If you register, you can do a lot more. And become an active part of our growing community. You'll have access to hidden forums, and enjoy the ability of replying and starting conversations.

Locomotive Front End Designs

Discussion in 'Steam Traction' started by ragl, Feb 19, 2016.

  1. Courier

    Courier New Member

    Joined:
    Dec 1, 2010
    Messages:
    176
    Likes Received:
    91
    Ell comment.jpg
    This is Ell discussing Tuplin's paper. Doesn't exactly answer your question but does show Ell was familiar with Young's work. Now if only Churchward had published a bit more than he did....
     
  2. Courier

    Courier New Member

    Joined:
    Dec 1, 2010
    Messages:
    176
    Likes Received:
    91
    Just posted that extract of Ell commenting on Young in his reply to Tuplin - and then saw you had said the same thing 5 minutes earlier.
     
  3. JJG Koopmans

    JJG Koopmans Member

    Joined:
    Nov 12, 2014
    Messages:
    268
    Likes Received:
    232
    Gender:
    Male
    Heritage Railway Volunteer:
    No I do not currently volunteer
  4. pete2hogs

    pete2hogs Member

    Joined:
    Oct 16, 2007
    Messages:
    530
    Likes Received:
    177
    To be fair Gresley had agitated for a stationary test plant for years, and of course sent Cock o' the North to Vitry for testing. (Although he seems to have done very little with the results)

    I do seem to recall some reference to the fact that the delay in adopting the Kylchap initially was waiting for the patents to expire, but clearly somewhere on the way (After Thompson's time? Because he clearly understood the advantages) there was a breakdown of communication as well.
     
  5. Lplus

    Lplus Well-Known Member

    Joined:
    Nov 24, 2011
    Messages:
    1,519
    Likes Received:
    571
    Location:
    Hampshire
    The patent apparently expired during WW2, possibly just in time for Thompson to use it.

    Not sure it was a breakdown of communication after the war so much as a preponderance of GWR and LMS people in the upper echelons of the BR design department. They didn't use it and possibly didn't see why it should be necessary. P Townend had to prove the economic advantages before permission was given to modify the A4s and A3s.
     
  6. John Stewart

    John Stewart Part of the furniture

    Joined:
    Sep 22, 2011
    Messages:
    4,206
    Likes Received:
    2,033
    Gender:
    Male
    Occupation:
    Retired
    Location:
    Hilton, Derby
    Heritage Railway Volunteer:
    No I do not currently volunteer
    I'm no engineer (as all will have noticed) but it did occur to me that, where it was suspected that some dimensional tuning might be required, it would be prudent to make both blastpipes and chimney at the largest size so that variable liners could be fitted to achieve the best results.
     
  7. John Stewart

    John Stewart Part of the furniture

    Joined:
    Sep 22, 2011
    Messages:
    4,206
    Likes Received:
    2,033
    Gender:
    Male
    Occupation:
    Retired
    Location:
    Hilton, Derby
    Heritage Railway Volunteer:
    No I do not currently volunteer
    I'm glad that some one has eventually said this. All the effort into the design of self-cleaning smokeboxes was to deposit the ash everywhere but in the pit! I don't think that anyone would start to pursue this nowadays.
     
  8. Courier

    Courier New Member

    Joined:
    Dec 1, 2010
    Messages:
    176
    Likes Received:
    91
    Swindon's advantage was more in ideas than equipment. They had methods of testing that could objectively compare locomotives or features of locomotives (such as front ends) and they could describe locomotive performance in simple graphs that were of use to the operating department.

    There was nothing inevitable in Swindon taking the lead on front end design - it could have been Doncaster if they had an similar culture of engineering (such as good communication between designer and operator).

    I believe regions had the autonomy to improve their locomotives (eg fitting double chimney to Kings didn't require a large budget) - and if Doncaster had decided to fit Kylchap to A4, A3, V2 etc in 1950 they could have done so.

    When there is a breakdown in communication (such as between running dept that knew the Kylchap A4 were superior and Doncaster who had forgotten it) you often find that both sides share the blame. I suspect L P Parker is as much to blame as Harrison and Cook.
     
    pete2hogs likes this.
  9. paulhitch

    paulhitch Part of the furniture

    Joined:
    Oct 12, 2008
    Messages:
    3,883
    Likes Received:
    1,712
    Gender:
    Male
    Location:
    Hayling Island
    Heritage Railway Volunteer:
    Yes I am an active volunteer
    Arguably it is somewhat less offensive than discharging effluvium from train lavatories over the track, which continued for decades after steam ceased to be used on the "real" railway. I take it you have never emptied smokebox char from a steam locomotive.

    PH
     
    maddog likes this.
  10. class8mikado

    class8mikado Well-Known Member

    Joined:
    Jun 1, 2009
    Messages:
    2,099
    Likes Received:
    394
    Occupation:
    Print Estimator/ Font of all knowledge (useful, or
    Location:
    Bingley W.Yorks.
    I havent had that pleasure but being any where near an open smokebox door with even a bit of the fire left is not nice but lets nit forget that the s.c smokebox apperatus also acts as a spark arrestor and im surprised that these arent compulsory nowadays...
     
    Spinner and paulhitch like this.
  11. JJG Koopmans

    JJG Koopmans Member

    Joined:
    Nov 12, 2014
    Messages:
    268
    Likes Received:
    232
    Gender:
    Male
    Heritage Railway Volunteer:
    No I do not currently volunteer
    In the posts about the 6023 I have announced some contributions on front end design to be posted here. Most of this is already discussed in my PhD thesis "The fire burns much better.." available from Camden. However, back in 2004/2005 writing the texts I used mainly contributions from railway publications. Since the web has grown rather a lot since those days there are quite a few contributions available from other sources that can be used as illustration of the arguments or, if you want, as proof. What you can expect is a contribution on single orifice exhausts. Next one on multiple exhausts and the background theory. The last one is an abridged version of my text in the SLS Journal about exhaust theory. This is not the place to show formulae, so you get a verbal explanation to get the gist!
    Kind regards
    Jos Koopmans
     
    Sheff, Chris86, Allegheny and 2 others like this.
  12. JJG Koopmans

    JJG Koopmans Member

    Joined:
    Nov 12, 2014
    Messages:
    268
    Likes Received:
    232
    Gender:
    Male
    Heritage Railway Volunteer:
    No I do not currently volunteer
    Sorry, long!

    We have to take a look at the combustion process first, to know the ratio of steam to combustion gases. The is discussed in textbooks like Phillipson and S.O. Ell wrote a contribution in “The concise Encyclopedia of World Railway Locomotives” p.386. In this he stated that this ratio was 1 to 1.75 including the extra air. Since 1 unit of steam momentum can only move the total 2.75 units of the ejected mixture at a theoretical maximum velocity of 1/2.75 of the exhaust velocity, the orifice to chimney diameter ratio reflects this value. Please note that we are already assuming a uniform velocity at the exit of the chimney. Ell suggested a chimney to orifice diameter ratio of 2.85 to 2.95. Different chimney diameter ratios result in other mass ratios, see for examples of this Kandakure: “Characteristics of turbulent confined jets”, available on the web.

    For this discussion the ratio 1: 2.85 for orifice to chimney diameter is regarded as well established UK practice.

    The standard exhaust system consists of a blast pipe with orifice on top exhausting used steam into a tubular or tapered chimney.

    The steam that is issued from the orifice starts entraining the surrounding combustion gases. By doing so it flows through a volume with the shape of a cone. Already in 1912 the German Trüpel measured the velocity distribution in such a jet. He found that this distribution had the shape of a bell-curve, a Gaussian or normal distribution as it is known in mathematics. It can be shown by repetitive calculation that the redistribution of the momentum of the steam jet onto the surrounding gases indeed results into a Gauss type velocity distribution. As such the free jet from the orifice is fairly well described.

    As stated in text books like that of Rajaratnams “Turbulent jets”, measurements have shown that the entrainment ratio Q/Q0=0.31x/d with Q the total, Q0 the steam mass, x the distance from the orifice with diameter d. Of course it is a useful approximation.

    At about 6 diameters from the orifice the entrained jet has about the same diameter as the chimney so it enters smoothly. A lower orifice produces a jet which is too wide on entering and a higher positioned orifice renders the lower part of the chimney useless. Given the entrainment ratio above, the total amount on entering the chimney is 6*0.31 or 1.86 times the steam amount, so that half the combustion products are already entrained at that location. The remainder is left to the actions of the chimney.

    Within the chimney wall boundary the redistribution of momentum continues of course, resulting in a peaked flow at the chimney axis steadily decreasing into a far more uniform one. The redistribution of momentum accelerates tiny volumes away from the chimney axis and the resulting “voids” are quickly replenished by the relative overpressure in the smokebox from the only direction possible, the chimney entrance. The chimney appears to perform a suction action.

    The transfer of momentum is complete when the chimney exit shows a uniform velocity profile. As long as this is not the case the chimney is underperforming. For a circular chimney uniform velocity is reached at a certain chimney length, however in the case of a tapered chimney the redistribution of momentum can continue for a longer time, hence the superiority of chimney taper. This process can also be verified by repetitive calculations.

    There are two additional remarks to be made about the exhaust jet. Firstly, all experiments show that there is a uniform pressure within the smokebox. The jet does not have to struggle with a pressure difference between orifice and chimney throat. So on entering the chimney the jet still contains all the original momentum because of conservation. There is no such thing as a “shock loss” which Giesl-Gieslingen promoted and in itself a 19th century concept, not found in any modern day textbook. Secondly since the entrainment ratio of the jet is linear with the distance and the peripheral area of the jet cone is clearly not, there is no relation between entrainment and the area of the jet, they simply do not fit together.

    A closer look at the other dimensions

    The distance of the orifice from the chimney throat is critical, this was shown in my thesis with proof from the Rugby tests. In 1933 Young showed this also in his Illinois tests, available on the web: https://www.ideals.illinois.edu/handle/2142/4435 see page 99.

    Also on the web the work of Singh et al. can be found : . Mixing and Entrainment Characteristics of Circular and Noncircular Confined Jets. They tested a 3mm orifice with a 39 mm tube with a length-diameter ratio of 4.2 Since this is a 35$ ASME publication I have redrawn the data, see below. They tested the orifice position from far below the entrance to well within the tube and the results show an optimal distance. The graph also shows that the total mass ejected is close to the ratio of tube to orifice of 13. Although the dimensions are outside those of chimneys these data from a random publication confirm the general pattern discussed.

    The length of the chimney is also an issue. Ell stated that 26 inches was ok, he did not use a ratio for the length unlike those of the other dimensions he gave. The thesis graphs showed that below a length to throat diameter ratio of 2 the performance started to degrade and in the case of GWR 6023 with a present ratio of 1.55 there is no performance at all.

    In fluid dynamics an orifice is called a nozzle, it converts (back)pressure into fluids velocity. A diffuser works the other way around, it converts velocity into pressure and a tapered chimney is such a diffuser.

    As a consequence diffuser data sheets are a proper means to show how a tapered chimney performs as part of a front end. In my thesis I used the graph that Fox showed in his textbook “Fluid Dynamics”. However there is also a text on diffusers on the web from the “Applied Fluid Dynamics Handbook” by Blevins. (Google: “Diffuser data book”) It discusses diffuser performance including the possible peakflow in the entrance. Fig. 7.12b shows the measured performance of diffusers with graphs of the pressure recovery coefficient Cp which is the measured pressure increase divided by the velocity pressure at the entrance of the diffuser. To avoid copyright problems I have measured the graph and redrawn the behavior of 1:7 and Ell’s 1:14 tapered diffusers, 8 and 4 degrees included taper respectively. As can be seen the graphs show fairly linear behavior when plotted against the logarithmic scale of the diffuser length increase.

    What can be learned from the graphs is that for the included angle of 4 degrees, for a L/R of 4 (L/D=2) the Pressure recovery coefficient Cp is 0.35 and increases by 28% to 0.45 if the chimney receives a length-radius ratio of 6 (L/D=3). So even small length increases pay off.
    Singh plot.jpg Diff data plot.jpg
     
  13. JJG Koopmans

    JJG Koopmans Member

    Joined:
    Nov 12, 2014
    Messages:
    268
    Likes Received:
    232
    Gender:
    Male
    Heritage Railway Volunteer:
    No I do not currently volunteer
    The case of the multiple orifice

    Already in 1863 Nozo and Geoffroy published their results with a model with 8 chimneys and compared these with results from a single chimney showing quite close results. They verified this with a test on a locomotive with 4 chimneys and found little difference with its single orifice sister locomotives. Even a system with 36 orifices and chimneys was tested.

    In the Bulletin of the International Railway Congress of 1900 it was reported that in Russia 2- and 4-orifice blastcaps were used.

    However it appears that the earliest use of identical double chimneys was reported by Legein of Belgium, closely followed by de Gruyter in the Dutch East-Indies. The Frenchman Chapelon developed his Kylchap unit and Lemaitre proposed his exhaust for the French Nord. As a result double chimneys, Kylchaps and Bulleid-Lemaitre units were used in the U.K. before WW2.

    In here we devote ourselves to the theoretical background of more orifices. Already in 1913 Buckingham wrote an article “On Physically Similar Systems” defining aspect of similarity. In his text he never used words like “model” or “scale”, he looked at systems. His work forms the solid basis of all model tests with airplanes, cars, trains, ships and the like. Basically he states that similar systems show similar performance related to the change of the variables. Translated to our problem it states that two proper scaled chimneys each take half of the performance of the original. However, since on locomotives the scaled chimneys are lengthened to the height allowed by the loading gauge, a double chimney is a scale model of an original with its height increased by a factor of square root of (2),1.41 or 141%. That this is a proper way of looking was demonstrated by Young (see the earlier reference). He tested a model chimney with a doubled length. A standard length chimney with 4 orifices showed identical performance, see pages 77, 81 and 104 . Young failed to give a proper explanation and one of the enigmas in front-end research for me is that this test was clearly missed by those concerned with later problems. Ell was aware of this study as he mentioned it in a discussion with Tuplin.

    The earlier discussed diffuser data sheet comes in handy now for the necessary quantification. Let us imagine that we have a poorly performing front-end with a chimney length ratio of 1.75(3.5). Its extrapolated pressure recovery coefficient Cp would be about 0.32. With a double chimney the L/D ratio would become 2.475 (L/R 4.95) and the Cp would be 0.39, 22% more. This however could be too much because the length ratio should be lifted into a comfortable range of 2 to 2.2. As a consequence all the diameters, including that of the orifice, could be increased. A 10% increase gives a decrease of the length ratio to just over 2.2 and this is precisely what Ell did in general with double chimneys like the those of the GWR.

    The principle could be carried further if double chimneys with each 4 orifices, like the Kylchap, were applied. The length ratio would rise to 1.75 *square root of (8) or 4.95 (L/R 9,9) and the pressure recovery coefficient would be 0.55, so it should be quite clear that the doubling of the orifice area of the LMS Pacifics compared to the original single orifice has substantial theoretical and experimental backing. The length-diameter ratio would be decreased by 1.41 and the pressure recovery coefficient would drop to 0.47, still very superior to the starting value of 0.32. Probably this might even be lowered further, but the decreasing chimney exit velocity might give trouble with the headwind.

    In the thesis the published values of the area increases were analyzed and a practical formula defined that appears to cover the possible increase: third root of the number of orifices. For two orifices this would be 1.26 or a 12% diameter increase. For the eight orifices the area increase would be twice and the diameter increase 1.41 as already demonstrated.

    It should be realized that the multiple front-end should be a properly scaled version, the increased orifice distances as found in the LNER Class 5s and Pacifics are prime examples of incorrect scaling, resulting in poor steaming at lower velocities. Scaled properly the surface area is not increased, contrary to what authors like Giesl and Nock attribute to it. In the case of 6023, the surface area at present is even smaller than the single orifice situation of early 2014.

    As for the application of four orifices in a single chimney, that chimney is the result of four scaled chimneys of which the superfluous metal is removed and the outer wall retained. Since the orifices of the four chimneys of course point to their respective chimney axes there is no reason why the removal of that superfluous metal suddenly needs an excessive inclination of the orifices.

    It should be realised that the chimney exit area from a double length chimney, as applied to a four-orifice front-end, results in a scaled chimney that has 1:7 taper. The GWR King class is a prime example where the most original chimney could have been used by Ell without the cost of the conversion to double chimneys. Even now there are a number of locomotives that could be adapted without any problem.
     
    Tim L, jma1009 and 3855 like this.
  14. JJG Koopmans

    JJG Koopmans Member

    Joined:
    Nov 12, 2014
    Messages:
    268
    Likes Received:
    232
    Gender:
    Male
    Heritage Railway Volunteer:
    No I do not currently volunteer
    The front end calculations

    Already in 1864 the German Zeuner published his formula for front-end calculations. It stated that the pressure difference between the smokebox and the outside atmosphere multiplied by the cross-area of the chimney, resulted in a force which was equal to the difference in momentum entering the chimney and the remaining momentum from the chimney exit. The momentum is the product of the gas mass times its velocity.

    Given the long parallel chimneys of the 1860’s it appeared fine for the days. However only a few years later tapered chimneys started to be used as these showed a better performance.

    The simple momentum equation appeared not to be valid anymore as calculations gave different results from the measured vacua. Next generations of well-known locomotive engineers like Strahl, Legein, Chapelon, Giesl-Gieslingen and Porta bent over backwards to supply proper theory. Their formulae proliferated vastly, but it appeared that all of them apparently felt they had little to publish as a valid proof, it just isn’t there. Of course they all developed working front-end systems, but this was mainly based on practical development towards a acceptable performance level.

    The first problem with a calculation of a tapered chimney system is the determination of the area to be used. In a tapered chimney part of the pressure is on the chimney wall and has to be taken care of. The second problem is the exit momentum. Measurements have shown that the real velocity distribution from the chimney exit differs from the ideal uniform one used in the calculations. Also the temperature and the density appear not to be uniform at that location.

    The third problem is a theoretical one. All the formulae from the designers above include the momentum of the secondary flow, that of the combustion products. I myself have always felt very uncomfortable with this approach. According to fluid dynamics theory momentum of flows have to be included whenever it flows from outside into the control volume under consideration. So if a control volume is used with a orifice at the chimney entrance and taken from there, the momentum of the secondary flow should be included. However, the choice of a control volume is one of convenience, so any other is also valid, like one extended to the bottom of the smokebox allowing only horizontal momentum. However the physical reality is always identical, independent of any chosen control volume. So the problem is now to develop a momentum calculation that is proper and good enough to show whether secondary momentum inclusion is justified or not.

    Fortunately, I have received from Dr David Pawson the test results of some 550 Rugby test with different locomotives and – classes. Since most of the other variables are now known from the test under consideration the question remains which area and which exit velocity should be used in the momentum equation evaluation.

    In the thesis (p. 462) Mr R. Saunders developed a formula for the area definition in the case of a tapered chimney. This is based on an ideal diffuser which has its pressure recovery coefficient Cp defined as a function of the entry- and exit areas squared. However, if we use measured pressure recovery coefficients from diffuser data the formula now represents the actually “working” area ratios and we could reverse the formula into one where the useful area is a function of the measured pressure coefficient. This approach was used in the renewed calculations.

    The next problem is the exit momentum. Realizing that both the ideal and the practical diffuser have identical entry conditions, the final determination of the diffuser performance is only due to the exit conditions. Since expressing the exit momentum in entry momentum terms, it appears that any pressure recovery coefficient would be hidden in de denominator of the exit momentum expression. If an efficiency would be defined as the ratio of the measured pressure coefficient to that of the ideal diffuser, the exit momentum value of any calculation needs only to be multiplied by the inverse of that efficiency. Basically it replaces the value of the ideal for that of the measured diffuser. The effect would be that the exit momentum increases, exactly what would be expected from a less than ideal system.

    I hope that the text explains it properly, for those with more affection to mathematical formulae the text in the SLS Journal of sept/oct 2017 is more suitable. It contains the explanations, formulae, calculations as proof and graphs. Attached is the resulting graph from the BR5 with 4.875 in. orifice. Graphs for the BR9 with double and Giesl chimney have already appeared in post #184 of “Draughting arrangements for Bulleid Pacifics including Giesl exhausts”.

    The calculations have been made without any inclusion of secondary flow. Since the results show a 1 to 4% difference with the test results only, the conclusion would be that this theoretical approach is correct for now. So all momentum equations from Strahl to Porta appear flawed.

    Anyone interested in falsifying this approach is more than welcome, it looks like a paradigm shift since a KISS adaptation of Zeuners formula performs best.
    upload_2017-10-25_15-31-35.png
     
    Herald, jma1009, Avonside1563 and 2 others like this.
  15. JJG Koopmans

    JJG Koopmans Member

    Joined:
    Nov 12, 2014
    Messages:
    268
    Likes Received:
    232
    Gender:
    Male
    Heritage Railway Volunteer:
    No I do not currently volunteer
    Most of this is covered in my thesis “The fire burns much better..” available from Camden.
    Back in 2005 proof was mainly acquired from sources covering exhaust research.
    However since people found some of it controversial, it contradicts their recycled perceptions,
    opinions and fairy tales, I have added this time additional proof which is nowadays freely available on the internet.
    Fluid Dynamics departments of different universities worldwide publish useful research when
    they compare classical tests with their Computational Fluid Dynamics calculations.
    Happy reading, looking for cover,
    kind regards
    Jos Koopmans
     
  16. Courier

    Courier New Member

    Joined:
    Dec 1, 2010
    Messages:
    176
    Likes Received:
    91
    Jos - many thanks for taking the trouble to share all of that. Two questions:

    - Could you explain the comment below a bit more, perhaps quoting the blastpipe orifice diameters for the different scenarios:
    - Is there a theoretical way to choose the right b/pipe diameter for a particular locomotive - or would it always be trial and error? If you had a target mass flow and were able to identify a target smokebox vacuum (which would depend on boiler tube resistance, grate resistance etc) that the front end would have to pump against - I presume you can predict the right b/pipe diameter (ie a particular front end could be characterised with a flow/pressure curve like a water pump)

    (presume the tricky part would be calculating the appropriate smokebox vacuum that will generate the right airflow for the that loco, that coal etc (ie thinking of how grate resistance will vary with coal type - large lumps/many air gaps, small lumps and dust/few air gaps)
     
    jma1009 and ragl like this.
  17. JJG Koopmans

    JJG Koopmans Member

    Joined:
    Nov 12, 2014
    Messages:
    268
    Likes Received:
    232
    Gender:
    Male
    Heritage Railway Volunteer:
    No I do not currently volunteer
    The most original chimney of the Kings had a throat of 16 inches which was reduced to 15 in. by the single system revision by Ell. Since it is wider and has a 1:7 taper it has more characteristics of a scaled version than the present chimney (15 in. and 1:14). As for the way to choose a blastcap diameter, that was statistics from realized systems. Ell produced a formula based on heating surfaces, probably from a plot of the data available to him (BTW I wish I had those!!), see the Encyclopedia mentioned or his 1953 lecture. From there the chimney diameter followed and so on. Nowadays we can calculate a little further using a pc. Note that in the days of hand calculations, and slide rules, calculations took a lot of time and trial and error was cheap, safe and quick.
    If the theory is proper, one can calculate a number of scenarios which from the given steam mass flow supply the possible vacua nowadays.
    The calculation of the vacuum that is needed for the coalbed and the tubing is complicated, I never ventured beyond the tubes and flues resistance. Others have tried and I have not seen the results.
    If I had to cope with a new design, I would test the locomotive with a makeshift single orifice chimney to establish the possibilities. From there on the final version which is an expensive casting in our time could be designed.
    I hope this is an explanation,
    Kind regards
    Jos Koopmans
     
    jma1009, ragl, nickt and 1 other person like this.
  18. Courier

    Courier New Member

    Joined:
    Dec 1, 2010
    Messages:
    176
    Likes Received:
    91
    Jos

    Do the cowls in a Kylchap have any benefit?
     
  19. JJG Koopmans

    JJG Koopmans Member

    Joined:
    Nov 12, 2014
    Messages:
    268
    Likes Received:
    232
    Gender:
    Male
    Heritage Railway Volunteer:
    No I do not currently volunteer
    Frankly, I just don't know. The problem is that all kinds and sizes of cowls have been tested by Goss back in 1902.
    Literally none of these systems with cowls outperformed a standard system in Goss' tests, so I am wondering why
    they kept on emerging in the years afterwards. I prefer to base my judgment on proof and as long as a Kylchap is
    not investigated I don't have the data.
    Kind regards
    Jos Koopmans
     
    Last edited: Oct 31, 2017
  20. class8mikado

    class8mikado Well-Known Member

    Joined:
    Jun 1, 2009
    Messages:
    2,099
    Likes Received:
    394
    Occupation:
    Print Estimator/ Font of all knowledge (useful, or
    Location:
    Bingley W.Yorks.
    From an observational point of view; certain locomotives that have happened upon good steaming characteristics without recourse to cowls or multiple blast pipes. These locomotives don't tend to be the ones that get re- equipped and so comparative tests are almost non existent. But since Kylchap, Lempor, Lemaitre and Geisl have all been shown to improve performance but only the Kylchap has cowls if the answer is 'yes' then its a 'yes but not by much'
     

Share This Page