With the recent successful launch of the Defiance-H "O" class hybrid rocket it's time to start getting serious...
This rocket will be a 6" diameter all aluminum rocket similar to the Defiance-H, only I'll use an all aluminum
fin can as well as a more elaborate nose cone with telemetry and deployment from an electronics bay in the nose
cone. I'll stick with the slip coupler N2O feed design plumbing through an externally activated ball valve. The
engine will be designated the HR-6, but I think I'll just include the engine development on this page since it's
an integrated design.
Here's a very early screen cap showing the expected flight parameters.
This is a screen cap from my Hybrid Calculator.
Here's a look at the first new piece of hardware, the Ganymede nozzle (on right) next to the Defiance-H nozzle.
It's still faster for me to draw things the old fashioned way. Here I've worked up the details on the nozzle and bulkhead.
I had originally thought perhaps I could use the old 6" nozzle I made for my first "O" class sugar motor, but I had used a stepped retainer on that nozzle, and I needed more expansion on this nozzle so I had to start with a new piece of graphite. It occurred to me while I was sanding the throat on the new nozzle that I could almost get my whole hand inside the nozzle. A few years back I was happy to get to the point my rockets were big enough to get my hand in the body tube, now the nozzle throat is getting to that point.
I really debated on how much expansion to use on this nozzle. At 350 psi I calculated an optimal expansion on the exit cone of 4.19" at sea level. Since this engine will be burning still at over 15,000' I could use more expansion at altitude. But, a hybrid normally drops thrust/chamber pressure as the burn progresses, and since the optimal exit cone diameter drops as chamber pressure drops, the 4.19" should still be pretty close even at MECO. Once I got to the last few cuts on the exit cone, I stopped just a little short of the 4.19". I wanted a large enough surface area to support the graphite, so I stopped at 4.1". I also reached the limit on the length of the exit cone cut I can do on my lathe. Well, at least in one pass. I had to cut in as far as I could go, then reset the compound and finish off the last inch or two.
I made 3 o-ring glands in the nozzle and cut a step in the convergent end for the fuel grain tube to slide in to.
Here's an adapter I made to turn down the outside of a 6" diameter tube.
I ordered a length of 6" diameter x .25" wall tube from Speedy Metals the other day, they are speedy as I had it in 2 days. I have two places I'll use this tube. The first use will be for a short internal nozzle retainer, the larger length will be used as a lower nose cone coupler section. I'll get into the nose cone design a little later... But what it boils down to, is that I need to turn down the outside diameter. My steady rest will only handle material to about 4.75" in diameter and my bull nose live center goes to 4". With the 5" x .25" wall tube I found 4" PVC pipe had an outside diameter of almost exactly 4.5", so I could ram the PVC pipe inside the 5" tube, use a reducer and turn it down using my bull nose live center.
For this 6" tube I needed another solution. So I found some 5" steel pipe which had an outside diameter of about 5.65". I cut off a 3" length and turned it down to just under 5.5". I left about 3/4" the full diameter of the pipe to create a stop point where the 6" tube would butt against. After both ends were surfaced true, I chucked up the pipe in the lathe and ran my center point into a hole drilled in a flat plate. Once I tightened things up with the tail stock, I tack welded the plate on the pipe section.
My lathe chuck can handle up to 12" (or more, I've never really checked) I only need this adapter on one end. Now I can turn any length tube up to the limit of my lathe bed. I'm sure at some point I'll wind up making more of these adapters for other sizes as well. I tested the tolerances and I'm within a few thousandths, that should be more than accurate enough for what I'm doing.
A little more now on the nose cone and the deployment strategy. I'm going to recover this rocket in two separate parts, all the primary electronics will be located in the upper nose cone section, this section will be ejected at apogee unattached to the lower body/engine section. The lower section will deploy a drogue chute at apogee at the same time the drogue chute for the nose cone is deployed. Both the upper and lower sections will have separate altimeters for main deployment at lower altitude. I'll try to size the parachutes to bring both pieces down at about the same rate, keeping the landing site of both pieces as close together as possible. This deployment strategy is intended to keep the apogee event as mild and safe as possible.
Here's the lower nose cone section getting the coupler area anodized.
Here's the lower nose cone body section after anodizing and final turning in the lathe.
The less shiny lower part is the coupler area. Since aluminum on aluminum can tend to bind, I hard anodized the coupler area, I'll also paint it with a graphite based low friction paint. The area above the coupler is the same diameter as the main body tube, and is really just an extension of the main body, the electronics will mount in a bulkhead above the coupler section. The coupler section will hold the main chute. One area of concern is the location of the ambient pressure ports. They will have to be located just under the nose cone on that 3" body tube extension. That's not the greatest location since since that area sees more turbulence, but I'll have to overcome that with code in the flight avionics.
The next area above that 3" extension will be the bonding area for a composite nose cone. Notice I made a couple of shallow grooves in that area, that's to give the epoxy a place to "key into". I'll likely drill a few key holes in it as well, for more added bonding strength. I thought long and hard about bonding the composite to the aluminum. But with adequate "keying" and the only real loading on the composite will be in compression, this should work fine even if the actual epoxy to aluminum bond fails. The primary reason for the bond to fail would be from temperature changes causing uneven expansion or contraction between the different materials.
Here's a first look at what I hope will be the nose cone plug.
I've been giving more thought to flight electronics and telemetry. Even more than in flight GPS data, I really
want real time telemetry from other sensors such as barometric pressure, engine chamber pressure and flight event
points. I also need the ability to upload commands from the ground such as "terminate thrust" and "manual
apogee deploy". I pulled out a Basic Stamp 2 last night, and starting thinking through what I needed to make
it work. The BS2 just didn't have the capabilities to do what I needed done, so I did a quick search on the net
and found the Basic
Atom. The Basic Atom 28-M has a lot of advantages over the BS2, it's faster, has
much more RAM and EEPROM, it has built in analog to digital convertors and the big thing is it does floating point
math up to +- 2.127 billion, where the BS2 was limited to numbers within a range of +- 65,563 and did not do floating
point math. Another added benefit of the Basic Atom is the code is written in a form of Basic similar to what I've
used in the Basic Stamp, so the learning curve should be somewhat shorter for me. So I ordered a Basic Atom 28-M
last night, I'll likely start a new page on just the electronics after I get a good start on the project.
The nose cone plug is about half done, I had to stop and fill in some cracks where the initial gluing didn't take between the boards. Everyone thinks this style nose cone is a true cone, but it isn't. It's a 5:1 ratio 3/4 power which should be optimum for a flight spending as much time above mach as this flight will. This plug is undersized to allow the lay up of fiberglass over the plug.
Oh what a mess I made turning this nose cone plug. The pile of chips is about 3" deep in places! This was all after I cut the corners down outside with a hand power plane. I have to admit I cheated a little too. I used the hand power plane to cut down the block of wood while it was turning in the lathe. I wasn't sure how well it would work, it left a very rough cut, but certainly was much faster than making the rough cuts with the lathe tools. I'm not going to get a much larger nose cone in this little wood lathe. I'm reaching the bed length and working the motor pretty hard too.
The raw nose cone plug installed in the aluminum part of the nose cone.
In the picture above you can see I have the lathe work on the nose cone plug done. I also painted the coupler area that I anodized earlier with a graphite paint. This paint is sold by John Deere and intended for heavy machinery. I'm sure it will wear off, but a quick re-paint should be all it takes to keep the nose cone loose and slick on the upper body tube.
To prepare for laying up the fiberglass, I filled any imperfections in the wood. Primed, spot filled and sanded twice. Then finished painted with one coat of the blue paint used on the Aestus, followed by three coats of parting wax and three coats of PVA. I was going to chuck up the aluminum base of the nose cone in the lathe, to make sure my nose tip was centered properly, but the graphite paint I used works so well, the chuck jaws wouldn't even hold. I guess I should have waited with painting that until I was about done... So I ended up checking the tip with a square. I didn't want the epoxy running down the inside, between the plug and the aluminum base. So I filled that little crack with parting wax.
I didn't have any illusions this lay up would be easy, and I've got a long way to go. But I did get the first 4 layers over the entire nose cone. The base area has 9 or 10 layers on it to bring the plug closer to the outside diameter of the "step line" of the aluminum. I considered not even putting a tip on the plug, since there's no way to lay up the cloth on that small of a surface. But I figured I needed something there regardless so I made the tip. Once the nose cone is mostly done, I'll cut the tip off and make a solid nose cone tip from epoxy and carbon fiber strands. I need the nose cone RF transparent, but that little carbon in the tip shouldn't affect RF that much, but it will add a lot of strength and thermal resistance to the tip where I need it. Tomorrow I should be able to try pulling the plug out, and work on future layers without the plug in place. I see a lot of lay ups and sanding in my future...
The plug de-molded very easily, a couple of sharp wraps on the counter top and the plug fell out the bottom.
After de-molding I laid up another 3 layers of glass. To speed up the epoxy cure, I'm using a couple of heat lamps and rotate the nose cone every 30 minutes or so. This should allow me to get in two or three applications a day. Of course there's a lot of sanding between applications to keep the ridges down and the surface roughed up for the next application to bite into.
I'm getting there...
After numerous layers of glass cloth (changing the direction of the weave and layout) I'm to the point now where I'm using epoxy filler to level out and fill imperfections. I thicken the epoxy to almost a paste with talc and apply it like body putty. I didn't want to use an automotive body filler in this nose cone, the body putty can (and does) bubble up when heated. It certainly takes longer this way due to the slow cure of the epoxy, and sanding is much slower, but in the end it will be a much stronger surface. The strength of the nose cone isn't in doubt, there is about 16 layers of glass at the bottom, tapering to about 8 layers at the tip. The composite doesn't even feel like a composite any more, it feels more like metal. I have a feeling you could drive a truck over this nose cone without imparting any damage.
I could have saved myself a lot of work by vacuum bagging, the vacuum process compresses the material so well, there would have been very few ridges to sand/fill from the layer overlap. But the end result will be the same, only perhaps a little heavier from the excess resin in the non vacuum lay up process. This is going to be one fantastic nose cone, I've already put many hours into it. I knew it would be a long build, that's why I started the nose cone right from the start of this project. It was one component of this rocket that has to be about perfect, and I didn't want to lose patience and get in a hurry.
After the first "putty" coat, I thinned the epoxy/talc mixture down to gel coat consistency and applied then sanded again. In this picture you can see I've formed the epoxy/graphite nose tip. I poured the mixture into the inverted nose cone, then tamped it in. The solid epoxy/graphite extends about 3" below the nose cone from what you can see of the black tip. This solid tip added 110 grams to the weight of the nose cone, but I think it's weight well spent. I roughed up the inside surface with a file and small wire brush to make sure the epoxy had a good "bite" on the inner surface. The wet spots you see in this picture are from a small 10 gram batch of the "gel coat" mixture to fill in any remaining low spots. After sanding this coat I'll do one more coat of just plain epoxy as a sanding surface for the finish sanding.
Aside from priming and painting the nose cone structure is done. I must say, while I spent a lot of hours on this (maybe 35-40 hours), it did turn out as well or better than expected. I've got a shipment of metal coming in this week, when it arrives I'll get started on the inside bulkheads and mounting system for the electronics.
Here's an almost to scale drawing of the upper rocket.
The deployment strategy is really quite simple, at apogee the electronics in the nose cone will fire the apogee charges. Three shear pins will release and the nose cone gets blown away from the main body of the rocket. There will be a drogue chute attached to each section, those chutes will be held together with a Velcro band (or similar). When they reach the ends of their shock cords the Velcro pulls apart releasing both drogue chutes and each section of the rocket descends on its own. Each section will also have a main chute tethered to a PIRM2 release mechanism. The engine section only needs an altimeter to fire the PIRM2 for main release since apogee deployment is all handled by the nose cone electronics.
I color coded the drawing:
Red indicates composite structure.
Light blue is the aluminum part of the nose cone.
Dark blue is bulkheads.
Purple is the main body tube section.
The green "data" line will send N2O pressure and chamber pressure to the telemetry board in the nose cone. I'll use "pull apart" or "break away" fittings on the wiring, once the nose cone is deployed this data is lost. Of course both channels should be reading 0 at that point. I'm not sure how I'll get the data line from the combustion chamber to the data line yet. I could run a tube through the N2O tank, with pressure fittings at either end and just push a wire through it. But that would cost me N2O capacity. What would be better would be if I could find a flat ribbon cable and bond it to the outside of the rocket. That would create a little extra drag, but I suspect the drag loss would be less than the impulse loss from a reduced N2O load. An internal tube is also another possible failure point, should the tube crack I'd vent N2O directly into both the engine valve compartment, and the lower electronics compartment.
If you looked over the drawing carefully, you may wonder how I intend to deploy the main chute on the nose cone, since the main chute is inside a bulkhead. I thought about just using more shear screws on the bulkhead, but that quickly looked like it could get messy. So I decided to run the drogue chute shock cord through a hole in the bulkhead, that way the bulkhead can slide out of the way when assembling the nose cone electronics. Once all the upper electronics and main chute are stowed, the bulkhead will slide up in place and I'll pin it in place using the shock cord to hold it. I'll run a small dowel pin through the shock cord and use some quick set epoxy or RTV to secure it in place. When the nose cone PIRM2 fires, it will drop the bulkhead out and it will stay attached to the shock cord under the drogue and above the main chute.
For the electronics board I used 1/4" plywood covered in 4 layers of glass cloth on each side. You can see the layout I drew up to the left of the finished board. This board is designed to be a snug fit in the nose cone, I'll install a slotted rail on each long edge of the board inside the nose cone. These rails will help support the electronics package and prevent flexing of the board. Lower center right is the nose cone bulkhead with 2 pieces of 1" x 1/8" angle aluminum, these will bolt to the bulkhead and to the board. To the far right is the main power switch.
Here you can see the board installed on the bulkhead using 3/16" bolts. It's hard to see because the switch is right in the bright spot of the flash, but it's installed on the lower right.
I have 3 ambient pressure ports 1/4" in diameter. You can see one of the ports in the center, below it I stamped the letter "E", that lets me know which hole is used to access the power switch. I really worked in the design phase to make sure I could easily access the switch. I hit the nail on the head because you can't miss the toggle lever on the switch. Just push something in there and it turns on. Just below and to the left of the port is the first retaining screw for the bulkhead. I'll use six of these 3/16" stainless bolts to retain the bulkhead.
On the left is the new 6" injector plate and bulkhead, on the right is the Defiance-H for comparison.
I haven't been idle lately, my new Basic Atom micro controller came in last week and I've been spending many, many hours (and ingesting many aspirin) working on the new mini telemetry system. I'm making progress, although at times I take one step forward in the coding and it takes me two steps backwards...
In the picture above you see the almost completed injector and bulkhead. The biggest problem I had was tapping the threads in the injector plate. I started by drilling a 7/8" hole in the bulkhead, then ran a boring tool in my lathe at 4 degrees to taper and enlarge the hole for the tap. I purchased a 3/4" NPT tap, but as I expected the tap was too big to fit in my lathe. So I tried running the tap into the hole by hand, as is usually the case, the tap ran in at a little bit of an angle. So now my fitting will be at an angle coming out of the bulkhead. That won't be a problem for static testing, but it may cause me problems in motor assembly in the final rocket. I'll deal with that when the time comes, but it may mean a new bulkhead down the road, and I'll probably have to make a jig for my lathe tail stock to hold the tap for future threading jobs.
I tried to design the injector plate and bulkhead with the tapered approach I used on the 5" engines, but I was limited by the thickness of the bulkhead and the much larger entrance hole the 3/4" pipe fittings require. So rather than the tapered flow path, I'm using a plenum design above the injector plate. If you look closely at the picture, you can see 2 steps on the bulkhead, the first step is the plenum and is .15" deep in the bulkhead, the injector plate has a corresponding .03" deep plenum cut into it. The astute observer may also notice I moved the ring of injector holes as far out as possible. On the Defiance-H flight, I had very low fuel regression. The only reason that really makes sense is that the Defiance-H had a larger port, which moved the injectors away from the fuel surface. So I'm moving these injectors as close to the fuel surface as is practical.
Speaking of injectors, these are bigger and there are more of them than even the Defiance-H used. In fact, I ended up adding a extra injector in the middle of the plate. That gives me 13 injectors at .155" diameter. If I went with 12 and the next drill size bigger, it was just a bit over what I wanted for N2O flow. The 13th injector came out just where I wanted the flow rate at. So, if something goes wrong, we can just blame the 13th injector ; )
Here's a new screen cap with the final, as built parameters.
I suppose I'm to the point now where I need to stop rating my engines by the hobby letter class. With a liquid propellant engine, ratings are usually given in pounds of instantaneous or maximum thrust. The same concept really applies to hybrids. I could run this engine with a light oxidizer load and it would fall into an "N" or "O" class depending on the N2O load, but will still maintain its initial thrust profile. That's a really nice feature of a hybrid. If I can't get a waiver to 45,000' I can just lower the N2O load and still have a great flight, with all the kick off the pad of a high impulse engine. That said, this should be a 1,500+ pound thrust engine. The initial kick at ignition will likely be closer to 1,800 or 1,900 pounds of thrust.
The nose cone is pretty much done now too. I installed two short "rails" inside the nose cone to retain the electronics board. I finished drilling and threading the 6 bolts to retain the bulkhead and I installed a PIRM2 and a forged eye bolt for parachute attachment. I'm kicking around ideas about how to test the recovery system. I might be able to install a large enough charge in the body tube section to blow it up there high enough to test the recovery system. Or I could throw the thing off a tall building or cliff, well, maybe not. I might just make a mini lower body tube with a cheap fin can to boost it a couple hundred feet in the air for testing.
I mentioned earlier I picked up a new micro controller, the Basic Atom. I've been spending a lot of time working on a simple telemetry system. I had some problems early on, then I found a post on a forum that mentioned the fact that the new release of the IDE (that's the software that programs the micro controller) had some changes in the way you write the code. Well, once I found the manual was not written for the new software, I switched to the old IDE and things started falling in place. I now have the Basic Atom gathering data and sending it to the PC application from a barometric sensor (for altitude), a chamber pressure sensor, a N2O pressure sensor and a sensor to detect apogee separation of the nose cone. I also have the PC application sending commands to deploy apogee and abort flight.
With the electronics going well, I got a little cocky and decided to see if I could get my OEM GPS module working. That was sort of a sticky subject with me, as I was perplexed and frustrated as to why I couldn't get it working when I first got the module. To make a long story short... I was using a Spark Fun shifter board to convert the CMOS levels of the GPS to RS232 levels the PC could understand, well, that's what I thought anyway. As it turns out, I was making a simple mistake. The shifter board doesn't convert CMOS levels to RS232 levels, it converts RS232 to CMOS levels. The shifter board converts RS232 down to whatever voltage level you power it with. So by powering the shifter board with the 3.3 volts the OEM GPS uses, I started getting good data out of the GPS.
I did a short page on the LM 317 regulated power supply . For those of you printing your own boards with a laser printer, there's an image you can use on that page.
Here's a shot of the Garmin 15 mounted to a power supply board and wired to the MaxStream radio via the Spark Fun shifter board. Notice how small the OEM GPS module is.
I'm not so sure I'll even use the GPS in the telemetry system for the Ganymede, but it is nice to have it working. The problem with integrating the GPS into the telemetry system is that I want two way communication with the rocket. Adding the GPS telemetry to the system really slows things down because the micro controller can only do one task at a time. If I'm gathering GPS data I can't be gathering/sending other data, or sending uplink commands. I do have a couple of options to give me both, I could run a separate transmitter dedicated to the GPS. That would be best, I'd get all my data and two way communications. I'd also have a second transmitter for locating the rocket after it landed if I lost one in flight. But that's another $500 for radios... The second option would be to only start sending GPS data after apogee. That gives me the data and two way communication that I need up to apogee. Once apogee deployment is confirmed, the micro controller could switch to sending only GPS data. That's not something I need to decide on right now, but I'll admit I'm leaning more towards a second transmitter dedicated to the GPS.
Moving on to propulsion. I picked up 1.5 gallons of polyester resin the other day, that should be enough for two grains in the HR-6 engine. I also machined a nozzle retainer last night, but I'm having second thoughts. A couple of pictures to help explain...
Here's the retainer sitting on the nozzle.
Here's the nozzle and retainer inserted into the casing.
This is a very similar system as to what I used in the Defiance-H, I had concerns about setting the nozzle so far inside the casing on the Defiance-H, and I'm having similar concerns here. Of course there weren't any problems with the Defiance-H, so I may be fine here too. If you extend the exit cone geometry out, it stays inside the retainer, so the exhaust should more or less be uninhibited by the retainer. I may know more about the Defiance-H flight here shortly. Mike Bennett discovered the reason for the recording altimeter's failure to download data was a glitch in code. One of guys the working on that project is supposed to be able to recover that data. So I may have performance data on that engine in flight, which would help with decisions to be made on this rocket. Of course, I'm planning a static test of this engine soon, so I'll have data from this engine as well.
This is a heavier retainer than used in the Defiance, although about the same length. What concerns me is the long retainer may disrupt the flow off the exit cone, leading to turbulence and loss of performance. There's no doubt in my mind the retainer will act as a secondary expansion area, but what I don't know is how that will affect performance. I've considered opening up the exit cone to meet the retainer, then taper the step on the retainer to meet the nozzle. That would give me a sort of bell shaped exit cone all the way out end of the retainer. Of course, that would also over expand the nozzle, which I think is what it would do the way it is, only with more turbulence.
Another option would be to drill and tap the retainer for attachment to the casing, then cut the retainer shorter. That would push the nozzle back another 3/4" closer to the end of the case. I wanted that external lip on the retainer to make drilling and threading easier and more precise, it also strengthens the exit area and provides some extra thermal mass. I may just have to sit and look at it for another hour or so before I make up my mind :)
My hour's up...
I decided to just taper the step. It turned out a 35 degree angle left the load bearing surface area intact, yet just met the inside wall of the retainer. Depending on how my first static test goes, I may end up changing the entire design. I can still shorten the retainer and/or bring the graphite exit cone diameter out to meet the retainer.
Here's the retainer with the tedious task of drilling a tapping the 18 holes for 1/4" bolts.
I had to be careful drilling the hole depth to .35" since I only had a material depth of .39". I started with a standard 1/4" tap, and went in a couple of threads worth with that tap to get the hole started. Then I finished with a bottoming tap to get the threads as near the bottom of the hole as possible. When I was done, I chamfered all the holes edges so they wouldn't catch when the retainer is inserted. There is maybe two holes I'm not real proud of, but all in all this important task went well. 18 1/4" bolts is over kill, I could have got by with half that many, but this case was used for a higher pressure solid motor originally and that's the hole pattern it had.
I need to drill and tap a forward bulkhead retainer and the engine should be ready for a fuel grain. If all goes well I'll have a fuel grain cast this week. Now I need to decide how I'm going to mount my test stand. I loaned out the trailer I had used for the HR5 engine tests and won't have it back until mid July, since I'd like to test before then, I'll have to come up with something else.
I calculated the density of my HDPE/Polyester resin at 0.03696 pounds per cubic inch. If I use the casting tube I have, they will only give me about 9.8 pounds of fuel. Which is fine for this first static test using only 30 pounds or so of N2O, but I may have to go with a larger casting tube, or no casting tube at all for the full blown "P" impulse flight. A full diameter fuel grain will give me 16.19 pounds of fuel, more than enough based on any fuel ratio from past tests. I'll base my decision on the regression rate of the fuel after the first static test. I've had the regression rate change by as much as 50% simply by changing the location of the injectors.
Here's the front side of the electronics board.
This is the battery side of the electronics board.
I mounted the mini telemetry system on the electronics board in the nose cone. The MTS board is to the right, a power supply board is in the middle and the radio on the left. I wanted some redundancy in my power supply, the small power supply board is a 30 amp Schottky diode diode connected to 2 separate battery systems. The primary battery is a set of 10) 1500 mAh NiCd batteries, the back up power supply is 2) 9 volt batteries hooked up in parallel. The diode will allow current to flow only from the highest voltage battery pack. If something happens to the NiCd pack, the dual 9 volt batteries will take over and provide at least enough power to handle deployment and radio transmissions for the flight duration.
A graph of the voltage vs. time.
I wanted to get a better idea of how long the telemetry system would transmit with the NiCd battery pack. So I set the system up on my work bench and let it go, I had the receiving radio/PC set up in another room. I recorded the voltage as sent by the MTS every 30 minutes. The nominal full charge voltage of the pack should be 12 volts. As you can see in the graph, after 9 hours I hadn't even dropped to the 12 volt level. It would seem the battery pack is more than adequate for the job at hand.
Here's a vid cap of the nose cone deployment test.
I performed a nose cone deployment test several days ago. This test used 6 grams of black powder in an aluminum charge canister. I used 2) #4 nylon shear screws to retain the nose cone. Click Here for a short video of the test. The test went fine with 6 grams doing a nice job of deploying the nose cone.
I performed my first recovery system test using the Ganymede nose cone on a quick, purpose built lower body tube. See the results on the LT 150 page.
I'm getting close to testing the HR-6 engine, so I had a few more details to work out. One of which was how to get chamber pressure data. The entire surface area of the bulkhead is either covered by the fuel grain or the injector plate. In the picture above you can see my novel approach to solving the problem. I drilled a small hole through the length of one of the injector plate retaining bolts.
On the outside of the bulkhead I drilled and tapped the back side of the bolt hole for a 1/4" NPT pipe fitting. The hole is pretty small, and probably will result in a slight delay in pressure build up. But it should do the trick and seemed like the easiest and safest way to get chamber pressure out of the bulkhead.
Here's the fuel grain installed in the motor. I used my old 5" ID cardboard tube to cast in. I cut up some irregular 5" tubes to make spacers to fill the casing, you can see them as a second layer of cardboard. For this static test I'll only be using about 50% of the engines ultimate N2O capacity, so I shouldn't need a full fuel grain. If I wind up burning through this case on the static test, that means I'll never get a full N2O load for the flight either.
I may be at the point now where the combustion chamber tube is single use. The 42" length of tube is one of the cheaper parts on the rocket, and there's no sense beating myself up trying to re-use it when the motor would be lighter and carry more fuel as a single use. I debated Casting the polyester resin directly in this case as well, but decided against it this time around. We'll see how the engine performs, then I'll make a decision on casting. The fuel grain weighs 12.6 pounds including the casting tube. The fuel grain is 26.83" long including the .25" nozzle step. The fuel grain extends all the way to the bulkhead, so the injector nozzle protrudes into the grain core about .45", leaving 26.13" of fuel exposed to burn.
The Ganymede engine has been static tested, the details of the test are on the ST 151 page.
On the left is a bracket made from mild steel, on the right aluminum.
My original plan for the fin attachment was to do a fin can made from aluminum to slide over the lower body tube, and the fins would simply be welded to the fin can. But I'm having second thoughts on that plan. I've been wondering if I could make some brackets to simply bolt the fins to the brackets and then bolt the brackets to the body tube. To test the idea, I made two sets of short fin brackets. The first set is mild steel angle, 1.5" x 1.5" x .125". I used my forge to heat the metal red hot, then pounded it to shape over a 5.75" solid round stock I use as an anvil. The second set is the same size but made from aluminum. Of course I didn't heat the aluminum, I just cold formed it over the round anvil.
Both processes seemed to work ok, the mild steel when heated is easier to form, while the aluminum is lighter. 1.5" angle stock may be overkill here. I suspect I could get by with 1" angle stock, and I may give that a try as well. Using brackets rather than a full fin can should reduce the Cd a fair amount and also allow replacement of a bent fin if such occurs at landing.
Here's one of the fin brackets in a homemade hydraulic press.
I've been wanting a hydraulic press for some time now, so I went to work putting one together. I had some 4" aluminum C channel and some 1.5" x .25" angle iron and the jack. All I needed was a couple of cross sections and some springs. After I had the press assembled I went about looking for good ways to get the fin brackets bent. The wood blocks worked pretty well. I used a piece of 2" round stock to apply pressure to the center. I was going to try to form the bracket over the 5.75" round bar I have, but aluminum springs back. So I'd need to over bend it on maybe some 4" stock to get the final 6" contour I need. It comes out pretty close just bending it over the wood blocks, then a few minutes with a hammer on the round anvil to finish the final contour.
This process is going to be time consuming, I'm only bending a pair of 4" brackets at a time, and I'll use 3 sets of brackets on each fin. Had I gone with 1" aluminum, I probably would have gone with full length brackets on each fin. But these brackets should be major overkill strength wise, so I'll space the brackets and just use three sets on each fin.
Here's a template I made to drill the bracket holes. This allows me to interchange or replace brackets as the need arises. On the right is a finished bracket.
Here's a bracket after forming. I made a steel form to help hold and shape as I apply pressure. The 2.5" round stock is what I use on top of the brackets to transfer the pressure.
I don't have any extra 6" tubing, but I wanted to test a bracket on some material so I used a piece of scrap 5" tube and bolted the bracket to it. I used a piece of scrap fin material as well. I clamped the fin material in my press and put all my weight on the tube, with over 220 pounds force the bracket didn't give at all. I may end up setting up a load cell and recording the actual load at failure. Keep in mind this is only one of three brackets on each fin.
Here's the complete (more or less) set of fin brackets. I've got about 15 to 18 hours in them just to this point. There's a lot more steps involved than a person would think. After cutting to length, de-burring the cut edges and drilling the 7 holes in each half bracket, I then had to screw them together with temporary bolts while I tapped threads in the lower holes, then cut the counter sink, bolt them together in the newly threaded holes, counter sink and thread the first two holes, install the upper two screws, then bend the brackets in the press, then cut the counter sink holes on the lower 6 holes. I read about a large project where the guy paid $1200 to have fin brackets made at a machine shop. Now I see why they were so expensive!
When I started with these brackets I intended to use tapped threads in one side of the bracket to screw the bolts into. I knew the 3/8" long bolts wouldn't make it into all the threads, but I really didn't want them sticking out too much either. After assessing the completed brackets, I decided to err on the conservative side and go ahead and use longer screws with nuts on the back side brackets. I ordered some flush captive nuts that will keep the profile as low as possible, and still give me good holding strength. I'll have to re-drill the back side of the brackets for the flush nuts, but that should be a quick job.
Sort of off the subject here. But I picked up a fresh roll of flux wire for my welder the other day and a set of new contact points. I've been using the flux wire and contact tips from Harbor Freight where I bought the welder. I'm not sure what the difference is, maybe the Harbor Freight contact points were the wrong size... But it's like a new welder now. I've had what I'd call hit and miss success with the welder in the past. One bead would be great, the next I couldn't keep an arc on. The new wire and tips are from KT Industries, sort of a generic brand I guess but boy am I happy with them.
The fins after the rough cutting.
I worked on the fins today, they are 3/16" thick 6061 aluminum. Luckily I have access to a large metal shear so cutting out the profile was easy. I wasn't so sure about the tapers on the leading and trailing edges though. I'd had nightmares about spending days sanding the edges down, using dozens of sanding belts in the process. But my fears were unfounded as the sanding went fairly quick, using only two belts and starting a third. Once the rough sanding was done with the 50 grit belts I used some 80 grit paper on a palm sander to take out the deep cuts from the 50 grit. I'll still have to finish sand and polish all the surfaces, which no doubt will take longer than the initial sanding, but the part I worried about the most is done.
I'm having second thoughts about using 3 sets of fin brackets on each fin. I think two will be more than adequate. Once I finish the recovery tests, I'll use the length of 6" tube on that test rocket to do some tests on the fin brackets. I have a feeling two brackets will be so strong I'll destroy the fin before the bracket budges.
I ran the math on fin loading in the event the rocket went into a tumble at peak velocity. Burn out is expected at 14,979' AGL so air density in slugs is .001398642. Peak velocity expected is 1970.8 fps. Cd of 1.0 and a single fin area is .5816 square feet. Using the drag formula that gives me a peak loading of 1,579.7 pounds force on a fin. While it's pretty unlikely that would ever happen, it's nice to know how much force could be exerted on a fin. At around 5,000' the maximum loading on a fin would be about 293.5 pounds force.
It would seem I'm a bit of gluten for punishment, I pretty much had all the fin brackets done, and I was starting to polish them... Then I decided they were just too big and heavy. Mind you, I like things strong, but this was overkill to the extreme. So I started all over again, this time using 1" x 1/8" angle aluminum rather than 1.5".
Here's the first fin attached using the new 1" brackets.
Here's the other side of the fin with the recessed flush mount nuts.
The fin attachment is going pretty well. The brackets aren't exactly perfect, I have a little rise on one end of the nut side bracket. But, I checked alignment it looks to be perfect. A quick push, pull and pound test on the fin also looks promising. The fin and the lower motor section feel like one, solid piece of metal. At some point I'll still do a real load test on a bracket to see what it's yield strength is, but I have a feeling I'll be fine.
I completed a series of deployment tests in static tests 150, 152, 153 and 154. I'm comfortable I'll have no problems with deployment. I had been planning on upgrading my mini telemetry board with a new one and some added features before the Ganymede flight. But things are working so well I almost hate to change anything.
The fins are all attached, so I dropped the nose cone on the combustion chamber. Sort of looks like one of those space ships from a bad 50's SciFi movie! Well, once I add another 12' of body tube the look should change...
I spent the better part of a full day working on the slip coupler and valve plumbing. For starters I had to make a tool holder for my 3/4" NPT tap. I've found I can't get threads perfectly straight by hand. I usually use my lathe tail stock and drill chuck to hold taps, with the part in lathe chuck and the tool in the drill chuck, you get perfectly tapped holes. But this 3/4" pipe tap was too big for the drill chuck, so I made a tool to hold the tap, with 2) 3/8" bolts on the square end of the tap to hold it in place. Then I turned one end down to fit the drill chuck.
Here's the tap holder in action as I cut threads in the new slip coupler.
Here's the new slip coupler more or less done. On the right is the combustion chamber bulkhead, on the left is a 5" bulkhead from the tank on the last static test. I need to order some new material for two more 6" bulkheads.
I almost had to replace the combustion chamber bulkhead. I tapped it earlier this summer for the static test, but it was out of alignment. Since the slip coupler is assembled blind, I need the parts to mate together with very tight tolerances. Threads being a little off wouldn't work. I was lucky in that the first time I tapped the bulkhead, I didn't go as deep with the threads as I could have. So I chucked the bulkhead up in my lathe and used the new tool holder to tap a new set of threads perfectly perpendicular to the face of the bulkhead. I ran these new threads deeper into the bulkhead, into new material so there would be no loss of strength.
6" x 12' Oxidizer Tank: 32.64 pounds
Nose Cone: 8.8 pounds
3) Bulkheads, Valve, Plumbing: 10.3 pounds
Combustion Chamber/Fin Can, Nozzle, Retainer, Bolts: 26.2 pounds
Parachutes and Recovery System: 6.1 pounds
Sensors and Hardware: 4.0 pounds
2) Couplers: 3.0 pounds
Dry Weight: 91.04 pounds
Fuel Grain: 14.0 pounds
N2O Load: 80 pounds
GLOW: 185.04 pounds
I've started the tedious job of polishing the fins, brackets and the combustion chamber. I've got the top two fins done. I have to remove all the bolts and brackets and polish each one individually.
You can see the drill with the polishing pad on it. Once I have a fin off I polish the body tube with the drill and pad. The fins and brackets are polished on a polishing wheel.
Here's one side of a bracket. I did make an attempt at angling the leading and trailing edges to make the brackets just a little more aerodynamic. All the bolt heads are counter sunk into the bracket.
Here's the other side of the bracket. You can see the flush mount nuts on the upright portion of the bracket.
I'm using this Loctite product to seal the bolts that go into the casing. It's a high temperature sealant that also locks the threads.
It's been a few weeks since I last updated this page. Unfortunately, there's bad news. The FAA has decided to limit amateur rocket waivers to 23,000' MSL in my area. It has to do with the FAA's reluctance to re-route east-west air routes, or allow me to "shoot between" flights even with real time contact with air traffic control. This puts a definite crimp in my abilities to proceed in step by step manor. I guess I understand what the FAA is trying to do, but I think it also leads people to take bigger steps, larger projects without previous tests to learn on smaller projects first. This no doubt, will lead to more failures, and I think, a less safe launch. In the long term, I'll be looking for a launch area in the upper Midwest, North Dakota is a likely candidate, and a location the FAA recommended to me.
In the short term, it means I've had to re-think the Ganymede project. I'll have to try to fly to only around 20,000' AGL. I ordered my N2O tank/body tube a couple of weeks ago, my supplier called today to let me know it's here. So I'll be picking it up this week. I ordered a full 24' length. I think I'll cut out a 9' length for this flight, and save the longer section for a future flight and possibly utilize an even longer N2O tank, should a higher waiver become available.
With the reconfigured rocket, I calculated the nose cone weight at 11 pounds, and the lower rocket section burn out weight at 75.1 pounds. I needed to get parachute configuration decided , with the weights known, I can now determine descent rates.
I'd like to keep the nose cone splash zone reasonably close to the main body. Although I think it would be better if they were actually a 1/4 mile or so apart. That would give me an extra chance at hitting a high spot in the terrain for radio direction locating. Even if I design the chutes to come down at exactly the same rate, I think they'll still be a good distance apart though...
Nose Cone:
18" drogue = 83.7 fps
48" Main = 29.4 fps
Lower Body:
48" Drogue = 85.4 fps
120" Main = 31.1 fps
5 October, 2008:
Work on the Ganymede came to halt 2 weeks ago. While working on my launch control trailer, I managed to run my left index finger into my table saw blade. I ran the finger into the saw blade from the tip, parallel and under finger nail. Half of the finger tip bone was cut out, and the remaining bone broke in two up to the joint. While it's just a finger tip, it's enough that I can't use my left hand, and healing is slow. The top of the finger may be dead and require more surgery to repair. At least the swelling has come down enough, that in the last few days I've been able to do some work. If all goes well, I still expect to have the Ganymede ready to fly next month. Typing is slow with a bad finger in the way, so don't look for detailed updates in the near future...
9 October, 2008
The finger situation has improved enough that I've been making some progress...
Here's the back side of the nose cone payload.
Above the large black battery pack is my Missile Works RRC2X 40K altimeter, to the left is a new Big Red Bee 70 cm locating beacon, at the very top is an LM 317 regulated power supply for the 70 cm transmitter. I'm using the 10 C cell battery pack to power everything, the (2) 9 volt batteries are wired in parallel and will supply about 2 hours of power to the entire electronics system should the nicad pack fail, or discharge below the voltage of the 9 volt batteries. A Schottky diode automatically switches to the battery pack with the highest voltage.
I've been using the Missile Works altimeter to test a new barometric altimeter based on a Basic Atom 28B chip. The new altimeter will only serve one function, that is to deploy the main chute on the main body section. I went ahead and wrote the code as if it were a dual deploy altimeter, that is, both apogee and main deployment. But for the Ganymede the nose cone electronics will deploy both drogue chutes, and the main body electronics will deploy the main at about 1,000 AGL. I used the same Honeywell pressure sensor as I used in the mini telemetry system, and it's rock solid. I ran about 20 tests of this new altimeter in my vacuum chamber along side the RRC2X altimeter, and it hit spot on apogee every time. One thing I noticed about the RRC2X, it fires the apogee charge at it's altitude limit of just over 42,000'. While my new altimeter continues to function all the way to almost full vacuum. I'll do a full web page on the new altimeter this winter, after I've had a chance to refine the features a little more. Right now I've got all the electronics working and installed in the nose cone in flight ready condition.
Here's all the bulkheads and couplers, ready to begin the "bolt up" of the components.
The tough part comes now, drilling and tapping all the threads to assemble the rocket. I'm not sure if I'll even wire up the pressure sensors for the telemetry system. I'd like to have them installed, but it is going to be a lot of extra work, and I'm running short on time. I'd like to have the rocket flight ready at least 2 weeks prior to launch date, so that leaves me just 2 weeks now to finish. I'll leave those pressure sensors as an option, if all goes well, and I have time, I'll work on them. If time doesn't permit, I'll leave them for the next flight.
12 October, 2008:
First Fitting of the complete rocket.
Another shot from the nose end.
Here's the bulkhead that drops in the top of the N2O tank. The main body electronics will mount to the aluminum plate under the bulkhead. The plate drops onto the internal retainer for the N2O bulkhead and is retained by a 3/8" threaded rod.
Here's how I made my oval tube round enough to get the bulkhead and bulkhead retainer in. Once the bulkhead and retainer are inside the tube, it holds its shape well enough to allow the nose cone coupler to fit inside.
Here's a first look at my new launch control shack mounted on one of my converted boat trailers. I designed the shack so I could unbolt it from the flat bed trailer, jack it up, and still use the trailer for other jobs if need be.
Here's a look inside the trailer.
This launch control shack is really something I wanted. I seriously considered buying an enclosed trailer, but I couldn't justify spending several thousand dollars on it, at least not when I could build something for several hundred dollars that would work just as well. It's not pretty, but should be functional. Now I'll be able to set up the trailer with everything I need for a launch, and not have to worry about loading up vehicles and setting up everything at the launch site. I even bought a PA system, so now spectators will be able to clearly hear the launch events.
I'll have 2 separate 12 volt power supplies; an externally mounted lead acid battery and an internal AGM battery pack for back up power. I still have a good deal of work to do, but it's useable now. Another nice aspect of this is that we can get inside out the wind and cold, making winter launches much more appealing...
15 October, 2008:
Looking into the engine/tank coupler. In the center is the slip coupler to allow N2O flow into the injector bulkhead, in front of the slip coupler is the 3/4" ball valve with the hex coupler to externally initiate N2O flow. To the right is the one way valve and nylon fill line.
I'm making progress, but it's still slow going. It seems like everything takes hours longer than I expect. Yesterday I cleaned everything that comes in contact with N2O. I use Dawn original dish soap and a variety of nylon brushes to scrub the parts, then rinse well in fresh water and air dry. The o-rings are Viton and I use Krytox grease, both are probably overkill with N2O, but better safe than sorry. I decided not to use the pressure transducers on this flight. It's going to take a lot of hours to configure the transducers, so I'll wait until this winter to work on the pressure sensors. In all likelihood, in one variation or another, this will be the only rocket I build for the next year and a half or so. So I'll have plenty of chances to test out the full electronics package on future flights.
Since I had extra room from not using the pressure sensors, I decided to loop the N2O fill line inside the coupler area. On the Defiance-H, I used a fitting the penetrated out the side of the rocket. But just running the fill line out the side allows for a closer cut and less drag. I chamfered both inside and outside edges of the N2O fill line hole, to prevent chaffing and weakening of the nylon fill line.
20 October, 2008:
Progress continues, albeit slowly. I performed a leak test of the N2O using dried, filtered air. I had a major leak at the lower bulkhead (valve end). So I removed the coupler, pulled the bulkhead and went inside the casing and relieved the casing through holes slightly with a counter sink tool using just my fingers. The idea is to keep the o-rings from getting cut as the o-rings on the bulkhead pass by the holes. After thoroughly cleaning things up again, I repeated the leak test. I use soapy water in a spray bottle to test for leaks, this time the o-rings sealed, but I noticed a small leak on the pipe threads going into the bulkhead. So I again removed the bulkhead, retaped and applied sealant to the threads and tested again. This time, everything sealed up with no leaks detected. I tested the upper bulkhead, and did get a few bubbles from the soapy water. But I'm going to leave the upper bulkhead alone. I was only using about 85 or 90 psi to pressurize the tank, and I think under the N2O pressure it will fully seal.
An added benefit to the design I'm using is that both tank bulkheads open into vented compartments, so even if there is a slight leak, it shouldn't result in any problems.
I cast the fuel grain last night, the grain came in at 12.4 pounds and is probably the best grain I've cast in polyester resin/HDPE pellets. The nozzle is installed now as well. Final assembly of the combustion chamber should be complete in the next 24 hours.
26 October, 2008:
Here's a look at the Ganymede with the painted nose cone and the body polished up a bit.
The last week or so I've been doing a lot of the little odds and ends jobs getting the rocket ready for launch. The fuel grain has been installed in the combustion chamber and both nozzle and bulkhead ends have been installed and are ready for flight. My fears that the slip coupler wouldn't align were unfounded, as the assembly went together without a hitch. As you can see in the picture above, I've painted the nose cone and did a light polish job on the body tube. I installed the launch lugs a couple days ago, and yesterday John stopped over to help me load the Ganymede on the launch rail and adjust the valve opening mechanism.
I finished up the parachutes the other day as well, then started having second thoughts about using a four chute recovery system. I know I'd be better off just using one chute on each section of the rocket, but if there's wind (and there usually is around here) I don't want 6 miles of drift, nor do I want the rocket sections landing at 100 fps. So the only other option is to dual deploy both sections. The added complexity of this system at times is frustrating, but in the end I think it will be worth the effort.
I have a few more retaining bolts to install on the nozzle retainer. When I built the case I was only thinking of static testing, not using the case for a flight. So I used 1/4" bolts in the nozzle retainer, those bolts are too long when I counter sink the case, so I've been grinding down each bolt, then cleaning up the threads and installing them in the retainer as I counter sink each case hole. It's slow going, it takes 15 or 20 minutes a bolt and there are 18 of them. I get bored quickly, so I only do a few each day...
Most of what's left is installing the recovery system. I'll test my new altimeter a couple more times before I do the final install in the main body tube. The nose cone electronics will have to wait until launch morning, as I'll want a fresh charge on the battery pack.
30 October, 2008:
I've been working on some the dozens of little jobs that need to be completed before launch day. Today I made a new PIRM2 to activate the Ganymede N2O launch valve. While I've got plenty of extra PIRM2's laying around, when I'm not space constrained, I like a longer unit that allows for a longer e-match. The small units require a short tip and are always at risk of breaking off the pyrogen, worse case scenario would be to expose the nichrome and short the e-match. So I went ahead and made a new, long version PIRM2.
The recovery system on the main body tube relies on a 3/8" threaded rod to anchor the chutes. This rod threads right into the upper bulkhead. I didn't want to take a chance of the threaded rod turning out under the chutes, so I used Loctite to glue the rod into the bulkhead. Earlier this week I ordered a 3/8" forged eye bolt to attach the shock cord. I was going to buy an eye bolt locally, since all I could find was the "wire" style, I was going to weld the eye closed myself. But in the end I decided to play it safe and order forged, high strength eye bolts from McMaster-Carr.
I've been kicking around what to do with the second pyro channel on the telemetry system. I've got the RRC2X for apogee and main, and channel one of the MTS as a back up on the apogee charge. I could use the MTS second channel for a back up to the main chute deployment, or I could use it as another apogee backup. Since apogee deployment is the most critical, I decided to use the second channel for additional apogee redundancy. It didn't seem to make much sense to me to use 3 e-matches on one charge, an extra charge could come in handy if for some reason the first charge doesn't get the nose cone off the rocket. So, I decided to make another charge holder for the second channel on the MTS system. This second charge holder will hold an oversized charge, about 1.5 times what I'm using in the primary apogee charge holder. Now I'll be able to use channel 2 as an all out "desperation" deployment attempt should the primary system fail for some reason.
Yesterday I cast some 20 gram APCP grains for preheaters. I also installed the mounting hardware for the N2O tank in the launch rail trailer box. I also made some mounting brackets to hold the N2O fill valve, made a through wall penatrator for the N2O line, and tested the entire wireless launch system again. The FAA called yesterday too, and told my waiver was on the way... At "T minus" 9 days, the forecast is calling for clear weather, but cold. With highs in the lower 40's, hopefully it'll be a touch warmer than that. But if we have clear skies and light winds, I'll take a cold day.