|
Post by rogsteam1959 on Jul 19, 2020 15:04:37 GMT
I saw a nozzle made by an Australian model engineer and I gave it a try. I also made a nozzle like Gary made. After I had them I made a try to see what it looks like if a send water through it. This is water going through the 4 hole nozzle youtu.be/Bzn4E_n4WV8This same with the starnozzle. youtu.be/_yI0-cgxb7gToday I made a test with the star nozzle on the rolling road. Let the engine running and watched what happened to the boiler pressure. youtu.be/EaNxRRqGaREHere you see the boiler pressure raised a little bit. Now I’m eager to see how it will be on the rails. After a few minutes running. At least the pressure didn’t go down😀
|
|
Gary L
Elder Statesman
Posts: 1,208
|
Post by Gary L on Jul 20, 2020 0:59:45 GMT
I saw a nozzle made by an Australian model engineer and I gave it a try. [All photos snipped] I also made a nozzle like Gary made. After I had them I made a try to see what it looks like if a send water through it. This is water going through the 4 hole nozzle This same with the starnozzle. Today I made a test with the star nozzle on the rolling road. Let the engine running and watched what happened to the boiler pressure. Here you see the boiler pressure raised a little bit. Now I’m eager to see how it will be on the rails. After a few minutes running. At least the pressure didn’t go down😀 I'm afraid your hosepipe test only proves that water will pass through both The point about the Jet Pump is that the whole device constitutes a System- it doesn't make sense to draw conclusions from tests on individual components. At a guess your star nozzle is crearting more back-pressure, hence the water shoots more powerfully. That said, I think the star nozzle would be perfectly valid in a Jet Pump blast arrangement, taking all dimensions from the outside circumference of the star. I emphasise though, both are in the nature of 'nozzles of last resort'. A plain nozzle should be adequate for the majority of installations. I only had to go that way because the proportions of the Bridget chimney meant (I think) that the blast nozzle ended up too big for the cylinders. -Gary
|
|
steam4ian
Elder Statesman
One good turn deserves another
Posts: 2,069
|
Post by steam4ian on Jul 24, 2020 11:21:21 GMT
Gary
Correct me if I am wrong. My understanding of jet pumps is that they comprise a fluid exiting a nozzle directed into a venturi cone. The propelling fluid fluid exits the nozzle at a prssure lower than before the nozzle creating a low pressure region. The propelling fluid then imparts motion to the pumped fluid which then undertakes an energy transformation in the venturi, motion to pressure. This is a brief explanation of a boiler injector or vacuum brake ejector. It also describes what happens in the smoke box of a steam locomotive, I drew attention to this in a previous post.
The problem with modern steam locos is the height restriction does not allow the venturi to be the optimum length for its diameter. The venturi diameter, some might call it chimney, funnel or stack, is critical to amount of air able to be drawn through the fire and thus the rate of combustion and steaming rate. To address this designers installed multiple chimneys (venturis) with multiple blast pipes (nozzles). An alternative is to enlarge to nozzle but this is counter productive to the jet velocity and the pressure drop experienced at the nozzle exit. Multiple nozzles into a single enlarged venturi were found to work, (Lemaitre). Alternately nozzles with complex openings also were found to work, the effective jet area was that appropriate for a single nozzle but the complex shape allowed a larger effective diameter and thus a larger venturi.
Multiple nozzles or irregular shaped nozzle allow more surface area available to the low pressure zone adjacent the nozzle(s) thus collecting more smoke box gas.
The much derided 1:6 and 1:6 rule of thumb is just a shorthand way of getting a workable nozzle and venturi relationship for a model. It is reasonable when the behaviors of jets are noted.
In full size it could be argued that it was Porta who finally got the nozzle/veturi relationship and dimensioning right.
Ian
|
|
Gary L
Elder Statesman
Posts: 1,208
|
Post by Gary L on Jul 25, 2020 0:35:13 GMT
Gary Correct me if I am wrong. My understanding of jet pumps is that they comprise a fluid exiting a nozzle directed into a venturi cone. The propelling fluid fluid exits the nozzle at a prssure lower than before the nozzle creating a low pressure region. The propelling fluid then imparts motion to the pumped fluid which then undertakes an energy transformation in the venturi, motion to pressure. This is a brief explanation of a boiler injector or vacuum brake ejector. It also describes what happens in the smoke box of a steam locomotive, I drew attention to this in a previous post. The problem with modern steam locos is the height restriction does not allow the venturi to be the optimum length for its diameter. The venturi diameter, some might call it chimney, funnel or stack, is critical to amount of air able to be drawn through the fire and thus the rate of combustion and steaming rate. To address this designers installed multiple chimneys (venturis) with multiple blast pipes (nozzles). An alternative is to enlarge to nozzle but this is counter productive to the jet velocity and the pressure drop experienced at the nozzle exit. Multiple nozzles into a single enlarged venturi were found to work, (Lemaitre). Alternately nozzles with complex openings also were found to work, the effective jet area was that appropriate for a single nozzle but the complex shape allowed a larger effective diameter and thus a larger venturi. Multiple nozzles or irregular shaped nozzle allow more surface area available to the low pressure zone adjacent the nozzle(s) thus collecting more smoke box gas. The much derided 1:6 and 1:6 rule of thumb is just a shorthand way of getting a workable nozzle and venturi relationship for a model. It is reasonable when the behaviors of jets are noted. In full size it could be argued that it was Porta who finally got the nozzle/veturi relationship and dimensioning right. Ian Hi Ian I don't want to be mistaken for any kind of expert here; I'm just a happy user passing on the principle as it was explained to me, and I can recommend it, because it works. And if it works for me, it will work for everybody, because only the scale of the drawing changes. Everything else must remain in exactly the same proportions and relationship as in the drawing. That is why you must start at the chimney top to get the scale factor, because it is the only part you cannot easily vary on a scale model. If it is the case that you don't have enough height available, then you will have to sleeve the chimney, but Mark tells me he has never needed to do this with any loco he has been consulted over, so the proportions must be broadly suitable for miniature loco draughting, even though that is not what they were developed for. So nothing you have said is wrong as far as I know, but it is pointless to try and relate the jet pump to Greenly's angles, Lemaitre's blast nozzles, or anybody else's individual components. It is the system, with everything working together harmoniously as one, that delivers the result. Vary from the drawing and the result will be unpredictable, and inevitably worse. Once optimised, as I am assured it is, you can go no further. Thus Greenly's angles are broadly right, but they take no account of the correct blastpipe sizing, nor the correct diverging taper of the chimney/venturi. Lemaitre's multiple nozzles should not be necessary*, because the Jet Pump gives a very soft blast and minimal back-pressure anyway. (I'm told there is a more or less straight-line relationship between the volume of the propelling fluid (steam) and the suction delivered, but it is not unreasonable to expect a drop off in efficiency at the high and low extremes of flow. This is where the real world, in the shape of chimney diameter prescribed by scale considerations might trip over the world of the theoretician, but let's not complicate matters). Neither Greenly nor Lemaitre has anything to say about the 1.5 diameters parallel portion in the nozzle either, all these factors must be present in the correct degree to get the optimum results. It sounds prescriptive (and it is) but the happy result is we don't need to understand any of it, just do it! Before I was introduced to this principle, I spent a lot of time trying to make sense of Ell's (Swindon) formula and Koopman's thesis. The former was unsatisfactory because I couldn't see how it could scale to our sizes, and the latter contained ambiguities that I couldn't resolve to my satisfaction. Since then, I've found that Ell's formula wasn't as unscalable as I supposed, and Koopmans has rewritten his conclusions but I haven't seen the results. There is only so much time available, and I'm content with what I now use, so I go no further. Hope this helps Gary *From my remarks earlier in this thread, you will see that I did need to diverge slightly from the theoretical, and my reasons (suppositions really) why this was necessary. I won't go over that again, because really it is a side-show, but don't imagine that multiple nozzles means that mine is any sense a Lemaitre nozzle. It isn't. Lemaitre's intention was to get more draw from a softer blast, my intention was the exact opposite; I had a very soft draught which I needed to accelerate without compromising the Jet Pump proportions in any way. It worked. -G
|
|
kipford
Statesman
Building a Don Young 5" Gauge Aspinall Class 27
Posts: 566
|
Post by kipford on Jul 25, 2020 9:30:41 GMT
Gary I designed jet pumps for 40 years for gas turbine air filtration systems, where every bit of performance was critical as we were bleeding air off the compressor and this has a marked effect on engine power loss. So every application was bespoke where we knew all the flow conditions. The design work, but there are so many unknowns with a model steam loco, like what are the actual primary and secondary flow rates, temperatures, pressure drops etc that it is probably as good as it gets without a lot of further work. I do have one query though. In all literature I have on jet pumps, our design procedures, the designs we did, CFD work on improving nozzle and mixing tube design. We never used a 1.5 dia parallel section in the nozzle throat. Normally 10/20 degree contraction to the throat with a very small flat so the diameter could be easily manufactured, then the all important sharp edge. If we were really going for absolute performance where cost was not so important, then we used a small divergent exit after the throat, again with a sharp edged exit. The most critical feature that effects performance is the relationship between the nozzle and mixing tube throat diameters. The other geometries all have an effect but are not so powerful. So personally I would use the geometry, but not the 1.5 parallel section, as this allows more flexibility in the layout. I doubt you will have any available test method that would be able to differentiate between having it or not. Regards Dave
|
|
Gary L
Elder Statesman
Posts: 1,208
|
Post by Gary L on Jul 25, 2020 12:16:49 GMT
Gary I designed jet pumps for 40 years for gas turbine air filtration systems, where every bit of performance was critical as we were bleeding air off the compressor and this has a marked effect on engine power loss. So every application was bespoke where we knew all the flow conditions. The design work, but there are so many unknowns with a model steam loco, like what are the actual primary and secondary flow rates, temperatures, pressure drops etc that it is probably as good as it gets without a lot of further work. I do have one query though. In all literature I have on jet pumps, our design procedures, the designs we did, CFD work on improving nozzle and mixing tube design. We never used a 1.5 dia parallel section in the nozzle throat. Normally 10/20 degree contraction to the throat with a very small flat so the diameter could be easily manufactured, then the all important sharp edge. If we were really going for absolute performance where cost was not so important, then we used a small divergent exit after the throat, again with a sharp edged exit. The most critical feature that effects performance is the relationship between the nozzle and mixing tube throat diameters. The other geometries all have an effect but are not so powerful. So personally I would use the geometry, but not the 1.5 parallel section, as this allows more flexibility in the layout. I doubt you will have any available test method that would be able to differentiate between having it or not. Regards Dave Hi Dave Don't mistake me for any kind of expert! I just do what I'm told and it works- and with the draughting of a steam loco, with, as you say, so many variables, that is what most of us want I think. This method gives a good result straight 'out of the box' which the average builder without specialist experience is unlikely to improve upon. From what I've seen, even the designers of miniature locos don't always fully understand what they are doing, and we know for a fact that full-size designers often got it badly wrong... though with less excuse if they were starting now than they had ¾ of a century ago! This applies to the parallel nozzles too. I have a very imperfect understanding of why it works, but I'm happy to accept that it does. I'm told that the reason is to smooth the turbulent flow in the throat, and I suspect that scale effects may make this much more important in the sizes we are dealing with; the ratio of surface area (inside the nozzle) to volume (of exhaust steam) is going to be far greater than in the kind of full-size applications that I think you might be comparing with; thus the influence of friction and other disturbances in the nozzle wall will be greater pro-rata. I have seen drawings which indicate quite a sharp constriction at the nozzle, often with a significant divergent taper too, and I think we will probably both agree that is unlikely to be helpful unless expertly configured and manufactured. There is a practical value in the parallel section, in that it makes manufacture, alignment and positioning of the jet very easy. Make it too high, skim a bit off, nothing else is affected. I doubt if that figures in the reasons for specifying it, but I found it helpful. Some designs of miniature blast pipe offer no other way of adjusting the height; if the inside is tapered this can't be done. Best regards Gary
|
|
kipford
Statesman
Building a Don Young 5" Gauge Aspinall Class 27
Posts: 566
|
Post by kipford on Jul 25, 2020 13:17:15 GMT
Gary Keep using what you know works for you. I only mentioned the 1.5 dimension as in my opinion it will not have any noticeable effect and hence need not be religiously adhered to. I agree with you a longer parallel section can make manufacture easier, equally its surface finish is much more critical and it should be a reamed finish. Out of interest we used nozzle throats from as small as 1 mm dia up to around 8 mm so not a million miles different. Hence scale effects do not come into it, nor will friction which effects the pressure loss as this only becomes a factor on very long pipe runs and the losses in the tortuous route from the cylinder to the nozzle will be orders of magnitude higher. My next project is a self designed loco (Victorian prototype) for which preliminary design is underway. On that one I am going to attempt to look the draughting from first principles and will post my findings on here when I have done the work. Regards Dave
|
|
Gary L
Elder Statesman
Posts: 1,208
|
Post by Gary L on Jul 25, 2020 17:36:43 GMT
Gary Keep using what you know works for you. I only mentioned the 1.5 dimension as in my opinion it will not have any noticeable effect and hence need not be religiously adhered to. I agree with you a longer parallel section can make manufacture easier, equally its surface finish is much more critical and it should be a reamed finish. Out of interest we used nozzle throats from as small as 1 mm dia up to around 8 mm so not a million miles different. Hence scale effects do not come into it, nor will friction which effects the pressure loss as this only becomes a factor on very long pipe runs and the losses in the tortuous route from the cylinder to the nozzle will be orders of magnitude higher. My next project is a self designed loco (Victorian prototype) for which preliminary design is underway. On that one I am going to attempt to look the draughting from first principles and will post my findings on here when I have done the work. Regards Dave Hi Dave, Thanks for that. Yes by friction I wasn't thinking about pressure drop in long pipe runs, only about the drag that occurs at the walls of the nozzle and thus contributes to turbulence. As you rightly say, the inside of the nozzle should have the best finish practical, to keep turbulence to a minimum. As a PS, the conditions at the blast nozzle have a lot in common with a vacuum ejector, as would the very small throats that you mention above. From curiosity I made my latest vac ejector as a tiny jet pump as exactly as I could to the drawing. It hasn't been tested yet under steam, but it seems to pull a water jet quite well at a modest pressure of air, for what that is worth. Gary
|
|
|
Post by keith1500 on Jul 25, 2020 22:18:12 GMT
Just out of curiosity if drafting is set up correctly would the loco have any soot (sooty oily gunge) deposits around the top of the smoke box, or should the smoke box be fairly clean except for a pile of ash in the bottom?
|
|
Gary L
Elder Statesman
Posts: 1,208
|
Post by Gary L on Jul 28, 2020 22:04:02 GMT
Just out of curiosity if drafting is set up correctly would the loco have any soot (sooty oily gunge) deposits around the top of the smoke box, or should the smoke box be fairly clean except for a pile of ash in the bottom? The latter. If you have gunge at the inside top of the smokebox, it can only mean that the exhaust is not exiting cleanly up the chimney. At a guess, either the blastpipe is not aligned well enough, or the blast nozzle is too low, or the petticoat is too small or missing. Possibly more than one of these. HTH Gary
|
|
|
Post by terrier060 on Aug 1, 2020 22:44:02 GMT
Out of interest, I made my blast nozzle a venturi. Difficult to describe without a drawing, but looking at Wilf's sketch above the nozzle tapers to the choke diameter and then tapers out to follow the 6 to 4 degrees. I felt that this would produce a cleaner and more direct draught, though I may of course be quite mistaken. Ed
|
|
kipford
Statesman
Building a Don Young 5" Gauge Aspinall Class 27
Posts: 566
|
Post by kipford on Aug 2, 2020 8:42:18 GMT
Ed This is known as a convergent/divergent nozzle. The contraction down to the throat and sharp edge is a convergent nozzle. It will work fine providing you have a sharp edge at exit of the nozzle. In the professional world you go this route only to extract the last bit of performance. You would not notice any difference between the two types in one of locos.
|
|
|
Post by rogsteam1959 on Aug 8, 2020 17:04:42 GMT
I had the chance to have a test run in www.huserland.de/With star nozzle. I am good with that. The engine is good at speeding, so the backpressure shouldn’t be that bad, and the steaming improved. I don’t need the blower that much. Almost closed and still enough steam.
|
|