Hybrids 2024

[cfb_rolley]

Sweet, thanks for the clarification on the sulphur dioxide.

For clarification of the CNVA chart, I was referring to this:

…for half of the rows there’s two throat ID columns? Two speed columns? Two injector columns? Some with different values in the same row, but some are the same? What?

For GSE I also was referring to just the bare bones electronics. I’ve decided on the new low cost 1/8” solenoids but was more concerned about building my own electronics to run them, because as nice as a Wilson FX setup would be, it’s astronomically expensive for Australians due to shipping and exchange rate, and I can likely build something for a fifth of the price a Wilson setup would cost by the time it’s landed in Australia.

 

[rocket_troy]

I've built my fair share of hybrid GSE systems - both wired and wireless. Much of the design is down to personal preference and the importance of various things. For me, simplicity, mobility an reliability are the 3 things I focused on with my last iteration.
https://www.rocketryforum.com/threads/wireless-hybrid-gse.180251/post-2442931
I use NOS solenoids as I was gifted them by a very generous member of the TRA community but I've also used other solenoid valves and ball valves in the past with success. Some smaller solenoid valves might have less a requirement for things like solenoid saving (NOS flavours are thirsty for electrons), nevertheless, a passive solenoid saver circuit generally comprises of 2 additional components per channel (cap and resistor) so it kinda makes sense to just include that protection in there irrespective.

That particular GSE only housed 2 channels for filling and dumping, but obviously add a 3rd firing channel would be just a case of an additional FET and connector.

TP

 

[A-ron]

I've finally managed to gather all of the components I need to build my first GSE controller, but before I start building it, I could use some constructive criticism of my circuit schematic...see attached PDF file.

My design was mainly inspired by launch controllers designed by Jim Yehle (Utah Rocket Club), Hieu Minh Nguyen (USC Rocket Propulsion Lab) and Michael Moncur (Rocket Team Vatsaas). These are relatively low-cost designs that make use of COTS components.

I've gone a step further by incorporating a 6-channel relay module. Note that each relay module input is actually connected in series to a 2200 Ohm resistor and opto-isolator, which drives a FET that drives the relay coil. These components have been omitted from the schematic for brevity, but in the case of channel 6, the input resistor limits the igniter continuity current to <5.5mA. A separate 4-channel opto-isolator module is included to convert relay input signals to 3.3V for controlling an external data acquisition system.

Power is supplied by a 11.1V 3S LiPo or 12V SLA battery located inside the pad control box. The remote control box connects to the pad control box with an ordinary ethernet cable. There's two independent solenoid saver circuits that have been sized for the valves I intend to use.

All buzzers and switches have a built-in LED that serves as a diagnostic indicator. SW1 is a safety key switch, SW3+SW4+SW5 are toggle switches with safety guards, SW7+SW11 are rocker switches and all others are momentary button switches.

Feedback from some you experienced GSE builders would be greatly appreciated! Thanks.

 

[RalfB]

Hello,
does it work with the remote control to switch several channels at the same time? I had the problem with my GES that the channels could only be switched one after the other.

 

[drewnickel]

Sweet, thanks for the clarification on the sulphur dioxide.

For clarification of the CNVA chart, I was referring to this:
View attachment 624831
…for half of the rows there’s two throat ID columns? Two speed columns? Two injector columns? Some with different values in the same row, but some are the same? What?

For GSE I also was referring to just the bare bones electronics. I’ve decided on the new low cost 1/8” solenoids but was more concerned about building my own electronics to run them, because as nice as a Wilson FX setup would be, it’s astronomically expensive for Australians due to shipping and exchange rate, and I can likely build something for a fifth of the price a Wilson setup would cost by the time it’s landed in Australia.

I believe the idea is the first column of throat ID is the recommended throat with CNVA (it enhances regression rate and atomization at the cost of flow rate). The second column saying throat ID is without CNVA.

 

[OzHybrid]

I've finally managed to gather all of the components I need to build my first GSE controller, but before I start building it, I could use some constructive criticism of my circuit schematic...see attached PDF file.

My design was mainly inspired by launch controllers designed by Jim Yehle (Utah Rocket Club), Hieu Minh Nguyen (USC Rocket Propulsion Lab) and Michael Moncur (Rocket Team Vatsaas). These are relatively low-cost designs that make use of COTS components.

I've gone a step further by incorporating a 6-channel relay module. Note that each relay module input is actually connected in series to a 2200 Ohm resistor and opto-isolator, which drives a FET that drives the relay coil. These components have been omitted from the schematic for brevity, but in the case of channel 6, the input resistor limits the igniter continuity current to <5.5mA. A separate 4-channel opto-isolator module is included to convert relay input signals to 3.3V for controlling an external data acquisition system.

Power is supplied by a 11.1V 3S LiPo or 12V SLA battery located inside the pad control box. The remote control box connects to the pad control box with an ordinary ethernet cable. There's two independent solenoid saver circuits that have been sized for the valves I intend to use.

All buzzers and switches have a built-in LED that serves as a diagnostic indicator. SW1 is a safety key switch, SW3+SW4+SW5 are toggle switches with safety guards, SW7+SW11 are rocker switches and all others are momentary button switches.

Feedback from some you experienced GSE builders would be greatly appreciated! Thanks.

For an RJ45 connection, the cabling twisted pairs are 1,2 3,6 4,5 7,8 There's a variation in colouring depending if you go with the A orB colour scheme but the pair locations remain the same.

I'd change 7 to gnd and 8 as a pair for Fire(ARM) to give the dangerous bit a ground protecting it in twisted pair and the connector and the fire on the outside.
Your continuity relay and your arm relay need to be on to provide 12v to the ignitor. Your ARM relay is really a FIRE relay. And your continuity, is really an ARM relay You might want to relabel to make the function clearer.
I'd also connect the Gnd to the NC connection of the Fire(ARM) to ensure shunting of the ignitor to ground until you fire it.
Your operational relays should all have Back emf diodes across them. Not just the solenoids they operate.
Put a robust bridge rectifier after the fuse as crowbar reverse polarity protection. If polarity is reversed, it puts a short across the input andforces the fuse to blow. Connect both AC inputs to + after the fuse and the negative to ground.
Use something like this one. Chunky and robust.

Regards
Norm

 

[A-ron]

Norm, thank you so much for your input!

I'll definitely change the RJ45 pin assignments as I hadn't considered shorts between wire pairs as a failure mode. The ignition circuit will also be reworked to include your suggestions, especially shunting the igniter.

The relay module actually does have a back EMF diode for each relay coil, but these and other on-module components were omitted from my schematic. I'll try to draw those inside the module outline for clarity.

Polarity protection is a great idea that I haven't seen in other designs and the "crowbar" is entirely new to me, so thanks for that bit of enlightenment.

There are also other changes I'm considering. In the interest of protection from welded relays, I'm thinking that there should be an extra relay to supply power to the FILL and DUMP relays. That would reduce the chance of either valve being stuck open. I'm also thinking that the ignition relays should be upgraded, since any igniter below 1.2Ω will exceed 10A at 12V.

 

[OzHybrid]

Crowbar is brutal. But it's your lowest loss solution. You can use a resettable fuse instead of a normal fuse, but it should be able to survive a short circuit otherwise they explode internally and are no longer resettable. Look for spec value isc=1000A or greater. ( short circuit current limit greater than 1000 Amps). This seems rediculous but please accept it's the minimum for a resettable fuse to survive. If you choose to ignore that recommendation don't say I didn't warn you.

 

[A-ron]

My first GSE build is going to have a blade fuse. If that doesn't work out, I'll definitely have a look at the resettable variety and pay special attention to the short-circuit current rating.

Thanks again for your suggestions, they certainly helped to make an improved schematic, crowded and crowbarred as it may be. See attached PDF.

I added the on-module relay control circuit that I had previously omitted, so the entire controller operation should be easier to understand. The revised ignition circuit is even more to my liking and the ARM+FIRE label combination fits really well. RJ45 pin assignments are still a work in progress and require more contemplation.

Update: To provide protection against a welded-on FILL relay, power for the FILL connector is now supplied through the DUMP relay's Normally-Closed terminal. Switching the DUMP relay on should automatically disable power to the FILL connector.

 

[hybridfreak]

Got my lathe running again after tearing it apart some 10 years ago and finished converting it to variable rpm drive. It now has a VFD, double the power (1,5HP) and an input for a limit switch (inductive) so I can have it stop on the spot when turning (latheing? ) internal threads.



The first part I chucked in the chuck was a aluminium/graphite nozzle I got for 38mm ex motors. I turned it down to fit my WCH I110. The original WCH nozzle was a bit fragile and lost a few chunks at some point.



Next will be a new injector! I think I'll static test it in stock configuration (straight injector, PVC grain) and then so some tests with ABS or ABS/wax and a more efficient injector design plus shorter grain to gain some nitrous volume. Maybe I can get it from 428Ns/112N (tot./avg.) to around 500/150? That would be nice

 

[cfb_rolley]

shorter grain to gain some nitrous volume. Maybe I can get it from 428Ns/112N (tot./avg.) to around 500/150? That would be nice

ooh I was going to ask about shorter grains and performance!

I have been static firing my micro-hybrid with different injectors and grain combos to see what happens lately, and have a bit of a theory that someone here might be able to explain more.

I’ve been testing a rifled ABS grain with very deep grooves, and a twist rate 1 turn per 28mm of grain length. This has a WAY higher surface area compared to straight port grains or other port geometries I’ve tested.

I tested on a .032, .027 and .019 injector so far, and the results have all been pretty good, but I noticed that exhaust of the .019 seemed to look like it was fuel rich from 0.5s in to the burn when compared to the larger injectors.

Now, I have no way of measuring the OF ratio during the burn, but here’s what I think is happening - with the geometry of this grain, the surface area increases as it regresses - but the surface area is so large at about 0.5s that the oxidiser flow through the .019 injector is too small, so… fuel rich.

This makes me think though, does this mean that for hybrids, you could potentially use a shorter fuel grain and get similar results with the same injector, provided that the two fuel grains have the same surface area?

…now i understand there’s *wayyy* more variables in hybrid burns than that, but I still wanted to know if that was a thing. If that’s the case, I could certainly see something like 12” x 54mm hybrid being used in place of a 16” x 38mm or something.

 

[hybridfreak]

ooh I was going to ask about shorter grains and performance!

I have been static firing my micro-hybrid with different injectors and grain combos to see what happens lately, and have a bit of a theory that someone here might be able to explain more.

I’ve been testing a rifled ABS grain with very deep grooves, and a twist rate 1 turn per 28mm of grain length. This has a WAY higher surface area compared to straight port grains or other port geometries I’ve tested.

I tested on a .032, .027 and .019 injector so far, and the results have all been pretty good, but I noticed that exhaust of the .019 seemed to look like it was fuel rich from 0.5s in to the burn when compared to the larger injectors.

Now, I have no way of measuring the OF ratio during the burn, but here’s what I think is happening - with the geometry of this grain, the surface area increases as it regresses - but the surface area is so large at about 0.5s that the oxidiser flow through the .019 injector is too small, so… fuel rich.

This makes me think though, does this mean that for hybrids, you could potentially use a shorter fuel grain and get similar results with the same injector, provided that the two fuel grains have the same surface area?

…now i understand there’s *wayyy* more variables in hybrid burns than that, but I still wanted to know if that was a thing. If that’s the case, I could certainly see something like 12” x 54mm hybrid being used in place of a 16” x 38mm or something.

Hybrids will work in a fairly large range of O/F-ratio, that's one reason "classic" grains work at all (round port all the way through). If you know the regression rate of your fuel, the oxidizer flow and the O/F you want to achieve you can calculate what surface area you need. That gives you a good starting point but in reality things are a bit different of course: tank pressure drops during the burn, regression rate isn't constant depending on the fuel and of course the grain's surface area will change depending on the geometry.

Regarding shorter hybrids, @ContrailRockets has done some very interesting stuff and posted about it here, for example:
Hybrids 2020

I actually have a 54mm 16" long (400mm) casing that I might machine some internals for to make it a hybrid

 

[drewnickel]

ooh I was going to ask about shorter grains and performance!

I have been static firing my micro-hybrid with different injectors and grain combos to see what happens lately, and have a bit of a theory that someone here might be able to explain more.

I’ve been testing a rifled ABS grain with very deep grooves, and a twist rate 1 turn per 28mm of grain length. This has a WAY higher surface area compared to straight port grains or other port geometries I’ve tested.

I tested on a .032, .027 and .019 injector so far, and the results have all been pretty good, but I noticed that exhaust of the .019 seemed to look like it was fuel rich from 0.5s in to the burn when compared to the larger injectors.

Now, I have no way of measuring the OF ratio during the burn, but here’s what I think is happening - with the geometry of this grain, the surface area increases as it regresses - but the surface area is so large at about 0.5s that the oxidiser flow through the .019 injector is too small, so… fuel rich.

This makes me think though, does this mean that for hybrids, you could potentially use a shorter fuel grain and get similar results with the same injector, provided that the two fuel grains have the same surface area?

…now i understand there’s *wayyy* more variables in hybrid burns than that, but I still wanted to know if that was a thing. If that’s the case, I could certainly see something like 12” x 54mm hybrid being used in place of a 16” x 38mm or something.

From tests I've done, you can definitely get away with a shorter grain by increasing surface area, and swirling the nitrous can help a ton (anywhere from 50-100% increase in regression rates doing so). When I was first starting off with hybrid motors at UTK, we were testing a 38mm x 16 inch Contrail motor. Adding swirling got us from 133 N-s to 187 N-s total impulse with no other changes, and we were also able to shorten the grain from 6 inches to 4 inches to increase the nitrous tank capacity. Eventually got that same motor to around 320 N-s total impulse. As for minimizing OF shifting, the OF shift is important more so for efficiency than for whether or not it works but if you have to shift one direction it's better to be more oxidizer rich. A circular port is nice in that if your propellant has a regression exponent around 0.5 you basically have a constant OF ratio. For lower regression exponents the motor tends to go fuel rich and for higher exponents the motor tends to go oxidizer rich. Typically you'll see something like the plot below for C* vs OF ratio when using nitrous, as nitrous oxide itself can work as a monopropellant and as you go oxidizer rich you are approaching that monopropellant performance.

 

[rocket_troy]

ooh I was going to ask about shorter grains and performance!

I have been static firing my micro-hybrid with different injectors and grain combos to see what happens lately, and have a bit of a theory that someone here might be able to explain more.

I’ve been testing a rifled ABS grain with very deep grooves, and a twist rate 1 turn per 28mm of grain length. This has a WAY higher surface area compared to straight port grains or other port geometries I’ve tested.

I tested on a .032, .027 and .019 injector so far, and the results have all been pretty good, but I noticed that exhaust of the .019 seemed to look like it was fuel rich from 0.5s in to the burn when compared to the larger injectors.

Now, I have no way of measuring the OF ratio during the burn, but here’s what I think is happening - with the geometry of this grain, the surface area increases as it regresses - but the surface area is so large at about 0.5s that the oxidiser flow through the .019 injector is too small, so… fuel rich.

This makes me think though, does this mean that for hybrids, you could potentially use a shorter fuel grain and get similar results with the same injector, provided that the two fuel grains have the same surface area?

…now i understand there’s *wayyy* more variables in hybrid burns than that, but I still wanted to know if that was a thing. If that’s the case, I could certainly see something like 12” x 54mm hybrid being used in place of a 16” x 38mm or something.

There are a few more variables than that and 1 important one in particular: GOX which is the Oxidizer flux rate. Basically, the regression rate of your fuel is directly proportional to that value ie. the flux density of oxidizer passing over the fuel. So, for a given mass flow of N2O passing through the injector(s), that GOX value will decrease as the cross-sectional port area of the core opens up, hence the regression rate of the fuel will reduce proportionally. Of course, it's even more than that in practice as there's a pressure drop in the supply feed as the N2O tank is depleted so the GOX value is reducing in practice by both the opening port area AND injector throughput, but in terms of O:F ratio, it's only the port cross-sectional area that primarily matters.
Saying that, yes, the O:F will typically decrease as there's a noticeable drop in feed pressure, so the exhaust plume will become richer. Because the theoretical performance (in terms of Isp or c*) of N2O hybrids doesn't change much between O:F ratios of 10:1 to 4:1, there generally isn't much concern within that band unless you start very rich.
You might notice more pitting in graphite nozzles as you push towards leaner ratios however.

TP

 

[cfb_rolley]

Excellent, looks like I'm at least on the page I'm supposed to be when it comes to understanding surface area vs. length of the grain. I'm assuming there is probably some point where the fuel grain can be too short regardless of how much surface area it has though, and if I had to guess, it's probably somewhere near the point where you start losing oxidiser straight out the mozzle before it can react with the fuel, and with the way swirl injectors work, they can offset that by a bit.

I did have thoughts about trying out a star grain where the star shape has very thin points that burn away early, regressing to a straight port grain - or in other words, a grain with a massive surface area initially but reduces to be the same as a partially regressed straight port after the "points" have burned up. The intention being to try and match up the surface area available, regression and oxidiser flow throughout every point of the burn. But after drewnickel and rocket_troy's info about how wide of a range the OF can be, that idea probably won't make a great deal of difference and you could just let the OF shift towards richer, as long as the OF starts off high enough.

....I reckon it's likely the case that for the .032 and .027 injectors I tested with this grain, the OF strated high enough, and the .019 just started somewhere lower and went deeper in to fuel rich territory than the other two. whether it went *too* rich and affected performance is something I'll have to do a bit more testing on I guess.

... but, those were all awesome replies! thank you!

 

[ContrailRockets]

Hybrids will work in a fairly large range of O/F-ratio, that's one reason "classic" grains work at all (round port all the way through). If you know the regression rate of your fuel, the oxidizer flow and the O/F you want to achieve you can calculate what surface area you need. That gives you a good starting point but in reality things are a bit different of course: tank pressure drops during the burn, regression rate isn't constant depending on the fuel and of course the grain's surface area will change depending on the geometry.

Regarding shorter hybrids, @ContrailRockets has done some very interesting stuff and posted about it here, for example:
Hybrids 2020

I actually have a 54mm 16" long (400mm) casing that I might machine some internals for to make it a hybrid

I have a 54 mm 11.25 inch long firing on stand you can see it on my YouTube

 

[hybridfreak]

Guessing from what you've posted the fuel grain on that motor is probably only around 2-3"? I have a 16" 54mm casing I'll convert to hybrid soon

Speaking of O/F: since Isp is a measure of impulse per mass wouldn't one want to aim for the lowest O/F with good Isp since that means greater mass (fuel) used per amount of oxidizer? Let's say 4:1 and 6:1 deliver the same ISP, burning the same amount of nitrous yields 7% less total impulse at 6:1 right?

 

[drewnickel]

Guessing from what you've posted the fuel grain on that motor is probably only around 2-3"? I have a 16" 54mm casing I'll convert to hybrid soon

Speaking of O/F: since Isp is a measure of impulse per mass wouldn't one want to aim for the lowest O/F with good Isp since that means greater mass (fuel) used per amount of oxidizer? Let's say 4:1 and 6:1 deliver the same ISP, burning the same amount of nitrous yields 7% less total impulse at 6:1 right?

This is usually my philosophy. There are some extremely dense fuels that have low optimum OF ratios like PVC, sorbitol, etc. TU Delft went the sorbitol route because the lower specific impulse was more than offset by running at an OF ratio of like, 2 to 3 with a fuel that has a density of around 1.45 g/cc. I've used PVC plastisol (same stuff used for casting soft fishing lures) with great success, and the bonus of the plastisol and its PVC content is the chlorine allows more aluminum to be consumed (aluminum is dense and very energetic).

The idea of dropping OF ratio being a good thing is largely due to the fact the nitrous has a limited density, around 0.8 g/cc, and so it takes up a lot of space for the impulse its delivering. The fuel can be far more dense, but to really have a meaningful impact on impulse density you want to maximize the fuel being consumed and minimize the amount of nitrous needed to burn it without dropping specific impulse too much. TU Delft was getting around 190 seconds delivered ISP on their sorbitol tests, though I've never managed to get sorbitol to light in smaller motors. The plastisol I've used in the past is a nightmare to cast because it's extremely viscous when heated and you have to heat it above 350 F for it to properly set, but you end up with an optimum OF between like 4-6, a regression rate as high as paraffin, and the ability to load very high percentages of aluminum (>30% aluminum by mass).

 

[cfb_rolley]

Another question that I’m abslolutely sure I’ve seen covered here but I can’t find it…

closing off vent lines once the motor is full - once you can visible see that the motor is full and venting liquid nitrous, is there any issue with using an on-board servo to close off the vent line before closing the fill valve? I know if the rocket is sitting on the pad for a while with everything shut, this would lead to a change in temperature so this is more for small motors that can’t internally vent where you’d be able to fill and fire rather quickly.

 

[drewnickel]

Another question that I’m abslolutely sure I’ve seen covered here but I can’t find it…

closing off vent lines once the motor is full - once you can visible see that the motor is full and venting liquid nitrous, is there any issue with using an on-board servo to close off the vent line before closing the fill valve? I know if the rocket is sitting on the pad for a while with everything shut, this would lead to a change in temperature so this is more for small motors that can’t internally vent where you’d be able to fill and fire rather quickly.

For most people it isn't really worth it to add a way to close the vent. For a properly sized vent, the amount of nitrous you are losing through the vent during a burn is negligible compared to the total amount of nitrous in the tank. Some people have used solenoids or pyro-valves to close the vents, and with motors like the ones through Contrail Rockets they vent through the combustion chamber so you could have a vent solenoid on the ground and not have to carry any electronics or valves on the rocket for shutting the vent. One VERY important note to remember is ullage: you need head space at the top of the motor for the liquid to expand if you shut the vent. If you have a completely liquid filled tank and it starts to heat up, the liquid will expand. When that liquid starts expanding its not going to care if there is a tank there or not, it's going to make room to expand (think soda can in a freezer, but a lot more violent). Video I attached was the first test we did at UTK trying to shut the vent before firing, and unfortunately didn't factor in ullage.
View attachment IMG_6546.mov

 

[rocket_troy]

Guessing from what you've posted the fuel grain on that motor is probably only around 2-3"? I have a 16" 54mm casing I'll convert to hybrid soon

Speaking of O/F: since Isp is a measure of impulse per mass wouldn't one want to aim for the lowest O/F with good Isp since that means greater mass (fuel) used per amount of oxidizer? Let's say 4:1 and 6:1 deliver the same ISP, burning the same amount of nitrous yields 7% less total impulse at 6:1 right?

Well, it's a point littered with nuances: if your delivered metric is basically focused on total engine specific impulse (as opposed to strictly propellant density specific impulse dIsp) then you have to remember to allow for the core in the fuel grain which can basically be treated as porosity for the sake of fuel bulk density ie. the void effectively lowers the fuel density per unit casing length (dry mass) which needed to be factored in.
Also, N2O density will vary with temperature, so those intentionally chilling their N2O to significant levels might even achieve higher densities than fuel choices such as HDPE for example, but generally fuel will be denser.

TP

 

[cfb_rolley]

One VERY important note to remember is ullage: you need head space at the top of the motor for the liquid to expand if you shut the vent. If you have a completely liquid filled tank and it starts to heat up, the liquid will expand.

Good point. I can see that situation would be would be pretty much guaranteed if the vent is just a hole at the very top of the bulkhead and you close the vent before closing the fill when the tank is full. I guess if the vent had a short dip tube that was the length of the head space you needed, that could work, also closing the fill, waiting a bit and then closing the vent.

I don’t think it’s necessary for what I’m doing but I asked because the solenoid I have here that I might use for filling is only 2mm orifice, and the vent on contrail motors is 0.71mm. I am a little worried that the fill solenoid is too small, and the tank won’t pressurise fast enough with the 0.71mm vent and might waste a ton of nitrous while filling, so I was just shower-thinking about a way to minimise losses.

Is there also a rule of thumb about the size of the fill tank for the size of the motor that it’s filling or is that another “it depends” question?

 

[rocket_troy]

Good point. I can see that situation would be would be pretty much guaranteed if the vent is just a hole at the very top of the bulkhead and you close the vent before closing the fill when the tank is full. I guess if the vent had a short dip tube that was the length of the head space you needed, that could work, also closing the fill, waiting a bit and then closing the vent.

I don’t think it’s necessary for what I’m doing but I asked because the solenoid I have here that I might use for filling is only 2mm orifice, and the vent on contrail motors is 0.71mm. I am a little worried that the fill solenoid is too small, and the tank won’t pressurise fast enough with the 0.71mm vent and might waste a ton of nitrous while filling, so I was just shower-thinking about a way to minimise losses.

Is there also a rule of thumb about the size of the fill tank for the size of the motor that it’s filling or is that another “it depends” question?

The question you're actually asking is whether it will *fill* fast enough, not pressurize.

2 things: (a) the area of a 2mm orifice is nearly 8 times that of a 0.71mm orifice and (b) the N2O feeding through the solenoid is in liquid phase (say 0.76g/cc @20 deg C) whereas the N2O *gas* venting out the 0.71mm orifice is 0.2g/cc @5.13 MPa @20 deg C; so the N2O mass throughput of the valve is on the order of 30 times that through the vent. So, in summary, you should be okay provided your fill lines aren't too small.

TP

 

[cfb_rolley]

Cool, I had a feeling I was just overthinking about the solenoid. I reckon I’m about a week or two away from actually ordering the motor now haha.

 

[rocket_troy]

Cool, I had a feeling I was just overthinking about the solenoid. I reckon I’m about a week or two away from actually ordering the motor now haha.

I think you'll be okay. I did neglect to account for fluid velocity which will be higher for the gas vent, but there's a fair whack of margin to chew through.

TP

 

[Tech 68]

Are there any tried-and-true GSE designs out there? I would love to eventually build my own.

I wouldn't hold out a lot of hope, but if these people are still active... maybe.
http://www.sorac.org/

 

[drewnickel]

Good point. I can see that situation would be would be pretty much guaranteed if the vent is just a hole at the very top of the bulkhead and you close the vent before closing the fill when the tank is full. I guess if the vent had a short dip tube that was the length of the head space you needed, that could work, also closing the fill, waiting a bit and then closing the vent.

I don’t think it’s necessary for what I’m doing but I asked because the solenoid I have here that I might use for filling is only 2mm orifice, and the vent on contrail motors is 0.71mm. I am a little worried that the fill solenoid is too small, and the tank won’t pressurise fast enough with the 0.71mm vent and might waste a ton of nitrous while filling, so I was just shower-thinking about a way to minimise losses.

Is there also a rule of thumb about the size of the fill tank for the size of the motor that it’s filling or is that another “it depends” question?

Having a bigger valve and lines leading to the tank can help. If your tubing is too narrow the velocity of the nitrous going through fill lines is high enough that it loses a lot of pressure and can flash to vapor (think Bernoulli's principle: a moving fluid exerts less pressure). Having lots of turns or discontinuities (changing to larger or smaller diameter flow paths along the feed system) can also drop the pressure. You want to minimize the speed of the nitrous as it is fed into the motor. Having larger valves and feed lines doesn't really mean a faster fill necessarily, as you will always be limited by the vent orifice, but it does allow the same mass flow rate at a lower speed. Something else I've seen is having a heat exchanger with ice or something similar to chill the nitrous going through the lines to prevent cavitation as its fed into the motor, but that was done to fill a motor as fast as possible (40kN-s O hybrid filled in like 3 minutes).

For most people its sufficient to say: really tiny vent, the largest diameter plumbing you can manage, the shortest and straightest possible path from supply tank to the motor, and as consistent a flow path as you can manage (i.e. keep the bore diameter consistent along the entire feed system). As for supply tank size, I'd never drain the bottle more than like 2/3rds in the process of filling the tank. So the supply tank should have >50% more nitrous than it takes to fill the motor, and more if you plan on accounting for possible aborts.

 

[G_T]

I use what amounts to a very large vent via a cryogenic oxygen certified pressure relief valve, so my fill rate can be quite high and I'm refrigerating the nitrous in the process. The cost turns out to be making the motor EXTREMELY hard to light quickly. Chilled nitrous is one heck of a fire extinguisher, on the order of 1/2MW cooling for my motor.

Speaking of which, I plan to launch it again at Potter this year.

Gerald

 

[Spacedog49Krell]

For most people it isn't really worth it to add a way to close the vent. For a properly sized vent, the amount of nitrous you are losing through the vent during a burn is negligible compared to the total amount of nitrous in the tank. Some people have used solenoids or pyro-valves to close the vents, and with motors like the ones through Contrail Rockets they vent through the combustion chamber so you could have a vent solenoid on the ground and not have to carry any electronics or valves on the rocket for shutting the vent. One VERY important note to remember is ullage: you need head space at the top of the motor for the liquid to expand if you shut the vent. If you have a completely liquid filled tank and it starts to heat up, the liquid will expand. When that liquid starts expanding its not going to care if there is a tank there or not, it's going to make room to expand (think soda can in a freezer, but a lot more violent). Video I attached was the first test we did at UTK trying to shut the vent before firing, and unfortunately didn't factor in ullage.
View attachment 628769

Nitrous BLEVE!

 

[hybridfreak]

I have started to work on a 38mm prefilled motor I'd call "AlphaTech" it will have an AlphaHybrids style injector (piston held back by BP) and Aerotech style casing/closures. Turning casings is, at least for me, the hardest part about making my own hybrids so I've long played with the ideas to just use available casings. Either in combination with a prefilled tank or a second casing as tank in a U/C-valve type setup.

Anyways, only got the closure done halfway so the pin valve will fit but need to do some more lathe-ing on it:





That's a 9oz tank and a 38/360 in the upper pic, second pic is with a 12oz tank which might be too large for this size casing.