Recent Changes - Search:

G3 Wiki Read First

* HomePage

* What is Gauge 3?

How To Contribute Content

Contact Administrator


* Building the LMS Stanier Black 5

* Build GWR dia. AA11 "Toad" from a kit

* Machining Steel Wheels

Thermo 201

Thermodynamics 201

Hopefully you will have read the 101 introduction to this subject. The basic three line model is shown again.

Energy Input.

Work Out. →

Waste Energy.


Depending on which fuel you burn the amount of energy is dependant on two main things. The calorific content of the fuel and the efficiency of combustion.

By this I mean what you burn and how you burn it.

Solid fuels are by their nature inefficient combustion sources. The burning only takes place at the exterior surface of the fuel plus air has to reach it and contact the surface. The finer the piece of fuel burning the more efficient the combustion -but the faster it burns. This means that the combustion layer of fuel should be “light and bright” spread across the fire grate. The heat energy from this burning is in the form of hot gases which heat the water via fire tubes with some conduction in the fire box sides.

Liquid fuels are more efficient as the flame can be easily fed from a source be it liquid meths wicks or sprayed oil. The fuel mixes well with the air and burns in a predictable manner. In the case of a meths burner there should be several small flames not one large flame to ensure a good air supply to the burners. Due to the poor calorific content of meths, direct conduction for heat energy to the water is desirable rather than using the hot fumes from the burners. This is why water tube boilers suit meths burners.

There are also the “hybrid” gas burners which sit in the half way between liquid and gas. The vapour burner which uses boiling meths vapour fed from a side burner, and the original Primus type which used paraffin in the same manner as blow lamps. These are both very efficient.

Gaseous fuels are the most efficient. The flame is clean neat and precise. It can be deployed in any manner of burners towards useful parts of the boiler. The following information is empirical and has never been proven, however it would seem that ceramic burners work better with fire tube boilers whilst bunsen type burners are better with water tubes.

All combustion systems require air… HOW the air is presented to the fuel alters the rate at which combustion energy can be extracted from it. The school bunsen burner gives a good example. If the hole at the base was closed the flame was a silent flickering yellow. If the air hole was part way open then the flame was a hissing bluish steady shape, with the air hole fully open the bunsen flame split into the two cones, the outer blue regions the inner blue cone and it roared. The amount of gas burned in all three stages is the same, the amount of energy we obtain increases with the amount of additional air… This principle is used for the starting fans and blowers used. If we group the small bunsen flames then we can generate an aerospike in which external air is drawn along the flame and carried with it. This is additional to the convectional effect of the hot gases.

What would happen if we had a coaxial tube to the burner and pulled air down the outside of the bunsen tube with a fan and then pushed it into the air and gas stream? The flame now burns very hot, the flame may elevate above the end of the burner tube and then burn in free space. This is then called an Elevated Flame or Muffle Burner and is a commercial heating method. If similarly we take air from the top of the fire box and direct it into the underside of the firebox to be burned by the fuel we gain a thermal advantage. This simple effect is easy to calculate.

For every 10°C increase in temperature, you double rate of the reaction, thus you double the output energy.

If you blast your fire with hot air -it gets hotter!


The primary source of Work in a steam locomotive are the cylinders and pistons. In a conventional full size loco the amount of steam that is admitted to the cylinders is NOT that required to fill them. The amount of steam is “cut off” to a rough 80% of the volume and this then expands to fill the cylinder. Doing this the steam cools and gives up the energy it has acquired to the piston.

Thermodynamically -is this the correct practice?

The answer -is no (!)

As the steam cools it takes energy from its surroundings -the cylinder walls and the piston. Plus there is the pumping loss due it forcing out the steam from the other side of the piston (in a double acting system). The exhausting steam on the other side of the piston is expanding and cooling as well…

So what is the solution?

Surprisingly the answer is that you admit MORE steam. If the valve system continuously admits steam until just before it hits full admission then the steam never cools during admission. The exhaust stroke has to be modified as well, rather than a simple sinusoidal slide open and then slide closed, it has to follow this rather odd-sounding pattern.

Just before the steam inlet valve closes the steam exhaust valve opens -just a bit. This creates a blast suction wave down the exhaust track and then the exhaust valve opens fully. Just before the steam inlet valve opens it closes to a crack THEN it fully closes. Needless to say this method is limited to very advanced valve gears such as the Cossart, Reidinger, and Caprotti types.

I very much doubt that there will be a G3 loco with any of these valve systems as a true working system -but what can we put to use? The square cube law states that as we halve the size of a cube the surface area to the volume ratio is lower than the original.


A 20cm cube has a volume of 8000cm3 with a surface area of 2400cm2 = 3.33:1
A 10cm cube has a volume of 1000cm3 with a surface area of 600cm2 = 1.66:1

Thus small objects lose heat faster than large objects. If we employ this idea then using a large cylinder and piston is more thermally efficient than using a small one. High pressure steam is more efficient than low pressure steam -but the generation of it requires stronger and heavier equipment. Superheated steam is a gas not a vapour, but can lose this additional energy to become a vapour without changing its temperature… Thus is does not condense into droplets. If the vapour is re-heated to a gas then this same energy can be used again. Champelon used this twice in his three stage compound. Large bore pipes retain more heat than small bore ones thus the “breathing” of the locomotive is improved by both having an ample supply of steam and it retaining its heat.


This is a work through of some of the principles espoused by the three designers shown in the 101 document. What are the major causes of Waste Energy? The principle one is friction. This can manifest itself as rolling resistance of the loco and carriages or frictional loses through gear chains. One of the lesser thought of conditions is the act of cornering. Here the wheel rides up on the coning and (in theory) this acts like the differential on a rear axle of a motor car. This is most often not the case and the wheel is “dragged” through the curve. This completely ignores the bearings on the wheel journals be they plain metal or ball races. Another source of frictional loss is strangely the wrong oil used in the bearings. Most people will use 20W/50 motor oil for their journal bearings but this is actually incorrect. The oil will never reach the required temperature to flow smoothly and will always act as a brake on the journal. A synthetic 10 grade oil would be better.

Da Porta found that changing the oil in the bearings of his ore trains from 1000 to 750 grade reduced “fuel”, (i.e. coal dust, sawdust and llama dung briquettes bound together with tar), consumption by 5%…

Frictional Losses

In a steam loco the principle frictional losses are the piston and valve gear system. In the mid part of the 20th Century the plain bushes of several US locomotives were replaced with roller races to very good effect. The question that everyone knows the answer to, is that roller races are more efficient that plain metal bushes. But the amount may surprise you. Given the same constant amount of force there is no difference between a wagon with plain metal bushes and one with roller races. IF however the amount of force varies, such as each piston stroke, THEN the roller wins -as it has no oil layer to be “squished” causing friction. Golsdorf experimented with boiler pressure fed oil to the main bearings and found much to his horror -that this was still not enough! His solution was to fit a high pressure oil pump and drill large oil ways everywhere.

Work Losses.

My steam loco has three main sources of parasitic waste energy. These are; the axle pump, the oil pump and the steam injector. The axle pump when not in use pumping water into the boiler is simply pumping water around the loop -thus absorbing energy and doing no useful work for it. The oil pump works only when the locomotive is moving but always pumps oil. The injector only works best with cold water which it heats thus expanding it into the boiler. Of the three of them, the injector is obviously the worst criminal thief of the locomotive’s energy, as it uses steam which has taken energy from the fuel to make.

The tender pump does not steal any energy from the loco -but only from the person pumping it!

Heat Losses.

The main problem with a steam loco is heat loss. This is normally lost through the sides of the boiler and the firebox. It is amazing how much the correct type of insulation helps with the conservation of the heat energy of a boiler. Da Porta experimented with 90 gallon oil drums and lagged them with various things before filling them with boiling water and waiting for them to cool. Despite all the “high tech” lagging that he tried -the winner was always wool. The problem with wool is that it burns -witness the many “Merchant Navy” and “West Country” aerostreamed casings with ripples from the burning of their wool insulation. (When reading accounts of da Porta at work at Rio Turbio, I often feel sorry for his fellow work mates…) After burning several sheep fleeces he found that the perfect fire proofing compound was either “Alum” KAl(SO4)2 or “Boracic Acid” H3BO3. Neither of which compromised the insulation properties of the wool fleeces and removed the combustion problem of wool due to its natural lanolin content. He also discovered the “multi layer effect”. This is where several thin layers of insulation actually work better than a single thick layer. The thick layers allow more conduction across their thickness and the hit and miss contacts of the thin sheets prevented this -thus the heat loss is along the length of the layer not across it.

Da Porta also perfected the use of steam lagging. As the loco moves through the air the wind cools exposed parts -such as cylinders. His approach was not to blast steam through them to heat them up -but never let them get cold in the first place! At night “smoulder pots” would be placed under the cylinders which would heat them until morning. The cylinders were heavily lagged except for where the plumes of the smoulder pots would keep them hot.

Your loco may, or may not, have snifter valves… These are very useful but placing them in the wrong part of the system can remove useful energy. I have followed the approaches of Champelon and placed my TWO snifter valves in the stated “correct” places for them to be. There should be a snifter at the wet heater of the superheater system. Thus when it opens, cold air is admitted to the coldest part of the system and conduction will soon heat the air up to temperature. Similarly the place for the boiler snifter is directly above the steam intake from the boiler. In both cases the bulk of the system will remain hot.

Feed water is a problem. In 101 we looked at the feed water heater as a means of scavenging heat energy from the exhaust gases. The traditional place to have the clacks deliver water to the boiler is at the side via a pipe that dips below the surface. The belief is that the cold water heats up and doesn’t disturb the mass of the hot boiler water. What happens is that it interferes with the circulatory movement of the water in the boiler and simply builds up as a cold eddy spot. This means that the boiler is not as efficient as it should be. The best place to have the water injected is actually into the steam space above the water in the boiler. This will condense some of the steam but the amount of surface movement causes rapid mixing of the cooler water without disturbing the heat flow around the boiler.

An Applied Thermodynamics Example…

The Irish State Railways (CIE) had a problem with the type of fuel that they were burning. This was peat not coal. The experimental engine they “cobbled together” at Inchecore was CIE 356. This started out as a conventional 2-6-0 tender locomotive. Pulverised peat was screw fed into the combustion chamber and air blasted into the powder spray, thus the mix burned at the best possible rate. The air blast was provided by a fan driven by a Leyland bus engine on a flat bed pulled behind the tender. The feed water was heated by a pair “Franco Crosti” style pre-heaters. Amazingly it all worked perfectly...

Edit - History - Print - Recent Changes - Search
Page last modified on April 30, 2018, at 09:03 AM