In the workshop at present, the LSWR 'Mazeppa' class 2-2-2 of 1847, designed by John Viret Gooch (older brother of the more famous Daniel). Once again, details are in Bradley's book, including a reproduction of drawings published in Railway Machinery by D. Kinnear Clark in 1855. While they are not works drawings, they do show all the internal details, so it's probably safe to assume that Kinnear Clark had the cooperation of the Nine Elms drawing office.
Like Liverpool, the construction is not straight forward. There is no obvious way to divide the body and chassis, so once again, the frames form a central "backbone" to which everything is attached in one way or another. Here is a photograph of some of the bits and pieces.
I'm using a form of compensation. The front axle pivots about the centre of the frames, and the driving and rear axles are supported on beams. The pivot point of the beam is biased towards the driving axle (the screw head is visible in the photograph below) so that more of the weight is transferred to the driving axle than to the trailing axle - a simple principle of mechanics that was pointed out to me by Brian Clapperton. The centre of gravity of the loco has to be between the leading axle and the beam pivot point, if it is behind the beam pivot, the loco just tips onto the rear axle and unloads the front axle completely. Thus, weight distribution will be critical.
The smokebox is tricky, because the assembly incorporates the cylinders which are set at an angle to the smokebox itself. Both the front and back plates are awkward, and different, shapes. They are held in place by two spacers that were carefully cut to shape. In fact the whole assembly involved lots of careful measurement of the drawings, marking and cutting out, and a few crossed fingers. The effort paid off, because it all went together as intended.
The smokebox wrapper wasn't as bad as I feared. I cut out an oversize blank, formed the curve across the top using my rolling bars, then eased the rest of it into shape around the front and back plates, using an assortment of round bars to form the curves. It was then soldered in place, cut back with the piercing saw to leave a millimetre or less to file away to make it flush. The back edge was easy because there is no detail there to mess up, but the front plate has a row of rivets that had to be avoided. I tidied it up with a sanding disk in my Dremel, used very carefully.
The valve gear and feed pumps between the frames are now complete. It all works, and the driving axle has no fewer than six eccentrics to drive everything! I had to thin out the frames of Brian Clapperton's gearbox to get them all in.
Crossheads, slidebars, connecting rods, and outside frames in place.
