History is littered with small locomotives built for industrial use and for light railways in both standard and narrow gauges. Few were intended to run at high speeds but were often required to work hard either because of the loadings or terrain. Like their larger brethren the valve gears, Walschaerts’ in particular, could suffer compaction by space limitations but this affliction is not a function of scale and excellence in design can be discovered.
In fact the Darjeeling ‘B’ Class gains considerably from a long and rather squat rectangular envelope. Two features strike the eye immediately: the anchor link is long and we can therefore expect to gain symmetry from the lap/lead function. The expansion link has no backset tail. To some extent the backset and pitch circle are governed by the inclination of the eccentric rod: as this decreases the pitch circle and lessens backset until the backset becomes zero as the tailpin hole coincides with the tangent square to the radius rod, beyond which the backset becomes negative, or ‘front set’. The latter condition is therefore frequently met when Walschaerts’ gear is applied inside the frames, where a comfortable eccentric size demands heavy inclination of the eccentric rod. Such an arrangement is of no detriment to good event equality.
At the opposite end of the scale, when the tailpin arc is more nearly coincident with the gear centreline the return crank angle approaches 90o, beyond which a 90o+ will become 90o- or vice versa according to admission type. As may be seen in much American practice a large tailpin and backset then become the norm.
Clearly, the Darjeeling pin is on the curved centreline of the expansion link slot, denoting front set, and yet the tailpin arc is very close to the horizontal gear centreline, with the possible advantage of good linear drive throughout eccentric rod and radius rod in full forward gear. What might a distribution analysis have in store? In addition to the squat envelope the Darjeeling 0-4-0 has in its favour outside admission slide valves and no requirement for long valve travels. The major advantages procured thereby are smaller angular swings of the combination link, reduced pitch circle and less angularity upsets as a consequence.
Relatively large backset values in distance do not induce large differences in angular swings of expansion link swings, particularly in outside admission cases of relatively conservative valve travel. The worst scenario comprises inside admission valves and long laps. In fact the angular swings of the Darjeeling engine comprehend a mere 1.4o difference. For the heavy climbing duties of the ‘B’ Class, with little need of fine equality of distribution in the shorter cut offs, one could make a case for more attention to equality in the longer cut off range at the expense of shorter running except in a model undertaking quite different duties, but the results are very good and not in need of such an approach.