I completed the draft of this post just before RMWeb went off-line, during its transfer to a new hosting service.  I have made more modelling progress during the last 10 days, so will upload a follow-up article, once I have seen that the current post has settled down in its new home.

 

Last year, in May 2021, I wrote a post in my blog called ‘From the Stars to Fire Fly’. I drew on contemporary illustrations by E.T. Lane to show some of the engines to which I referred. I now find myself thinking about those engines that came before the ‘Stars’ – usually written off as ‘freaks’ and total failures.

 

A few of the engines were, indeed, ‘freaks’, since their designers tried ingenious but ill-fated approaches to meeting Brunel’s specification, but most of them were simply too small and too lightly built for reliable service. A selection from ‘The Engineer’, 1910, shows how the outlines of several of these engines compare with ‘North Star’ (top left):

 

Selection of Early GWR Engines

 

Although Brunel was a great visionary and created many wonderful engineering works, his knowledge of early railway locomotives was severely limited. At the time when he was planning his ‘Great Western Railway’, locomotive design was still in its infancy. ‘Rocket’ had succeeded at Rainhill in 1829, less than a decade earlier, and it looks as though Brunel had not been following the rapid developments that had happened over those years, Thus, when he drew up specifications for his locomotives, he specified dimensions and weights that were far too conservative. For example, he specified a maximum weight of 10½ tons, already far exceeded by contemporary engines, He also specified a ridiculously low piston speed of 280 feet per minute at 30 miles per hour. This was the sort of speed that might have been used on stationary engines but contemporary locomotives featured speeds of over 500 feet per minute.

 

The low piston speed required by Brunel implied the use of very large diameter driving wheels – the engine ‘Ajax’ originally had 10’ diameter wheels – which were necessarily very heavy. This left little weight in the budget for the boiler and other components, which were forced to be very small and insubstantial. It was a combination of all these factors – lack of steam raising ability and components that could not take the stresses and strains of regular operation - that made these engines so unsatisfactory and caused great difficulty for his locomotive superintendent, Daniel Gooch, in trying to keep them in working order..

 

I decided to examine one of these under-sized engines in detail, to learn why they were such failures and, for my first example, I have chosen ‘Aeolus’, one of three ‘large’ engines supplied by Tayleur. According to Brian Arman, in his comprehensive survey of ‘The Broad Gauge Engines of the GWR – Part 1’, ‘Aeolus’ appears to have been the best constructed and, initially, the most reliable. It did not, however, fare very well when entrusted with the first GWR public train on 4th June 1838.  It seems that this engine had difficulty in maintaining a speed of 6 mph with a loaded train.

 

As originally built, ‘Aeolus’ had 8’ diameter driving wheels, with a very small firebox and boiler. We are fortunate to have a copy of a works drawing of the engine, as built:

 

Aeolus Works Drawing

 

The very small diameter of the boiler is vividly displayed in the above drawing and, out of interest, I overlaid the head-on view from this drawing with a drawing of ‘North Star’ to the same scale. The difference is dramatic!

 

Aeolus and North Star Boiler Comparison

 

Despite its initial poor performance, Gooch took ‘Aeolus’ in hand and it was later accredited with some fast runs with very light trains but it was soon laid aside, before becoming one of the first engines to be re-built at the new Swindon Works in 1843.

 

Once again, I must pay tribute to that young pupil at Swindon, E.T. Lane, who sketched many of those early engines, which he saw as he walked around the works. Some of his sketches were worked up into detailed drawings but I find that it is often the sketches that reveal much more character. Sadly, Lane died at the age of 20; a reminder that life in the early 19th century was far more precarious than most of us experience nowadays.

 

His sketches of Aeolus, made during or shortly after its re-building at Swindon. show that the driving wheels were reduced to 6’ diameter and the proportions of both boiler and firebox were altered, to allow for a much larger firebox. Other details remain unchanged from the original works drawing. Lane’s sketches are annotated with dimensions of the various parts, which I have used as the basis for a 3D model.

 

First Steps towards a 3D Model

 

Lane’s sketches were probably made in a notebook as he stood with the engine in front of him. They are far from being accurate drawings but capture the ‘character’ of the engine very well. The value of these sketches lies in the copious annotations of dimensions, which I assume he must have measured himself from the actual engines.

 

My challenge was to use the combination of the overall appearance and the detailed dimensions to create a drawing from which I could make a 3D model. For this, I started by using my usual method of importing the Lane sketch into ‘Fusion 360’ as a ‘canvas’. I then calibrated the canvas so that the overall length of the engine, as annotated by Lane, corresponded to the dimension of a 4 mm scale model.

 

Outside Frames

 

The first part of the engine that I addressed was the right-hand outside frame. I started by drawing a rectangle corresponding to the annotated dimensions of the frame, then adding the wheel mountings, as shown in the screen-shot below:

 

My dimensioned outline of the Frames in 'Fusion 360'

 

Once I had the main outlines of the frames correctly scaled and positioned, relative to the defined wheelbase and wheel diameters, I could use the ‘push/pull’ tool in 'Fusion 360' to create 3D ‘bodies’ of the frame and splashers (rivet detail and springs will be added later.), as shown below:

 

My 3D Model of Frame and Splasher Fronts for ‘Aeolus’

 

Firebox-Boiler-Smokebox

 

The information on these parts is even more limited, as Lane’s sketches only show the side elevation. We know that the boiler retained its original 3’ 6” diameter and, when rebuilt, had a length of 7’ 11”. The cylinders, after rebuilding, measured 15” bore by 18” stroke.

 

The end-view in the Works drawing shown above, gave me the relative positions of boiler and cylinders. I used this information to extrude the smokebox from the Works drawing, as shown below. The depth of the smokebox is given by Lane as 2’ 2”.

 

Extruding the Smokebox in 'Fusion 360'

 

The locations of the cylinders are determined by the positions of the cranks on the driving axle. Because these are widely spaced. It is necessary to increase the width of the smokebox to accommodate them. I took guidance from a photograph of the sister engine ‘Vulcan’, which shows the sides to have been ‘bulged’ out from a point close to the mid-height of the boiler.

 

‘Vulcan’ – rebuilt as a tank engine 1858

 

I extruded the firebox, similarly to the smokebox, to a length of 12 mm. My model boiler is simply a tube, 31.67 mm long. of 14 mm inside diameter and wall thickness 0.75 mm, to represent the cladding.

 

Boiler Fittings

 

The design of the boiler fittings – dome, inspection cover, and chimney – followed my usual methods of using a combination of the ‘push/pull’ and the ‘rotate’ tools in 'Fusion 360', to create these cylindrical structures.

 

The inspection cover and a plinth for the combined dome and safety-valve cover were joined to the relevant main components, with the ‘chamfer’ tool being used to create a smooth transition at the join. I created the rather complex outline of the upper part of the safety-valve cover from a cross-section sketch, using the ‘rotate’ tool to convert this sketch into a solid body.

 

The drawing by G.F.Bird, produced early in the 20th century for ‘The Engineer’, shows fluting around the periphery of the dome. I was not sure about this, as it is not visible on the photo of ‘Vulcan’. I decided, however, that some short lines on the contemporary Lane sketch were probably intended to indicate fluting, so decided to add this feature. I created a single slot in the centre segment of the dome and then used the ‘circular pattern’ command to repeat the feature all around the periphery, as shown below:

 

Creating the fluted dome segment in Fusion 360

 

After adding the fluting, the complete dome and safety valve cover appeared as shown below:

 

My 3D-model of the combined Dome and Safety Valve Cover

 

In contrast, the chimney was straightforward but, from the Lane sketches, it appears to have been fitted over a separate plate at the top of the smokebox. I created this plate first, as a wrapper over the smokebox, and then produced the chimney itself by using the ‘rotate’ tool from a cross-section drawing. Finally, I created a fillet at the joint between these two parts, as shown below:

 

Creating the Chimney and its Plinth.

 

Now I had all the parts needed to put together the main body of the engine in ‘Fusion 360’. I have not yet added details such as rivets, boiler bands, etc. but the main outlines are clear. Because of the small diameter of the boiler, there is a lot of ‘open space’ between the frames, which will need filling with a representation of the motion!

 

Assembly of the major components in Fusion 360

 

Next steps

 

I now need to pay attention to the chassis. From the sketches by Lane, it is surprising to see that this engine retained its primitive ‘loose eccentric’ valve gear, even after re-building in 1843. The layout of this gear includes the use of a ‘weigh-bar’ in front of the smokebox, which looks very similar to that which Stephenson had used on his ‘Planet’ locomotive, designed back in 1830.

 

Since I only have a side elevation view plus sketches of a few details by Lane, I shall have to spend some time considering the probable layout of the visible components in front of the smokebox. This will be the main subject of my next post.

 

Mike