SEVERAL ARTICLES in recent Gazettes by Harold Jones¹⁾ and me have highlighted three considerations that must be borne in mind when using tighter curves (PECO Setrack curves and points have a radii of 3ft 4.5in or 1028mm).

1) Rolling stock used must be capable of traversing tighter curves. Mainstream RTR locos and stock are engineered and stated to accept R2 as a minimum radius. Even the forthcoming Ellis Clarke Black 5 4-6-0 manages this challenge, albeit with adjustments to the cylinders. For bespoke manufacturers of large express finescale locos, buyers are likely to have to check first. When I enquired at a Guild show as to whether a company’s 0-6-0 Dean Goods loco could traverse R2 curves I was met with a look of horror! Given the prices of some of these models, maybe their intended owners can also afford spacious layouts with sweeping curves? Kit and scratch builders can allow for radii clearances, with sideplay of driving wheels and bogies and attention to steps, cylinders, and anything else that may foul. I do find that kit sellers often fall back on advising 6ft minimum radius for their models, even shorter, inside cylinder locos that could be easily built to cope. Clearly this helps avoid customer complaints, but may also put potential customers off even trying. Ian Kirk has pointed out on the Guild Forum that vehicle overall length also matters, regarding inner and outer overhangs.

Personally, I prefer a maximum vehicle length of scale 60ft, or 440mm, so Mk 1 coaches at 63ft 6in or 470mm are only just acceptable. Many Co-Co Diesels are much longer and, unlike steam tender engines, do not articulate on curves. Even though Heljan advise an 800mm minimum radius, I would find the optics poor. The forthcoming Class 153 single railcar scales out as 76ft or 23m, so I admit a modern image problem, though EU and USA modellers with longer stock frequently show layouts with R1 (914mm) and R2 curves.

2)Track laying standards, dog-leg rail joints, and non-level track laying, can try even the most tolerant rolling stock, and are likely to cause buffer-locking or derailments in the immediate vicinity, hopefully leading to corrective measures being applied. Reverse curves also magnify the effect of transitioning between radii. For instance, many of us use PECO medium (6ft) radius points for crossovers, particularly in station loops. The sideways displacement of coach buffers at the reverse is more severe than coming off an R2 radius onto a straight section; the eye sees a sweeping reverse transition in the former.

Obviously with sharper curves, sometimes within goods yards, care to avoid designing-in an over abrupt reverse must be exercised, which brings us to how to do this.

3)Transition curve planning, and some helpful Setrack templates. Harold Jones has made the point that buffer locking while propelling stock is not a problem on relatively constant radii curves, where adjoining vehicles are of similar length (eg a rake of bogie coaches, or 4-wheel goods stock). The inner buffers will be doing all the work while the outer buffers are apart.

Longer or fixed connecting couplings could be used, though my experience on R2 curves is that this is not necessary. In fact if, couplings are tight on the curve and the adjoining vehicles are of dissimilar length (eg tender and first bogie coach), the coupling exerts a helpful centring effect, minimizing the chance of buffer locking.

However when coming off an R2 radius when propelling, there will be sideways displacement of even similar length vehicles; the trick is to transition the radii over a track length able to accommodate the likely longest vehicles.

Thus I have provided templates kindly drawn up by Stuart Davison, in response to a post I started in the Guild Forum Modelling section, on October 3rd 2020. The Gazette editor has kindly agreed to make them more easily accessible Transition Curves for PECO Setrack User.

A 900mm length of PECO Flexitrack, cut midway at 460mm and 440mm will provide a left and right hand transition curve to deploy at each end of an R2 radius return curve and minimise the risk of buffer locking