, I am sure he'll get this project to full success and set new records with innovative ideas such as he tried here. In fact, I think Adrian's already set records in 24mm altitude, officially or not (could not believe he was penalized for using his own altimeter).

Anyway, thank you Adrian for this project's write-up--you're inspiring and assisting us all.

Thanks, Mark.

Adrian A on 12-01-2009 11:14 AM

Comments 1-10 of 31 shown. Click here to read comments 11 through 31.

Registered users can add comments and discuss this article. To participate, please login or register.