|
Post by Roger on Feb 27, 2019 20:37:25 GMT
Condensing cone...As mentioned earlier, the Condensing cone is where the steam is condensed by the water into a solid column of water + condensate. We have already fixed the length of the Mixing cone (L2) and the gap between the Condensing and Mixing cones. Bob Bramson states that nearly all injectors have been made with the two cones being the same length, ie L1 = L2 He goes on to state that although this is fine for small injectors, ones over 25 fl oz/min (710CC/min) or running at over 90PSI should have a Mixing cone that is longer than the Condensing cone. The argument is that in larger injectors the relative losses in the Mixing cone due to surface friction is less of an issue. This is because the volume increases by the cube of the diameter whereas the area only increases by the square. Therefore there are proportionately less losses in a larger injector and the Mixing cone can therefore be longer. Personally I don't think this makes sense, although I understand the logic. The fact is that there's no wriggle room on the length of the Mixing cone when you have to tie it down using the method described earlier. On the other hand, you certainly can change the relative lengths of L1 and L2 but not by changing L2. The length of the Condensing cone L1 can be changed freely to be as long or short as you like from the point of view of the geometry. Of course there are limitations imposed on this by what we first looked at when designing the ejector portion. So the question is, how long should the condensing cone be? Bob Bramsons assertion about the relative lengths is hard to verify from the dimensions in D.A.G Browns book because he uses Annular regulation. That means the effective length of the Condensing cone is less than it's physical length due to the Steam cone sitting inside it by some distance. In that case, it seems to me that the effective length of the Condensing Cone in all of the D.A.G Browns examples is shorter than the length of the Mixing cone. I've been told in the past that the proportions of the lengths should be in a specific ratio, but I don't seem to have written that down anywhere. Perhaps someone could remind me of that and the source of the information? From what I remember, the length of the Condensing cone (L1) is always less than the length of the Mixing cone. Again, the effective length of the Condensing cone (L1) is hard to assess. Note:- 1 Imperial Fluid Ounce = 28.41 Cubic Centimetres So in summary, there are more questions than answers when it comes to the length of the Condensing cone. What seems to be agreed is that it's the same length as the Mixing cone or shorter, never the other way round. Condensing cone length by Anne Froud, on Flickr
|
|
kipford
Statesman
Building a Don Young 5" Gauge Aspinall Class 27
Posts: 566
|
Post by kipford on Feb 27, 2019 20:49:57 GMT
Reg You correct the whole basis of injectors and ejectors is reliant on Bernouilli. It applies to any fluid flow, air water etc in any flow situation. Roger I spent 35 years involved in the design of Jet Pumps using bleed air from gas turbines for various aerospace applications. Whilst not a full steam injector, the basic fluid dynamics are very similar. I would be happy to collaborate with you on this project if you want. However I cannot contribute until the beginning of April due to being in New Zealand on holiday! Couple of points to note. Ignoring the gaps between the cones, the dominant factor that controls the velocity and pressure through the injector is the area ratios as the changes in diameter have a square law effect on velocity. Hence small changes in diameter(manufacturing tolerance) can have a significant effect on how well a unit performs. Conversely cone angle determines how efficiently we change velocity or pressure. Small changes in angle say 1or 2 degrees will not have any significant impact on performance. This is useful as it allows the length of unit to be varied. Hope this is of use. Dave
|
|
|
Post by delaplume on Feb 27, 2019 21:02:02 GMT
Interestingly, he says that the jet of water contracts slightly as it leaves the Condensing cone so it ought to be able to enter the Mixing cone without touching the edge where the cone starts. I understand that breaking that edge (the Linden secret?) can really help make injectors work. This may not be necessary if the geometry of the cones including the gap is good enough. I believe this phenomenon is the 'Vena Contracta' as seen in connection with orifice plates. I've a feeling that Bernoulli's mass flow equation etc. are applicable to injectors as well as orifice plates. There's plenty of light reading on the subject out there for the dedicated / insomniacs. Reg Hello all, Just a few observations}------- 1) You can see an Orifice Plate at work on the last carriage of a train with steam heating fitted......... 2) Injectors are based on a "Conservation of Energy" flowline ie}-- Total Energy available before = Total Energy available after ( Assuming no frictional losses )...So it is the FORM of that energy which changes.. 3) Here is Bernouli's Equation which states the same thing but is expressed in velocities}----- en.wikipedia.org/wiki/Bernoulli%27s_principleFrom my Marine Engineering days I think the basic equation is}---- P1 x V1 divided by T1 = P2 x V2 divide by T2 where P=Pressure, V = velocity and T = Temperature. 4) You'll find this in Carburetors, Aircraft Wings, perfume sprayers, Loco blast pipe calculations, etc.... a reduction in cross-sectional area = increase in velocity = reduction in pressure. 5) LBSC used to make reamers from round silver steel with an angle machined on one end..This was halved and then heat treated... 6) No matter how accurate you design and make the various components it is ESSENTIAL that ALL CONES maintain their relative distance to one another and ALL are IN-LINE !!.......So an accurate assembly jig will be needed as will sound silver soldering techniques.....something Roger has already ably demonstrated.. Keep it going Roger !!
|
|
|
Post by Roger on Feb 27, 2019 21:04:48 GMT
Reg You correct the whole basis of injectors and ejectors is reliant on Bernouilli. It applies to any fluid flow, air water etc in any flow situation. Roger I spent 35 years involved in the design of Jet Pumps using bleed air from gas turbines for various aerospace applications. Whilst not a full steam injector, the basic fluid dynamics are very similar. I would be happy to collaborate with you on this project if you want. However I cannot contribute until the beginning of April due to being in New Zealand on holiday! Couple of points to note. Ignoring the gaps between the cones, the dominant factor that controls the velocity and pressure through the injector is the area ratios as the changes in diameter have a square law effect on velocity. Hence small changes in diameter(manufacturing tolerance) can have a significant effect on how well a unit performs. Conversely cone angle determines how efficiently we change velocity or pressure. Small changes in angle say 1or 2 degrees will not have any significant impact on performance. This is useful as it allows the length of unit to be varied. Hope this is of use. Dave Hi Dave, Any contributions are gratefully received. I'm not well versed enough in fluid dynamics, despite studying it as part of my Degree many moons ago. This business of how long the Condensing cone is seems to be largely empirical as far as I can see. What I'm hoping for is to come up with some clear cut proportions that will work within the range of sizes we typically use. Anything you can come up with that gets closer to that goal would be most welcome.
|
|
|
Post by Roger on Feb 27, 2019 21:06:54 GMT
I believe this phenomenon is the 'Vena Contracta' as seen in connection with orifice plates. I've a feeling that Bernoulli's mass flow equation etc. are applicable to injectors as well as orifice plates. There's plenty of light reading on the subject out there for the dedicated / insomniacs. Reg Hello all, Just a few observations}------- 1) You can see an Orifice Plate at work on the last carriage of a train with steam heating fitted......... 2) Injectors are based on a "Conservation of Energy" flowline ie}-- Total Energy available before = Total Energy available after ( Assuming no frictional losses )...So it is the FORM of that energy which changes.. 3) Here is Bernouli's Equation which states the same thing but is expressed in velocities}----- en.wikipedia.org/wiki/Bernoulli%27s_principleFrom my Marine Engineering days I think the basic equation is}---- P1 x V1 divided by T1 = P2 x V2 divide by T2 where P=Pressure, V = velocity and T = Temperature. 4) You'll find this in Carburetors, Aircraft Wings, perfume sprayers, etc.... a reduction in cross-sectional area = increase in velocity = reduction in pressure. 5) LBSC used to make reamers from round silver steel with an angle machined on one end..This was halved and then heat treated... 6) No matter how accurate you design and make the various components it is ESSENTIAL that ALL CONES maintain their relative distance to one another and ALL are IN-LINE !!.......So an accurate assembly jig will be needed as will sound silver soldering techniques.....something Roger has already ably demonstrated.. Keep it going Roger !! Hi Alan, Thanks for that. Just one point on the last one, I think the cones are usually a press fit in a smooth body, the distances being set by a pusher with a setting length.
|
|
|
Post by delaplume on Feb 27, 2019 21:21:00 GMT
Hi Roger,
I'm sure that'll be a "No Brainer" for "The Micron Kid" !!----- LoL !!
OK, for those who fancy a go it's time to sharpen your tools and don those magnifying goggles !!
Good luck, Gentlemen....
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Feb 27, 2019 21:59:03 GMT
Hi Roger,
The proportions of your 'condensing cone' to 'mixing cone' are dealt with in Bob's EIM articles. They can also be analysed from the DAG Brown drawings in his invaluable book, taking into account the steam cone being inserted a bit into your 'condensing cone' for which all dimensions are provided as part of the 'annular' gap for repetitive work doing batch production.
There are also other sources such as the Eric Rowbottom injectors described by Basil Palmer in ME. And Bill Carter's 'weeny injector' described by Laurie Lawrence in ME as part of his 'standard size' series, which Steve P has referred to in his post - Laurie's standard injector being a copy of the Arthur Grimmett and Ted Linden injector.
To latch onto Bob Bransom, without considering others who went before, seems to me to be ignoring other evidence. The Eric Rowbottam injectors described in ME are quite relevant to a 16 oz per minute injector, and DAG Brown knew all this. So did Gordon Chiverton, who no one has yet mentioned, who arguably provided by the trade the most reliable and successful injectors in the UK, and his products are much sought after these days, as are the Arthur Grimmett injectors and Ted Linden injectors.
Cheers,
Julian
|
|
|
Post by Roger on Feb 27, 2019 22:43:47 GMT
Hi Roger, The proportions of your 'condensing cone' to 'mixing cone' are dealt with in Bob's EIM articles. They can also be analysed from the DAG Brown drawings in his invaluable book, taking into account the steam cone being inserted a bit into your 'condensing cone' for which all dimensions are provided as part of the 'annular' gap for repetitive work doing batch production. There are also other sources such as the Eric Rowbottom injectors described by Basil Palmer in ME. And Bill Carter's 'weeny injector' described by Laurie Lawrence in ME as part of his 'standard size' series, which Steve P has referred to in his post - Laurie's standard injector being a copy of the Arthur Grimmett and Ted Linden injector. To latch onto Bob Bransom, without considering others who went before, seems to me to be ignoring other evidence. The Eric Rowbottam injectors described in ME are quite relevant to a 16 oz per minute injector, and DAG Brown knew all this. So did Gordon Chiverton, who no one has yet mentioned, who arguably provided by the trade the most reliable and successful injectors in the UK, and his products are much sought after these days, as are the Arthur Grimmett injectors and Ted Linden injectors. Cheers, Julian Hi Julian, I only have those two books, so that's why I've used those as the source so far. Hopefully I'll be able to collate more information from many sources over time. I'll be surprised if there's much difference in any of these, the fundamental physics can't be changed, so there isn't much room for modification. Once you decide on the cone angles, the ratio of the ejector throat sizes and the gap, the item with the most variation appears to be the length of the Condensing cone. If you look at all of the different designs in the middle of the D.A.G Brown book, it's appears that the dimensions have all been arrived at empirically. The sizes seem pretty arbitrary, not following any kind of formula. Maybe that isn't the case, but if they have been designed by a set of rules, he doesn't appear to mention what those rules are. The attraction of Bob Bransoms book is he tries to be as analytical as possible. I'll be interested to see whether any of the other sources do this, or whether they just present injector dimensions that they've managed to make work. Although that's interesting, it doesn't help when you want to take a clean sheet and design an injector from scratch. I'll be interested to see what else is out there by way of explanation as to how their designs have been arrived at. I suspect there's a lot of trial and error and opinion rather than hard facts, but I might be wrong.
|
|
JonL
Elder Statesman
WWSME (Wiltshire)
Posts: 2,902
|
Post by JonL on Feb 28, 2019 9:30:24 GMT
Will you be using a test boiler and flow rig? Sounds like a lot to be learned
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Feb 28, 2019 9:43:13 GMT
Hi Roger,
To quote Bob's table further on he specifies the following lengths for
12-16 oz pm, and 18-25 oz pm injectors respectively :-
Condensing cone 0.130" , 0.140"
Mixing cone 0.150" , 0.160"
I have just measured up a Gordon Chiverton 16 oz pm (actually more like 18 oz pm) injector that had the combining cones pushed out, and the first half is 0.125" long, and the second half is 0.165" long.
Incidentally I can discern no 'Linden chamfer' on the start of the second half - just the merest of de-burring of perhaps 2 thou.
Obviously, having decided a taper of 9 degrees, and having decided the various throat sizes of all the cones, then the 2 halves of the combining cone should be a certain total length. Bob then suggests where the gap between the 2 should be for the overflow; determining their respective lengths.
Going back to your analysis, it is self evident that if the gap is say in the middle, the ratio of steam cone throat to outlet from the first half of the combining cone does not meet the requirements for this section of the injector to act as an ejector - sucking in any air in the water supply pipe and sucking in the water, and expelling both through the gap between the 2 halves of the combining cone before the injector starts to feed.
The above is a requirement for a self starting and self re-starting injector and for the injector to be of the lifting type.
Bob states that the D/d ratio for the cones should be between 1.38 and 1.45. However, in the EIM articles the explanation is very confusing, and the drawing makes very little sense to me. Your copy of his book may be clearer and fuller, and not have suffered from perhaps poor editing and reproduction of the drawing.
Cheers,
Julian
|
|
|
Post by Roger on Feb 28, 2019 10:01:58 GMT
I'll go back and amend any errors or add more information to each section as necessary so that each part of the injector design can be seen in one place rather than fragmented by updates. Delivery cone...The Delivery cone throat diameter is driven by the Steam cone diameter both shown in red. These are taken from Bob Bramsons graph mentioned before. You can see that the Delivery cone throat is slightly inside the entry to the cone. I've shown the lead in as an extension to the angle of the mixing cone so you can see how the position of the Delivery cone throat is driven by that. Bramson says it's not critical but suggests 6-8 degrees, but Brown uses anything from 6-13 degrees, increasing with size. Both seem to agree on a minimum, but that's all. If 6-8 degrees works on everything, 7 degrees seems a plausible angle to use for all sizes until proven otherwise. The intersection of the Mixing cone and Delivery cone angles can be drawn at the Delivery cone throat. That, along with the Delivery cone and Mixing cone diameters decides the position of the throat. The gap is decided by how much lead in you want to add to the front of the Delivery Cone. This usually takes the form of a bell mouth. The longer the lead in, the smaller the gap. Bramson says the gap isn't critical, and gives figures of 0.762mm - 1.016mm (30-40 thou). I'm guessing that the gap has two purposes. Firstly that it allows the full flow of water being delivered to exit the overflow so that starting can be established. If it was too small, the velocity of the jet coming from the mixing cone would be impeded. Secondly, most injectors use the overflow from the gap between the Condensing and Mixing cones to pass through that area, so it must be able to pass the steam and air drawn in while it's starting. Bob Bramson also gives a rule of thumb for the length of the Delivery cone, the length being 10-12 times the throat diameter. The diagram below shows the situation when the Injector is delivering water. The non-return valve between the Condensing and Mixing cones is now closed so air can't be drawn in at that point. However, air can be drawn in between the Mixing and Delivery cones which is unwanted. Most commercial injectors are made like this, and that's poor design in my opinion. Air is compressible, and this appears to limit the performance when the temperature of the feed water is high. It's also bad news for Steel boilers as it causes corrosion. Presumably, drawing air into the boiler from the overflow reduces the amount of water being delivered and decreases the efficiency too. The injectors fitted to 1501 have a large non-return valve on the overflow, so this is equivalent to the two valve miniature version. Some designs have two non-return valves, one for each gap between the cones, and the claim is that these work with feed water up to 37C. If this is the case, it seems bizarre that most locomotives aren't using them. Some users leave the feed water running continuously or bathe them in cold water to make them work. I can't help but wonder if any of this would be necessary if the two valve variant of the injector was fitted. Injector running by Anne Froud, on Flickr I understand that no water or steam should be coming from the overflow while the injector is working. Bob Branson states that the 'chirping' noise often heard is a sign of instability. If this is the case, it would seems that many injectors aren't working as well as they should.
|
|
|
Post by Roger on Feb 28, 2019 10:03:54 GMT
Will you be using a test boiler and flow rig? Sounds like a lot to be learned Yes, I don't think it's practical to experiment without something like that. There certainly is a lot to be learned.
|
|
|
Post by Roger on Feb 28, 2019 10:20:17 GMT
Hi Roger, To quote Bob's table further on he specifies the following lengths for 12-16 oz pm, and 18-25 oz pm injectors respectively :- Condensing cone 0.130" , 0.140" Mixing cone 0.150" , 0.160" I have just measured up a Gordon Chiverton 16 oz pm (actually more like 18 oz pm) injector that had the combining cones pushed out, and the first half is 0.125" long, and the second half is 0.165" long. Incidentally I can discern no 'Linden chamfer' on the start of the second half - just the merest of de-burring of perhaps 2 thou. Obviously, having decided a taper of 9 degrees, and having decided the various throat sizes of all the cones, then the 2 halves of the combining cone should be a certain total length. Bob then suggests where the gap between the 2 should be for the overflow; determining their respective lengths. Going back to your analysis, it is self evident that if the gap is say in the middle, the ratio of steam cone throat to outlet from the first half of the combining cone does not meet the requirements for this section of the injector to act as an ejector - sucking in any air in the water supply pipe and sucking in the water, and expelling both through the gap between the 2 halves of the combining cone before the injector starts to feed. The above is a requirement for a self starting and self re-starting injector and for the injector to be of the lifting type. Bob states that the D/d ratio for the cones should be between 1.38 and 1.45. However, in the EIM articles the explanation is very confusing, and the drawing makes very little sense to me. Your copy of his book may be clearer and fuller, and not have suffered from perhaps poor editing and reproduction of the drawing. Cheers, Julian Hi Julian, Thanks for that, it certainly seems that the Condensing cone is usually made shorter than the Mixing cone by roughly the same proportion. My guess is that a rule defining the relative proportions would be satisfactory. Presumably the Condensing cone only has to be long enough to fully condense the steam under all operating conditions. If it's any longer, you're just wasting energy due to friction on the walls of the cone. Again, my guess is the Linden chamfer only becomes useful if the alignment isn't that great. I find Bobs description of the the cones in Fig.22 on Page 22 completely baffling. It's not clear to me what cone or cones he's talking about. I don't find Bobs book to be any clearer about the ejector either. He clearly knows what he's trying to get across, but the diagrams and text don't combine to make it clear. It took me ages to figure out that the ejector portion of the design refers to the ratio of the Steam Cone throat and the Condensing Cone throat. It's obvious to him, but not necessarily obvious to the reader. In my opinion the book needs to be reworked in conjunction with someone who knows nothing about the subject so these things can be made crystal clear. That sounds ungrateful, but it's not supposed to be. The book gives a real insight as to how injectors are designed, if only you can understand the way it's presented.
|
|
jma1009
Elder Statesman
Posts: 5,896
|
Post by jma1009 on Feb 28, 2019 10:42:28 GMT
Eric Rowbottom, via Basil Palmer (both of whom knew a great deal about injectors) provided a table of the lengths of the 2 halves of the combining cone for different sizes of injectors. It is on p.440 ME 7th May 1976.
You can work out the respective lengths of the range of injectors shown on p.52 and 53 of Derek Brown's book (the length of the second half of the combining cone is not dimensioned as such, but can easily be found by doing a bit of subtraction).
In all the above, the first half of the combining cone is shorter than the second half, and the gap between the 2 is as a result not central.
Hi Roger,
You might like to consider the following, which is not covered by Bob...
The steam cone has a divergent taper on it's exit from the throat diameter. The final diameter of this exit point is quite important when discussing all the above because the steam expands to a wider diameter than the smaller throat in this divergent taper. As far as I am aware the jet of steam exits the steam cone the size of the exit diameter rather than the throat diameter.
This has a bearing on the ejector characteristics. The length of the diverging taper on the steam cone should not be too long, and the combining cone throat size can be on the generous side.
Cheers,
Julian
|
|
|
Post by Roger on Feb 28, 2019 11:07:47 GMT
Eric Rowbottom, via Basil Palmer (both of whom knew a great deal about injectors) provided a table of the lengths of the 2 halves of the combining cone for different sizes of injectors. It is on p.440 ME 7th May 1976. You can work out the respective lengths of the range of injectors shown on p.52 and 53 of Derek Brown's book (the length of the second half of the combining cone is not dimensioned as such, but can easily be found by doing a bit of subtraction). In all the above, the first half of the combining cone is shorter than the second half, and the gap between the 2 is as a result not central. Hi Roger, You might like to consider the following, which is not covered by Bob... The steam cone has a divergent taper on it's exit from the throat diameter. The final diameter of this exit point is quite important when discussing all the above because the steam expands to a wider diameter than the smaller throat in this divergent taper. As far as I am aware the jet of steam exits the steam cone the size of the exit diameter rather than the throat diameter. This has a bearing on the ejector characteristics. The length of the diverging taper on the steam cone should not be too long, and the combining cone throat size can be on the generous side. Cheers, Julian Hi Julian, Unfortunately I don't have any ME articles, I'd be interested to see those. What I'm hoping to do is to find a simple ratio or formula that gives the relative lengths of the Condensing and Mixing cones rather than a table. There must be a relationship that defines the practical dimensions of these, even if that's derived from experimental data. The length of the Mixing cone looks to me like it ought not to be variable given a Steam throat diameter and the angle. The only thing that can change its length is the gap between the Condensing and Mixing cones once the Ejector ratio and throat sizes have been taken from the table. The Condensing cone is a different matter, because that can extend as far to the left in my diagram as you like from the point of view of the geometry. In principle, it looks like you could make the Condensing cone say twice as long as the Mixing cone, not that you would do that. It seems to me that the decision is only a question of how short you can make the Condensing cone while still achieving complete condensation of the steam. Making it a bit too long probably isn't going to matter, but it would make the assembly longer and less efficient. Agreed, the steam jet is ideally parallel and the same size as the outlet of the nozzle. As far as I can tell, the only restriction on the size of the outlet is when the flow ceases to be supersonic. If you go too large, the comes a point where that happens and then the velocity will decrease instead of increase. It would be interesting to know if that situation is anywhere near close to the size that's used.
|
|
|
Post by delaplume on Feb 28, 2019 11:18:51 GMT
My understanding is that the "Chirping" noise is air being drawn back up through the overflow ??.....You can simulate this yourself by drawing air backwards through pursed lips...........Keep your lips at a constant, small setting and all you'l hear is a steady flowing sound.....now alternate the setting and that familiar "Chirping" becomes apparent..........
Air in Boiler feedwater is NOT required and obviously will lengthen the time taken to pump through the amount of water needed......
Yes, as with Engineering Drawings then Engineering literature needs a good proof reader before being published....
There's a definite art to Instructing, either verbally or in print, as I found out when I took up the post of "Apprentice Master" at ABRO Donnington.....I ( we ) had to attend an "Instructors Course" before being let loose on the 2nd year Apprentices.........
I did pose the question to one of our Instructors as to who instructed them and so on....but he declined to reply !!...LoL !!
|
|
|
Post by Roger on Feb 28, 2019 11:57:50 GMT
Regulation...In the diagram below, water (blue arrow) is being drawn into the injector by the action of the Steam jet. You might imagine that you could allow as much water as you like to be drawn in, but this doesn't appear to be the case. If there's nothing to restrict the flow of water, anything in excess of what can pass through the Mixing cone is spilled through the gap between the Condensing and Mixing cones. The amount of water allowed to be drawn into the injector is regulated by restricting the water passage at the point of entry to the Condensing cone. The restriction is correct when it limits the flow of water to the point where no more spills from the overflow. There are two ways the regulation is done, but usually an annular gap is left between the Steam Cone outlet and the inside of the Condensing cone which it protrudes inside. This poses two problems. Firstly the nose of the Steam cone is wafer thin making it prone to damage and erosion, and secondly it's not easy to measure what the gap is. The method I've shown is End Regulation which controls the flow with a pinch point between the ends of the cones. Bob Bramsons book shows that sometimes the gap (G) is zero, and that's where the diameter of the outside of the Steam cone is smaller than the inlet diameter of the Condensing cone. I'm not sure why that would ever be necessary though. Fig 10 on page 10 shows the same proportions as my model and it's at odds with the Fig 24 on page 25. In the diagram below, the area is simply the circumference Pi * D * G. The diameter D is chosen because any diameter larger than that results in a bigger circumference ie a larger area. The limit is the smallest diameter that the water has to pass through. The figures of area are less easy to come by, but they are the same for both methods of regulation. Bob Bramson does show another figure 24 on page 25 which again I find bewildering. I've used the areas that you can calculate from the Annular regulation data in Fig. 23 on page 24 to calculate the area needed for End regulation and hence the gap. Presumably someone with a better understanding of fluid dynamics would be able to give a simple formula for working out the cross sectional area for a given flow rate in this arrangement. We know the expected delivery, so this ought to be pretty simple to work out. End regulation by Anne Froud, on Flickr I don't really understand why Annular regulation is the norm, End regulation seems a much more practical arrangement. Bob Bramson says he always uses End Regulation and I can see why.
|
|
kipford
Statesman
Building a Don Young 5" Gauge Aspinall Class 27
Posts: 566
|
Post by kipford on Feb 28, 2019 18:30:22 GMT
Roger Pity I am on holiday so cannot help with calls, values etc at present, but another point on re-reading one of your posts. Like you I disagree with the comment on surface friction loss being a factor. The loss factor you apply is based on the length to diameter ratio plus allowances for surface finish and Reynolds number. However you will need order of magnitude longer cone length to diameter to get any significant difference due to friction loss. Dave
|
|
JonL
Elder Statesman
WWSME (Wiltshire)
Posts: 2,902
|
Post by JonL on Feb 28, 2019 18:45:38 GMT
I should imagine the surface friction might be a factor in the parts of the injector where laminar flow isn't occurring, I suspect most of the fluid moving through is quite turbulent? Also the rougher the finish the more the Reynolds number will come into things.
I've not touched fluid dynamics in some time...
|
|
kipford
Statesman
Building a Don Young 5" Gauge Aspinall Class 27
Posts: 566
|
Post by kipford on Feb 28, 2019 19:07:54 GMT
Jon Having designed high pressure jet pumps for 35 years, we never took friction loss into account, I will need to check at home but the simplified formula for the loss factor is something like l/D x 0.06, which in the context of the cones even using the smallest end diameter is insignificant and its effect would be lost in test scatter. Roger Thinking outside the box, if I was designing at work we would consider a couple of scenarios. First you could combine the mixing and condensing cones and use annular slots to allow entry, this would get over some of the concentricity issues. The other thing is using sharp edged entries to the various cones immediately gives you 50% efficient (K factor of 0.5) entry loss. If the entry has a radius then the loss significantly reduces,a radius of 1/6 the diameter gives virtually zero loss, hence bellmouth entries on gas turbines etc. Any radius will help and will also help reduce concentricity effects. A small change in cone angle can be used to good effect here, as the radius benefits would outway the any increase in loss due due to the angle change. Anyway must stop now as my wife is demanding breakfast before we drive from Napier to Wellington. Dave
|
|