Curve Radii

[Soar Valley Light]

Hello All,

As the planning of my line is now into the detailed stage I'm seriously having to pin down my track plan. This means finalising the alignment of the line too so that the terra-forming can progress in the right way in the right places (hopefully at the right time too!)

My outline design was based around four, six and eight foot curve radii. During the more detailed planning it's struck me that I could take realism a stage further be designing 'prototype' radii. Maybe I should explain. A good deal of railway design in the heyday of railway construction revolved around the chain as a unit of measurement. It wasn't exclusively so but definitely figured large. If you don't believe me just look how many gradients and radii (measured in yards) are divisible by 11 (half a chain).

So, my plan is to work with 1.25, 1.5, 1.75 and 2 chain radii curves which work out at 1320mm/4'-4", 5'-2", 6'-1" and 6'-11" respectively. You might think that this is taking realism several steps too far but it keeps me happy!

My query here is to ask if anyone knows what sort of radii our prototype narrow gauge railways were built to. I dare say they varied wildly from railway to railway and even along individual railways but they were all engineered to one extent or another and therefore someone made a decision about this when drawing them up, setting them out and constructing them.

Happy steaming,

Andrew

[Peter Butler]

I don't pretend to know anything about the engineering side of narrow gauge railway construction but, one of the reasons they were built, other than cost, was because of the nature of the terrain.
Is it not the case then that the landscape itself dictated the radius of many curves on a line? Where the line followed contour lines on hillsides it must have led to many compromises where radii were concerned? Engineers must have tried to ease the curves as much as possible, however, to begin with a standard minimum in mind surely would not work in practise?

[GTB]

My query here is to ask if anyone knows what sort of radii our prototype narrow gauge railways were built to.

How long is a piece of string?

Apart from the topography of the country being traversed, the minimum radius was also determined by the size of loco needed to work the projected traffic.

As to locos, the following is from the Hunslet ng loco catalog

- quarry Hunslet 0-4-0T, min. radius 20 feet.
- ROD 4-6-0T, min. radius 100 feet.
- Sierra Leone 2-6-2T, min. radius 132 feet.
- Beddgelert 0-6-4T, min. radius 300 feet.

My knowledge of narrow gauge civil engineering is limited to the VR 2'6" lines and they ranged from 5 chain minimum on the Whitfield line, 3 chain minimum on the Gembrook and Walhalla lines and 2 chain minimum on the Beech Forest line, which had several miles of continuous 2 chain curves on a 1:30 grade. The Powelltown Tramway my modelling is based on had 7 chain minimum on the steam worked part of the line.

When the first G class Garratt was tested on the 'Beechie', it stuck solid on one curve, which was found to have moved over the years and was down to 1.8 chains.

Curvature also sets the speed limit of trains. The max. speed on all VR ng lines was 20mph, but it was reduced to 10mph on sections with 2 chain curves. The VR speed limits were enforced with recording speedometers on all locos and an army of clerks in head office checking the charts.

Will you be laying the track with transition curves and superelevation as well?

Regards,
Graeme

[Soar Valley Light]

Hi Graeme and Peter,

You are right that many factors inform the final design of any railway. The terrain is very important, particularly if 'contour hugging' is a high priority. The actual construction details of the railway will be an important compromise between 'best line' and least amount of earthworks. Another very important factor is the balance of cut and fill. Sufficient material needs to be dig out of cuttings to create the fill required to build the embankments. Sometimes it would be necessary to 'overdig' a cutting to provide this, a narrow gauge example of this can be found above Bratton Flemming on the L&B I believe (To supply Parracombe bank). Just occasionally surplus material had to be tipped in spoil heaps or over widened embankment structures. This was to be avoided wherever possible though as the effort and cost of doing so was extra to that required for the actual railway construction.

Companies generally engaged their own engineer quite early on in the process of promoting a line. Preliminary surveys would be made and drawings produced. There must have been some remit from or discussion with the company board to establish the type and nature of railway their ambitions were aimed at constructing in order for the survey and drawings to produce a line of route. No doubt further (extensive!) discussions then ensued to agree what was required; I'm sure costs became much more of a priority at this stage when the engineer began to indicate the costs which various standards of construction were likely to incur. I'm sure that every company, no matter how focussed on 'cheap' or 'light' construction, wanted the very best that could be achieved for their line, but the costs of doing so must have necessitated compromise of these ideals in a great many cases.

Further, more detailed surveying probably took place next to enable the drawing up of detailed plans that would be necessary for Parliamentary approval. It's likely that the bulk earthworks calculations could be worked up at this point (although even more detail was probably produced later for tender and construction purposes). It would certainly be at the 'Parliamentary plans' stage that the ruling gradients and radii would be established in principle, these would be quoted in the Act authorising construction. These may have been tweaked later as the project moved through final design and into construction but the fundamentals would be unlikely to change.

The design of locomotives would be much more likely to be informed by the alignment and gradients than the other way round. Consideration of the type of railway and the way the traffic was to be worked would certainly have included a consideration of loco design, and that might have had some influence of the various stages of planning, but it must be remembered that the railway itself was the most important (and expensive) element, loco's were only one of the pieces of equipment required for it's operation and wouldn't have enjoyed quite the same status we shower on our own little treasures.

Graeme's description of the VR railways construction and operation is very interesting. Long 2 chain curves on gradients like that must have made for some spectacular railwaying! I would have loved to have seen that. It's easy for curve radii to wander, alignment errors are much harder to see in curves than they are in straights. Where curves and gradients are severe I can easily see how any minor tightening could cause big problems.

Thanks for the responses chaps. It's amazing where this hobby of ours leads us!

Andrew

[IrishPeter]

The ever helpful E. R. Calthrop suggested a minimum radius of two chains in 2'6" which is about 7'7.5" in 16mm.  The Indian Government recommended a minimum of 100' for 2' gauge which is about 5'3" in our scale.  There is rather a nice shot of the Calthrop-Jones couplings doing their job on a 150' radius curve in 'Light Railway Construction.'

Cheers,
Peter in AZ