|
|
Post by Johansson on Nov 22, 2012 16:04:26 GMT -5
I am a great fan of the Armadillo blog, very interesting test results, pics and videos!  |
|
|
|
Post by ernie wrenn on Nov 22, 2012 17:03:00 GMT -5
I have several aluminum bottles that are new, out of date. If any one needs some let me know and I will send them to you. mostly 2.5 lb small tanks.
ernie
Compucar Nitrous systems
|
|
|
|
Post by britishrocket on Nov 23, 2012 4:23:00 GMT -5
Hello Anders, I just checked the site after returning home from working in Africa. Congratulations on an excellent set of test runs! It was very gratifying indeed to see everything work. It ran, you got shock diamonds and it didn't melt! Your tests have given me confidence that I am working along the right lines with my own design. I would agree with Steve and others observations here that you seem to have a low frequency instability. I would also agree that this is almost certainly attributable to low feed pressure causing oscillations coupled to the feed system. If you are looking for a decent pressure transducer that is reasonably inexpensive you could do worse than look at the Parker ASIC Performer range. I got some of these for my own project. You can see a picture of one on the British Reaction Research blog, here:- britishreactionresearch.blogspot.co.uk/2012_04_01_archive.htmlThe business end of the sensor has a 1/4 inch BSPP thread on it. Incidentally the response time of the sensor is better than 1mS, so it is good for 1kHz. It should pick up your low frequency pressure oscillations. Here is a link to the data sheet for the unit:- www.rotec.net/pdf/section%209.1%20asic%20performer.pdfAt the moment I am (still) working on the shear coaxial injector results and I'm hoping to get some preliminary work done on the gas centred swirl coax injector. Thing is, I'm back from work but straight onto the run up to Christmas, so spare time is likely to be short! Congratulations again, Carl. |
|
stevep
Veteran Member
Joined: November 2012
Posts: 120
|
Post by stevep on Nov 23, 2012 8:53:22 GMT -5
Hey Anders, Re all the calcs--you're most welcome. I actually happen to find this to be fun and it's prodding me to do some things with my software that I've been meaning to do for a while, so I'm happy to help. Trial and error is great, but there's a lot of paths to go down with rocket motors (just ask Armadillo!)--with some theory, you can at least choose the most fruitful paths instead of just shooting in the dark. The other day you asked: What would the theroretical pressure drop over the injector be then when we know that 10 bar gives 0.1L/s?
I'm guessing you're thinking we could deduce chamber pressure from the pressure drop given that we know the tank pressure is 60 bar. That's a clever idea and one that I'd not thought of. It's easy to do--we just use the orifice flow formula that I gave before, but rearrange it to solve for the pressure drop based on mass flow: deltap = ((mdot / (Cd * A)) ** 2 / (2 rho) I wrote a short program to run through the numbers based on various run times and it gives:
time mass press chamber
flow drop pressure
15 .077 22 38
14 .082 25 35
13 .085 27 33
12 .096 34 26
11 .105 41 19
10 .115 49 11
9 .128 61 ---
(The pressures are in bar, flow is kg/sec through one orifice.) The way to read this chart is that for a given time, there is a certain mass flow. We know that because we have the mass that was in the tank and we believe that the tank fully emptied during the test. That mass flow could only happen for a certain pressure drop ("press drop" column). To get that pressure drop, given that the tank pressure is 60 bar, the chamber pressure would have to be what's in the 4th column (60 - "press drop" = chamber pressure). As you can see, the results are very sensitive to the time and the problem with that is we probably don't have a really accurate way of determining that. That's good news on your digital pressure sensor--hope it isn't too difficult to hook up. Believe me, you'll appreciate having it, rather than having to count frames of video. I did that exactly once, then built a digital data acquisition unit! Carl: Nice to hear from you! I am wondering why you think the feed line pressure would be the problem--he's pressurizing both the fuel and the nitrous with 60 bar of nitrogen. For any reasonable chamber pressure (calculated based on propellant flow and Cstar) there's at least a 50% pressure drop over the injector. This is confirmed by the numbers I've listed here--even at a duration of 15 seconds, the pressure drop had to be at least 22/38 = 58% of chamber pressure. Have I totally missed something? My belief at this point is that it is poor combustion--drop size and/or mixing and/or insufficient L* being the likely culprits. --Steve |
|
stevep
Veteran Member
Joined: November 2012
Posts: 120
|
Post by stevep on Nov 23, 2012 9:32:35 GMT -5
Anders,
Re the tank filling question: one thing you could do (though I'm not sure I'd advise it without some sort of remote control of the valves), is to insert a "T" in the feed line that runs from the bottom of the nitrous run tank to the valve (the double valve that sits between the tanks and the motor). Connect the branch of the "T" to your scuba tank (with a valve, of course) and fill the nitrous tank from the bottom. While doing this, leave the valve at the top *open* until you've filled the run tank with the desired amount of nitrous.
There are problems with this: if the top valve is manually operated, you'll be standing over it breathing nitrous fumes while the tank is filling--not good. Second, you have to be very accurate with the amount of nitrous that goes into the run tank because you need to leave a bit of space at the top ("ullage") for vapor so the tank doesn't overpressurize. If you wait until nitrous *liquid* comes out of the top vent, you've put in too much nitrous and you need to let some evaporate out before you close the vent. Judging the right amount could be tricky.
The other problem with this is that when you insert the "T", you've now created a "dead end" space which I think is discussed in the Aspire safety paper.
As far as your idea of adding air (I hope you meant nitrogen) to the scuba tank before filling the run tank, the physics should work, but I have no idea about the advisability of it--I don't have any personal experience loading nitrous tanks and pretty much plan to follow the standard hobby hybrid process (vent hole in the run tank).
According to the Aspire "Physics of Nitrous" paper, simply elevating the fill (scuba) tank well above the run tank *should* work without any chilling--the nitrous liquid flows downhill into the run tank and the air/nitrous vapor from the run tank flows back up into the fill tank. This is exactly the same process as you'd have if you were using water instead of nitrous--the air from the run tank simply bubbles back up the fill line into the fill tank as the water runs down from the fill tank to the run tank. We know from practical experience, however, that this can take some time if the fill lines are small diameter because the bubbles tend to block the flow of liquid, so I don't know how practical this would be for you. Aspire claims it works for them, but they don't give details on fill line size, etc.
Sorry I can't be of more help on this one...
--Steve
|
|
|
|
Post by Johansson on Nov 23, 2012 17:10:35 GMT -5
Carl, Thanks for the praise! A friend of mine has a bunch of pressure sensors similar to the ones you linked to so before the next test I will have everything set up for computer logging! Steve, Sorry but I do not follow you, what has time to do with calculating propellant flow through an orifice? We tried to use gravity to move the NOX from one tank to another but that only worked as long as there were a pressure difference between the two tanks, once the pressures had stabilised the fill tank didn´t drop a single gram in weight for 30 seconds so it would have taken days to get all of the NOX out of it. Perhaps our fill line is too narrow for this to work but a larger fill line means more loss of liquid NOX when the tanks are disconnected. I see no reason why adding air (yes air, I have a scuba compressor at home so its free compared to nitrogen which would more than double the cost to run the engine) would cause any problems, as long as I keep close check on the pressures it should work just fine. Adding a vent hole would be another option, too bad I didn´t think of this during construction since drilling a hole in the tank would most likely add debree inside it that could clog up the lines and injectors later. Cheers! /Anders |
|
jdw
Member
Joined: January 2012
Posts: 41
|
Post by jdw on Nov 23, 2012 19:01:48 GMT -5
You could dill the hole with low pressure in the tank. Have drilled exhaust manifold on turbo diesel with engine running to blow out debris. This was done to install EGT sensor for edge power tuner.
Salute
Justin
|
|
stevep
Veteran Member
Joined: November 2012
Posts: 120
|
Post by stevep on Nov 24, 2012 6:37:55 GMT -5
Anders,
Flow through an orifice is a *rate* -- like km/hour is a rate of speed, but in this case we're measuring/calculating rate of mass flow: kg/sec.
You started with 2.3 kg of nitrous. If the tank emptied in one second, that would be 2.3 kg /second (through 2 orifices or 1.15 kg/sec through each orifice. If the tank emptied in 5 seconds it would be 2.3/5 = .46 kg/sec through 2 orifices or .23 kg/sec through each. If the tank emptied in 10 seconds it would be 2.3/10 = .23 kg/sec through both orifices, .115 kg/sec through each orifice.
What I did was exactly that calculation for emptying times of 9 through 15 seconds to determine the flow rate for one orifice for one second. Those are the numbers (time, flow rate for one orifice) in the first two columns. Once we have the flow rate for one orifice, we can calculate the pressure drop that would be required to push that amount of nitrous through an orifice in one second. That's the third column. Once we have the pressure drop, we can calc the chamber pressure as tankPressure - pressureDrop = chamberPressure.
Re the air: I was worried about contaminants; if it is air for breathing then I guess there wouldn't be any chance of oil or fine particles in it. If you have a scuba compressor, you must be a serious diver!
Re filling: Sorry to hear the gravity feed didn't work...I get your point about a larger fill line wasting nitrous.
If you're willing to deal with a vent hole, then perhaps you might be able to make the *equivalent* of a vent hole by inserting a small tube down through the top inlet extending down into the tank 50 mm or so. You would fill from the bottom as I previously suggested. The small tube would be in place of the valve at the top and would be sealed within the existing fill tube (run the tube through a fitting, weld the tube to the fitting, test for leaks, etc., then attach fitting to existing fill line). As the nitrous flows into the tank only vapor escapes through the tube, but eventually the liquid level rises to the bottom of the tube, and tank pressure forces it out the top of the tube where you can see it. Then you know to stop filling. A tube large enough to be able to work with would be too large for a vent (you'd lose too much nitrous because it would vent too fast), so you'd want to plug the bottom, then drill a relatively small hole in the tube (quite close to the bottom). The Aspire "Physics of Nitrous" paper mentions 0.3mm (I know you're fond of drilling tiny holes :-) ).
You'd still have the problem of being around nitrous venting to the air during filling unless you made a remotely operated fill valve. Second, at your near (or below?) freezing temperatures, there is some chance that the moisture in the air will freeze around the top of the vent tube; as nitrous escapes it cools rapidly and can freeze flesh and anything else it comes into contact with, including the moisture in the air, the point being that it could freeze the vent shut. I have no idea how much of a problem this would be for you. In fact, when the nitrous *vapor* escapes through the vent, what you see as a small white cloud is actually water vapor being condensed, not the nitrous itself.
If you're concerned about losing too much nitrous through even a tiny vent hole, you could arrange for a valve in the vent tube, but, here again, you'd probably want it remotely controlled. Because you're pressurizing to 60 bar, you'll lose more than the average rocket will at normal nitrous pressures.
Hmmm...now that I think about it, you probably want to solve this venting problem anyway, because when it comes to putting this in your vehicle, I don't think you want to be sitting in your vehicle with nitrous venting nearby.
--Steve
|
|
stevep
Veteran Member
Joined: November 2012
Posts: 120
|
Post by stevep on Nov 24, 2012 7:48:43 GMT -5
Anders,
Regarding your next test, I would suggest that in addition to adding your digital pressure sensor, you *leave on* the analog gauge and video tape the gauges as you tried to do for the first test as a backup for the digital stuff. These tests are difficult/time consuming to run, so rather than possibly come away with no data because the digital stuff failed, you would at least have the analog data.
--Steve
|
|
|
|
Post by ernie wrenn on Nov 24, 2012 9:29:23 GMT -5
How about installing a nitrous purge kit on the tank. It can hold the pressure and opened for a short time to vent the pressure and dump unneeded content when you are finished. I can send you the parts needed. ernie Compucar Nitrous Systems www.compucarnitrous.com |
|
|
|
Post by Johansson on Nov 24, 2012 15:04:41 GMT -5
You could dill the hole with low pressure in the tank. Have drilled exhaust manifold on turbo diesel with engine running to blow out debris. This was done to install EGT sensor for edge power tuner. Salute Justin Thats a good idea, thanks! |
|
|
|
Post by Johansson on Nov 24, 2012 15:19:22 GMT -5
Hi Steve, Aha, then I get it. The diving compressor air is very clean with an active coal filter that catches all moisture and debree before it goes into the tank, I don´t dive but my father used to do it for a profession years ago and still had an old 270 bar compressor around that I managed to get working after some valve overhaul. The idea of venting the run tank appeals more and more to me, this way I won´t have to weigh the fill tank which is very tricky to do out on the race track. I can make exchangable vent tubes with different lenghts so I can choose if I want to fill the tank 1/2 or full (except for the head space of course). I won´t have to sit in the vehicle while filling the run tank so any NOX vapour coming through the vent won´t bother me, after all it takes a lung full of pure gasseous NOX before things starts to blur out. ;D Good idea about leaving the analogue video logging in place, if the logger fails to work even this time I will at least know how high the thrust chamber pressure was. Cheers! /Anders Anders, Flow through an orifice is a *rate* -- like km/hour is a rate of speed, but in this case we're measuring/calculating rate of mass flow: kg/sec. You started with 2.3 kg of nitrous. If the tank emptied in one second, that would be 2.3 kg /second (through 2 orifices or 1.15 kg/sec through each orifice. If the tank emptied in 5 seconds it would be 2.3/5 = .46 kg/sec through 2 orifices or .23 kg/sec through each. If the tank emptied in 10 seconds it would be 2.3/10 = .23 kg/sec through both orifices, .115 kg/sec through each orifice. What I did was exactly that calculation for emptying times of 9 through 15 seconds to determine the flow rate for one orifice for one second. Those are the numbers (time, flow rate for one orifice) in the first two columns. Once we have the flow rate for one orifice, we can calculate the pressure drop that would be required to push that amount of nitrous through an orifice in one second. That's the third column. Once we have the pressure drop, we can calc the chamber pressure as tankPressure - pressureDrop = chamberPressure. Re the air: I was worried about contaminants; if it is air for breathing then I guess there wouldn't be any chance of oil or fine particles in it. If you have a scuba compressor, you must be a serious diver! Re filling: Sorry to hear the gravity feed didn't work...I get your point about a larger fill line wasting nitrous. If you're willing to deal with a vent hole, then perhaps you might be able to make the *equivalent* of a vent hole by inserting a small tube down through the top inlet extending down into the tank 50 mm or so. You would fill from the bottom as I previously suggested. The small tube would be in place of the valve at the top and would be sealed within the existing fill tube (run the tube through a fitting, weld the tube to the fitting, test for leaks, etc., then attach fitting to existing fill line). As the nitrous flows into the tank only vapor escapes through the tube, but eventually the liquid level rises to the bottom of the tube, and tank pressure forces it out the top of the tube where you can see it. Then you know to stop filling. A tube large enough to be able to work with would be too large for a vent (you'd lose too much nitrous because it would vent too fast), so you'd want to plug the bottom, then drill a relatively small hole in the tube (quite close to the bottom). The Aspire "Physics of Nitrous" paper mentions 0.3mm (I know you're fond of drilling tiny holes :-) ). You'd still have the problem of being around nitrous venting to the air during filling unless you made a remotely operated fill valve. Second, at your near (or below?) freezing temperatures, there is some chance that the moisture in the air will freeze around the top of the vent tube; as nitrous escapes it cools rapidly and can freeze flesh and anything else it comes into contact with, including the moisture in the air, the point being that it could freeze the vent shut. I have no idea how much of a problem this would be for you. In fact, when the nitrous *vapor* escapes through the vent, what you see as a small white cloud is actually water vapor being condensed, not the nitrous itself. If you're concerned about losing too much nitrous through even a tiny vent hole, you could arrange for a valve in the vent tube, but, here again, you'd probably want it remotely controlled. Because you're pressurizing to 60 bar, you'll lose more than the average rocket will at normal nitrous pressures. Hmmm...now that I think about it, you probably want to solve this venting problem anyway, because when it comes to putting this in your vehicle, I don't think you want to be sitting in your vehicle with nitrous venting nearby. --Steve |
|
|
|
Post by Johansson on Nov 24, 2012 15:23:55 GMT -5
Hi Ernie, Thanks a lot for your offer, but it feels a bit like overkill for this application when I can use a simple ball valve and a metering jet to do it. I am no huge fan of overcomplicating things so I will stubbornly keep using the simplest possible sollutions. Cheers! /Anders How about installing a nitrous purge kit on the tank. It can hold the pressure and opened for a short time to vent the pressure and dump unneeded content when you are finished. I can send you the parts needed. ernie Compucar Nitrous Systems www.compucarnitrous.com |
|
|
|
Post by ernie wrenn on Nov 26, 2012 8:51:39 GMT -5
Glad to help...
ernie
|
|
|
|
Post by britishrocket on Nov 26, 2012 9:42:28 GMT -5
Hello Anders/Steve,
Pleased to meet you Steve. I am the one who is missing something. I didn't realise that the feed pressure was 60 bar. In that case, there will be no feed system coupling. You are quite right therefore to suspect poor combustion/atomisation and or insufficient L*.
Anders:- Fully agree with the idea of maintaining the analogue gauge along with the digital sensor. It is possible to get some excellent very small CCD cameras these days at very reasonable cost. I have a lot to do with remote telemetry systems in my work life. You can have all the sensors and DAQ systems you like, but you still can't beat a camera pointing at a gauge.
|
|
