Yet Another Liquid Motor

[stevep]

Thanks, Anders, but I have yet to make flames, so I'm still far behind you. I eagerly await your next test, BTW. We should all learn quite a bit from it.

--Steve

[britishrocket]

Hello Steve,

Agree with you regarding Anders' efforts. Anders, you have already achieved what I am still striving towards.

Steve, I hear what you are saying regarding the choking of gaseous flow in the injector. However, I thought that the two governing criteria were the critical pressure ratio and the ratio of orifice to pipe diameters, or beta ratio. As you say the critical pressure ratio for air is 0.528, based on a Cp/Cv of 1.4. However, this is only valid when the orifice diameter ratio is less than or equal to 0.2.

In my experiments, the beta ratio of the annulus was close to 1. Even when I flowed the air through the orifice in the central post, the beta ratio was 0.45 (1.8mm orifice with a 4mm feeder). So I don't think I was seeing choked/sonic flow in my injector.

As you have pointed out, momentum ratio, mass ratio and velocity ratio are important performance governing parameters in this type of injector. As far as the pulsing goes, I believe that it may be influenced by the setting up of a recirculation zone at the interface of the air and the water. There are ways and means to minimise this, mostly centred around specifics of injector geometry.

I am still tied up with work at the moment, however I am assembling all the bits and pieces to have a go at shadowgraphy with the shear coaxial injector. I also just got a new compressor, so once I am able to return to my workshop I hope to have something interesting to share.

[stevep]

Carl,

Thanks for reminding me about Beta. I picked up my CO2, but am now swamped until the end of the month and don't have any time to play with it! I feel like a little kid waiting for Christmas.

I am certain your flow was nowhere near choked because your design process (based on Leroy J. Krzycki's book) would have sized things to keep the flow in the incompressible range. I only cranked my flow up in order to raise the momentum ratio (since I didn't have any denser gas to play with) and in cranking it to the max available, I unwittingly ran into the choked flow issue. Krzycki makes a point of keeping the flow rate low to avoid compressibility issues (which you'd run into before you ran into choked flow). Sorry I don't have time to respond more fully--maybe this weekend.

--Steve

[britishrocket]

Hello Steve,

Yes I was certain that there was no choked flow in my design. It was something that I tried hard to avoid. The Deimos project at TUDelft, in the Netherlands, had a lot of problems with flow choking in their feed system (mainly due to control valve sizing, not enough Cv). Their test report made interesting reading, and was in one of the Journals of the British Interplanetary Society, I forget which.

I know exactly what you mean about being like a kid waiting for Christmas. I am very busy with work at the moment. Meanwhile all the gear for my shadowgraphy attempt has either turned up or is on its way, and I am eager to get it set up. My better half informed me today that my new compressor has now been delivered too.

With regard to CO2, one of the things to watch out for is it freezing when discharging from an orifice. I used BOC Argoshield light for some of my early experiments. This is a mixture of Argon and CO2 and I had trouble with the CO2 freezing and blocking my orifices. CO2 does this when discharged from anything around or above 8-900psi into a pressure lower than about 60psi.

Hope that you get some free time soon, looking forward to seeing your results.

Best Wishes,

Carl.

[britishrocket]

Hello Steve,

Hi, hope you are well. Would you be able to give me a link to the online copy of Settles that you found? Thanks in advance,

Carl.

[stevep]

Hey Carl,

I think I got the Settles book on Schlieren/Shadowgraphy here:
ge.tt/3u5qn8P/v/0

Thanks for the heads up re frozen CO2. I'll watch for it.

So I think I may have gone over the edge on this drop size stuff. I was driving on wet pavement the other day and the tires of the other cars were spraying mist all over my windshield. Before I turned on the wipers I caught myself analyzing the sizes of the drops on the window! Maybe it's time for a break :-)

--Steve

[britishrocket]

Hello Steve,

I understand. I am often accused by friends and family of being unhealthily preoccupied with rocket engines.

Thank you very much for the link to Settles. The book is ridiculously over priced and it seems to be the case in the UK and the US. In the UK the Kindle edition is going for over £80. I'm not that obsessed.

Hope you get some time to devote to droplets in the near future,

Carl.

[ernie wrenn]

Stevep

Regards to CO2 vs N20 ... Both gases operate at the same pressures and both will have a refrigeration effect due to their boiling point of -128 f'. They should give you real time conditions.
The nozzle is going to frost using the liquid of either but the gasious pressures should not have any major frosting effect.

ernie

[britishrocket]

Hello Ernie,

I bow to your experience with regard to liquids. However, as far as CO2 gas goes we will have to agree to differ. High pressure C02 (above 900psi) will freeze when expanded through an orifice to a pressure less than 60psi. When I say "freeze" I mean the outlet will be a mixture of gas and CO2 slush or "snow".

[ernie wrenn]

Ahhhhhh.. Now we must agree to agree..

ernie

[britishrocket]

Nice one Ernie. Steve, I now have a copy of Settles, so thank you very much for the link. I am back from work now and I am waiting on my speedlight flash unit arriving. I hope to begin the shadow sizing experiments by the end of this week or at least the start of next.

At the very least I will have the old beast fired up and producing sprays this week.

Hope you have managed to free up some time to continue your investigations,

Best Wishes,

Carl.

[turbochris]

what temp does N2O freeze? I've seen dry ice come out of CO2 liquid flashing to vapor too.

[ernie wrenn]

-128 ' f.

ernie

[stevep]

I finally got some time to play with my bottle of CO2. I made a wooden stand to hold the bottle upside down so I would get liquid CO2 coming out of the valve. Yes, I turned the bottle right side up (with just a short nipple, no real orifice) and watched it come out to see if there was a difference--there was. Right side up I got clear gas, upside down I got a big plume.

First test with an orifice was just with a short bit of tube and a .021" (.5 mm) orifice:



What you can't see in the photo is that the CO2 comes out in a stream at the same size (roughly) as the orifice for about 1mm then rapidly expands into the plume. The stream is white, just like the plume. I'm assuming that what I'm seeing in the plume is very fine particles of solid CO2.

I then tried my coaxial injector from the last water test. No photos, but the results weren't instantly successful. With the CO2 coming through the post, and some water flowing through the annulus, I was able to make a mist. That's the good news.

The bad news is that the back pressure (or whatever it is) that prevents the water from flowing until the water pressure is turned up makes dialing things in very difficult. For one thing, it is hard not to get the pulsating effect I saw with shop air at high pressure, and for another it is hard to get the water pressure just right to get *some* water flow, but not huge amounts--I have a needle valve I'm going to install to see if that will help on the water pressure side. And I'll stick a gauge on the CO2 side in hopes that it will help me dial that part in.

If I run the injector for more than a few tens of seconds or so, it cools down to the point where the water freezes. So far, though, I haven't seen any clogging that I could absolutely attribute to solid CO2. I also ran the orifice arrangement in the photo for some tens of seconds without clogging. The tube developed frost on the outside (as I expected) but the orifice itself didn't clog. However, there was quite a bit of solid frost (water ice, not solid CO2) stuck to the face of the orifice--I suspect what is happening is that air is getting circulated right around the gas flow which is constantly bringing in fresh air laden with humidity which is then frozen by the cold CO2. I would guess that if I ran it for a significantly longer period of time, it might develop quite a bit of ice around the orifice, but for short runs, it seems fine.

I'm hoping to play some more over the next few days...
--Steve

[stevep]

I can't call what I've been doing "experimenting"--that would overly dignify the thrashing about that has actually gone on, so let us say after thrashing about some more, I am beginning to understand some of the issues.

The first test was to simply verify that the mass of CO2 that was flowing through the post was approximately what it should be: this was done by weighing the tank, flowing CO2 for 15 seconds with a wide-open valve, and re-weighing the tank. The goal was 0.2 lbs/second (about .9 kg) and the actual flow was just about that. That was helpful, but what was truly astonishing is just how vigorous a flow that is--the plume shot out about 2 meters and obviously had a great deal of force behind it. This was pretty much at tank pressure (600 psi, 40 bar, or so). Compared with shop air, this is like comparing a garden hose to a fire hose!

The other observation I made was that there was no sign of freezing during that 15 seconds--in fact, there was no sign that the tubing/injector cooled much at all, indicating that most of the phase change (liquid -> gas) was taking place within the injector post itself.

This corresponds to what I know about nitrous hybrids of the type flown by hobbyists. In those the only pressure drop that takes place is in the (usually single) orifice of the injector. There is typically no valve between the tank and the injector, just a very short run of tubing with a diameter much larger than the orifice. The problem with putting a valve between the tank and the orifice is that if the valve is only slightly opened, it creates a sizable pressure drop and within a short period of time (tens of seconds or less, depending on how large the pressure drop is) things start to freeze up. To my knowledge, those folks who *do* put a valve between the tank and the orifice use the valve not to throttle, but simply as an on-off gate thereby avoiding the pressure drop/cooling issue.

But I had planned to throttle the nitrous and the fuel, not just turn it on/off. With air/water, this isn't a problem as long as the air flow is kept non-choked. But nitrous (and my simulant, CO2) are just barely subcritical which means that a drop in pressure (from the 800 psi or so tank pressure) causes them to partially turn to vapor which in turn drops their temperature--even to the point that some of the liquid starts to freeze.

I played with "throttling" and discovered that significant throttling introduces enough of a pressure drop that some (maybe all?) of the liquid flashes to gas right at the valve. The immense cooling that takes place at that point creates pellets of CO2 which soon clog the orifice. To understand this a bit better, I trained a heat gun on the valve but it was no match for the CO2--I didn't do any timing, but the freezing-up still took place in short order. This obviously has implications for throttling during long flights (way, way in the future), but it also has implications for getting a coaxial injector to work.

Some results (post orifice is 0.073", 1.85 mm):
600 psi measured between valve and injector) flowed .22 lb/sec (0.1 kg/sec) with no sign of frost on tubing
400 psi, flowed .060 lb/sec (0.028 kg) with slight frost on tubing
200 psi, flowed .026 lb/sec (0.01 kg) with more frost on tubing
100 psi, flowed .013 lb/sec (0.006 kg) with lots of frost, "flutter sound" indicative of choked flow, and the flow of CO2 started to diminish toward the end indicating icing up internally thereby restricting the flow


Other experiments identified a couple of problems with coaxial injectors and nitrous:

1. Getting the pressures of the liquid and gas correct relative to each other is extremely tricky unless you have flow meters and pressure regulation for both the liquid and the gas.

The tests that confirmed this for me were the following: with shop air, I could regulate the pressure (and hence the flow in mass/second) with a valve. I could create a reasonably fine spray if I adjusted the air pressure and water pressure just so, but the two flows were not independent--increasing the air pressure through the post influenced the water flow by creating a counter-pressure resulting in less water flow unless I increased the pressure behind the water.

I confirmed this with CO2 and water by doing the following: some time ago I built a 1-liter "fuel" tank that could be pressurized and would hold a measurable amount of fuel (water, in this case). I filled it with water, pressurized it with shop air, and measured the flow through the injector (with nothing flowing through the post) over 15 seconds--turned out to be about what I had expected--close to 0.07 lbs/second. That was nice to know in and of itself, but then I repeated the test with CO2 flowing at the required 0.2 lbs/second and measured a water flow of 0.0026 lbs/second! Yup, almost *no* water flow, even with 100+ psi (7 bar) behind it.

The next step would be to use the CO2 to pressurize the water and see if I can get enough water to flow at higher pressure. However, the drop size depends on there being a certain relationship between the two flows--it is not clear to me that I can establish that relationship while at the same time maintaining the correct relationship for combustion. It is also not clear that I can do it without some sort of pressure regulation of the nitrous (see the freezing problem above) which would negate one of the beauties of using nitrous: you can run it straight from the tank and "regulate it" with orifize size alone. So adding a heavy regulator to the mix isn't desirable for a flyable system. It isn't even clear to me that you can run nitrous liquid through a regulator even if you wanted to without it quickly freezing up just as it does with a partially opened ball valve.

2. The second problem has to do with the cooling aspect of running nitrous through an orifice. The post of my coaxial injector is about 1" (25 mm) long. That is much longer than the narrowest portions of most hybrid injectors and it means that the nitrous is flashing to vapor somewhere within the post and making it very cold in the process. My earlier tests were run with just nitrous flowing through the post and no water flowing through the annulus. Once I started flowing some water (or trying to), it froze very quickly (a few seconds at low nitrous flow). OK, no big deal, because my fuel is going to be diesel and that doesn't freeze. So perhaps during hot-fire tests/flight the injector would work fine, but testing it is going to be messy--water is a great simulant because it is cheap and not harmful to the environment. Spraying diesel all over the place, or trying to capture it, is a major pain. Alcohol is less of a problem (at least it evaporates), but at $12/gallon, it's going to be pretty expensive and it does contain additives that one shouldn't be dumping into the environment.

So here's the deal: The coaxial injector may work well with gas/liquid if the pressures can be independently manipulated, and flows closely monitored, but I am fairly sure it's next to impossible to get right with a subcritical liquid (N2O or CO2) with a high and uncontrolled pressure. It is looking like the fabrication simplicity of a coax injector is (for my situation) offset by its testing and tuning complexity.

At this point I'm ready to suspend work on a coax injector and try what I was originally going to try: a simple injector with relatively small fuel holes impinging on a central stream of nitrous. One of the attractive aspects of non-coaxial injectors is that there are more degrees of freedom and no coupling between the oxidizer and fuel flows--you can regulate the fuel flow by changing the number of orifices as well as changing their size. With the coaxial injector, you're limited to 1:1 (number of ox orifices to fuel orifices). I will also avoid (I think) the problem of freezing the water. The downside, of course, with the non-coax injector is having to drill some small holes, but as long as I stay away from compound angles, it should be relatively easy to make. If so, the problem of small holes may turn out to be a problem more easily solved than trying to get a finicky coaxial injector to work.

I'll know in a day or two....
--Steve