Experimental Sounding Rocket Association

Tropos 2 Hybrid Rocket

For pictures, click here. For videos, click here.

Update 9/22/09: We conducted a static firing of the Tropos 2 3" nitrous oxide/sugar motor on September 14 to test a new data acquisition system, main valve, and as a recruiting opportunity for the Rocket Club at Utah State University (smoke and fire!). The motor ignited well, indicating that we have the ignition problems solved, but "surged" repeatedly during the burn (see the data and a discussion here). We think it is because the main nitrous oxide valve we tested has too small a port, causing a pressure drop ahead of the injector. This probably caused three-phase flow (gas, liquid, and solid), causing intermittent partial clogging of the injector. We may go back to the old 3-piece valve, but it had some problems with leaking and being a bit too stiff for the spring to open consistently. It was also a bit heavier than the new valve. We may switch to a 2-piece valve--that decision has not been made yet. A new load cell ($1 university surplus!), in-house instrumentation amplifier ($10 in parts), and $50 data acquisition module were used with good results. It is interesting to note that the chamber pressure was less than in previous tests, yet the measured thrust was greater. We think part of this is because previous tests were done at an angle on the launch rail (with friction), whereas this test used a new horizontal "trapeze-style" test stand with minimal friction.

The USU Rocket Club decided they would like to rebuild the Tropos 2, and began work on new bulkheads immediately. We will work on getting an FAA waiver for a launch in early December.

Update 9/1/09: With the start of the new school year, the Utah State University Experimental Sounding Rocket Club is starting work on the Tropos 2 again. Yesterday we tested the igniter design with 2-micron aluminum powder instead of magnesium. The igniters worked well, igniting by nichrome wire alone and burning with a 2-3 inch flame for approximately 12 seconds. We plan to do another static test of the 3" motor with the candy fuel grain on Sept 10. We plan to use the new horizontal test stand to gather more accurate thrust data.

Update 8/6/09: In preparation for the static firing at the 4th IREC, we reconfigured the motor with an electric valve and mounted the motor and tank on a horizontal test stand. Turns out that when we took apart the motor, the O-ring from the fill check valve had come off and was partially blocking the injector. So that would explain the apparent reduced thrust during the April flight. It may be that there was a combination of that plus the igniters being inadequate...it's possible that on the first ignition attempts during the launch attempt, the igniters couldn't ignite the full flow (the O-ring was still on the check valve). Then on the last attempt, the O-ring came off, restricted the flow, and the reduced flow rate was easier to ignite. We probably screwed up, because we noticed that the check valve was no longer sealing well on the last attempt, and should have put two and two together that the O-ring had come off and perhaps it could get into the main flow path. (The O-ring came off on a previous test. We got a new valve with a different O-ring material, but evidently that didn't help. The next check valve won't have an O-ring!). For the static firing at the competition, we had no check valve installed--just left the fill line connected. We are confident that the current igniter design works, though we may switch from magnesium in the igniters to aluminum because aluminum is easier to get.

The static firing demonstration at the competition went well. It was a demonstration only--no thrust data was collected.

Update 5/1/09: We made some redesigned igniters yesterday, along with "standard" ones for comparison. It was interesting that the standard ones (made the same as the ones for the flight) burned for 33 seconds instead of 17-18 as they had during initial tests. And the flame didn't seem as energetic. We are scratching our heads as to what we changed in the formulation or our process to cause this. Some of the redesigned igniters had a little more flame, but we don't really have a handle on the whole igniter problem apparently.

Update 4/28/09: We disassembled the motor yesterday, and it was in good shape. Very little fuel had burned--only about 0.3" radially and only about 1 lb was consumed. About 3 lbs was consumed during the last static firing. So we obviously didn't have a full burn. But the port was consistent along the length of the motor--it didn't seem to burn just at the aft end, for example. We may static fire this motor again (to test whatever redesigned ignition approach we settle on) without recasting it.

Update 4/26/09: We launched the Tropos 2 yesterday on the third try--we evidently still have ignition problems. We think we have a good idea what happened. On the first attempt, apparently only one of the two igniters fired, and the motor never ignited even though the nitrous valve worked fine. We had a spare igniter, so we replaced one of the igniters (we happened to remove the one that did fire, not realizing that the other was a "dud"), refilled with nitrous, and tried again. This time, the valve didn't open on time--we think it was too cold from the previous attempt's nitrous flow. We were trying to figure out how to re-pack an igniter with some leftover sugar/KNO3 pyrogen that we had brought with us, and when we removed the other igniter we saw that it had never fired. It had an open wire inside it, so we rigged up a bypass wire through the nozzle (kudos to the students for being creative!) and decided to try one more time. The igniter fired, but a good amount of nitrous came out before the motor itself ignited. It may not have had full combustion, too. The thrust was insufficient--it climbed quite slowly and flew straight for a couple hundred feet before nosing over and flying basically horizontally for about 500 feet. As it descended, the recovery sequence started (the nosecone pivot worked) but it didn't have time or altitude to fully deploy (it's not clear yet if the main parachute door opened or not). The rocket impacted at a shallow angle and at high speed, mostly destroying it. The tank, plumbing, motor, and parachutes seem to have survived, but the airframe and avionics were destroyed. We will perform a post-mortem next week, and look at the burn pattern in the motor to try to see how it burned.

We saw that we had delayed ignition on the second static firing, where we fired only one igniter. It is apparent in the video and the data. The idea was that if that igniter didn't fire, we wouldn't open the valve and we'd go switch the igniter wires to try with the second. Apparently one igniter is borderline capable of producing enough heat to ignite the motor reliably and quickly. After seeing the delayed ignition in the second static firing, we decided to fire both igniters simultaneously for the launch (just to be sure). We were on the right track, but didn't make the full realization that one igniter probably wasn't enough to have reliable and fast motor ignition. As luck would have it, only one igniter fired on the first launch attempt (and the motor did not ignite). We had checked continuity right before the launch attempt and both igniters checked out fine. We think the nichrome wire physically burned out before it ignited the "bad" igniter. If both igniters had fired, we probably would have had a good launch, So instead of having redundancy in the igniters, we actually needed both igniters to fire, which reduced reliability significantly.

The delayed ignition during the second static firing.should have been a warning to us that we didn't have our ignition problem totally solved after all. Test, test, test! And think about the potential implications of non-nominal events!

We have some ideas to make it so one igniter will reliably ignite the motor (but we will still use 2 for redundancy). We will decide on a solution and hopefully test again in a static firing. We will probably do the static firing in a horizontal configuration on a flexure stand (with an elbow to keep the tank vertical) to minimize guesswork about weight and strap resistance affecting the measured thrust.


Update 4/22/09: We recast the motor and tested the parachute extraction sequence today. Everything is on track for the launch on Saturday.

Update 4/21/09: The overall thrust level was less with the new corn syrup/sugar fuel (plots posted here), but we also had some leaks where the igniter wires pass through the igniter caps. Nothing was damaged, but chamber pressure and thrust probably did decrease as a result. The nozzle throat eroded somewhat, but not nearly as much as with the caramel. The data does show a much smoother burn. The corn syrup/sugar has several advantages over caramel (easier to cast, smoother burn, less nozzle erosion), so we will use it for the launch on April 25.

Update 4/20/09: The static firing went well, though it doesn't look like the performance of the sugar/corn syrup was any better than the caramel. It did seem like a smoother burn, but total impulse appeared to be about the same. Peak thrust was actually less, though we haven't compared the thrust curves side-by-side yet. The peak chamber pressure was 498 psi (design pressure = 500 psi). The nozzle appears to have fared better, but it's hard to tell because the fuel dribbled out after the firing and the throat is partially obscured. We will know more tomorrow when we disassemble the motor.

Update 4/18/09: We cast the motor with the sugar/corn syrup formulation and will static fire it on Monday. This formulation has less water in it than caramel, so hopefully it will behave closer to the thermochemical predictions for nitrous oxide/sugar. We also hope that it will have less of a tendency to slough off the surface--it may be that the nozzle erosion was caused by liquid droplets or charred caramel impinging on it at high speed. It it also possible that a significant amount of caramel exited the chamber without combusting fully (reducing thrust and specific impulse), but this is just speculation without having a means to identify the exhaust constituents. Another benefit of this formulation is that it should be less likely to melt at high ambient temperatures and ooze out of the motor after a firing. Some potential problems with the sugar mixture is its brittleness (will large chunks break off during the burn?) and its transparency. Some hybrid motor builders put carbon black in their fuel to block IR radiation to increase the temperature at the surface (to increase vaporization and regression rate) and to protect the case wall from radiant heat. We are dubious if these effects are significant, and did not include any opacifiers in the fuel.

Update 4/17/09: We continue to tie up loose ends on the rocket: assembling the parachute, installing the camera and mirror, and sanding and painting. The erosion of the nozzle and poor thrust and specific impulse are problems. We are making a new nozzle for the launch, and are considering an alternative fuel (see below). Postponing the launch isn't an option because the launch area will be used by military aircraft in the following weeks. The launch will test our altitude prediction capability, stability, and recovery.

We made a small-scale motor out of sugar/corn syrup and tested it with gaseous oxygen. It burned quite well. It seemed easier to ignite and more energetic, but we don't have any instrumentation for the small motors so we don't know for sure. We may recast the 3" motor with this formulation and static fire it Monday.


Update 4/15/09: We disassembled the motor yesterday, and it's in pretty good shape. The injector plate looks perfect, but the nozzle throat has eroded significantly (from 0.62" to about 0.81" diameter). This is unexpected and may indicate that caramel should not be burned in a motor with a graphite nozzle. Previous graphite nozzles with HTPB fuel (and nitrous oxide oxidizer) showed no significant erosion after many burns. We burned about the right amount of caramel (2.51 lbs vs 2.63 predicted). We are continuing to look at the test data. The thrust and specific impulse do not match predictions (much lower than predicted). We aren't too concerned at this point, since the predicted thrust and specific impulse would result in a peak altitude close to 20,000 feet according to RockSim Pro (we're shooting for 10,000 ft). The launch is planned for April 25. Remaining tasks include recasting the motor, doing a parachute test (nosecone and parachute door release, and manual chute extraction to check the deployment sequence), installing the camera and mirror, and sanding and painting the airframe.

Update 4/13/09: Success! The motor fired successfully, though the valve opened much more slowly than desired. We will continue to work on that. Plots of thrust and chamber pressure data and a short discussion are here.

Update 4/11/09: A frustrating morning. In our first attempt, the igniters fired well, but the nichrome wire to release the main valve was too short, and burned up before melting the Dyneema. So the valve never opened. We had a spare igniter, so we installed it and tried again. This time, the Dyneema melted just fine, but the valve didn't open! It was very stiff--it took quite a push to manually open it. So we are looking at getting a stronger spring and/or disassembling the valve and relubricating it with Krytox. But, it didn't leak!

We will try again on Monday afternoon/evening. We may still try for launch on the 18th, but things are getting tight.

Update 4/10/09: All systems are "go" for tomorrow's static firing.

Update 4/9/09: We built igniters for the static test on Saturday. We also installed the fins into the boattail.

Update 4/7/09: We tested the latest igniters, and they were still unsatisfactory. We made new igniters using the "sugar rocket" formula used by hobbyists (sugar, corn syrup, water, and potassium nitrate) since we were able to find potassium nitrate locally and at low cost (stump remover from a gardening store). These worked quite well. We used them to test the "shorty" motor with a small amount of nitrous oxide, and it burned well. The valve also opened more slowly because we tightened the packing nut to eliminate leakage. This may have helped with ignition as well. With the successful ignition test, we plan to attempt a static firing on April 11.

We also marked and cut the slots in the boattail for the fins. The fins will be bonded to the boattail and the inner motor sleeve for strength and stiffness. A cant angle of 0.3-0.5 degrees will be used to give the rocket a 1-2 rev/sec roll rate.

Update 4/5/09: We tested the "stepped opening" valve and the valve performed as hoped. However, even the reduced initial flow of nitrous blew the burning igniter pellet out of the "shorty" motor without igniting the caramel. It appears that using an igniter pellet held only in place by the igniter wires will not work. We started work on the pyrogenic igniters which would thread into the injector plate and fire into the forward chamber of the motor. We have a formulation (from previous work) that works (at least with HTPB fuel), however it requires specialty chemicals only available from fireworks/rocketry suppliers. We are working on a design that uses locally available ingredients but have not gotten one to work well yet. Testing will resume on April 6.

Update 4/1/09: We held a brainstorming session on the ignition failure and decided that the most likely causes were 1) the igniter pellet did not provide enough energy, 2) the nitrous flow rate blew the igniter out of the motor before it had a chance to really ignite the motor, and/or 3) the caramel requires more ignition energy than other fuels. We came up with three possible fixes that we hope to test on April 4. We will test them with a "shorty" motor simulator that includes the forward end of the motor, injector, and some caramel, but no nozzle. We will also test with just a smaller amount of nitrous (~1 lb instead of ~8) so we don't waste as much. The three concepts are 1) use two pyrogenic igniters (an in-house design) mounted in the injector plate, firing into the forward mixing chamber (this approach was successfully used with previous nitrous/hydroxyl-terminated polybutadiene motors), 2) try a "stepped opening" for the valve using two loops of Dyneema, 3) coat the inside of the caramel with a more-easily-ignitable fuel: paraffin wax. We will try options 2 and 3 first since they are easier than making the pyrogenic igniters.

We also had some nitrous leakage in the fill line (apparently a seal was missing on the cylinder adapter), and some leakage developed around the stem packing on the main valve on a third firing attempt on the 28th. This froze the valve, and it didn't open during a third static firing attempt. We also noticed that the check valve in the fill line was no longer sealing. We took apart the valves and found that the packing on the main valve appears to have manufacturing flaws, and that the seal in the check valve is gone. We think we can make the main valve work by tightening the packing nut, but may have to buy a new check valve. We think the Viton seal in the check valve may have gotten too cold and broken because the nitrous was leaking too much (it gets extremely cold if there are leaks larger than a small bubble leak). We may switch to a Buna (nitrile) seal as apparently it has better low-temperature properties.

Update 3/28/09: We attempted the static firing of the caramel/nitrous oxide motor this morning but had ignition problems. We are currently in "head scratching" mode. The igniter fired but the motor did not ignite in two attempts. But after the nitrous ran out on each attempt, the motor had a small flame coming out the nozzle. The test objectives of verifying ignition, sustaining stable combustion, verifying airframe structural integrity under thrust loads, and gathering chamber pressure and thrust data were not met. The test objectives of validating the nitrous oxide fill procedure, verifying proper operation of valve and ignition controls, and verifying proper operation of the valve release mechanism (while loaded with oxidizer) were met.

Update 3/25/09: More flight simulations are being conducted, and they now predict a peak altitude close to 20,000 feet! The loaded weight of the rocket is actually closer to 54 lbs (instead of 65--that was an error). We may have to reduce the amount of nitrous oxide to be loaded for launch to keep it to 10,000 feet. This is all contingent on the actual thrust profile (from the upcoming static test) matching the predicted profile--not a guarantee by any means.

Update 3/24/09: The parachute door was installed in the airframe, and the release mechanism is complete. The motor sleeve was installed into the boattail, but the fins have not been installed yet. The nosecone release mechanism was tested. Plans are on track for the static firing on 3/28/09.

Update 3/23/09: The recovery/oxidizer bulkhead was installed. The parachute door was cut into the airframe. The parachute attachment U-bolts were installed in the thrust plate and parachute strap pass-through holes were cut into the recovery/oxidizer bulkhead. The ring knife to open the main parachute deployment bag was fabricated. The access port for the nosecone tensioner (part of the release mechanism) was cut. A port was cut into the airframe to allow connection to the nitrous fill hose and for the valve opening wires. A test of a "doubled" igniter pellet showed that instead of a longer duration (the desired result), it just had a bigger flame. So a single pellet will probably still be used for ignition. All completed components were weighed and the rocket is currently 15 lbs over its original estimated weight (current loaded weight estimate 65 lbs). This is not a major issue because the flight simulation program predicted a peak altitude of 15,000 feet (10,000 is the goal) with the lower weight. More simulations will be conducted, especially once we have actual thrust data. The static firing has been postponed again (to 3/28/09 hopefully) so that it can be conducted in concert with a cryogenic test on the plumbing for the "Procyon" high-altitude (100 km) rocket also under development. Instrumentation ground support equipment for both tests is still being completed and the load cell and pressure transducer have not been calibrated yet. Assuming a successful static test and parachute deployment test, the test launch is tentatively planned for 4/18/09 in the west desert of Utah.

Update 3/13/09: Holes were drilled in the airframe for mounting the thrust plate. The plumbing was finished and assembled. The motor was assembled and mounted to the thrust plate. The valve opening mechanism was successfully tested.

Update 3/11/09: The mandrel was successfully extracted and the fuel looks good. The load cell mount for the static firing was fabricated. Finish work and painting remains. The static firing has been postponed until 3/21/09 to allow time for calibrating the load cell and pressure transducer.

Update 3/6/09: The motor is cast, with the nozzle in place. It is cooling and the mandrel will be removed in the next few days. The injector plate and thrust plate have been modified for installation of a pressure transducer. The valve opening mechanism needs a shorter spring but otherwise is finished. Remaining work for the static firing includes testing the pressure transducer, finishing the tube portion of the plumbing, performing a leak check, installing bulkheads and launch rail buttons, installing the load cell on the launch rail, and testing a modified ignition system designed to give a longer ignition duration. The static firing is planned for 3/14/09.

Update 3/3/09: The graphite nozzle is complete. The foam core fiberglass fins have been laid up (but not installed) and are curing. The valve handle and spring mechanism is in work. Motor casting should occur in the next few days. The injector plate and thrust plate (bulkhead) need to be modified to allow installation of a pressure transducer for the static firing. The static firing will measure thrust, chamber pressure, and total propellant consumption data to determine total impulse, average specific impulse, verify predicted chamber pressure, and look for combustion instability or other problems.

Update 2/25/09: The motor tube, end closures, and injector are complete. The motor case held pressure during its pressure test but did require minor redesign to reduce point-loading at the retaining screws. The graphite nozzle plug is ready to be machined into an actual nozzle. The casting mandrel is complete. Once the nozzle is finished, then we will be able to cast the motor. The fiberglass body tube is complete (less sanding/painting), all bulkheads are complete, and bulkhead fitting/installation is starting. The foam cores for the fins are ready for fiberglass overlay. Actual fin installation into the boattail will wait until we conduct an analysis to determine the desired fin cant angle. The motor static firing (with the motor installed in the airframe) will probably occur without the fins in place. Testing of the non-pyrotechnic nosecone, parachute door, and main valve release technique with the flight avionics is complete. All oxidizer components (tank and plumbing) are on hand but not assembled yet.


Overview

ESRA is developing the "Tropos 2" hybrid rocket with help from the Utah State University Experimental Sounding Rocket Club ("USU Rocket Club" for short). The Tropos 2 is intended to be a baseline design at future Experimental Sounding Workshops.

The Tropos 2 is designed to launch a 10-lb payload to 10,000 feet, and participants in Workshops will get documentation to use it as a "jumping off point" for their own designs. The idea is to get over the initially high learning curve associated with designing and building a rocket from scratch, but not "bypassing" the learning curve that can happen with buying a pre-engineered high power rocket and motor. The Tropos 2 will be an open-source design as opposed to the proprietary designs of commercial vendors.

And finally, the Tropos 2 is designed to be low-cost (<$1000) and easy to build, but will also require some composite layup, machining, fluid connections, and wiring in order to give the students a real hands-on manufacturing experience.


Tropos 2 Basics:

Dimensions: 8" diameter, 88 1/2" long
Loaded weight: 49 lbs
Airframe: fiberglass/polyester composite with plywood bulkheads, foam-core fiberglass fins
Propulsion: nitrous oxide/caramel hybrid, 3" diameter motor, 500-cubic-inch nitrous tank, graphite nozzle, ball valve for nitrous
Recovery: pivoting nosecone at apogee, 15-ft main chute deployment at 1000' AGL, activated by PerfectFlite HA45 altimeter
Payload: digital video camera and T-shirts

GDL-Propep predicts a specific impulse of 237 seconds for nitrous/sucrose (table sugar; the closest thing we could find for caramel. Since caramel includes milk in addition to sugar, we expect the actual Isp to be less). We have had good success with caramel and gaseous oxygen in our 2-liter hybrid rockets developed with the USU Rocket Club for outreach to local K-12 students. Using an estimated thrust curve and the above Isp, a simulation program predicts an apogee of over 15,000 ft. We will perform a static test and get a real thrust curve and plug that into the simulation program for a better estimate of peak altitude.


For pictures of Tropos 2 components, click here.

More pictures and CAD drawings will be posted soon.