(err yes....)

While there is a widely held belief amongst friends and family that I am "less than sane" I have been studying this problem for over 6 months now and even I have to question myself about this one.. There currently exist two functioning G/T locos at 5 inch gauge and one at 3 1/2 inch gauge. The two at 5 inch gauge are the BR GT3 and the GWR "Kerosene Castle" -the 3 1/2 inch model is of a Union Pacific GTEL. All of these use commercial gas turbines originally designed for model helicopter use by WREN turbines of Rotherham. So in practice there is at least a precedent for building a gas turbine locomotive.

The three G/T built for the UK railways were the the GT3, and the two for the GWR -the last one of which was converted to AC traction and used as a training vehicle in the Manchester area. The GT loco I am looking at was never completed -but it did run on the test bed at North British Locomotive

Information about this loco is at best, non-existent. I thought that the problems of finding information about the Golwe locomotive was difficult -this can only be classed as several orders of magnitude harder... No one seems to be sure about the configuration of the locomotive, some sources give it as: 1A1A -A1A1 or C-C and others as 1C0-C01. One source gives it as pure mechanical drive and another as an electric traction bogie. The drawings for this loco that I commissioned show it to have an "English Electric Type 4" bogie as would have been found on a Class 40.

Very little is known of the actual results since quite a lot of data was destroyed as "waste paper" when NBL went bust -but a model of the locomotive is thought to exist in one of the Glasgow Museums. What sets this loco apart from the other G/T locos was the choice of fuel...

This G/T burnt COAL.

The coal seems to have been burnt as a primary source of heat and then the hot exhaust gasses passed through a secondary intermixer which simply used pure air. The heated pure air then turned the rear turbine, there are reports that some of the exhausted air was re-circulated back to the coal furnace to provide draught for it.  The exhaust gasses from the "producer turbine" then turned a "power turbine" which ran the AC side of the loco. A similar system was used by Pratt and Whitney in their "Nuclear Turbine" for the USAAF "Atomic Bomber" programme during the 1950's. The turbine itself is somewhat ambiguous, as one source comments on coal dust pitting the exhaust blades, whilst another drawing I have shows an intercooler system in which the hot exhaust gases from the coal would never have come in contact with the blades -which would have been powered by hot pure air(?)

I normally work out EVERYTHING before I even pick up a screwdriver. I don't expect anyone to actually build one, (other than me) -but who knows...

AFAIK there are no drawings of this loco or pictures of it on the net -mine will probably be the first available!!! It is 70 feet long (94.5cm at G3 scale) and is built to the std BR loading gauge. It is not symmetrical and it only has a driving cab at one end. It was, I believe, to be a passenger hauler -rather than freight. The drawings I have are on six A2 sheets -two of which show the loco in profile (1/2 per sheet).

Everything so far is worked out as a collection of modules, (I am after all a designer!), so the collection of standard modules reads as follows:

The Power Bogie Modules -COMPLETE
The Chassis Module  -COMPLETE
The AC Generation Module -COMPLETE
The Power Control Module -COMPLETE
The R/C Module -COMPLETE
The Turbine -WORK IN PROGRESS....

All of the modules are applicable to Live Diesel and Electric powered locos. I don't in all honesty expect this to be even begun for about 18 months -by which time most of the problems inherent in the last part will be understood both Mathematically and Physically.

The following shot is of the GWR "Kerosene Castle" which I took myself at the 8th Barrow Hill Beer Festival. Please Note the slight blurring of the images is due to auto focus of the camera getting confused with the amount of coal smoke in the air -not the amount of liquid consumed by the user...

See Picture 1.

It is unfortunately a "stuffed" exhibit as the gas turbine was removed several years ago... I do not know whether the auxiliary Diesel motor and Dynamo still function(?) Anyway the model will be built from standard units that can be "out sourced" to other models. So what I am hoping is that although there will be one gas turbine model at the end of the process -there will be other locos that have been built using the same modules. This produces the interesting concept of "MU ing" models together with a version of "BLUE STAR" working...

What I had in mind was actually building my own turbine... Not of the impulse reaction type -but of the Tesla Disc type. The maths for this is far easier to work out than that for a reaction turbine. The problem being that the exhaust side of the turbine will have to be made of alloys, I presume some type of stainless steel. I had supposed that a 300 series Stainless steel, (304 or 316), would be my first choices. My little lathe and mill only have 150Watt motors so I had worked out that it could possibly take 400 series steels as a maximum...

I am fortunate to live in a city where there are plenty of cut up and sectioned gas turbines for display in the local Industrial Museum...

My "Bible" that I am using in designing my Tesla is "The TESLA Disc Turbine" by Cairns. The design there at the back of the book uses Aluminium and is designed for Air and Steam use. I have seen a couple built and running. My son has one made for him by me that uses tap water and scrap CDRs.

I am working on 8cm discs as a maximum within my "loading gauge" and I plan to use Pure Butane for the fuel with three combustion chambers.

Well I am a little further along the path and as expected I have hit a confusing area... Now the Tesla turbine works on the friction between the disc and the "fluid" between them to compress / expand to completion. I had expected the distance required for the compressor discs to be different from those of the exhaust discs -due to the differing temperature and pressure of the exhaust gases. This has proved to be correct and the formulae that I am using are "interesting" to say the least.

However this is not the main problem.

What has been the problem are the combustion chambers. I have decided now to go to FOUR combustion chambers and to use "doped input" for my exhaust gases. This means that I will spray a small amount of water into the combustion chamber to increase the amount of exhaust gases. This lowers the exhaust temperature -but gives me far more exhaust gases to use. Now the problem is to design a shape that slows the incoming air, allows it to burn, and then allows the hot steam/air mix to smoothly flow into the volute shape that is the exhaust output to the exhaust turbine...

All of this would be fairly easy if the thing ran at a constant temperature and (of course) the "Universal Gas Equation" is almost useless for the type of calculations required -other than to give me a "ball park figure" to see whether I have screwed up my maths along the way with the more "wilder" of my calculations.

What I have decided to do is simply GUESS and work from that. So, my intake temperature will be 400K with an exhaust temp of 700K. I still have no idea of which shape is the best for for a combustion chamber though!!!

Well after "some discussion" with people in the know about gas turbine locomotives -there are some(!) Most of of my formulae that I have evolved and derived have turned out to mostly correct. The worst thing about the "beer and pretzels" session was the guy who produced a book containing most of the formulae that I had very painfully had to work out from first principles over the past 6 months....

Using the Universal Gas Equation [p1 v1 / t1] = [p2 v2 / t2] gives me an input temperature, volume & pressure, and an exhaust temperature for my turbine, this gives [1 x 1 / 400] = [p2 v2 / 700] = 1.75 volumes of exhaust gas

The boundary layer for my intake turbine (i.e. half the distance between my turbine discs) can be found by the working out the Kinematic Velocity of the air at 400K and the exhaust gases at 700K

This is given by the formula PI x Square root of: (1,376 x Kinematic Viscosity @ K Kelvin / RPM) where the KV is in sq ft per second(?)

Which gives me a disc separation of: 0.33mm for the compressor discs and 0.38mm for the exhaust discs for a turbine running at 25,000 RPM.

The discs will be made from 25 sheets of 2mm thick 8cm sq. 316 stainless steel, with 10 for the compressor side and 15 for the exhaust side, the gaps simply made by lathing off the correct amount of metal from one face and then the stack of discs bolted together to the dumb bell shaft faces. As to how I am going to get this to start up I have pondered on and worked out a solution that does use my typical "left handed, inverted -and shoved in sideways" approach... Since the turbine will have to rotate at 25k RPM this means that a DC motor cannot be used as the commutators would be in a state of permanent arcing. So what we need is an AC motor with no moving contacts(!!!)  Ideally this means a 3 phase AC motor of the "outside runner" type, i.e. the field winding are stationary and a tube with the field magnets revolves around it due to the AC flux. Fortunately 3 phase ESCs are very common. This also has the added advantage that the "outside runner" motor can be then used as a 3 phase generator for the power bogies. I can feel the "itching" in me to start work on this, but I have a definite path through my models and now I know that when it comes time to build this, I will have all the parts in place.

See Picture 2.

Having consulted most of the sectioned engines on display and (when no one was looking!) poked a steel rule down various venturi and measured diameters I have come to the conclusion that a simple flame tube and expander system will suit me best. If the turbine looks like half of a "Welland" engine -there is a good reason for this....

A few days have passed and unbelievably I have some accurate data on the loco in question. I even have a low res JPEG of the model in the Glasgow Museum which at the moment I cannot show you for copyright reasons...

Working cycle

There were two separate, but linked, circuits - the combustion circuit and the turbine circuit.

1.Combustion circuit. Pulverized coal and air were mixed and burned in a combustion chamber and the hot gases passed to a heat exchanger where heat was transferred to the compressed air in the turbine circuit. After leaving the heat exchanger the combustion gases entered a boiler to generate steam for train heating.

2.Turbine circuit. Air entered the compressor and was compressed. The compressed air passed to the heat exchanger where it was heated by the combustion gases. The heated compressed air drove two turbines - one to drive the compressor and the other to power the locomotive. The turbine exhaust (which was hot air) then entered the combustion chamber to support the combustion.

Specification

The locomotive was never built but the specification was as follows:

▪Wheel arrangement: C-C, later changed to 1A1A-A1A1

▪Horsepower: 1,800, later reduced to 1,500

▪Weight: 117 tons, later increased to 150 tons

The projected output was:

▪Tractive effort,

▪30,000 lbf at 72 mph

▪45,000 lbf at 50 mph

▪Thermal efficiency,

▪10% at 1/10th load

▪16% at half load

▪19% at full load

The transmission was to be mechanical, via a two-speed gearbox, giving a high speed for passenger working and a lower speed for freight. The tractive effort figures, quoted above, look suspiciously high for the specified speeds. It seems more likely that the figures quoted are for starting tractive effort and maximum speed in high gear and low gear respectively.

However it does confirm several of the drawings that I have sourced for this loco and (of course!) wildly contradicts other drawings for it. The model shows a C0-C0 loco with typical Deltic or GG-1 “whiskers” and it is also a Dark Powder Blue colour.  The commissioned drawings say that the loco was Black. All my information prior to this has produced a 1C0-C01 or 1A1A-A1A1 configuration. The bodywork agrees with the drawings (mostly) and the position of the turbine and ducting also agree. The bogies are I think a; "we have a pair of these going spare -so we might as well use them" from the model builders dept and to be honest they look like they belong on a French ANSEA designed loco. Come to think of it the entire front end looks very 50's continental and an OBB, DB or SNCF badge would not look out of place on the front of it.

See Picture 3.

Ok, having examined the available plans, the model, and all the details, it would seem that I have bits of what I am going to call “A” “B” and “C” versions of the loco. The Glasgow model is pure version “A”, the technical specs above are a mix of “A” and “B”, the drawings I have of parts of, are MOSTLY version “C”. However I do have exact matches for parts of “A” and “B”... The obvious question is: “Do I have enough complete drawings of any of the versions to start work on a Working Drawing of this loco?” The sad answer to that is of course NO...

What I can do however is take the “common” information from all three versions and see what I can graft to fit together. The bodywork is common (bar the odd few holes) to all three versions. The turbine is unique to all three versions. The bogies are also unique to all three versions. The more I look at my commissioned drawings the more I feel like scrawling dimensions and potential parts numbers on them -which is normally for me a sure sign that the design is cooking within me.

Well after another hard weeks evenings work at the calculator, (spreadsheets are fine for large scale number crunching but you do not get a "feel" for the numbers), I have some idea as to what sort of shape I need for a flame tube. In a way I am lucky as while I was at college my burner that I used for my chemistry lab work was a Mills Vortex burner rather than the standard lab Bunsen burner So I was able to "experiment" with the various air holes and tube lengths to get the right flame for the beaker / flask that I was cooking at the time.

It works out that from my researches -the flame tube is the most critical part of the design.

There are three basic types, The Swirl, The Toroidal flow, and The Back flow. I have opted for the swirl design as I think this will be easiest for me to build, although not the most efficient -it does present the simplest maths....

The combustion is split into three parts. Air is mixed with the fuel from the centre of a small set of vanes, one third of the way down the length of the flame tube there are additional air holes to complete combustion and then the last third of the tube has lots of holes to admit cool air to the flame. It is this flame heated air that then expands into the exhaust turbine.

See Picture 4.

Another problem that reared its head and is currently laughing at me -is ignition. I need spark plugs that are small enough to fit within the flame tube but not large enough to intrude into the flame path to cause turbulence. My best bet is, (I think), a 10mm racing plug -as this will have 4 annular electrodes and no projecting central pin. The other alternative is the "Tazoon" system of a conductor with no insulation from the cool side of the  turbine. A fellow forum member suggested a simple cheat that might well work. I “Blow” a glow plug and then use a Tesla coil for ignition across the gap.

Due to the nature of “Life the Universe and Everything” the main sources of information to help me design this loco -have been other locos... I have found a rich seam of drawings for the GT3 including a draft of the gas turbine assembly.

See Picture 5.

You can see the “dog Bevel” arrangement of the reversing gear quite clearly and a set of dimensioned drawings.

See Picture 6.

Here is some basic information about the GT3...

“During English Electric's quest to advance rail traction the company designed and built a gas-turbine powered loco, No. GT3 between 1959-61. The physical appearance of the loco very closely resembled a 4-6-0 steam locomotive, the tender of which carried diesel fuel in place of coal.

Over the years other gas turbine locos were built, but GT3 was significant in having a mechanical transmission. GT3 emerged from English Electric's Vulcan Foundry in May 1961 painted in a red oxide livery, with a cast GT3 numberplate on the cab side, and an English Electric roundel replacing the BR lion and wheel logo on the tender. Once the loco was mobile from its builder's it was moved to Mary-le-Bone and displayed in the Institute of Locomotive Engineers exhibition. After it was accepted by BR for operation, GT3 was transferred to the Rugby testing station where stringent proving and comparison tests were carried out. This included taking the loco through its paces up to 97mph on the famous Rugby test rollers. After departing from Rugby the loco was deployed on further track testing which included gradient and brake evaluation, and during the course of these a number of problems were identified. This led to the loco operating trials' on the Leicester to Woodford Halse route of the former Great Central. In early 1962, GT3 was used for testing over the arduous Shap incline with trains of up to 16 coaches. These tests proved very successful and on one run a speed of 43mph was recorded at the summit. By the time GT3's testing was almost complete, the decision had been taken by the BTC to invest in diesel and electric traction and thus the interesting gas-turbine principle had no place in the modernisation of the railways (at that time). By late 1962 GT3 was returned to English Electric and many components were removed, with the chassis eventually being sold to Wards of Salford for scrap.

On a number of official English Electric papers of the period, GT3 was referred to as BR No. 19000 and on one artist's impression of the locomotive, dated August 1958 the name Lord of the Isles was added.”

Similarly I have on order from Amazon a book I saw on Ebay “Great Western Railway Gas Turbines” by Kevin Robertson. I bought it off a seller on Amazon -as it was cheaper that way.... (The nerve of some people!)

Well I got my book and I got rather a bargain thank you very much!!! Someone obviously had not looked at either the inside cover or the contents of the book, as inside was the signature of Kevin Robertson, (along with another signature and message), inside the book were two limited edition numbered and signed post cards of 18000 and 18100. I don’t know how much these would have added to the total cost of £10.07p, (inc P&P), but I would willingly have paid a little more for it if it had known, (and I bet the price would have been somewhat higher).... Am I going to buy books from this seller again? (OH YES!!!)

The books also had addendems on the other gas turbines -including mine. There in front of me was the first set of complete drawings for type “A” showing the C-C configuration with the central gearbox and the propshafts to both power bogies -along with the conrods between the axles... It was very interesting to read the history and background of these two locos and I would fully recommend it to any potential builder of these locos. What did surprise me was the fact that the main troubles of the first loco were not as I expected the turbine -but related to the combustion chamber and the train heating boiler. The info gleaned from the back pages has pointed me in the direction of another book that I will have to buy now! It is called “Locomotives that Never Were”. (Can I possibly get lucky twice....?)

The answer (shame!) was no I couldn’t... However I have been busy with the X-Acto knife and sticky tape. The set of A3 drawings that I had commissioned from PDH are now joined together and (finally) I can sit down with them and start doodling all over them. I am sorry for the quality of the following shot -but the only way I could get them all into the same shot was to put them on the hall floor and then hang over the bannister...

See Picture 7.

You can see the front roof air intakes, the central gas turbine heater plant and at the rear the roof exhaust vents

After a suitable pause (for someone to "acquire" some equipment over the weekend...) testing of the burner began at just after lunch. The mark two combustion chamber has survived a 20 minute burn without parts of it melting, glowing (too much!) or running out of fuel!!! The initial design DID WORK -but it was felt that it was just too hot for use in a model loco. So, the new design uses a back flow combustion chamber like the original Whittle designs. The exhaust temperature was fairly constant at 850C for most of the run. The new design will mean the engine is more compact than the its original design and will run cooler. It will also have to draw air from both sides of the intake turbine to cool the combustion chambers -another feature of the early designs!!! (WHO said "History Is Repeating Itself???")

See Picture 8.

The maths behind this is simply horrifying and all I can say is that Whittle armed only with a slide rule and a Simplex Calculator must have been very patient -or have known someone of the calibre of J.T.Latimore or P.Snodgrass to do the work for him... Hooker when he was taken on by Rolls Royce was derided for being “Not much of an Engineer” and he admitted that he wasn’t -he was just a great mathematician and designer!!!

To Be Continued.