Yes, but it is not really about whether it is working or not. It is about whether it is working with the least effort and lowest backpressure.
kind regards
Jos Koopmans
Stack exhaust design revisited: the 1 in 6 taper myth?
-
JJG Koopmans
- Posts: 342
- Joined: Thu Oct 01, 2009 10:01 am
- Location: Netherlands
Re: Stack exhaust design revisited: the 1 in 6 taper myth?
Henry Greenly was indeed a very complex and awkward guy!
He had flashes of brilliance, and yet was 'blinkered' in many other respects.
All evidence suggests he was a bit of a prima donna, and not easy to work with (an understatement). He fell out with anyone who questioned him.
He fell out with Capt John Howey of the RHDR in a fabulously big bust up, and clash of personalities.
The Greenly formulae/tapers and cylinder to blast nozzle ratio are referenced by Young in the 1930s Illinious tests and paper. They were not actually formulated by Greenly but had more to do with a famous UK model engineer called James C. Crebbin, who was a wealthy model engineer and founder member of the SMEE, the oldest model engineering society. 'Uncle' Jim was a great experimenter, and paid people to make bits for him to experiment with, and paid them to build his ground breaking locos circa 1900-1910. He carried on experimenting for many years after.
Greenly favoured large diameter cylinders and large grates for his loco designs. These were heavily criticised by both LBSC and later with a bit of maths and science by Jim Ewins, whom I knew. However LBSC stuck to large cylinder diameters on his designs so that the 'tyro' or begineer could potter round on 30 psi if the fire wasnt ok.
Greenly had some very odd ideas about superheaters - or not fitting them, and relying on useless grids in the smokebox. Some of his valve gear designs are quite awful.
If you have a loco with a free steaming well proportioned boiler, a good valve gear, and avoid dustbin size cylinders, and use efficient superheaters, you will have a loco far better in performance than anything Greenly designed in the UK. He died in the 1940s, and no one would these days build a Greenly design.
The GWR used Goss as a basis for experiments in fullsize in the early Churchward period to establish a number of 'standard' smokebox draughting dimensions. This was then fine tuned later on by Sam Ell, though the original Churchward standards remained untouched by Ell, such was the excellence of GWR understanding of these things 40 years earlier. Ell knew of Young's work, but took it to another level with practical experiments on fullsize locos.
So to summarise Greenly used large (too large) cylinder diameters, and this is a factor to be noted in this debate. His boilers were not free steaming nor did they have proper superheaters, and the valve gears made bad use of the saturated steam supplied. Had his locos been 'efficient' a much larger blast nozzle could have been used. A smaller blast nozzle in Greenly locos was also required because of the large grate area in proportion to everything else.
Cheers,
Julian
He had flashes of brilliance, and yet was 'blinkered' in many other respects.
All evidence suggests he was a bit of a prima donna, and not easy to work with (an understatement). He fell out with anyone who questioned him.
He fell out with Capt John Howey of the RHDR in a fabulously big bust up, and clash of personalities.
The Greenly formulae/tapers and cylinder to blast nozzle ratio are referenced by Young in the 1930s Illinious tests and paper. They were not actually formulated by Greenly but had more to do with a famous UK model engineer called James C. Crebbin, who was a wealthy model engineer and founder member of the SMEE, the oldest model engineering society. 'Uncle' Jim was a great experimenter, and paid people to make bits for him to experiment with, and paid them to build his ground breaking locos circa 1900-1910. He carried on experimenting for many years after.
Greenly favoured large diameter cylinders and large grates for his loco designs. These were heavily criticised by both LBSC and later with a bit of maths and science by Jim Ewins, whom I knew. However LBSC stuck to large cylinder diameters on his designs so that the 'tyro' or begineer could potter round on 30 psi if the fire wasnt ok.
Greenly had some very odd ideas about superheaters - or not fitting them, and relying on useless grids in the smokebox. Some of his valve gear designs are quite awful.
If you have a loco with a free steaming well proportioned boiler, a good valve gear, and avoid dustbin size cylinders, and use efficient superheaters, you will have a loco far better in performance than anything Greenly designed in the UK. He died in the 1940s, and no one would these days build a Greenly design.
The GWR used Goss as a basis for experiments in fullsize in the early Churchward period to establish a number of 'standard' smokebox draughting dimensions. This was then fine tuned later on by Sam Ell, though the original Churchward standards remained untouched by Ell, such was the excellence of GWR understanding of these things 40 years earlier. Ell knew of Young's work, but took it to another level with practical experiments on fullsize locos.
So to summarise Greenly used large (too large) cylinder diameters, and this is a factor to be noted in this debate. His boilers were not free steaming nor did they have proper superheaters, and the valve gears made bad use of the saturated steam supplied. Had his locos been 'efficient' a much larger blast nozzle could have been used. A smaller blast nozzle in Greenly locos was also required because of the large grate area in proportion to everything else.
Cheers,
Julian
Re: Stack exhaust design revisited: the 1 in 6 taper myth?
Incidentally, The RHDR 15"g locos designed by Greenly, and his 7.25"g LMS 'Royal Scot' design do not follow the Greenly 1:3 and 1:6 tapers, and were all indifferent steamers. The RHDR loco have all been divested of Greenly errors over many years, and are now quite different to how originally designed and built with considerable performance improvements. I was asked to look at the Greenly 7.25"g Royal Scot draughting and was amazed how bad it was!
I dont think Greenly understood these things at all - or at least with large diameter smokeboxes and short chimneys 'bodged' some of his most important and well known designs. Expediency or ignorance?
Cheers,
Julian
I dont think Greenly understood these things at all - or at least with large diameter smokeboxes and short chimneys 'bodged' some of his most important and well known designs. Expediency or ignorance?
Cheers,
Julian
-
David Powell
- Posts: 534
- Joined: Thu Mar 13, 2003 10:38 pm
- Location: Pickering Ontario Canada.
Re: Stack exhaust design revisited: the 1 in 6 taper myth?
I cannot say much about Greenly designed locos, because I do not have any. However, a friend has a 2" scale version of the ME traction engine of 1934 designed by Greenly. It has almost exactly the same performance as my 2" scale "Minnie" traction engine. I have several of each model in 1" scale, many of the parts are interchangeable, Indeed I suspect that the " Minnie" design may well be a plagiarism of the Greenly. I hope this is of interest David Powell.
Re: Stack exhaust design revisited: the 1 in 6 taper myth?
Hi,
I have Jos's book and I read it, but that was a few years ago. Anyway, I go with what works and if it doesn't work the way I think it should, I go back to the theory (and other things) to try and discover why.
This is leading into Jos's comments about momentum transfer. He's absolutely right in this. The following explanation is for you non-engineer types. For engineers, this is fundamentally obvious, or at least it should be. The following isn't meant to be a rigorous proof with equations, its just a simple explanation of what should happen.
Although it pulses, the exhaust has a fairly high velocity. The whole idea of an effective front end is to take this velocity (read energy) and combine it with the mass of steam coming out of the nozzle to get momentum (mass x velocity). With the proper design, you can use this energy to create the desired suction to pull air through the fire and get the products of combustion out the stack.
The goal is to take this exhaust and mix it fairly thoroughly with the smoke, etc. in the smokebox and then get the whole mess out the stack using the energy in the exhaust, preferably with the best efficiency you can manage. As much as possible, you need to take the blast momentum (blast mass x velocity) and transfer it to the combined momentum of the mixture (mixture mass x velocity).
One of the best ways to do this is with a multiple jet blast nozzle because each jet entrains (sucks) its own share of the smokebox contents and the combined flow is headed towards the stack. If you take this flow and direct it into a secondary petticoat, you can then use it to entrain more smokebox contents into the primary petticoat which then directs it into the stack. Another advantage of the two petticoats is that you are pulling from two different levels in the smokebox and this is more effective.
There have been a number of tests done by model engineers in England (Ewins and Lawrence come to mind), with as many as 7 nozzles (6 around a 7th central one) but Lawrence found that 4 worked reasonably well (better than 3 or 5), so that's what I used and I found he's right. It helps, but by itself, the change doesn't make a fantastic difference. There needs to be a little space between the individual nozzles for this setup to be effective. However, when you combine the multiple nozzle setup with other things, the combination becomes quite effective.
If things aren't set up right (too great a distance between the blast nozzle and the petticoat), you can have the smoke form a torus (donut) in the smokebox which just sits there and rotates about itself with the blast going out the stack separately in the donut hole with very little of the smokebox contents mixed in it. The blast velocity creates turbulence which promotes mixing of the blast with the smoke. For the best efficiency, this turbulence must be effectively managed. It is also important that the resulting flow is fairly uniform in velocity across the section and that it fill the stack crossection to form a "seal," so outside air can't creep in between the blast and the inside of the stack. Too big a stack ID. can allow this to happen.
The stack taper also helps to create a vacuum because the increasing cross sectional area of the stack slows the flow down, thus converting velocity into increasing pressure which winds up at atmospheric (just like the recovery cone in an injector). Thus it follows that the pressure inside the smokebox must be lower. The long Lempur stack is more efficient in this than the short stacks usually seen in models and in full size.
To sum up, if the stack "barks" then the blast pressure is higher than it needs to be for good efficiency. However, if that's what you like, go for it, it just means that the efficiency will be less. If the bottom tubes plug up, the draft is too high and/or you are pulling from the bottom of the smokebox. On the other hand, if the Locomotive won't steam without a lot of blower, you don't have enough draft and you need to figure out what the problem is and how to fix it. As I said before, I found out how to make a wide firebox locomotive perform like a pussycat and what I learned can be applied to many others.
This is pretty long, but its just what I've found and I hope it helps others.
Richard Trounce.
I have Jos's book and I read it, but that was a few years ago. Anyway, I go with what works and if it doesn't work the way I think it should, I go back to the theory (and other things) to try and discover why.
This is leading into Jos's comments about momentum transfer. He's absolutely right in this. The following explanation is for you non-engineer types. For engineers, this is fundamentally obvious, or at least it should be. The following isn't meant to be a rigorous proof with equations, its just a simple explanation of what should happen.
Although it pulses, the exhaust has a fairly high velocity. The whole idea of an effective front end is to take this velocity (read energy) and combine it with the mass of steam coming out of the nozzle to get momentum (mass x velocity). With the proper design, you can use this energy to create the desired suction to pull air through the fire and get the products of combustion out the stack.
The goal is to take this exhaust and mix it fairly thoroughly with the smoke, etc. in the smokebox and then get the whole mess out the stack using the energy in the exhaust, preferably with the best efficiency you can manage. As much as possible, you need to take the blast momentum (blast mass x velocity) and transfer it to the combined momentum of the mixture (mixture mass x velocity).
One of the best ways to do this is with a multiple jet blast nozzle because each jet entrains (sucks) its own share of the smokebox contents and the combined flow is headed towards the stack. If you take this flow and direct it into a secondary petticoat, you can then use it to entrain more smokebox contents into the primary petticoat which then directs it into the stack. Another advantage of the two petticoats is that you are pulling from two different levels in the smokebox and this is more effective.
There have been a number of tests done by model engineers in England (Ewins and Lawrence come to mind), with as many as 7 nozzles (6 around a 7th central one) but Lawrence found that 4 worked reasonably well (better than 3 or 5), so that's what I used and I found he's right. It helps, but by itself, the change doesn't make a fantastic difference. There needs to be a little space between the individual nozzles for this setup to be effective. However, when you combine the multiple nozzle setup with other things, the combination becomes quite effective.
If things aren't set up right (too great a distance between the blast nozzle and the petticoat), you can have the smoke form a torus (donut) in the smokebox which just sits there and rotates about itself with the blast going out the stack separately in the donut hole with very little of the smokebox contents mixed in it. The blast velocity creates turbulence which promotes mixing of the blast with the smoke. For the best efficiency, this turbulence must be effectively managed. It is also important that the resulting flow is fairly uniform in velocity across the section and that it fill the stack crossection to form a "seal," so outside air can't creep in between the blast and the inside of the stack. Too big a stack ID. can allow this to happen.
The stack taper also helps to create a vacuum because the increasing cross sectional area of the stack slows the flow down, thus converting velocity into increasing pressure which winds up at atmospheric (just like the recovery cone in an injector). Thus it follows that the pressure inside the smokebox must be lower. The long Lempur stack is more efficient in this than the short stacks usually seen in models and in full size.
To sum up, if the stack "barks" then the blast pressure is higher than it needs to be for good efficiency. However, if that's what you like, go for it, it just means that the efficiency will be less. If the bottom tubes plug up, the draft is too high and/or you are pulling from the bottom of the smokebox. On the other hand, if the Locomotive won't steam without a lot of blower, you don't have enough draft and you need to figure out what the problem is and how to fix it. As I said before, I found out how to make a wide firebox locomotive perform like a pussycat and what I learned can be applied to many others.
This is pretty long, but its just what I've found and I hope it helps others.
Richard Trounce.
-
JJG Koopmans
- Posts: 342
- Joined: Thu Oct 01, 2009 10:01 am
- Location: Netherlands
Re: Stack exhaust design revisited: the 1 in 6 taper myth?
I intend to add as little as possible since a full explanation takes two pages.
Once the jet exits from the orifice it starts entraining and the velocity distribution has the shape
of a Gauss curve (bell shape).After entering the chimney the transfer of momentum does not really change its behaviour, the high velocity momentum around the axis of the jet is transferred to the outside perimeter. Since any tiny accelerated volume near that perimeter leaves a void behind, this void is quickly replenished from below, the only possibility as the chimney wall blocks flow from outside. As a consequence the jet in the chimney is surrounded by an induced secondary flow of combustion products which the chimney appears to inhale. This behaviour was already described by the American Prof. Goss when he reported on his systematic tests from 1890 onwards.
Since momentum is transferred in the direction of the chimney wall, the velocity profile becomes more flat and the ultimate velocity distribution is a uniform one, an equal velocity on exiting the chimney. For such a case it is possible to make an estimation of the final entrainment ratio and simple math will show that theoretically triple the mass exits the stack at 1/3 of the orifice velocity if the pressure difference is neglected for simplification.
This explanation has two consequences, firstly since the momentum transfer needs a path length a longer chimney is advantageous, secondly a parallel stack will have a limiting value compared to a tapered stack where the velocity profile can get a lower value and hence is more efficient.
As for the case of more orifices, people should realise that double chimneys and the like, but also
four or more orifices, are scale models of single chimneys with an increased length. A double chimney is a scale model of an original of (square root 2) 141% length and a four-orifice type is a model of a double length chimney. Since momentum transfer in these longer chimneys lead to more uniform exit velocities they perform better. The improvement, if executed properly, is huge, not marginal.
If you want proof of the theory, Young gave it already in 1933. If you still are a disbeliever, see
http://www.6023.co.uk/news/news.htm and read it from 18.1.2015. It was possible to revive a dead locomotive this way.
As for the comments on petticoats, I stick to the tests of Goss since they were of a comparative nature, the system without any petticoat performed best.
Kind regards
Jos Koopmans
Once the jet exits from the orifice it starts entraining and the velocity distribution has the shape
of a Gauss curve (bell shape).After entering the chimney the transfer of momentum does not really change its behaviour, the high velocity momentum around the axis of the jet is transferred to the outside perimeter. Since any tiny accelerated volume near that perimeter leaves a void behind, this void is quickly replenished from below, the only possibility as the chimney wall blocks flow from outside. As a consequence the jet in the chimney is surrounded by an induced secondary flow of combustion products which the chimney appears to inhale. This behaviour was already described by the American Prof. Goss when he reported on his systematic tests from 1890 onwards.
Since momentum is transferred in the direction of the chimney wall, the velocity profile becomes more flat and the ultimate velocity distribution is a uniform one, an equal velocity on exiting the chimney. For such a case it is possible to make an estimation of the final entrainment ratio and simple math will show that theoretically triple the mass exits the stack at 1/3 of the orifice velocity if the pressure difference is neglected for simplification.
This explanation has two consequences, firstly since the momentum transfer needs a path length a longer chimney is advantageous, secondly a parallel stack will have a limiting value compared to a tapered stack where the velocity profile can get a lower value and hence is more efficient.
As for the case of more orifices, people should realise that double chimneys and the like, but also
four or more orifices, are scale models of single chimneys with an increased length. A double chimney is a scale model of an original of (square root 2) 141% length and a four-orifice type is a model of a double length chimney. Since momentum transfer in these longer chimneys lead to more uniform exit velocities they perform better. The improvement, if executed properly, is huge, not marginal.
If you want proof of the theory, Young gave it already in 1933. If you still are a disbeliever, see
http://www.6023.co.uk/news/news.htm and read it from 18.1.2015. It was possible to revive a dead locomotive this way.
As for the comments on petticoats, I stick to the tests of Goss since they were of a comparative nature, the system without any petticoat performed best.
Kind regards
Jos Koopmans
Re: Stack exhaust design revisited: the 1 in 6 taper myth?
Jos, I recently started running my new Milner Hunslet. It steams well and runs well. It does throw a huge amount of ash out the stack which rains down heavily on the passengers.
I went back to the earlier posts to have a look at my front end. Finding this paragraph from you I ran the numbers.
"I promised some info on Buckinghams Pi-theorem. Most info can be read in a text book on fluid dynamics, if you are interested look it up. However, applying his simple recipes, the functional relationship for front end dimensions can be found. If the diameter of the blast orifice is d, the diameter of the chimney throat (narrowest) D, the distance of the orifice to the throat x and the length of the chimney between throat and exit L, the the following ratios can be used: D/d, x/d, L/D.
This is all that is needed for comparisons between front-end dimensions. For 1/1 locomotives the huge amount of test data gathered by Sam Ell in the 1950's can be analysed to get the really working ratios which are D/d=2.85, x/d= 6 to 7, L/D greater (>) 2. Never, never, never make this last ratio smaller!!! "
The print calls for a blast dia. of .386" which then gives a stack dia. of D=1.1", an L=2.2".
I took D/d=2.85 so D=.386 X 2.85= 1.1"
Am I doing something wrong to get a stack length of 2.2"? and a stack diameter of 1.1"??
Thanks,
I went back to the earlier posts to have a look at my front end. Finding this paragraph from you I ran the numbers.
"I promised some info on Buckinghams Pi-theorem. Most info can be read in a text book on fluid dynamics, if you are interested look it up. However, applying his simple recipes, the functional relationship for front end dimensions can be found. If the diameter of the blast orifice is d, the diameter of the chimney throat (narrowest) D, the distance of the orifice to the throat x and the length of the chimney between throat and exit L, the the following ratios can be used: D/d, x/d, L/D.
This is all that is needed for comparisons between front-end dimensions. For 1/1 locomotives the huge amount of test data gathered by Sam Ell in the 1950's can be analysed to get the really working ratios which are D/d=2.85, x/d= 6 to 7, L/D greater (>) 2. Never, never, never make this last ratio smaller!!! "
The print calls for a blast dia. of .386" which then gives a stack dia. of D=1.1", an L=2.2".
I took D/d=2.85 so D=.386 X 2.85= 1.1"
Am I doing something wrong to get a stack length of 2.2"? and a stack diameter of 1.1"??
Thanks,
Fred V
Pensacola, Fl.
Pensacola, Fl.
-
JJG Koopmans
- Posts: 342
- Joined: Thu Oct 01, 2009 10:01 am
- Location: Netherlands
Re: Stack exhaust design revisited: the 1 in 6 taper myth?
Fred,
Your exhaust proportions look ok to me. However, I have no idea how the orifice diameter was
determined. What are the stack dimensions given on the original drawing? If the stack is wider
and shorter it may diminish a fierce blast. Right now my suggestion is to try and increase the
orifice diameter. It appears that exhausts are quite tolerant on a too small chimney diameter as opposed to one too large.
Kind regards
Jos Koopmans
Your exhaust proportions look ok to me. However, I have no idea how the orifice diameter was
determined. What are the stack dimensions given on the original drawing? If the stack is wider
and shorter it may diminish a fierce blast. Right now my suggestion is to try and increase the
orifice diameter. It appears that exhausts are quite tolerant on a too small chimney diameter as opposed to one too large.
Kind regards
Jos Koopmans
Re: Stack exhaust design revisited: the 1 in 6 taper myth?
Hi Fred,
Your lovely Milner Hunslet (which I have watched in action on youtube) has a few peculiarities. Very long steam passages between cylinder and steam chest, and an indifferent valve gear which we have discussed together on the modeleng.proboards thread some time ago.
You have a long chimney. If you apply Jos Koopmans' fomulae to your smokebox draughting I am quite sure you will notice a significant improvement. I dont think you have quoted the correct extracts of Jos's formulae in miniature, nor understood it's simplicity in application in a 2 cylinder loco in miniature. You could also have a look at Don Young's Hunslet 5"g design as Don used the Ell proportions which are similar but not as correct as Jos's.
Cheers,
Julian
Your lovely Milner Hunslet (which I have watched in action on youtube) has a few peculiarities. Very long steam passages between cylinder and steam chest, and an indifferent valve gear which we have discussed together on the modeleng.proboards thread some time ago.
You have a long chimney. If you apply Jos Koopmans' fomulae to your smokebox draughting I am quite sure you will notice a significant improvement. I dont think you have quoted the correct extracts of Jos's formulae in miniature, nor understood it's simplicity in application in a 2 cylinder loco in miniature. You could also have a look at Don Young's Hunslet 5"g design as Don used the Ell proportions which are similar but not as correct as Jos's.
Cheers,
Julian
Re: Stack exhaust design revisited: the 1 in 6 taper myth?
Thank you Jos and Julian. The bore is 2.281" and the print called for the .386" blast size. The stack ID is 2.32" , 12" tall and has the bottom set just inside the smoke box with no petticoat. The exhaust is 3" below the stack.
I added a superheater in the firebox due to the long steam passages so now the stack is totally dry. I can't tell if I'm getting any superheat or not. I changed the valve gear as was discussed and the engine runs great.
What do you think? Would it help to change to a 4 port exhaust nozzle? The stack is a bit large so maybe that would help.
I added a superheater in the firebox due to the long steam passages so now the stack is totally dry. I can't tell if I'm getting any superheat or not. I changed the valve gear as was discussed and the engine runs great.
What do you think? Would it help to change to a 4 port exhaust nozzle? The stack is a bit large so maybe that would help.
Fred V
Pensacola, Fl.
Pensacola, Fl.
-
James Powell
- Posts: 508
- Joined: Fri Oct 28, 2011 10:42 pm
Re: Stack exhaust design revisited: the 1 in 6 taper myth?
It almost certainly will improve both engine power and n to go to 4 port exhaust vs 1 port. If you get Micheal Guy's calculator, and run the full engine #'s, that will give an ultimate output of the size of the exhaust cones which should be used. You have a huge advantage, in that the chimney is high enough to make the 1:7 slope easy to obtain. If anything, NG engines with long stove pipe stacks have too much stack in relation to diameter, rather than not enough. One trick we have used in traction engines is to use the area above a choke as "open" space, with the choke diameter being set by the design (see the Lempor calculations to calculate the size). So, there are possiblities as to how to deal with the exhaust. The more efficient the entraining of exhaust gasses in the chimney, the less power that will be used, which will leave more energy to be extracted by the cylinders into useful work, rather than hotel load.
You might find that you end up with a design for a chimney shorter than the cab roof- remember that the quarry hunslets had a reason why the cabs are set at the height they are set at- and that has to do with short welshmen. I'm tall...no good for quarry locos unless they are cabless. The stack is going to clear the cab of exhaust, if possible- that's one of the reasons why smoke deflectors are as important as they are on high speed designs. I know that a little bit of work was done on them to make the full sized engines vision areas clear, even at the expense of power at speed. There are all kinds of trade offs made in designing a full sized engine from scratch.
James
You might find that you end up with a design for a chimney shorter than the cab roof- remember that the quarry hunslets had a reason why the cabs are set at the height they are set at- and that has to do with short welshmen. I'm tall...no good for quarry locos unless they are cabless. The stack is going to clear the cab of exhaust, if possible- that's one of the reasons why smoke deflectors are as important as they are on high speed designs. I know that a little bit of work was done on them to make the full sized engines vision areas clear, even at the expense of power at speed. There are all kinds of trade offs made in designing a full sized engine from scratch.
James
-
JJG Koopmans
- Posts: 342
- Joined: Thu Oct 01, 2009 10:01 am
- Location: Netherlands
Re: Stack exhaust design revisited: the 1 in 6 taper myth?
Hi Fred,
If I understand you correctly, you kept the .386 orifice, decreased the stack diameter from 2.32 to 1.1 in. and its length from 12 in. to 2.2 in. As a consequence there is a huge difference of the velocities in the stack.
In your present case at the exit the theoretical total mass would be 2.85 that of the steam with a velocity of 1/2.85 of the velocity at the orifice. Apparently this allows for ash ejection which will only happen above a certain critical velocity, dependent on particle size/shape and the like. The drawing situation would suggest a mass of 6 times that at the orifice and 1/6 of the velocity of which I assume that it is not throwing ashes. I do not think that the original exhaust system would draw a huge amount of extra air, so a total ejected mass ratio of 6 is quite probably not realistic.
My suggestion right now is to go for the 4-fold orifice with a 60% increased area, if properly positioned in relation to your present stack this will take care of fairly uniform exit velocities, so the stack will suck properly, but with a 60% lower velocity which might(!) be below the critical ash entrainment velocity.
Since we do not have any experience with your type of coal this may or may not work.
Probably best for starters would be if you could systematically test your present situation first with varying train velocities, with lower steam use by regulator throttling, and see whether there really exists a non-ash throwing situation.
Kind regards
Jos Koopmans
If I understand you correctly, you kept the .386 orifice, decreased the stack diameter from 2.32 to 1.1 in. and its length from 12 in. to 2.2 in. As a consequence there is a huge difference of the velocities in the stack.
In your present case at the exit the theoretical total mass would be 2.85 that of the steam with a velocity of 1/2.85 of the velocity at the orifice. Apparently this allows for ash ejection which will only happen above a certain critical velocity, dependent on particle size/shape and the like. The drawing situation would suggest a mass of 6 times that at the orifice and 1/6 of the velocity of which I assume that it is not throwing ashes. I do not think that the original exhaust system would draw a huge amount of extra air, so a total ejected mass ratio of 6 is quite probably not realistic.
My suggestion right now is to go for the 4-fold orifice with a 60% increased area, if properly positioned in relation to your present stack this will take care of fairly uniform exit velocities, so the stack will suck properly, but with a 60% lower velocity which might(!) be below the critical ash entrainment velocity.
Since we do not have any experience with your type of coal this may or may not work.
Probably best for starters would be if you could systematically test your present situation first with varying train velocities, with lower steam use by regulator throttling, and see whether there really exists a non-ash throwing situation.
Kind regards
Jos Koopmans