I have not forgotten about my plans to create a diorama, as described in my Broad Gauge blog. 

 

While I was researching the early days of the GWR, however, I found myself thinking about what had gone before – back to the ultimate pre-grouping period!  This led me to Trevithick’s engines, of which I photographed a replica at Blists Hill Museum near Telford:

 

Trevithick replica at Blists Hill Museum

 

Trevithick was a remarkable inventor but, sadly, not a good businessman. He not only demonstrated, by 1802, that a compact high-pressure steam engine could be adapted to both road and rail uses but also that such a vehicle could be propelled by adhesion alone. In his Penydarren locomotive of 1804, possibly his most famous design, he turned the exhaust steam up the chimney, so creating the ‘blast pipe’ which was to prove an essential component of later locomotives. He sold all the rights to his designs and emigrated to South America to seek his fortune. He eventually failed there and died in poverty, back in England.

 

His son, Francis, wrote a two-volume biography ‘Life of Richard Trevithick’, published 1872, which describes his father’s life and inventions. In Vol.1, page 181, there is an illustration of the Tramway engine that was built for South Wales in 1803:

 

 

Next, as so often seems to happen when I start a new research project, an entire ‘can of worms’ began to open up!

 

The book also notes, however, that “particulars are taken from Trevithick's letters where other evidences are conflicting . It is probable that more than one tramroad - engine was constructed in Wales at that time.” The dimensions listed above do not correspond with the associated drawings! For example by scaling from the drawing, if the wheelbase is 4’ 1” then the boiler is only about 5 feet long, while the flywheel is about 8’ 9” diameter, so there is no way it could have fitted through the Plymouth-works tunnel on the Penydarren tramway!

 

Later in the book, there are references to three different engines, including the one that I saw (in replica) at Blists Hill museum, Coalbrookdale.

 

“We have no account of the railway work done by the Coalbrookdale travelling engine of 1802. The Welsh tramroad - engine of 1803 took a gross load of 25 tons, at the rate of four miles an hour, over a bad road, with sharp curves and stiff inclines, and without load ran at a speed of sixteen miles an hour. The Newcastle locomotive of 1804 was, in general outline, similar to the Welsh locomotive, but in detail superior. The wheels were to run on rails instead of trám-plates, and were 9 inches farther apart than the Welsh locomotive, giving increased steadiness. The boiler and return tube were wholly of wrought iron; the fire -door and chimney were at one end of the boiler, and the cylinder and guide-rods at the other end, giving more room to the engineman than on the Welsh locomotive, which had all those things at one end of the boiler. The cylinder of the Newcastle locomotive was of the same size as the Coalbrookdale engine of 1802 , being 7 inches in diameter, with a 3-feet stroke, and therefore was probably made at Coalbrookdale, from Trevithick's drawings and patterns of 1802, with its regulating blastpipe and steam of from 60 to 145 lbs . on the square inch .”

 

At this point, there is even a Swindon connection, as shown by correspondence between Joseph Armstrong and Francis Trevithick:

 

“F. TREVITHICK , Esq . May 10th, 1870

DEAR SIR,

By this post I have forwarded a tracing of the old locomotive for Mr. Trevithick, and likewise some copies of an extract from the Gateshead Observer.' They would have been sent ere now, but I have had some difficulty in procuring an original from which to get a tracing, and have also spent time in hunting up this extract, and having it reprinted .

Yours truly,

JOSEPH ARMSTRONG Esq, Swindon.”

 

In Plate V. , a is a steam-cylinder 7 inches diameter, 3 -feet stroke , fixed in the boiler ; b , piston-rod ; c, crosshead ; d, guides ; e , stay ; f, connecting rod ; g, crank ; h, driving axle ; i, fly-wheel ; j, gear-wheels ; k , four driving wheels, 3 feet 1 inch diameter, 4 feet 8 inches from centre to centre ; 1, four-way cock ; m, lever for working cock ; n, plug-rod ; o , cylindrical wrought- iron boiler, 4 feet diameter , 6 feet 6 inches long ; p, firegrate ; q, return fire - tube of wrought iron , 2 feet 3 inches diameter at the fire -door end, 1 foot diameter at the chimney end ; r, chimney ; s , fire-door ; t , railway of longitudinal timbers, 3 inches wide, 4 inches deep ; u, cross-sleepers, 4 inches wide , 3 inches deep ; 1 foot 1 inch apart ; gauge between wood rails , 4 feet 10 inches ; weight of engine in working order, 4 tons.

 

 

 

 

Trevithick's 'Newcastle' engine

 

After searching for more information, I found Anthony Dawson’s much more recent book ‘Before Rocket: The Steam Locomotive up to 1829’, published 2020, in which Chapter 1 reviews Trevithick’s work between 1802 and 1814.

 

This text agrees that there were at least three similar locomotives. Dawson points out, however, that the Penydarren locomotive “was built to be convenient for the driver to work the engine and for the witnesses to see what was going on” He suggests that “this would probably rule out the cylinder being at the same end as the chimney. Thus, it is more likely that the Penydarren locomotive had the cylinder at the opposite end to the chimney and firebox, somewhat analogous with the Gateshead [Newcastle] Engine.”

 

I do not know why he discards Francis Trevithick’s description and statements in reaching this conclusion, since the earlier work appears to be clear about the differences between the two engines.

 

There is also the matter of a ‘replica’ built by The Museum of Wales (Waterfront Museum, Swansea), completed in 1981.  Their guidebook states “that the replica was built working from Trevithick’s original documents and plans (now in the National Museum of Science and Industry).” This leaves the question “which plans?”. Their ‘replica’ places the chimney at the opposite end of the boiler from the cylinder, unlike the illustration in Francis Trevithick’s book but similar to the later Newcastle engine

 

Perhaps I am missing some recent piece of scholarship but it is hard to see why the layout of the later Newcastle engine was preferred for the Welsh replica over Francis Trevithick’s 1872 description of the Welsh engine. Francis T. did, however, note that there was conflicting evidence and that “It is probable that more than one tramroad-engine was constructed in Wales at that time”. As I noted above, the engine, as drawn in his book, could not have run through the tunnel on the Penydarren tramway, although reducing the size of the flywheel would be a relatively simple modification.

 

It seems that, after a lot of reading and thinking, the can of worms remains open!

 

Creating a Model

 

Naturally, I felt the need to create my own model but then I found a 3D printable design, created for the Prusa printer on the ‘Printables’ website  This work is licensed under a Creative Commons (4.0 International License) Attribution—Noncommercial—Share Alike @ThorinOakenshield.

 

This model, however, is a much larger and more sophisticated model than my own usual designs but I thought I might re-scale and simplify the design. Then, there is the question of ‘which engine does it represent?’

 

I tried my method of overlaying the drawing on p.181 of Trevithick’s book over a still from the video of the ‘Printables’ model and … in all the major respects they fit well together! (note that the video image is not quite ‘side-on’)

 

Overlay of Trevithick drawing (red) on ‘Printables’ model

 

So, whether or not this is actually the Penydarren engine or another Trevithick prototype, I feel that I am close to the appearance of one of the very first locomotive engines ever to run on rails!  I decided to reduce the dimensions of my version, to fit a 4’ 4” gauge trackbed in 4 mm/foot scale, as the track gauge seems to be the only firm figure we have!

 

To do this, I imported the drawing on p.181 of Trevithick’s book into ‘Fusion 360’ as a ‘canvas’ and then scaled it to the correct gauge at 4mm/foot scale. This confirmed the wheel diameter as 15 mm (prototype 3’ 9”).

 

Drawing from Trevithick's book scaled in Fusion 360

 

Next, I loaded the .stl file for the wheels, downloaded from the web, into ‘Fusion 360’ as a ‘mesh body’. I laid this body over the above drawing and adjusted the size until the wheel diameter matched the drawing. This needed a scale factor of 21.34%.

 

I imported all the .stl files for the web model into the ‘Cura’ software and re-sized them all by the same scale factor.  Then I loaded all the re-scaled files into ‘Fusion 360’ as mesh bodies.

 

I could then move each of the parts within ‘Fusion 360’ into its appropriate location around the boiler, as shown below against the ‘canvas’ backdrop:

 

‘Printables’ files, re-scaled and loaded into Fusion 360

 

This confirmed that the re-scaled parts should all fit together as required. It also indicated, however, that some parts would be too thin to print successfully at the reduced scale. In particular, the boiler casing would be far too thin, so I decided it would probably be easier to design a new part for this component and, maybe, for some other parts as well.

 

The next step will be to review all the parts for their suitability for printing at 4 mm scale and make modifications where necessary.  In the meantime, I show a couple of 3D views of the model, rendered in ‘Fusion 360’:

 

Two rendered views of 3D-model assembled in Fusion 360 from ‘Printables’ files

 

Mike