Links to Other Pages of Interest on the Defiance Rocket:
This project has been in the back of my mind for some time. I was originally going to use the 6" diameter "O" motor I built and tested in the summer of 2005. But as I got into the simulations, it turned out the motor wouldn't meet the requirements of the flight. As you get into larger diameter rockets, the drag penalty increases dramatically with diameter, so I decided to go with a 5" minimum diameter rocket. This rocket will be very similar in design and construction to the earlier A2MD rocket, with a few exceptions.
I'll use 5" OD 6061T-6 aluminum for both the motor section, and the upper body tube. Calculated burst pressure of the tubing is 2,250 psi for .125" wall tube, well within safety limits as maximum chamber pressure should be around 500 psi. I'll use a carbon/glass fin can as in the A2MD, and an aluminum coupler/bulkhead retainer will be of similar design as well.
Some differences from the A2MD will be a glass composite nose cone, and I'll have to retain the nozzle (graphite instead of steel) with a bolt pattern in a retainer as my lathe isn't long enough to machine a snap ring groove in the casing.
I'll go glass composite with the nose because it's RF transparent, so I'll locate a transmitter in the nose cone as well as a redundant drogue system. I don't think the rocket will travel at high speeds long enough to melt the nose cone, but I'm pretty sure I'll lose some paint.
Am I excited about this project? Let's put it this way, my toes tingle when I think about it! So, sit back and follow my progress. It may be a long ride but should be worth the price of admission.
Here's a preliminary drawing from an AeroLab capture.
Here's a screen capture from FPRED showing the motor parameters.
Here is the thrust/time curve as predicted by FPRED.
This is a capture from Winroc showing the predicted rocket performance.
This capture from Winroc shows the batch mode calculating optimum weight. As you can see, 23.8 pounds would give me peak altitude.
22 June, 2006:
I ordered the aluminum tube for the Defiance today. As you can see by the information above I decided to go with an 8' motor section. I had been debating going with a 10' section, but that was an extra $120 shipping... And the length to diameter ratio would have been much higher than anything I'd built before. In the end, it seemed the 8' motor section was more practical from a safety standpoint as well as monetary standpoint.
I needed a mandrel for making casting tubes. I need 4.25" propellant diameter, and I've been unable to find a cardboard tube that size, which was as I expected. So I'll have to make the tubes myself. I found thin walled 4" PVC pipe has an outside diameter of 4.25", I would have liked something a little heavier duty, but the PVC is readily available and inexpensive so I thought I'd give it a try. I'm using plain Kraft water activated 2" tape from U-line. A layer of clear vinyl was wrapped over the PVC pipe, then 8 layers of tape were spiral wound over the mandrel. I used an iron on medium setting to help remove wrinkles and set the glue.
This is the first casting tube made from the PVC mandrel (right).
The 8 layers of paper tape didn't come close the thickness I wanted, as the tube wall was only .045" thick. I'll have to try it again, only using about 15 layers next time.
Here's a rough drawing of the nozzle for the Defiance.
The nozzle will be turned from graphite, with a 1/2" steel retaining ring. The ring will be bolted in place with 16) .25" stainless steel screws.
I've been playing with some recovery ideas, so I started a page just for the recovery system here.
The Defiance body/motor parts. Long tube at right is the motor/lower body tube, shorter tube to the left is the upper body tube, next to the left is the drogue mortar tube which needs to be cut into 2) 24" lengths, next are 2) 24" aluminum 1/4" pipe nipples for deployment charge holders, 5) 2"" aluminum 1/4" pipe nipples for deployment charge holders, 2) 1/4" forged eye bolts and 1) 5/16" forged eye bolt, 2) 5" diameter x 1" aluminum disks for a motor bulkhead and an upper body tube bulkhead, and lastly is a 5" x 1/4" x 12" aluminum tube I'll turn down for a body tube coupler.
14 July, 2006:
After spending most of my time working on the recovery system for the Defiance, I decided if I wanted to test the motor in the near future I'd better get at motor making. I've also been spending time doing some workshop remodeling, despite the mess of construction I decided it was time to work in the shop, rather than on the shop.
First order of business was to turn 2 casting fixtures, the base and centering cover for casting into my custom made paper casting tubes. Each fixture has a slot cut .2" deep so the cardboard tube can slip into the recess. This allows me to cast directly into the cardboard tube without an outside support, it also creates the gap between grains in the motor so no spacers are required.
Here's the casting set up. The picture was taking on top of a glass case, just in case it looks confusing.
Speaking of casting tubes. I've been making a few more of those for the motor too. I settled on 14 layers of paper for a tube thickness of about .1". I learned one thing, making these tubes is very tedious. I spend a good hour making a pair. That is, I wind up a tube about 18" long and cut 2 good 7" casting tubes from it. There's about 15' of paper in each strip, every other layer I use a hot iron to help flatten the paper and remove any wrinkles. The iron also helps to set the glue and really helps keep the tube smooth. If I could buy these tubes, I sure would...
Despite the oppressive heat and humidity, I forged (pun intended) ahead on the Defiance motor. First item today was the one thing I'd been dreading most of all, the steel nozzle retainer... It is starting out as a 6"x6"x1/2" plate. To get it chucked up in the lathe I'll drill a 3/4" hole in the center, then run a bolt through the hole and secure it with a nut. The threaded end of the bolt can then be chucked up in the lathe and the piece can be turned from a plate to a disk.
Here is the plate after being turned down to 4.74" in diameter. I can now remove the bolt, chuck it up in the lathe jaws and turn out the inside. I'll wait with that though, until the nozzle is done and I'll match the retainer to the nozzle divergent cone.
Here I've started on the nozzle. This is left over 6" graphite from the SBS-6250 motor, so I had a little extra work to bring the outside diameter down to 4.74".
I've been working on the fin can too. This is the tube part of the fin can. I laid up 5 1/2 layers of 6 oz. glass and 2 1/2 layers of carbon on the tube.
For the fins, I was going to lay them up with a 1/4" plywood core, as I had done in the past. But I couldn't find any good flat plywood... So I decided to use a foam core, but after I bought the foam, I really had second thoughts about that. This rocket should fly at speeds over mach 2, and I worried the fins may ablate into the foam core. I wanted the leading edge sharp, and strong. So I decided to try a lay up of all glass and carbon fins. I set about making a composite sheet, then I could cut the fins from the sheet. For the first lay-up, I used two layers of carbon in the center and fours layers of glass on either side of the carbon, a total of 10 layers. I'll lay-up another 10 layers, 2 more carbon and 8 more glass for a finished thickness of about .15". Another 3 layers on each side of the fin to attach them tip to tip to the tube, and I should have fins of about .19" thick and very strong. That will be 26 layers of glass and carbon when done!
I made good progress today. Here is the finished nozzle and retainer. The steel retainer continues the 30 degree exit cone to increase the divergent exit diameter.
I also cast the first grain tonight. Each grain will weigh just under 5 pounds. I ordered more erythritol today as well, after checking my supply I found I was down to just over 10 pounds and I'll need over 20 pounds just for this test. I need 12 grains in all, and I'll be pushing to get all the grains cast if I want to test the motor next Friday. My order of erythritol should be here Tuesday, if I get 6 grains cast in the next few days, I'll have to cast at least two grains a day on Tuesday, Wednesday and Thursday. Doable, but considering the time it takes for each grain to cool, I won't have any extra time...
Work on the motor is almost complete. The forward closure is now complete, it's a one inch thick aluminum disk with a hole drilled and tapped for a pressure transducer hose. I almost screwed up the bulkhead, I was going to use my "bolt through a hole technique" to chuck the disk in my lathe, but the 3/8" bolt wasn't heavy enough. So I had to chuck up the disk on the very edge of the lathe chuck jaws, then turn the disk down from one side, then reverse the disk and do the other half. I was going to use thicker o-rings as well, but I didn't have the needed distance to get a tool into the work for the wider grooves the big o-ring required. So I ended up cutting the glands for my standard .103" o-ring. Which shouldn't be a problem as I used the same o-ring width in my previous 6" motor.
Here you can see I've drilled and tapped (12) 1/4" holes for nozzle retention. The holes are in clusters of 3 with extra space between the holes for the fins on the fin can. If I needed to, I could add 4 more screws and ride the fin can above the bolt pattern. But the idea is for these 12 screws to retain the nozzle by passing though the fin can first, retaining it as well. The stainless steel machine screws are rated at 80,000 psi tensile strength, which should give me a working pressure maximum of 2,250 psi. Actually I'll have more margin than that, because the nozzle doesn't have as much loading as a bulkhead would, and I calculated the load as if it were a bulkhead.
Speaking of the bulkhead, I'm planning on retaining it with the coupler, as I did on the A2MD. The problem is that I have to turn a piece of aluminum tube down from 5" to 4.75", and my steady rest won't handle 5", nor my bull nose live center. So what I'm going to try to do is ram a length of 4" schedule 40 PVC pipe into the aluminum tube, then use a reducer on the PVC the to allow me to use my 4" bull nose live center. Right now I got the PVC into the aluminum tube by freezing the PVC,then pounding it into the aluminum tube. Later tonight when it cools down a few degrees, I'll try turning it down. If this fails for some reason, I'll retain the bulkhead with a conventional bolt pattern and worry about the coupler later.
17 July, 2006:
Here is the rough turning of the tube using the PVC pipe mandrel. I tried to avoid deep cuts as to not stress the PVC too much.
Here is the finished coupler after a little sanding and cutting to length.
I did see one big problem last night when I started turning the coupler. I didn't have an accurate device to measure the diameter of a 5" tube such as this. The largest micrometer I have is 3", and the calipers I have don't have long enough tabs to reach over the tube. So I ended up buying a large outside caliper, you know, the tear drop shaped thing that opens at the bottom. I was a little concerned about it's accuracy, but it seemed to work fairly well in the end. I just had to be very careful and precise in the measurements.
The coupler turned out well, better than I expected considering I used PVC as a mandrel and the outside caliper to measure. In fact, I may have gone for too much accuracy, as the coupler is a very snug fit in the tubes. Just as well I guess, I can polish them up a little more, and I'd rather a tight fit than a loose fit. The PVC pipe mandrel I pounded into the aluminum tube wasn't coming out without a fight. I had expected that and had a plan in place. I preheated an oven to 250 degrees, hot enough to soften the PVC but not so hot as to anneal the aluminum. After about 30 minutes I pulled it out of the oven, the PVC was like soft rubber, and pulled right out.
The coupler has been installed on the motor tube in this picture.
I was going to retain the coupler into the lower motor tube with 1/4" stainless steel machine screws, but I decided to go with 3/16 so I could counter sink the holes and keep the screw heads flush with the surface. Instead of 16 larger screws, I went with 30 of the smaller screws. I know 30 screws are overkill, the casing has a burst pressure of about 2,250 psi and my 30 screws will handle well over 3,000 psi. But if I have a failure, I'd rather not have the forward bulkhead ejecting into the electronics...
So after several hours of drilling, tapping threads and blistering fingers I had the holes all drilled and tapped. I still need to countersink most of the holes, my rechargeable drill battery died so I decided to finish the countersinking later. I'm really pleased with the coupler, I put it back in the lathe and filed it slightly, then sanded again and it's about as close to a perfect fit as could be hoped for.
One thing that kind of "gets my dander up" is the condition of the aluminum tube. It comes pretty scratched up, and in the case of the 4' upper body tube there was a slight dent in one end, the end the coupler slides into. It wasn't much of a dent, only a couple of hundredths of an inch, but enough that the coupler wouldn't go in. To remove the dent as best I could, I made a rod from some 1" by 1/2" walnut wood, and sanded a curve in one end. Then I laid the aluminum tube on a piece of curved lead, and pounded out the dent with the wood tool and lead anvil. I cleaned it up enough that it now fits in the coupler, but the other end (nose cone end) is an absolute perfect fit. It slides in smoothly with no play whatsoever.
Casing Length: 96.125"
Forward Bulkhead: 1.00"
Coupler: 4.42"
Nozzle: 5.84"
Nozzle Retainer: .5"
Nozzle End Space: .75"
Internal Casing Length: 83.625"
For heat protection of the forward bulkhead I applied a layer of RTV then a disk cut from EPDM rubber. You can see the short pipe extending into the combustion chamber. This pipe has several holes drilled in it as well as the open end. This is the source for the pressure transducer tap into the combustion chamber. The idea is to try to keep it open by using the extension with multiple openings.
Here is the bulkhead with the EPDM layer glued to it. I allowed it to cure overnight before motor assembly.
Here are the last three grains cast.
All the grains have been wrapped in a layer of heavy foil tape, and the pyro mix has been painted on all burning surfaces. To get 7' and over 56 pounds of propellant grains into the motor could have been a challenge. But I've been here before, so I rolled out the EPDM thermal liner on the floor, then used a straight edge to align all the grains. I made sure the EPDM sheet was square to the row of grains.
Then the EPDM was wrapped around the grains. I used two layers of EPDM, I had the extra diameter in the casing so I figured I might as well make use of it. The seam on the EPDM was closed with foil tape. Once the grains were "packaged", I used talcum powder on both the EPDM liner and inside the casing. Then slide the "package" into the motor. Once the grains and liner were in proper position in the casing, I used high temp RTV to seal the ends of the EPDM liner to the bulkhead and nozzle as they were installed.
Here's the pressure transducer I'll use for data acquisition on this test. The pressure port in the transducer and the first fitting were filled with silicone grease to help protect the transducer from hot combustion gas.
Here is the transducer installed in the forward bulkhead.
I made a thrust plate from 2 layers of 3/4" plywood, between the layers I cut a groove so the data cable isn't cut or exposed.
This was taken after the first test and you can see the indent the 3,400 pounds of thrust left in the plywood. I used a few wraps of electrical tape to hold the plywood plate in place for the static test.
27 July, 2006:
The first test of the Defiance motor was performed on July 21, the results can be seen on the Static Test 129 page. The motor burned about half as long as expected, and looked like it was going into a nice flat, stable burn, then the burn accelerated and created a big hump on the pressure data curve. I believe the odd burn may have been caused by one or more of three things. First, I think core to throat ratio needs to be increased to reduce erosive burning. Next I think I may have had some inhibitor failure and may need more robust casting tubes. Lastly I had pushed the Kn a little higher in this motor in the first place, considering the long length to diameter ratio I probably should have kept the Kn a little more conservative, and I will on the next test.
It looks like I'll be getting some custom made casting tubes from Yazoo Mills, I had purchased stock tubes from them in the past, and I was pleasantly surprised I could get custom tubes for a reasonable price. These tubes will be 4.375" inside diameter with a wall thickness of .12". This is both larger in internal diameter and thicker than I used on the first test of the Defiance motor. I'm increasing the core diameter of the propellant grains to 2", so with the increased diameter of the grains I should still be able to get a full propellant load of 58 pounds.
I sort of wonder if in the first test the grains were too loose of a fit in the casing. If the grains offset somewhat, that could lead to erosion of the propellant or breaking off of small bits of propellant in the core gas stream. I don't usually like to run grains too tight though, as I like to keep them loose enough for the grains to pressurize from behind to prevent cracking. This new grain configuration will run about as tight as I ever have, with only about .045" overall clearance.
One other aspect of the burn I've been thinking about. With such a long length to diameter ratio I'm going to get some erosive burning no matter what I do. So what about evening out the burn by speeding up the burn in the forward grains with some iron oxide... Perhaps .5% for the first couple of grains, then .4% for the next two, .3% and so on until the last 4 or 5 grains had no iron oxide. The problem is, I haven't done enough tests with lower amounts of iron oxide to determine how much a given percent would speed up the burn. If I do decide to try this, I'll likely play it safe with a .25% iron oxide in only the first few grains. But this does seem to be another practical solution to the problem at hand.
31 July, 2006:
I picked up a bottle of sodium silicate solution today to see if it would work as an extra inhibitor layer on cardboard casting tubes. I understand it is very fire resistant so I thought a test was in order.
Here you can see one of my homemade 5" casting tubes cut in half with the one on the right painted with the sodium silicate solution. The solution is about like a heavy sugar syrup or almost honey like consistency. I allowed the painted tube to dry about 1 1/2 or 2 hours, then proceeded with the burn test.
Tube on the left was coated with sodium silicate solution, the one on the right was uncoated.
I used a small hand held propane torch set to high, 1.5" from the surface of the tubes. Each tube was subjected to 10 seconds of flame from the torch. As you can clearly see in the picture, the tube coated with sodium silicate performed much better. Only the first layer of paper was cracked and no charring occurred, the tube showed no signs of continued burning after the torch was removed. The uncoated tube burned through at least 4 layers of paper, and continued to smolder after the torch was removed.
This quick test would seem to indicate the sodium silicate solution could be an easy and practical additional inhibitor layer for cardboard casting tubes and thermal liners. The solution leaves a smooth, shinny surface on the cardboard, so I don't think I'd use it on the inside of the casting tube because it would likely lead to poor grain to tube bonding.
Other developments today. I placed the actual order for my custom made casting tubes from Yazoo Mills, to my amazement I was told the tubes would ship out Wednesday! Wow, that's great service. I was expecting a couple of weeks at least. I also received my graphite today, with the heat wave finally predicted to break this week, I should be able to get back into my shop and do some work again. My 2" Delrin coring rod material also arrived today, so it looks like I'll be testing the motor again in the near future! Armed with custom made casting tubes, an additional inhibitor in the sodium silicate and a larger core in the grains I really expect this next motor burn to more nominal than the first.
3 August, 2006:
This is the start of the layout for the fins. I'm using 1/4" oak plywood for these, as this was the best quality and flattest plywood I could find. I decided to go 1" wider on the fin semi span, there is evidence that rockets with long length to diameter ratios need a much higher stability margin than shorter, stocky rockets. So better safe than sorry I added an inch to the span.
Here you can see the new graphite nozzle and the fins with a layer of carbon fiber cloth.
The new graphite nozzle turned out very well, in fact it only took 2 hours to turn on the lathe. Well, ok really 3 hours because I made a big mistake when I started on the nozzle. I was about half way done with the nozzle and I was boring out the throat, and I just plain got carried away and bored the throat to 1.65" rather than my intended 1.55". I ran the numbers in Kn Calculator and decided I just didn't want to go that low on the Kn, so I started over with a new piece of graphite. Luckily, I didn't screw this one up because it was the last piece of large graphite I had!
The plywood for the fins was reasonably flat, but to insure absolutely flat fins I decided to apply a layer of my wide weave carbon fiber. This fabric has more strength than a tight weave fabric, but is hard to use and is really best for large, flat surfaces. I laid out the fins on a sheet of 1/2" acrylic, then laid down some vinyl sheet as a release, once the epoxy and carbon were applied to the fins I laid another layer of vinyl over the fins and used a piece of 1/4" plate glass to press the fins flat. I added about 30 pounds of weight on top of the glass for added compression.
The fins came out the next day absolutely flat, and quite stiff. It's a real job trimming the excess carbon fabric, and after cutting the loose material from the edges (as in the left 3 fins), I used my 10" disk sander to clean up the edges (as in the far right fin). I'll have to rough sand the fins to get the next layer of carbon/epoxy to adhere.
5 August, 2006:
Here's the start of the fin can assembly. I used my standard alignment technique, the can and fins are drawn out on paper then squares are used to hold the fins in place while epoxy "tacks" are allowed to cure.
Here's the last fin being tacked in place. Notice I used a couple of 1/2" thick blocks of acrylic to move the fins up 1" from the bottom of the fin can. I decided I wanted to add more retaining bolts to the nozzle and had to get the fins "out of the way" to accomplish this.
Just as in the A2MD I'm laying up four layers of carbon and glass on each fin surface tip to tip. This is the third surface to be covered, once this surface is cured I'll move on to the last surface. Notice how I left the excess cloth hang over the leading edge of the fins. The weight of the excess material holds the cloth against the leading edge making for a nice clean composite leading edge once trimmed and sanded.
The lay-up of the fin can is done, and a little coarse sanding done to take the rough edges down. The A2MD fin can is next to the Defiance fin can to give a perspective of scale. I doesn't seem possible, but I used over 2 yards of carbon fabric and 6 yards of glass fabric in the lay-up of the fin can. This fin can is too big to fit in my curing oven, so I used my car's dark interior and the summer sun to help heat cure the epoxy.
10 August, 2006:
It's time to start thinking about a nose cone. My initial plan was to turn a plug out foam, and use the foam as a core to lay up a "one off" nose cone. But after getting started on the foam core, I really didn't like the way the project was heading, that is, a lot of work for a single nose cone. So I decided to give two part mold making a try. This is a rather involved process, but fairly straight forward, so I decided to detail the work on it's own page.
Click Here for the Defiance Nose Cone Page.
This is another one of my couldn't resist moments. I just had to see what the rocket was going to look like assembled! The rocket is 3/4" shy of 14' long.
I bought a couple of these folding plastic saw horses the other day, then made a couple of wood brackets that snap over the top. The acrylic plastic was scrap left over from cutting out the casting stand base and top. This rocket was a little too big for my 3/4" PVC rocket stands.
7 October, 2006:
After many months of construction and testing; two full load motor tests, two test flights and several changes to the deployment system, the Defiance is ready for a full propellant load flight. A waiver application has been sent in for an October 28 or 29 launch, as it sits now it looks like the waiver will be granted to 25,000'. Assuming a rocket Cd of .61, the estimated altitude should be about 24,754' assuming a nominal flight.
Here's the fin can after the most recent paint job. After fussing around with many different brands of spray paint. I decided to throw all my Krylon paint away, poor nozzles that are hard to push and throw spatters everywhere, and just not the finish quality I'd like to see. One brand I do like is Valspar, not real well know but the paint seems to be first rate. But best of all is DupliColor, it's an auto grade paint with spray nozzles that are easy to use and give excellent coverage. The paint also dries fast and is very tough.
Here's a neat shot of the nose cone looking down the body. The nose cone of course isn't bulbous like it appears here, that's just a result of the camera angle. What looks like dust on the nose cone in places is really just reflections of the floor and ceiling. I have also done an initial polishing of the aluminum body tube.
This is shot of the where the nose cone shoulder meets the body tube. I wanted to give you an idea of how well these two parts mate up. I radiused both the inside and outside of the upper body tube opening so it would be less likely to cut the shock cord. If a rocket is moving at any speed at all when the drogue deploys, the drogue trails behind the rocket until it reaches the end of the shock cord. At that moment there is a lot of force on the shock cord as it's making a 180 degree turn over the edge of the upper body tube. Of course, with cardboard or even fiberglass rockets this is what causes a "zipper". In the case of my aluminum upper body tube it could easily slice the shock cord.
Here I'm holding the nozzle retainer. I purchased a set of transfer punches so I could accurately punch, drill and tap threads into the steel retainer. This should make nozzle retention stronger, as well as adding a little extra length to the useable casing length.
Update: 6 November, 2006:
The Defiance rocket was flown on October 28, 2006 to an altitude of 29,389' with perfect deployment of all three chutes and a soft landing. You can see the information on the flight at the Launch Test 134 page.
Here's the Defiance back home and on the wall. Looks like I'll need a bigger wall in the near future!
I'm assuming this flight was an altitude record for a sugar propellant flight, feel free to correct me if I'm in error. This rocket and flight have been the culmination of a step by step process I've been following for several years. This rocket not only met, but exceeded my expectations. When I started this process, I was looking to find a slow burn sugar propellant. That challenge was accomplished with the development of KNER propellant. I wanted to break mach with a sugar propellant, while I have done that a number of times before, this flight to mach 1.9 was certainly icing on the cake. I wanted to fly a sugar propellant higher than anyone has flown before. I believe I met that goal at 29,389'.
While this rocket and the flight were not truly ground breaking in terms of EX rocketry. It certainly set a bench mark for sugar propellants. I'm really not sure if most people even fully understand what goes into something like this. More time was spent on the nose cone than most people spend on an entire rocket. Countless hours went into flight simulations, calculations of stress loads, thermal loads, structural loads, it goes on and on. Again, while this isn't earth shattering and is done routinely all the time, keep in mind. Everything was done by one person. And that's what makes it so special to me, it was my project. I alone would hold responsibility for success or failure. That's why I take such a deep sense of personal satisfaction in this rocket, it was my project from start to finish.
I can't say I'm done with large sugar propellant projects forever, but I am done with them for the time being. It becomes increasingly more difficult to go higher on a sugar propellant. With an increase in altitude of only 5,000', my propellant load would almost quadruple. That's assuming I go with a wider diameter casing on a single stage. It does seem possible to push the propellant length even more, so it may be possible to squeeze another 5,000' out of a similar rocket. At this moment I'm undecided as to a next project...