Redraughting
locomotives
N.A.H. Day
Over the last 14 years I have redraughted many steam locomotives. I am not aware of any other person who has done as many locomotives in the world currently. The objective here is to clarify some of the reasons for doing what I have and some of the conclusions from this work.
The major reason for converting the first locomotives on Snowdon Mountain Railway (SMR) was to enable them to keep up with the then new diesels. This was achieved after a number of years work. There was little or no reference to text books on design and thus much was learnt before the system I now use was developed.
There are two basic principles involved in the draughting of steam locomotives:
Both the above are relevant to advanced draughting systems but only the first is valid in traditional systems.
Fundamentally it is almost impossible to put high mathematics to draughting except as an approximation. This is simply because the whole of a draughting system is turbulent flow. Such flows are totally random and thus impossible to model mathematically. No better collaborating evidence of this can be found than Porta's paper.
There are now, with the advent of computer based data loggers, two ways of recording data generated. It is wrong to compare tests made by the traditional equipment against that of the data loggers. Until a direct duplicate test is made on an engine such comparisons are misleading.
Comments on Lempors
The principles of the Lempor have been well documented and thus there is no point in me repeating them.
I have made my nozzles by a machined fabrication. This has worked well and many of the nozzles have little or no signs of wear. I did consider making nozzles by the method of sheet metal forming but the fabrication is just as effective and long lasting.
Comments have been made about the limited gas flow around the base of the nozzles. True, everything is a compromise.
The fundamental forms and proportions as described by Porta and Wardale are correct. The one requirement which I misunderstood and know is true is the angle of inclination of the nozzles. This has to be the 7.25 degree angle. It is what causes the boundary layers to work. Failure to comply with this will only result in a traditional level of draught. There is an optimum size of nozzle which is around 2 inches diameter. It is only when this diameter is exceeded is an increase in their number justified. The reality is that the flue height dictates that a double arrangement occurs before the number of nozzles being increased is reached. To have an additional vertically placed nozzle in the centre is of no benefit as it is the boundary layer that is of importance with the Lempor.
Differing fuels require different chimneys, and more specifically flue, sizes. This can be illustrated by the fact I had to change the mixing chamber dimensions on SMR No.4 when converted to oil. The main reason being the effects of the atomising steam. It is possible to alter the heat absorption characteristics for a boiler by altering the proportions of the draughting system.
The results I have had are dramatic. In some instances the fuel consumption has dropped by 50% and back pressures by up to 40psi. Smoke and spark emissions are also dramatically reduced. To give more detailed comments would be misleading and open up misinterpretation of the results which has been know to happen.
Would I build other types of chimney? Yes if there was one known to be better but there is nothing proven to be better than a Lempor. There are other forms to be fully investigated BUT they are not public knowledge until they are fully evaluated. Some claims made for other forms of draughting are dubious and only so far justified by different forms of testing which do not relate to the accepted methods previously used.
Why do this and not build new locomotives? Simple, the cost of building new engines is high. It is better to do something rather than nothing and let steam die.
July 2004