Locomotive Front End Designs

It would seem to me that the effect of the Kylchap cowl dividing the exhaust from the blast pipe and drawing in and entraining
some of the combustion gases in the lower smoke box is to get the effects of both a petticoat pipe and a multijet in the front end.

NB Obtaining an even draught all over the tube plate is beneficial: it was one of the advantages of the Adams annular blast pipe
which drew in from a lower level to inside the ring of the blast opening and the tubes in the Kylchap A4s kept distinctly
and beneficially cleaner through all of them.

 

My theory about the Kylchap, which I have absolutely no data to back up, is that the pressure difference between the blastpipe and atmosphere is divided by the three stages. Then from Bernoulli's theorem, the velocity of the steam jets at each nozzle would be much lower, and the kinetic energy lost during the mixing of the steam and combustion gases would therefore be lower, resulting overall in a lower back pressure on the cylinders.

By similar reasoning you could argue that it would be much more difficult for a Kylchap to reach sonic velocity in the first few milliseconds after the exhaust valve opens, compared with an exhaust with a single stage of entrainment.

 

Since the vacuum in the smokebox is uniform, measured for over 100 years now, there is not really a flow from the lower or
upper part of the smokebox. The Kylchap acts as a very long chimney and as long as there is no investigation about its
flow regime I myself would like to abstain from any fine detailing.
As for the Adams, its superiority was not because of its specific construction but simply because it generated two separate
jets.
Since Chapelon showed in his earliest article that the smokebox vacuum was tripled in his test locomotive, it is no wonder
that the tubes are cleaner in the case of the A4 where the same must have happened.

As for the three stages in a Kylchap, such an assumption does not explain why Goss measured poor performance in his tests
with intermediate cowls. Please note that Bernouilli is not applicable, a chimney with its temperature and density differences
is excluded from his theorem!
Kind regards
Jos Koopmans

 

I stand corrected.
I had thought that the Kylchap cowls were convergent, giving an increase in velocity and corresponding backpressure at each stage. After studying the diagrams in "The fire burns much better..." I realize they were parallel, and as such I agree, I don't really see the point in them.

 

Apart from the fact that they worked better than an orthodox double chimney.

 

8 jets (in a double Kylchap) instead of 2 would explain that.

 

Exactly!

 

The question is whether or not a double chimney Kylchap is superior to two four jet blastpipes exhausting into two conventional chimneys with no cowls - ie are the cowls useful or just an unnecessary hindrance to cleaning tubes etc.

That would be an interesting CFD study.

 

Townsend said they were a hindrance - so they didn't bother with cleaning the obstructed tubes

 

The Kylchaps do have the largest orifices around, however, I am always wondering whether they do no loose
a lot in the flow to the outside world. Needing elephant ears is also an item. Plain chimneys with more orifices
could do the trick were it not that these orifices have to be positioned higher in the smokebox, you win some, you loose some.
As for a CFD study, where is that eager student, please come forward! I have been trying to do something with
a CFD student version, and some years ago I acquired an axisymmetrical version which allowed 3D. Modelling
extra orifices then became the problem so I tried a ring orifice. Anyway it needs someone of a younger generation.
Note that CFD studies have to be verified by tests, could be models, I would like to suggest 1:1, who dares!
Kind regards
Jos Koopmans

 

Would 6023 be sufficiently representative of a Kylchap without cowls? If it is eventually possible to test it, it may be interesting to compare the data with Rugby test data for a Kylchap equipped locomotive, if suitable extrapolations can be made to compensate for any differences.

If anyone does a CFD study on a Kylchap, I'd still be curious to know what would happen if the cowls were made convergent instead of parallel. I'm also wondering if Chapelon tried this and found it didn't work!

 

On the contrary, it is the Question. To improve the steaming of ex-LNER V2s a few were fitted with plain double chimneys which made little or no difference. Later, a few more were fitted with double Kylchap chimneys which did make a difference. For more on this see Peter Townend's book 'Top Shed', for it was he that was instrumental for not only this conversion, but also the A4s and A3s too.

 

There is a fundamental and little explanation of the difrence between any simple chimney weather it has one blast pipe nozzle or multiple nozzles like 6023 has now and any form of Kylchap, Kylpor, Lempor and some of the derivatives. This is where you give the steam ejected from the nozzle at a slight angle most notably around 7.25 degrees at a surface like the internals of a Kylchap cowling or the wall of a mixing chamber of aLempor.

The pressure increase decreases the viscosity of the boundary layers agenst the wall. The result is a highly fluid layer which allows the general mass of gasses to travel at a higher velocity. The result is a greater vacuum. The effect is the same as an ice skate on ice which melts the ice and the water provides the slippery boundary layer.

Much of what I see here on draughting is pure academic debate of historic documents which ignore this principle and a few others like delarvel nozzles, internal stream lining and the shaping of nozzle forms which even when the maths is done right will affect the overal performance of the draughting. There is no need to repeat here the well documented eveidence of the above when it's evidence is in the work of Wardale and Porta, which is freely avalibe.

 

Mr Day, you have been telling this fairy tale for 15 years or so. Prove it or stay out of a scientific discussion where proof is submitted instead of opinions.
Kind regards
Jos Koopmans

 

I did the experiment in 1992. Straight and angled. The difrence was twice the vacuum. It is you who dose no want to understand.

 

I am not aware of a Rugby test of a Kylchap equipped locomotive. Without it I am hesitant about comparing the present 6023 layout, which not measured but appears to perform, with a Kylchap.
Kind regards
Jos Koopmans

 

Is the experiment written up somewhere?

 

Not currently, but the information is going in my book all in good time. What I describe about the change in viscosity is a known scientific principle, another exsample is a hill side mud slide. No messing, no fairy tales. Porta's Kylpor, Lempor Paper has an opening statement to the effect that such a system is 100% more efcient than conventional systems and the Red Devil book by Wardale confirms this and experimental tests done in China about characteristics of nozzle position and defusers. My experiences with many locomotives confirms their statements. There writings put it all into words better than I currently can.

 

Heres an odd question. if part of the function of a Lempor is that the Blast nozzles are angled to enable a significant interaction with the mixing chamber walls , why for a 4 nozzle system is the bell -mouth of the diffuser round and not a Quatrefoil / round cornered square as this would be a more logical interface... the cross section could flow to circular at the junction with the diffuser cone if required... Worth an experiment perhaps ?

 

Yes there is always room to use practical experiments. People have tried this Wardale and Girglestone but I have seen no documented evidence of this. I have used truncated cones in a similar vain and these allow for more consistent construction standards.