THRP-1

[G_T]

So, some things worked well, some were improvements, and some things didn't work.

I have not grabbed any data from the boards yet. This is just a short report of what happened and what I observed, with not much in the way of analysis.

Turns out there was no long 1010 rail available at the time. John Derimiggio volunteered his tower for the launch with his long rail, but it is 1515 not 1010. He had a 1010 rail for it, but not on the field at that time. 1515 could work, but it is not a wise approach. There is some chance using 1010 on 1515 can cause the rocket to fall off. The RSO team nixed that! So I ended up on a short rail that only extended a couple feet past the upper rail guide. The nosecone was past the end of the rail. But conditions were pretty calm with occasional gusts due to thermal activity. The rocket has some mass, and a good expected thrust to weight ratio. So we all thought it should be ok.

I had been busy with the construction, and had not familiarized myself with the Featherweight Tracker, or the RunCam 2 4K video. So I got help from the rest of our local group with those items. The tracker is in the nosecone (important detail for later) so once the rocket was on the rail but horizontal, the tracker was turned on and the nosecone buttoned up and pinned on. The video was turned on and the rail raised to vertical. We had previously adjusted the rail for a very slight tilt into the winds at altitude.

My recollection of events has it that I turned on the video camera, but didn't start it recording. I hope I'm wrong about that! I'll know when I pull the video card and see what's there. This is where my lack of experience with the camera bit.

We basically manhandled the rail via the rocket. The rail guides stayed solidly in place, so the wet sanding prep work did its job. I can state from experience they would have popped off without that prep. Been there, done that.

The rest of the prep was done, the altimeters were happy, the doubled igniter (each ematch augmented with some BKNO3) was installed on a thin dowel to hold the igniters against the forward bulkhead. I THINK the igniters were against the forward bulkhead, but I didn't mark the dowel. Again, rookie error due to being brain fried. Being all the way at the top is a lot more important with this motor than with most AP motors. If the dowel was a half inch short that would negatively affect the startup. It was probably correct, but I don't have proof.

The nitrous tank valve was opened, and we were ready to start the flight tank fill procedure.

We all retreated to the remote location with the remote control box for the fill and dump solinoids. We had one of the range crew with us for communication with the RSO station. The RSO authorized starting the fill, and we commenced operations.

Temperature was somewhere in the upper 80's I think. This was warmer than during the static test, so I expected to boil off a little more nitrous for chilldown. The cryogenic pressure relief valve popped open sooner than had been the case during the static test, and more frequently. But then when the valve was closed, I thought I was hearing some high pressure gas hiss. We were a couple hundred or so yards away so it was hard to be sure.

I got a pause and went closer to investigate. There were three possibilities. (1) venting at the supply tank plumbing. That doesn't have to be an abort. (2) venting into the combustion chamber. That's an abort. (3) venting at the cryogenic pressure relief valve.

When I got closer I could confirm the high pressure hiss was coming from the area of the rocket and not the supply tank, and there was no visual evidence of haze from the nozzle region of the rocket. So I concluded it was (3). From the sound and the lack of a visible cloud, I thought I could fill the flight tank faster than the venting. I guesstimated I might lose 10% rough order of magnitude total impulse, but elected to continue with the launch. The valve was not a problem I could fix short of replacement and I didn't have a replacement. Nor could I get one in time for this launch.

I went back to the remote station and continued with the filling procedure. When the flight tank was near full and the temperature of the tank lowered, the jet of fog comming from the vent confirmed it was not fully closing when it was supposed to close. So it really was (3). I completed the fill and transferred control back to the RSO for countdown and launch.

There was approximately two seconds of cold flow out the nozzle before ignition occurred. This was better than in the static test, but even this beefed up preheater was not sufficient for instant-on. The loss of two seconds of oxidizer was worth about a third of the oxidizer supply, roughly, I estimate.

This may actually have been caused by the valve failure. Since the valve wouldn't close, the flight tank provided nitrous at a lower temperature than planned, due to the continuous visible boiloff. Lowering the temperature of nitrous has a disproportionate affect on the required ignition energy. The flight tank was definitely well below the dew point on Friday and it wasn't a humid day.

But it did light, producing a beautiful straight climb with visually approximately the expected acceleration. The motor was operating normally once it got going.

The burn was again super smooth, just as in the static test. It is rather loud for its size, but does not have the typical pressure oscillations of a hybrid. That aspect of the design, mostly due to the high turbulence fuel, appears to be working perfectly.

Of course with the boiloff and the delayed pressurization, the motor likely only produced about 2/3 the total expected impulse, if even that. But what it did, it did well.

We witnessed deployment at apogee, but it was too visible. And the tracker confirmed via descent rate that the main was out. For the main to go out, the tether must have failed. I've assigned no cause for that yet. It could have been mis-wiring, mis-configuring an altimeter, gas venting into the electronics bay causing a pressure spike and fooling an altimeter, or hardware failure of some sort. Cause TBD.

We observed the slow descent and slow drift to an off-field landing. We also had the GPS coords from the tracker. Since this is a freebag deployment, the nosecone was not expected to be with the rocket.

When we went to the GPS coordinates for the nosecone, and got permission from the farm owners to retrieve, we found the whole rocket. Turns out the tether bolt in its travel up and out of the upper tube, managed to snag and flip itself into a loose knot that stuck right under the tape I'd put on the kevlar main line to prevent chafing at the top of the tube!

Nosecone under drogue was expected to have similar descent rate as the rest of the rocket under main, so there was essentially no stress to pull out the loose knot. This was a fortuitous occurance!

That concludes the quick look report. I'll take pics and post analysis over the next few days.

 

[G_T]

After we observed the landing, it was time to disassemble and remove the nitrous supply tank and GSE. I put a large crescent wrench on the hex cap for the fill stem and loosened it. Then when I went to pull the wrench off the stem, it wouldn't come off!

Turns out the wrench warmed up the cold metal of the stem, and thermal expansion locked it on tightly. We literally couldn't get the wrench off the nut!

In the end, I donned some gloves and vented nitrous for a bit to chill the nut back down, and not freeze my hands. That worked.

It was annoying at the time, but humerous now!

 

[rocket_troy]

Congats on the flight; despite the obstacles it appeared reasonably successful ie. the important boxes were ticked and that's no mean feat for that stage of a complex experimental project.

TP

 

[waltr]

Yes, big congrats on the challenging flight.
Thanks for the report and looking forward to pics, video and additional analysis.

 

[G_T]

From watching Bill Schworer's launch video, GPS altitude 8822ft. The time between preheater igniting and a vent line burning through was a fraction of a second. The time from then to the rocket climbing the rail under thrust was right on three seconds. The first two seconds of that were cold flow, and the last second appeared to be hot flow (at least above the dew point).

Hot flow would generate additional chamber pressure compared to cold flow. The start of pressurization would slow the loss of nitrous. Nitrous is lost fastest under cold flow - faster than when the motor is burning, due to the greater pressure drop across the injectors.

There were only three seconds of thrust produced (based on video) before the vapor phase tail-off. So roughly half the nitrous vented before the rocket even left the pad. With a full burn, the theoretical altitude of 15150ft might have been achievable. The predicted burn duration was 6.2 seconds, which is roughly consistent with the above data.

The preheater apparently did not ignite completely as there were no visible flames coming out the nozzle at any point before thrust. With this preheater, there should have been. My speculation is the burn-thru partially quenched the preheater. The fill line and the stubs need to be inhibited to slow burn through to allow preheating to occur before nitrous flow.

The visible flame was smaller with this generation of fuel formula than was the case for the static test. The main change was substituting Gran 16 Mg for fine Ti. The flame was cleaner. My speculation is the Mg is burning predominantly in the combustion chamber, unlike the Ti that seemed to do a lot of burning outside the motor. That change is likely an improvement. Metal that does not burn in the combustion chamber just sucks out heat to no benefit. Both are theoretically good metals for this application, but this motor is demonstrably too small to use Ti as a fuel.

 

[G_T]

I did wipe off the exit cone of the nozzle at the field, to clean off the soot. The residue is essentially soot by appearance, like lamp black, and wipes off readily.

Unlike an AP motor, the motor after the burn does not stink. That's one nice thing about this type of motor.

After the rocket was launched, during the cleanup I gave the fill line a squeeze test. The line was more rubbery after use than before. After the line has had a couple days to sit around, it feels hard again. Nitrous is a good organic solvent. I suspect that long-term exposure to nitrous would dissolve the line. It is better to treat the fill line as single-use, just to be safe, for motors such as these which can take a couple minutes or more to fill.

 

[rocket_troy]

IIRC N2O doesn't permeate Nylon (which is typical for fill hoses) or any plastic for that matter with any rigidity to it. The mention of the fill hose feeling rubbery is totally foreign to me (I always use nylon). What material was it and why did you choose it?
If you're really set on chilled N2O, have you considered top filling and employing an injector more akin to liquids eg. pintle? Yes, I'm well aware of the trade-offs of going down that path (been there, done it).

Irrespective, I'd still consider your launch a success given the challenges, project maturity and available resources.

TP

 

[OzHybrid]

IIRC N2O doesn't permeate Nylon (which is typical for fill hoses) or any plastic for that matter with any rigidity to it. The mention of the fill hose feeling rubbery is totally foreign to me (I always use nylon). What material was it and why did you choose it?
If you're really set on chilled N2O, have you considered top filling and employing an injector more akin to liquids eg. pintle? Yes, I'm well aware of the trade-offs of going down that path (been there, done it).

Irrespective, I'd still consider your launch a success given the challenges, project maturity and available resources.

TP

To me a rubbery feeling would mean the hose had absorbed N20 into the surface internally and then ballooned partially once the pressure was off. I use either nylon or steel braided over teflon. Is it possible you are using polyurethane air hose instead of nylon. Are there any markings on the hose?
I did a fair bit of work with liquid CO2 (which has a lot of similar properties in handling terms) in SFX and if you used the wrong O ring type, it could expand into a donut..... Once it sat for a bit it would shrink back down. My gut says you've got polyurethane hose.....
With the solenoid getting hot, it's pretty easy to make a solenoid saver. This restricts the current for holding it open but gives full power for the time to fully open it.

 

[G_T]

I have some suspicions about the fill line not being entirely nylon. When heating the ends of the stub lines used to plug two of the injectors, there was some evidence of some boiloff, as if there were some small percentage of a plasticiser or other volatile present. This produced surface bubbling which would go away upon continued exposure to heat. I would not expect that behavior of pure nylon.

With so many products coming from China now as the root source, and China factories having a tendency to make unapproved substitutions and adulterations when not being watched, who knows what the real composition might be. I'm pretty sure it was mostly nylon. Polyurethane would not have allowed me to produce the fused ends via heating. It had to be a thermoplastic of some sort, or a blend of such.

I do not recall such behavior from the static test. That was using a different batch of fill line, obtained from a different source. I did not keep the bag it came in since I just used the whole line for better standoff of the GSE. There wasn't enough for good standoff for two burns, though it was more than enough for one.

I think the armored line running out of the GSE I borrowed is steel over possibly teflon. I know it is rated for nitrous.

Nylon is rated only fair for nitrous exposure. Short term is ok, but not frequent and not continuous exposure. Rubbery might be over-stating it a bit. At a moderate squeeze, it had more give to it than before using it, and it re-hardened over time.

PS - Thanks for the circuit diagram Norman.

PPS - The cryogenic relief valve on the flight tank is not rated for nitrous, but it is rated for oxygen service and is supposed to be a Teflon seal. That's why I felt confident using it. I'm hoping that when I tear it apart I find that it was just a bit of contamination that got in the seal. That is, that I just got unlucky. I'd rather not replace the valve every burn. I'm not sure if I can get those any more. When I last looked they weren't available, but with Covid raging at that time... They may be available now.

 

[G_T]

I've been giving the preheater problem some more thought. Since it seems rather difficult to solve at the preheater, what I should probably explore is shielding the fill line and stubs to render them slower to melt through. The goal would be to give the preheater time to get fully going before nitrous is injected.

That's for the current motor design. I might do something differently with a redesign. Since this was not a fully successful burn, I'm rather tempted to give it another shot before redesigning for an N class motor or larger. Whether flight or static, I don't know yet.

 

[G_T]

It seems the startup to full thrust might have been delayed a tiny bit on this burn. Judging from the pic in flight, by this point the motor was generating full thrust. The flame was most of the length of the rocket (flame about 7' long), and appeared to get even larger in the next second. But leaving the pad it was somewhat shorter. It might be the O:F ratio was higher at startup while the core was smaller, due to ineffective preheating, lower fuel surface area, and reduced residence time. I'll know more from the Raven data, presumably.

Here are some pics from the launch.

I do think this is the coolest rocket I've flown! It is what I would have wanted my L3 rocket to be, if TRA would have allowed an EX motor for cert.

I'm in the blue shirt. Al Anderson is in the yellow shirt. Jerry O'Sullivan is in the white shirt, and you can barely see a bit of Mitch Guess behind Jerry. Bill Schworer is holding the camera. The very helpful field crew person didn't show up in these pics, unfortunately. URRF crew was very helpful with this unusual project!

 

[G_T]

Looks like I can still get the pressure relief valve, REGO PRV9432T550. It even costs less than it used to! It also looks like I could also get REGO PRV9432T600 which would make ignition easier. That didn't used to exist. I'd have to re-run all the numbers to see if a new nozzle or other change would be required, and to make sure the initial fuel port diameter is not too constrictive for the additional flow. Total ISP would reduce some, burn duration would be a bit shorter, and thrust would be a bit higher.

 

[jderimig]

Gerald, is the long length of fill tube you used to allow a pressure drop from supply tank pressure to motor tank pressure?

 

[G_T]

No, just to get the tank and GSE farther from the rocket in case something went wrong. I borrowed those items, so I didn't want to place them in harm's way! Chilled nitrous is much harder to detonate though.

I guess this place can't handle a video file very well. Sorry folks; can't post one.

 

[jderimig]

No, just to get the tank and GSE farther from the rocket in case something went wrong. I borrowed those items, so I didn't want to place them in harm's way! Chilled nitrous is much harder to detonate though.

I guess this place can't handle a video file very well. Sorry folks; can't post one.

So how/where does the pressure drop occur if you have a hot tank and the cryovalve is doing its thing? Across the injectors and the fittings along the way?

 

[rocket_troy]

The pressure drop comes from the PRV ie the adiabatic expansion of the gas on the supply side of that orifice. That process ultimately lowers the kinetic energy of the gas molecules within the motor tank thereby lowering its temperature and vapour pressure.
The difference between a spring loaded PRV and a static orifice is the PRV can *regulate* the venting by the pressure it sees thereby also regulating the kinetic energy of the gas molecules within hence the temperature of the gas within hence the thermodynamic equilibrium conditions within the motor tank (within the limits of its orifice and opening).
So that's the thermodynamic "source" of the pressure drop you'll see along the length of fill hose and associated chokes, valves and plumbing that lays between the N2O in the supply tank and the N2O in the motor tank. All those parts "enables" it to occur (ie. provide the flow restriction).

That might not correctly answer your question - depending on what angle you're coming from I suppose.

TP

 

[G_T]

Troy, you posted it better than I was going to! So after some erasure, here's the fill process:

Pulse in nitrous, let it boil. Cryovalve opens on excess pressure, venting gas. Cryovalve closes, so open fill valve again. Repeat until sputtering indicates fill point is reaching the top of the tank and some liquid is being spit from the cryovalve. then give it longer between pulses of nitrous, and use shorter pulses. Goal is to approximate equilibrium. As the flight tank warms up it will increase in pressure and vent. After venting it is a little low in fill and probably slightly under pressure and temperature due to hysteresis at the valve. So give it a small pulse.

In equilibrium state, the pressure drop is across the fill solinoid which is closed.

The cryovalve set open pressure is several hundred PSI below the supply tank pressure.

It doesn't show in the pictures or video, but the flight tank was dripping in moisture from condensation by the time the filling was well under way.

In the static test I didn't know for sure how it would go. But you get a feel for the rhythm of the fill process pretty quickly. The worst that can happen is (1) thinking venting means full and stable, or (2) wasting more nitrous than necessary. Sputtering is VERY visible and audible as such. Venting just sounds like a high pressure vent. Both are audible from a long distance. Nothing about this motor is quiet. It's loud on venting, and loud when burning. I've posted a pic of what sputtering looks like, when the tank is full.

Sorry it's rather fuzzy, a zoom in. The tent row is much farther away than it appears.

BTW, sputtering venting acts pretty effectively as a cold gas thruster. It sways the launch rail a few inches. Gas venting doesn't generate nearly as much thrust so the rocket sits still. If the vent is placed 90 degrees from the rail lugs, give or take, the swaying during sputtering is quite visible.

PS - The effective vent orifice when open is probably a little larger than I'd prefer for this size motor (M). I'd recommend a flow constriction orifice on the input side of the valve for smaller motors.

 

[jderimig]

The pressure drop comes from the PRV ie the adiabatic expansion of the gas on the supply side of that orifice. That process ultimately lowers the kinetic energy of the gas molecules within the motor tank thereby lowering its temperature and vapour pressure.
The difference between a spring loaded PRV and a static orifice is the PRV can *regulate* the venting by the pressure it sees thereby also regulating the kinetic energy of the gas molecules within hence the temperature of the gas within hence the thermodynamic equilibrium conditions within the motor tank (within the limits of its orifice and opening).
So that's the thermodynamic "source" of the pressure drop you'll see along the length of fill hose and associated chokes, valves and plumbing that lays between the N2O in the supply tank and the N2O in the motor tank. All those parts "enables" it to occur (ie. provide the flow restriction).

That might not correctly answer your question - depending on what angle you're coming from I suppose.

TP

Yes thank you. So the venting mass flow will depend on the restrictions in the fill line to maintain an equilibrium pressure drop, right? So if there are low level of restrictions in the line more nitrous would be wasted than if there were "optimum" restrictions?

 

[rocket_troy]

Yes thank you. So the venting mass flow will depend on the restrictions in the fill line to maintain an equilibrium pressure drop, right? So if there are low level of restrictions in the line more nitrous would be wasted than if there were "optimum" restrictions?

Good question. I would (with a moderate degree of confidence) say not generally, but that's coming from the perspective of purely utilising the vent as a means to transfer N2O from the fill tank to the rocket. In that instance, it's the heat in the system and the heat being transferred into the system from the surroundings that determines the specific cooling the vent needs to produce hence how much wastage of N2O that needs to flow out from that orifice per unit of N2O in the system. For a static port (vent) the equilibrium pressure and temperature within the motor tank will be proportional to that of the supply + the transferred heat and inversely so to the throughput of the orifice. So, increasing the supply side will also increase the receiving side however increasing the orifice throughput reduces the receiving side (at a cost of wasted N2O) and also increases the delta P from supply to receive - however I'm seeing that as a consequence of the conditions, not in itself the cause. That's where my thoughts are right now as I type this at least

TP

 

[OzHybrid]

I've been giving the preheater problem some more thought. Since it seems rather difficult to solve at the preheater, what I should probably explore is shielding the fill line and stubs to render them slower to melt through. The goal would be to give the preheater time to get fully going before nitrous is injected.

That's for the current motor design. I might do something differently with a redesign. Since this was not a fully successful burn, I'm rather tempted to give it another shot before redesigning for an N class motor or larger. Whether flight or static, I don't know yet.

RattWorks used a fiberglass insulation for the tube it wanted to protect in the fill line. This is similar, but looks to be a bit thinner. Good luck with the next one.
Norm

 

[G_T]

There is only an equilibrium pressure drop in the fill line under steady-state conditions. With this sort of a fill process, that really doesn't exist.

The supply tank starts at a certain temperature and pressure. As it provides liquid, the volume of ullage will increase - which is liquid phase converted to gas phase, taking heat away from the remaining liquid. The temperature of the supply liquid drops, as does the pressure. So the supply conditions are not constant regardless of the system it is connected to, once filling commences.

The fill solinoid can be either open or closed.

If closed, the flight tank can be either venting, or not. If venting through the PRV, the pressure on that side is dropping as is the temperature. If not venting, the pressure on that side is increasing as is the temperature. I'm presuming the target temperature is lower than ambient, and the target pressure is lower than the supply tank pressure.

If open, the flight tank can be either venting or not. If venting through the PRV, the results are complicated, depending on the rate of supply vs the rate of exhaust through the PRV, among various other things. If not venting, then the flight tank pressure and therefore temperature are increasing as liquid is added and ullage is squished.

There isn't an equilibrium. But one can creep up on an approximation using short fill bursts and waiting for results to settle down, once the flight tank is essentially full.

In the end, the pressure will be some unknown lower value than the PRV set pressure, or the PRV would be venting.

In the end, the volume of liquid in the flight tank will be some unknown value lower than the full volume, or the PRV would be venting due to thermal expansion of the nitrous. After all, the flight tank is at a lower temperature than ambient and is absorbing heat from the surroundings.

The goal is to have the pressure close to the PRV set pressure, and the volume of liquid in the flight tank close to max (very little ullage). That does depend on how one uses the fill solinoid valve. Each pulse adds mass, and increases pressure and therefore temperature. Each venting decreases mass, pressure, and temperature.

And at least at the moment, the pulses are under human control. An automated control system with some encoded physics could of course be superior - as in get closer to desired results and waste less nitrous while doing so. I'm doing it on the cheap - by cycling a switch.

If via appropriate control of the fill solinoid one achieves a condition close to the PRV set pressure (and therefore close to a planned temperature) with a nearly full flight tank, one will have achieved a more consistent nitrous supply to the combustion chamber than would have been the case for using a pure vent in the flight tank. The performance of the motor should be much more repeatable.

 

[G_T]

Norm, that's an interesting idea, at least for the parts of the tubes I don't want to burn through. I was figuring on doing some torch vs time tests with various options. The exact delay desired is a tough call. If the delay is too short, then the motor will reach proper combustion only after some cold venting of nitrous through the combustion chamber. If the delay is too long, then proper combustion may never occur.

 

[maddmaxx11]

I have a question about the solenoid saver in the picture above.
I have a launch system that a guy on here built but and the solenoid saver has a 11000 uf in parallel to a 15 ohm resistor this is connected to ground on one side and the other is going to the relay. The other side of the solenoid is tied to +12 v. From what I understand the cap acts like a short at first across the resistor until the cap starts to charge. In the picture above the 2 ohm resistor is always limiting the current. Is this circuit wrong?

 

[rocket_troy]

There are 2 main ways of doing it with passive RC - using the cap to short (like you said) and transition to the resistor (solely) as it charges or discharge from a charged up cap for your opening pulse and bleed onto relying solely the resistor as it discharges. The 1st method is probably the most common I'd say.
Then there's other active methods using PWM, but the RC methods work pretty well.

TP

 

[waltr]

Where I work we use the parallel RC in series with relay coils. This does work well to provide a higher 'pull-in' Voltage (current) and a lower 'hold' Voltage (current).

 

[G_T]

I haven't had time to pull the data yet, or go through more of a disassembly/assessment. But here are some pics post-flight.

There is a lot of fuel left, but that is to be expected with only half the burn. The internal burn profile goes rounded triangular, due to the three injectors. It's possible that if a full burn had occurred, there may be some areas of exposed liner. I'll make more of a judgement about the regression rate after disassembly and weighing.

The intended irregular burn surface is quite evident. That's what I mean by high-turbulence fuel.

Nozzle cleanup was wiping with a paper towel.

The PRV looks normal on the outside. Nothing indicates what was going wrong with it. Disassembly required...

The liner of the preheater is barely making it. On the first burn I used a doubled liner and the liner had hardly any damage. So this time I used a single layer liner, but a more energetic preheater. That liner is crispy toast.

FWIW, this is the second time I used those O-rings in this motor. No, you probably shouldn't. Yes, I did. They don't get as much heat as would be the case for an APCP motor. Viton is used on the side where nitrous exposure is possible.

I did not disasseble the flight tank after the first burn, and it sat for a couple years after the static test. I'll assess the internal condition when I clean the precombustion chamber and remove the injectors, and the PRV. If it still looks very clean inside, I might not disassemble it this time either.

 

[OzHybrid]

I have a question about the solenoid saver in the picture above.
I have a launch system that a guy on here built but and the solenoid saver has a 11000 uf in parallel to a 15 ohm resistor this is connected to ground on one side and the other is going to the relay. The other side of the solenoid is tied to +12 v. From what I understand the cap acts like a short at first across the resistor until the cap starts to charge. In the picture above the 2 ohm resistor is always limiting the current. Is this circuit wrong?

They are effectively the same. One is in the ground line. The other (mine) is in the positive. The value of the resistor determines the hold open current. The value of the capacitor needs to be large enough that it fully opens when being used at your cylinder pressure.
You might want to also have a back emf diode across your solenoid and relay.

 

[G_T]

Very nice paper here, for those interested in some more detailed background of typical N2O hybrid rocket motors flight tank behavior during filling and firing: http://edge.rit.edu/content/P07109/...ocuments/AspireSpace - The Physics of N2O.pdf

THRP-1 takes an approach that puts it a little off from the situation described in the papar, in that the flight tank will end up nearly full of chilled liquid nitrous due to the top mounted pressure relief valve (PRV), but unfortunately at a lower pressure. I've given up ISP in exchange for density, to improve the overall motor performance (total impulse available from that volume of motor). I also do a fast fill rather than a slow fill, encouraging boiloff to chill and densify the nitrous. It's a bit different sort of beast.

If I make a THRP-2, there is a lot of instrumentation I'd like. Pressure sensor, temperature sensor, and video into the top of the tank. That combo would tell me a great deal about what is actually going on in the motor, rather than just thinking I'm getting what I'm intending to get!

 

[G_T]

Screenshot from the Raven 3 data. Looks like this burn went resonant unlike the static test. So much for a smooth burn. The ramp-up due to cold flow start is pretty evident as well. About the only things I can say is (1) it dumped a lot of nitrous before it got going, and (2) the acceleration once it is going is right around what it should have been. About half the total impulse was lost due to cold flow and leaking vent.

That cold flow start totally killed the performance. It turned an M into an L. The 10g of BKNO3 I threw at that problem didn't help at all. In fact, it was probably worse than on the static test. Disappointing.

It would be nice if the FIP program could use sane scales. It would make reading the data off the graph easier.

It turns out the "new" Raven3 I got wasn't new at all. There was another flight recorded on it. That had me confused for a few minutes, since the data didn't look very close to correct for this flight! Or do they come with some pre-loaded test nowdays?

The wiring for the tether and CO2 appear correct to the altimeters. The Raven3 programming looks correct. Haven't checked into the RRC3 data yet.

Does the Raven3 record when it fired events?

I did see a much larger G shock on nosecone deployment than I would have thought - roughly 50G for a very brief moment. The baro data from the Raven3 showed hardly a wiggle at that point, so there wasn't gas blowback into the electronics section to speak of.

The knot I invented for attaching the Kevlar to the eyebolt for the main worked fine. I'll have to pass the knot on sometime. It's pretty cool.

Also on casual inspection no gasses from the ematches vented back into the electronics area, which is good. So I've eliminated a number of the possible reasons for the tether deploying the main at apogee.

I might have lost a cable I need for the RRC3.

 

[G_T]

Judging from the accelerometer data, it looks like the main came out perhaps about a second after the drogue, and likely there was at least a little bit of horizontal velocity at the top or else apogee was off by a few seconds one way or the other. Enough to cause a few G's when the main opened. There wasn't enough time between the two openings to build up enough speed for the G's otherwise. That's if I'm interpreting it right.

Looks like I'll have to borrow a friend's RRC3 cable to get the data from that altimeter, and to check the programming. If I can't find an issue there, then I'm reduced to concluding one of the altimeters glitched, or the tether failed mechanically for some reason and released at a fraction of its rated load.

Perhaps there was some bit of debris incorporated into the assembly that prevented the ball cup from moving fully upwards. That would have resulted in the balls possibly holding the pin against casual loads but not securely.

Unfortunately the landing scooped up some dirt from the freshly harvested field, and some silty dirt got in there. So if that's the reason, I'll never be able to prove it.

Unfortunately there was no onboard video captured. I chalk that one up to operator error most likely. I turned the video on, but probably didn't get it started recording. That video could have been useful figuring out what happened with the tether.