Improve that exhaust!

J.J.G. Koopmans Venray The Netherlands   14/04/2001

Model steam locomotives often use a blastcap orifice, with a diameter that is determined by the points
of view of Greenly, as used by LBSC or Martin Evans,  which is 1/7 of the cylinder diameter or determined by the cylinder volume. The diameter of the chimney is determined by cones with a 1:3 ratio at the bottom and 1:6 at the top of the chimney.
In general these dimensions are too large, but they appear to work properly, so that the model engineer is quickly satisfied.

These dimensions can be criticised however.
A fixed ratio exists between the amount of steam evaporated and the amount of smoke-gas that is
produced by burning the necessary fuel. Even if only half the heat released from burning the fuel is used to convert the boiler water into steam, about 4 kilo of steam is being evaporated by one kilo of coal at the pressures used in model engineering. To burn that kilo of coal, theoretically about 11,5 kilo of air is needed, which produces a total of 12,5 kilo of smoke-gas while burning. To prevent the formation of
carbon-monoxide, excess air is needed which is kept  around a minimum of 20%, giving 15 kilo of
smoke-gas as the final result. The earlier mentioned 4 kilo's of steam is added so that 19 kilo's of mixture are ejected out of the chimney. This ratio is more or less independent of the size of the locomotive.

To eject this 19 kilo the jet from the orifice, containing 4 kilo of steam is being used. The momentum, the product of the velocity and the mass of the jet, is transferred to the amount of smoke-gas allowing the
mixture to flow from the chimney. It must be clear that 4 kilo of steam can move 19 kilo of mixture,
however with about 4/19 of the velocity of the steam from the orifice. The 19 kilo mixture passes the chimney with 4/19 of the velocity of the steam-jet and it will be clear that the size of the chimney must
allow for this amount at a lower velocity.

As such we have defined a ratio of the chimney diameter to that of the orifice obviously having the same ratio as that of the steam to the mixture mass, 4:19, nearly 1:5. In real size this ratio is about 1:3 which is not a large difference.

In practice one finds a lot of deviation from this ratio.
For instance the drawing of LBSC's 5" gauge 0-6-0 "Speedy" shows an orifice diameter of 6,35 mm and a chimney throat of 34,84 mm, a ratio of 1:5,5. With this ratio,  the chimney is so large that far more air than the necessary 20% excess is sucked through the boiler, all of which must be heated and ejected from the chimney. Even worse examples can be found on models of contractor engines.

Regarding model chimneys as too large is not a novel point of view. The Model Engineer of July 1989 contains an article about this subject completed with changes.
Among others, a badly steaming  "Speedy " was adapted to receive a 6,73 mm orifice with a narrowed chimney of 26,2 mm, a ratio  of 1:3,89 and as such closer to real life proportions.

If you are interested in a more economical locomotive, that uses a lower blast to push the used steam through the orifice, do consider some changes.

However we have be aware of the various possibilities. If you have a locomotive with an exhaust orifice having the same diameter as the blastpipe, and a correctly sized chimney, there is little room for
improvement unless this locomotive is short of steaming capabilities.
In general we should aim at using the full area of the blast pipe in preference to reducing the orifice
diameter. This can be done in two ways. If the chimney diameter is too large it can be reduced with an
inserted sleeve, the diameters of orifice and the chimney diameter must remain between the 1:3 to 1:5
ratio, and the smaller the better. However a very elegant solution is to change towards a multiple exhaust. A chimney that is too wide is ideal for this.


First a little theory. If the single, well performing, chimney is replaced by two chimneys, each with half the original throat area, and the two orifices also reduced to half the original size, then the scale of both equals 1/2Ö 2, or about 71% of the former situation. This scale is valid for all linear dimensions, both orifices should also be arranged to be closer to the chimney throats, at 71% of the earlier distance. However the piston does not "feel" any change,  the total area of the orifices remained the same and the steam flows with the same  velocity as in the single system, but now through the double chimneys! We have created a situation where in a 71% model steam flows at 100% of the original velocity. This new combination will perform better because the time taken by both parallel chimneys to entrain the smoke gas is less and the smoke gas has to be resupplied faster, producing an improved vacuum. Because, in this mental exercise,  we are only interested in regaining the original vacuum value we must adapt the velocity of the steam jets to the scale of the chimney. The entrapment process should take the same time as originally with a single orifice. The scale of the velocity should then be identical to the scale of the chimney systems.
This happens when the scale of the velocity * the scale of the diameter * the scale of the diameter is equal to ½  because of the calculation of the volume of the steam flowing through each orifice.  We have as such scale ³ = ½  and the matching scale of both velocity and chimneys is   or 79,4%. Without any problem a single orifice of 10 mm can be replaced by  2 orifices of 7,94 mm each!
If you tend to regard this calculation with disbelief, the experimental verification is given by none other then S.O. Ell, supervisor of the BR locomotive tests in the early fifties. It can be found In his chapter on the tests in  "The concise encyclopaedia of world railway locomotives" Ed. P. Ransome-Wallis, Hutchinson of London, 1959, p. 395. He gives dimensions of both a single and a double chimney system as deduced from the tests showing a 78.45% diameter value of the double system compared to the single diameter.

There being no objection to a continuation of this way of reasoning, we can deduce the general equation for n chimneys as , so that  for instance  with 7 orifices, similar to a Giesl exhaust, so that the original 10 mm orifice can be replaced by 7 orifices of 5,2 mm each. This is a huge increase in area resulting in an equally large decrease in counter pressure on the piston during the removal of the steam through the
orifices.  The combination of exhausts+chimneys can be simplified, each orifice does not need its own chimney. Arranged in a proper pattern in a single chimney with matching total area the flow of each orifice acts as if it flowed in its own chimney.

If we were to use the revised layout of Speedy as an example, and retain the chimney, the total orifice area could be increased as follows:
1,6 times with 4 orifices of 4,24 mm , drill ##19, or 1,7 times with 5 orifices of 3,94 mm, drill 23, or 1,9 times with 7 orifices of 3,5 mm ,drill 29. Because the chimney area has to follow the increases, the area as proposed by LBSC can be used again.
The orifices have to be arranged in a proper pattern and the new blastcap looks like a showerhead. In the case of the 7 orifices, 6 on the circumference, one in the middle. The distance from the throat of the chimney will be 52% of the original. The centre of each orifice can be taken from a drawing in which in a circle with the diameter of the chimney 4, 5 or 7 circles are drawn touching each other and the circumference.

The locomotive will retain the same vacuum but the orifice area has been increased from 100 to 191%
in the case of the 7 orifice system. The power of the locomotive will increase.

This example can be used as an example for almost all model locomotives.

The drawings of Princess Marina and little Tich(!!) show the same pattern.
Almost the only restriction is the ratio of your blastpipe to the chimney diameter. So if your chimney
diameter is way over 3 times the orifice diameter and your orifice diameter is smaller than that of the blastpipe, improve that exhaust!