Building a narrow gauge locomotive on 5" Gauge.....2 1/2" scale. By John G. Richardson

Part 2 ....Chapters 8 to 16

For more illustrations see ' Building the locomotive PIXIE on 5" Gauge ' click here pixiesbackground.htm

Chapter 8

Coupling rods and bearings. Brake hangers and blocks. Lubricator and mounting plate. Coupling blocks. Footplates, support brackets and angles.

Motionwork on prototype


' PIXIE ' has round-section coupling rods terminating in marine bearings at each end. The coupling rods for ' Pixie ' are made from lengths of silver steel, shouldered and entered into the bolting flanges at each end. These joints were brazed and the faces cleaned up. The bearings were simulated on ' Pixie 'using blocks milled from mild steel incorporating the profile of the prototype castings. Each block is scribed deeply to represent the joint between the two parts of the bearing and each crankpin hole is fitted with a pressed-in phosphor bronze bush. The length of the coupling rods was checked against the wheel centre template made earlier. Small adjustments were required at this stage, due no doubt to minor differences in dimension at each joint. Adjustment was achieved by turning-off the inner ends of the bearing blocks in the four-jaw chuck. To complete each assembly, matching keeper plates were filed to shape, clamped in place on the coupling rod ends together with the bearing blocks and the whole assembly was then drilled for bolting. mm size bolts were found to be very near to the required diameter and these were used, the heads being turned to the shape of those on ' PIXIE '. The thread ends were turned away and the bolts drilled for split pins. Oil cups are fitted to each bearing block. For ' Pixie ' these were machined from brass hexagonal material, which provided the flats by which the cup is screwed into the block. A small form tool ground up from a slip of toolsteel ensured the same 'antique' profile to each cup. The cups are topped off by a threaded cap screwed into place, these were reproduced by turning down the heads of BA bolts and knurling the periphery of the resulting head. It was noted at the time of the first survey that any missing caps were replaced by corks, as yet no caps have gone missing from Pixie so this has not been necessary! Each of the upper bearing bolts has a turned groove coincident with the oil cup to allow lubricant to penetrate to the bearing surfaces.


On ' PIXIE ', braking arrangements are simple in the extreme, there being only two blocks acting on the front of the rear wheels. The brake hangers are pieces of bar drilled for fixing to the frames, for the brake block pivot pins and attachment to the brake rod. The brake blocks we r e cut from a ring of mild steel turned to match the wheel section internally and externally to a section which could be milled to provide the hanger slot. The hangers pivot on shouldered studs, captive in the frames and backed up by large square washers. The pivot pins securing the blocks to the hangers are fitted with split pins. OILBOXES \par \par Four brass 'trays' or reservoirs for axlebox lubricating oil are mounted on the stiffening angles. The reservoirs and their hinged lids were milled from some anony mous brass from the scrap box, probably parts of electical switchgear dismantled in the ongoing ' salvage and alternative uses campaign ! Rubber tubing forming minute pipes conveying oil to the bearing surfaces of the axleboxes was obtained from a suppl i er of laboratory equipment, a useful source of speciality items such as rubber and plastic tube. The pipes are attached to small diameter copper pipes silver soldered into the oil trays. The feed to the bearings is via similar copper bends which are solde red into the axlebox castings. Old fashioned fluffy wire pipe cleaners pushed into these bends stems the flow of oil to ensure a constant flow of the necessary lubrication at the critical points.


PIXIE has a mechanical lubricator mounted on a steel plate fixed using two bolts through the right hand frame angle. On the model this is replaced by a proprietary lubricator from Norman Spinks. The lubricator was seen in an advertisement, catching the eye because it incorporates roller clutches rather than a ratchet and pawl mechanism, this allows an infinately adjustable stroke. The oil case is not too unlike that on the prototype, except that it lacks an oil level gauge. On Pixie, as on the prototype, the drive is taken via. a small diameter rod clamped to the valve rod, adjustment to the flow being achieved in the model form by selecting one or other of the holes in the pump actuating lever. The geometry of the action can be adjusted by nuts set on a threaded portion of the actuat ing rod. When first seen PIXIE was carrying diesel coupling blocks, and, as this is how she was to be recorded, diesel coupling blocks were next on the agenda. The dimensions from my survey were set out in sketch form and the scrap box yielded the majority of the small pieces of mild steel required for fabrication. Access to TIG welding equipment provided the opportunity to assemble the parts ' in the square ' and a session with the angle grinder, used offhand, introduced the necessary geometry. A study of the prototype front plate revealed the fact that a variety of attachments had been made during the life of the loco but currently the blocks are attached by 4 large diameter bolts, these are replicated by BA bolts.The coupling pins were turned fom scraps of mild steel rod. These were cross-drilled whilst clamped onto a vee-block. A line was scribed on the periphery of the head, a centre finder was used to scribe a diameter coinciding with this line. The centre of the cross hole was marked with a punch. The pin with the diameter set vertical was clamped in a vee block and a pointed wiggler used to locate the punch mark below the centre bit in the drill press. An overlever clamp remains on the drill press table at all times providing a simple means of retaining material and jigs during drilling operations. The fact that the clamp is ' resident ' avoids situations where one might otherwise be tempted to ' take a chance ' by holding something for a one-off drilling operation, putting ones fingers in jeopary at the time of the drill breaking through when snatch is liable to take place.


Two support brackets were fabricated from plate and angle steel, these are important as they support the weight of the bunkers. They are rivetted back to the main frames using dummy rivets as previously described. The rails are sections of angle rivetted to the main frames and the drawbeam at footplate level, those supporting the outer plates being at a lower level than for the c e ntre plate. The centre footplate is slotted for the damper handle and provided with a cranked angle which engages in the handle to provide several settings of the damper opening. There is also a checkout for the drain pipe from the water gauge. The rail which trims the outer footplates is cut and bent to form a curve matching that of the bunker later mounted above it. PIXIE'S footplates are replicated in the model by stainless steel plates which are bolted to the angles and rails.


On top of each of the frame stiffening angles are mounting blocks on which are seated the bearing flanges on the boiler. On the prototype the brackets are fabricated from steel angle, these are reproduced in the model although to ensure that the boiler would sit true to level within the frames, final machining to height was was left until the boiler could be offered up into position and the appropriate dimension established. Cutouts were required in the stay immediately behind the boiler, these provide clearance for blowdown valves, sited just above the foundation ring. Details not to be found on the full-size locomotive are the holes for dumping rods or pins which allow the ashpan to be dropped in case of emergency. Whilst the downward movement of the ashpan is restricted, there is still sufficient room for the grate to drop and the fire to disperse in the event of the pins having to be pulled in case of emergency These details just about c omplete the frames. Of course detail varies from loco to loco and there is a lot of enjoyment to be obtained from identifying details for inclusion in ones model. It may be that a particular prototype has been modified at some time in its life, something a dded, a piece of plate fixed at a potential crack position or some other ' adjustment ' made in service. For example, in the case of PIXIE there are some plates bolted to the front of the weatherboard where at some stage the angles have been cut and subsequently rejoined. These are yet to be incorporated in the model. It is well worth while researching such detail which will go a long way to improving the authentic atmosphere of the finished model. On PIXIE there is a small piece of plate attached to the front frame member, this provides a purchase for the bolting attachment of the drain cock return spring. Small details such as this which were perhaps an afterthought on the part of the designer, or an innovation on the part of a restorer, help to promote the atmosphere so beloved of the narrow gauge enthusiast.

Chapter 9


Piston rod gland and slide bar fixing


Smokebox saddle. Cylinders, ports, passages, covers and glands. Pistons and piston rods.


The smokebox saddle is a simple casting, produced from a rudimentary pattern which governed the profile with allowances for shrinkage but incorporating no fine detail. The casting was clamped to the topslide and the bolting faces were machined using a fly cutting tool. This useful tool simply produced in the home workshop comprises a substantial section of 2 1/2 " diameter bar with a drilled hole for a circular tool and a grub screw set in the perimeter to enable the toolsteel to be set at the correct angle. The mass of metal comprising the body of the tool when CHAPTER NINE gripped in the three jaw chuck serves to induce a 'flywheel' action reducing the effect of the intermittent cutting, a feature of flycutting operations. in order to increase the versatility of the attachment it is provided with three further holes which match the centres of the tapped holes in the faceplate on the milling machine quill. This arrangement is always used when heavy cutting / planing operations are carried out in the mill.

The seating for the smokebox and the small rebate which conceals the joint between the saddle and the smokebox were fly-cut whilst the saddle was fastened to the vertical slide, set parallel with the axis of the lathe. When setting a fly cutting tool to radius it is essential to remember that too small a radius will result in metal being cut away from the depth of the concavity, i.e. with too great a rise on the chord. Cutting should be started with the tool set to a slightly larger radius than that finally required, the tool being adjusted inwards, slowly reducing the radius of the cut until the finished radius is achieved! This needs a little thought but a few lines on a piece of paper will soon verify the facts, and may save a valuable casting from destruction!

On completion of the machining operations the small gussets which stiffen the bolting flanges on PIXIE were silver soldered into position, leaving an generous fillet of solder to replicate those on the casting. At this stage a template was prepared for the location of the cylinder / frame fixing bolts The centre line of the cylinder was transferred to this template and the template set aside for later use.


The cylinder castings were made from simple wooden patterns which followed the main outline of the pieces. In view of the writer's limitations in the pattern making department, it had been decided that such items as seatings for the valve spindle and piston glands would be added after the main machining operations had been carried out. First the top face of the casting was fly cut to the correct inclination. This angle was determined from a trial set out in marking blue on the frame, using the dimensions from survey and the template previously prepared. As well as providing a plane surface to assist in setting out this face would later determine the correct angle for boring the cylinder. The next operation was that of machining the steam chest. PIXIE has the steam chest integral with the cylinder casting, and this called for some massive excavations in the model casting ! A new term was added to the vocabulary at this stage. Conversation with a ' real ' engineer evinced a description of an operation known as ' hogging-out ' This elegant term describes the process of making multiple drilled holes within the area which would eventually become the steam chest cavity in the casting, before finally the remaining few pieces of metal were milled out, using an end mill, to leave an accurate cavity. This procedure was duly followed and the resulting steam chests proved to be acceptably accurate.

Having cleaned up the perimeter of the steam chest, attention was then turned to boring the cylinders. This was carried out with the castings secured to the topslide using straps of stock material in conjunction with suitable packings and tie bolts to retain the casting securely during the fairly substantial cuts of the boring operation. A word of warning...... unless ties such as used in this instance are set absolutely vertical, relative to the topslide, with, in this case tapered packings properly anchored into position to provide a firm purchase, they may loosen in the course of boring, with dire results. The tendency is, due to the cutting effort applied together with vibration in the system, for the tie bolts to loosen as they assume a position normal to the topslide and the tee-nuts therein. At this stage the cross travel of the saddle must be locked by tightening the clamp nut. On my lathe I have fitted a short lever here to enable me to exert stronger pressure on the clamp bolt without the need to use a spanner.

Once lined with the axis of the lathe and with the correct inclination determined by the previously fly cut face, the ties checked and the cross-slide secured against movement, a start was made on drilling for the piston rod. PIXIE only has one cylinder cover on the front end of the cylinder, the piston-rod passing through a gland on the the closed end of the casting. Drilling for the piston rod was done using bitts held in a headstock-mounted drill chuck. This operation was started using an end mill to provide a flat land on which the centre bitt could bear without being deflected, after the centre bit a succession of changes of drill size, up to clearance size for the rod, ensured a correctly located and accurately sized hole. The hole was not reamed because reaming to finished size can, in this instance, when the gland nut is inserted, form what is in effect a 'self locking clutch'. Any misalignment of two reamed holes occurring in the course of taking-up the gland causing the piston-rod to bind.

The cylinder bore was cut using a specially made, short, stiff boring tool set up in the four jaw chuck which in this instance was used as a boring head. On the Myford this required some thought. Finally reference to, would you believe, a page by ' Geometer ' in the 6th September 1962 issue of MODEL ENGINEER provided the solution ! Two pieces of 1/4" plate were cut and inserted between the jaws. This allowed the boring tool to be moved laterally across the axis of the machine, within the central hole in the chuck, by the amount required to bring the bore to finished size. The metal plates ensured that at all positions of the tool it was still supported by all four jaws. A proper boring head would have simplified the operation but, bearing in mind the few occasions on which such a tool would be used, budgetary considerations determined that other accessories were more important.

In machining bores which are to have a cylinder cover at one end only it is necessary to guard against producing a bell-mouthed bore. This is best achieved by bringing the bore nearly to finished size then taking only very fine finishing cuts. LBSC called this ' taking the spring out of the tool '.

Cylinder boring completed, the writer encountered a piece in MODEL ENGINEER by Keith Wilson relating to the profile of cylinder bores. Given the benefit of Keith's experience earlier in the course of building Pixie the cylinder bores would have been made according to his ' words and music '.

Next the position of the ports was established and the milling machine brought into play to cut them. These were not the working ports, as the port faces were to be in stainless as described earlier. They were cut within the outline of the eventual port positions to a depth which would permit the steam ways to intercept within the metal between the bottom of the steam chest and the cylinder bore. The steam inlet and the exhaust passages from the steam chest terminate on the top face of the cylinder block where the flanged pipes seat. These passages were drilled from the bolting face of the cylinder block and were, after the vertical parts of the passages had been drilled and swarf cleared from the holes, subsequently sealed by small turned and threaded plugs with slotted heads similar to grubscrews. The plugs were anointed with liquid gasket prior to being screwed home. The slotted heads of the plugs enabled insertion using a screwdriver and were of course concealed between the cylinder casting and the frame after erection. No excuses are offered for the use of liquid gasket. Joints do not need to be taken-up too tightly, the synthetic rubber hardens as it cures, excess is easily stripped away with a sharp knife, and a good seal results each time. This technique may not be in line with the practice adopted in the prototype but I feel sure that, had Engineers of yore been given access to liquid gasket, they would have used it. Indeed it is interesting to ponder, given the range of materials available to us today, what the Trevithics and Stephensons would have selected for their constructions were they here to make the choice !

The passages between the valve chest and the cylinder ends were drilled in the drill press, care being taken to achieve the correct inclination by packing and clamping to the drill table. First the open end of the cylinder was prepared with the bevel which allows steam to pass into the cylinder, this was done by filing , using a scrap of brass to shield the bore should the file slip. As well as permitting the passage of steam in service, this bevel allows the hole positions to be spotted with a fine punch to ensure that the smallest centre bitt can be started correctly. LBSC's tip on grinding the drill bitts used to drill the channels with one lip wider than the other to enable a broken bitt to be extracted is a winner, as was proved in the process ! There is nothing quite like the panic that one experiences when a drill or a tap breaks in an expensive casting, possibly the more valuable because a lot of time has been spent working on that casting up to the stage at which the breakage occurs. I suppose as a last resort spark erosion could be used but fortunately thanks to LBSC this has not yet been necessary.

The inclination of the passage from the port to to the blind end of the cylinder was such that the holes could be drilled using an extended drill from the steam chest. An extended drill was necessary to avoid the drill chuck coming into contact with the bolting face of the steam chest. Extended drills can be simply fabricated by grinding a stepped joint on the drill shank, cutting a matching step on a suitably sized rod, applying flux, binding with soft iron wire and then silver soldering the two together. A collection of various sizes of drill treated in this fashion will build up as work proceeds and provide a useful armoury for future use.

It had already been decided that, as the valve events of the model would in the early stages, prove to be somewhat tentative, the valve port faces would be made replaceable. Stainless steel was used for the faces and this proved to be simple to machine to the required degree of accuracy as it could be set up flat on the bed of the milling machine. Those without access to a mill could equally well machine the piece on the vertical slide. Final fixing was carried out within the steam chest bedding the stainless steel onto ARALDITE and securing the plate using purpose made, stainless steel countersunk screws into tapped holes in the bottom of the steam chest. In the event that subsequent modification to the porting may be required, the ARALDITE bond could be broken down by local application of heat.

Flanges were prepared for the inlet and exhaust pipes. These were used to locate the drilled and tapped holes for their fixing onto the top of the cylinder casting, these were then set aside for assembly onto their respective pipes in due course.


Having bored the cylinders and prepared the steam passages, the next step was to produce the cylinder covers, in the case of PIXIE these are of fairly substantial section with a nice domed centre. Machining this centre called for a lot of knob twiddling, a thoroughly enjoyable business. Achieving the correct profile was aided by the production, on the drawing board, of a card template which could be applied to check the profile between bouts of twiddling. Experience has shown that such profiled surfaces generally turn out to be less than spectacular due to a conservative tendency on the part of the machinist, reluctant to take those last few cuts that would make all the difference to the form of the piece. This springs from the memory of those ' last few cuts ' taken in the past which have resulted in a scrapped part and lots of rework ! The use of a template as described promotes the development of the striking shapes which are a feature of, and provide much of the charm of, the narrow gauge locomotive.

A glance through any of the books which describe the motive power employed by the various narrow gauge and light railways evidences the fact that engineers in the heyday of steam were quite happy to spend money on ornamentation which would be frowned upon by the ' value engineering ' oriented designer of today.

The covers were mounted in the three-jaw and the spigot prepared a good fit in the cylinder. The spigot was carefully filed away to provide a clear steamway to, and from, the steam passage. On a previous locomotive this step had been omitted and only remembered when on trial with air, when unbelievably there was no action on switching on the air supply. this was of course simply rectified, once the cause of the problem had been identified, although at the cost of unbolting the cylinder covers and breaking the then carefully made gaskets (that was in the days b.i.g, before instant gasket). Drilling for the studs followed using a template from thin plate material. The hole positions were transferred from the cover clamped into position and then the first hole was drilled and tapped to allow the insertion of a stud to immobilise the cover during the subsequent drilling of the remainder of the holes. The use of a template whilst apparently laborious is a worthwhile precaution as otherwise although the holes for the studs can be set out on the cover and transferred to the cylinder, with a gunmetal on gunmetal joint slip can easily occur. Again harking back to a previous model, the draughtsman had drawn equally spaced studs although with the cover located as drawing, one of the studs would have coincided with the position of one of the steam passages, small errors like this sometimes slip into details and are worth watching for. One cannot help but wonder what is the effect on the modeller following instructions in a periodical when he reads an announcement such as ' we regret that owing to a printers error in our last issue etc. etc.!' Perhaps its advisable to work a few issues in arrears in which case printers errors will have been accounted for by the time work commences on a particular part. In the same vein one magazine, now defunct, published an apology some 16 installments into a series on building a particular model, a note to the effect that they had lost touch with the writer and were seeking someone to complete the series! This must have had disastrous results !


Next came the preparation of the steam chest covers. This called for a template which was duly produced from a scrap of mild steel. The template was used to drill each cover plate in turn, including the small tapped holes for the metal threads used to break the gasket joint when access was required to the valve chest for maintenance purposes on full size PIXIE.

Once the cover plates were drilled the template was again used to locate the tapped holes in the cylinders, one hole being first drilled and tapped and a bolt inserted to secure the template against displacement whilst the remainder of the holes were spotted through. PIXIE has a union set in the right hand cover providing a take-off for a pipe to a steam chest pressure gauge, this was produced from a chucking piece of material from the scrap box.

The studs securing the steam chest cover are, on PIXIE, set out uniformly about the centre line of the cover. At each end is a small bolt, set into a tapped hole on the centre line. These bolts serve the important purpose of enabling the steam chest / cover joint to be broken when carrying out adjustments to valve settings. It was a delight at the time that the covers were required to be removed to be able to slacken off the lock nuts on these small easing bolts at the ends of the cover, then to break the cylinder/cylinder cover joint by screwing down the bolts. Although on Pixie are they are only 8 BA. they did the job famously. Those old engineers certainly knew their stuff !


Seatings for the valve rod gland and tail cover on each cylinder casting were milled and shaped from bar stock. They were then silver soldered into place on the cylinder casting, correct location being ensured by the use of short lengths of carbon salvaged from the bushes of a vacuum cleaner, turned to a push fit in each assembly. The carbon resists the solder whilst retaining the correct location and is a good example of the recycling activities which form a part of the writer's modelling activities! A good fillet of solder was formed at each joint to emulate those on the prototype casting. After the soldering operation each gland was faced up true to the main body of the casting.

The glands for the valve rods were turned up as circular, flanged bushes prior to being sawn and filed to the required shape. The body of the piston-rod gland is interesting, incorporating as it does the seating for the slide bar. Production of the body involved a considerable amount of hand-work, carried out once the the hole for the rod had been drilled and counter bored for the gland. The piston rod gland has an oilway for lubrication and is adjusted by two shouldered studs as is that for the valve rod. Rather than use conventional packings in the glands, ' O ' rings are preferred as these provide tightness whilst causing minimum friction between the parts.

A nice little detail seen in service is an oil-soaked length of tow which is tied to the easing bolt on the valve chest cover and looped around the valve rod to provide lubrication. Very practical, and a useful bit of ' colour ' to assist with the narrow gauge image !


Pistons are turned from gunmetal, The piston blanks being mounted in the three jaw, the holes for the rods drilled and tapped with a 1 / 4 in M.E Thread. The thread was cut away using a 1 / 4 inch diameter D-bitt for half of the piston thickness to allow the piston rod to be drawn into position by its thread in the later operation of uniting the two parts.

Piston rods of silver steel were set into a simple collet in the three jaw. These collets are home produced from suitably sized slugs of metal by drilling and reaming to the finished diameter of the spindle or rod to be gripped. After marking by centre punch at No. 1 jaw to ensure accurate re-location the collets are then slit with a razor saw. As work progresses a number of useful collets accumulate and are be stored with the tooling ready for re-use as the occasion arises.

The thread for the pistons was cut with extreme care, using an M.E. die and ensuring that at all times the die stock was firmly supported square to the axis of the rod by the nose of the tailstock chuck. A tailstock-mounted die holder is of great assistance in such work and can be manufactured in the home workshop from a proprietary mandrel and some offcuts of round bar. A secondary bush which can be set into the space for the die, and which permits longer adjustment screws to pass, allows the use of smaller sizes of die within the main sliding die holder.

With the rod set in the collet within the three jaw, the ends of the rods were lightly centre drilled prior to oil being placed on their threads and into the the tapped holes in the pistons. The pistons, gripped in a hand protected by rag, were then screwed onto the rods, care being taken that the plain portion of the rod entered well into the smooth, ' D-bitted ' part of the hole. Once the pistons were secured in this fashion the tailstock centre was engaged in the small centre and the pistons turned to finished diameter with reference to each cylinder bore. Each piston was then marked to ensure subsequent matching of the sets. It is helpful if a uniform system of marking is followed, centre-pops on an unimportant surface will suffice, say one pop for items on the right side of the loco and two on the left, looking forward from the footplate, or perhaps simpler still one pop for one side, and no pops for the other !

With each piston assembly still mounted in the collet and supported by the tailstock centre, grooves for ' O ' rings were machined using a parting-off tool. Grooves for these rings must be produced to the sizes recommended by the makers, this being critical to achieving a sound seal whilst reducing frictions to a minimum. The edges of pistons and grooves should always be 'broken' using a fine file, to avoid damage to the rings as they are installed.

On final assembly, the piston rods are retained in the crosshead bushes by wedges passing through slots in both parts. In the case of Pixie these wedges are, in fact, bolts with the heads filed to simulate the ends of wedges. Holes for these simulated wedges were not drilled until their position could be accurately located at the time of trial assembly of the engines. Builders may wish to replicate the full size arrangement which may call for slotting the piston rod and crosshead and providing appropriate cotters and wedges.

Chapter 10


Marine bearings on Pixie

Motion plate. Slide bars and crossheads. Valve gear. Weigh shaft and reversing gear. Valve setting. Brake column.


The motion plates are of course a handed pair, those on PIXIE are quite conventional and those for Pixie were produced by cutting plate to the required profile. Then the flange for fixing to the frame was attached by silver soldering. The lugs which support the trailing end of the slide bars were soldered into place at the same heating. Drilling the holes for the slide bar fixing bolt was left until final erection. The motion plate was trimmed with a half round fillet on the front face, simulating the detail on the prototype, this was attached using soft solder. For the purpose of trial assemblies the slide bar was be held in place using a toolmakers clamp. The location of the motion plate and thus the position of the fixing holes in the frames is best established at the time of setting out the motion work.


Each engine on ' PIXIE ' has a single slide bar the front end of which takes support from the body of the piston rod gland on the back of the cylinder casting. The trailing end is bolted to a bracket on the bottom of the motion plate. The slide bar itself is a piece of profiled steel, in the case of Pixie ground gauge plate. This was milled to shape and the fixing holes and countersunk lubrication holes drilled in conventional fashion in the drill press.


On PIXIE the crossheads are substantial steel boxes, assembled around the slide bars. Lining each box is a set of slippers, so shaped as to remain captive once the crosshead assembly is bolted up. The box is fabricated from three plates and a fabricated mounting, assembled by bolting. Incorporated in the bottom plate of the crosshead assembly is the mounting for the piston rod / coupling rod joint. This comprises a flange for the gudgeon pin and a boss into which the piston rod end is secured. The boss was first turned onto a mild steel blank, and was left overlength for subsequent adjustment when the motion was assembled. The crosshead blanks were then milled to profile and section, incorporating the bottom plate of the crosshead. Twelve bolts were prepared. The threads were turned from the ends of the bolts which were then drilled for split pins. In course of erection on PIXIE the leading two bolts of the assembly are entered from above the crosshead, the remainder from below. It was important that the assembly on Pixie replicated that employed on PIXIE, otherwise with all six bolts entered from the top, the projecting ends of the trailing bolts would have interfered with the movement of the coupling rod. Detail such as this must be studied carefully at the time of survey as each locomotive is likely to have some such idiosyncrasy. Omission of such information from the notes made at the time of survey could result in problems at the time of erection or worse, damaged parts at the time of air testing.

The hole for the gudgeon pin was then drilled and a phosphor bronze bush pressed-in. To ensure a good fit the gudgeon pin is machined from silver steel, brought to a high finish in the lathe. As is general practice on PIXIE the ends of the pins are turned down and fitted with split pins for safety purposes.


Slippers are of gunmetal and provide a nice exercise in milling, either in the milling machine or the vertical table on the lathe. As with many thin or slender components it is worthwhile soldering, or bonding the material to a more substantial section of metal. This backplate can be more readily gripped in a vice, or if necessary, bolted to a faceplate or bed. In this way machining can be carried out without fear of a nearly finished part moving under the cutting effort and becoming scrap. This technique was adopted for the slippers which were brought to within a few thousandths of the calculated finished size prior to filing and burnishing to fit within the crosshead, ensuring a nice easy action. There is a tendency to make parts too slack a fit at this stage. It must be bourn in mind that the running-in process affects every part of the motion and parts which seem quite tight at first, or trial, assembly soon free-up as the running in process proceeds ( using plenty of lubricant of course! )


This is a critical part of the construction. Parallelism (if there is such a word) between the piston centre line and the slide bar being most important for free running. Fortunately in the case of Pixie there was an opportunity in the course of the erection process to maintain this parallelity ( ? ) by adjusting the level of the bearing surfaces at the top of the glands and if necessary insertion of shim washers at the bracket on the lower edge of the motion plate.


PIXIE is fitted with Hackworth valve gear. This gear is sometimes derided by pundits because vertical movement of the axles upsets the valve events to some extent. This has not proved to be too noticeable in the case of Pixie and indeed the Hackworth gear is employed effectively on what must be the hundreds of Sweet Pea locomotives, designed by Jack Buckler, running on club tracks about the country. Jack Butler contributed an excellent article to ENGINEERING IN MINIATURE May 1986, Volume 7, No. 12 examining some of the criticisms and recommending methods of optimising performance. These include adjusting the height of the weighshaft and ensuring the correct ride position of the axle to ensure that the die block pivot moves equally about the weighshaft centre. In a more recent edition of ENGINEERING IN MINIATURE Vol 15 No.1 June 1993 a highly technical article, almost of Ph.D. standard by Nigel Bennet describes the computer simulation of Hackworth valve gear for Edward Thomas and incorporates a lot of useful detail on the motion work for Sweet Pea.

In the case of Pixie the critical dimensions of the prototype were replicated as accurately as possible and in the air trials these worked satisfactorily. no doubt the events could be fine tuned in light of Messrs Bucklers and Bennetts discussions, and maybe they will, when time permits perhaps. Meanwhile Pixie's performance is quite satisfactory with her replicated sizes incorporated.

Hackworth gear takes its motion from the return crank fixed directly opposite the crank pin i.e. on a line passing through the crankpin and centre of the axle. The vibrating lever attached to the return crankpin connects with a pivot pin on the die block which runs within straight guides set upon the weighshaft. With the guides in line with the centre of the axle the valve rod pivot on the vibrating lever is driven through an oval path of which the length of the horizontal axis equals twice the valve lap plus the lead. Rotating the weighshaft alters the inclination of the guides and changes valve events accordingly.

The size of the components of the valve gear was scaled down from the prototype and cardboard templates were employed at the drawing board to ensure that all would work when assembled onto Pixie. Whilst a spare weighshaft and guides was produced having extended slides for experimental purposes, to date these have not been installed.

One end of the return crank blanks was drilled and bored to match the shouldered crankpin. The other end was drilled for the vibrating lever pivot pin. The blank was milled to shape and the pivot pin set into place using retaining fluid. The return crank itself was secured to the crankpin by a slotted and bolted connection, aided by a further touch of retaining fluid.

The vibrating lever on PIXIE has a small marine type bearing at the lower end, and two bushed holes for the die block and valve rod pivot pins. As with the connecting and coupling rods, the marine bearing permits some small adjustment to be made to the geometry at the time of final erection. The vibrating lever blank was cut from mild steel, the pivot holes drilled and then milled to final shape. When milling bosses, or indeed when shaping the ends of any of the rods or levers by means of an an end mill in the lathe or milling machine with the material set upon a pin, it is important to remember that the material must only be moved against the direction of the cutter. Failure to observe this will result in the cutter attempting to ' climb ' onto the material, the resulting snatch causing damage to the part or, worse, the operators fingers ! Phosphor bronze bushes were pressed into the pivot pin holes.


Oil cups are fitted at each pivot point and adjacent to the bearings. These oil cups like those on other narrow gauge locomotives add a touch of individualism, they are of brass, of interesting shape and can be polished, and are ' perched ' somewhat precariously it seems, on the vibrating levers and die blocks. Those on the vibrating lever were milled to shape prior to fitting the tubes which convey oil to the bearings. These tubes are fitted into drilled holes with just a touch of bonding fluid, sufficient to fix them without blocking the tubes. Each cup is topped with a screw plug turned from a BA bolt, the head being brought to profile then knurled prior to the arisses being removed. A small centre bitt was used to drill the countersunk breather holes in each plug.


The weigh shaft is of silver steel, shouldered to enter the back plates of the guide assemblies. The shaft is rotated from the reversing lever by a reach rod which is connected to a lug on the bottom of the right-hand guide. The cheeks of the guide are milled from gauge plate, later hardened by heating,and are attached by bolts and a locating pin passing through the backplate and the weighshaft.

The weigh shaft brackets were fabricated from steel angle and plate, the lower section being welded-up. The top plate has small seating pieces sweated into place to accomodate the stud nuts when the two parts are assembled. Once the two parts are assembled the hole for the weigh shaft can be drilled and reamed. The brackets are assembled onto the frames by bolts passing through the frames and the frame angles. Ideally the position of the brackets should be determined when the whole of the motion work has been completed and the valves inserted onto the rods within the valve chest. At this stage the correct location can be established by trial and error with the brackets clamped to the frame angles.

A modification to the guides has been suggested by Jack Buckler in describing ' Sweet Pea ', ENGINEERING IN MINIATURE. In his redesign the guides are milled from one block of mild steel, a slotted cover plate being attached by countersunk screws onto the front of the guides so formed. This appears to be a simpler form of construction and may commend itself for some models.


Die blocks are of tee section and have a shouldered pin for connection with the top of the eccentric rod. this pin is entered from the back of the die block, a further shoulder seating in a counterbore in the back of the block. An oil cup is mounted on the top of the die block. With a good sliding fit achieved within the guides, the die block and pin were case-hardened by heating and treatment using ' Casenite '


The lever was prepared by sawing and filing from a piece of mild steel. The handle, with a pin for insertion into the top of the lever, was turned from a piece of square material and soldered into position. The latch block prepared by filing was soldered into position at the same heating. The latch itself was prepared by soldering a blade onto a square steel shaft with the top milled and drilled for connection to the latch lever. A small spring maintains the latchblade in contact with the sector plate. A good source of high quality springs has been found to be the plastic video tape case. On opening a defunct case one finds a number of small stainless steel springs also some stainless steel rollers as well as several self -tapping screws, all useful candidates for the scrap box !


The slide valves were machined from gunmetal, heeding LBSC's advice that dissimilar metals should be used for slide valves and portface. The blocks of gunmetal were turned to overall size in the four jaw, leaving a little on the length for final finishing work. then the cavity for exhaust steam was milled out using an plain end mill and finished with a bull-nozed bitt, care being taken to work within setting-out lines lightly scribed on marking blue.

The slots for the valve rod and the recesses for the collars which are used in place of a buckle were milled then filed to shape, care being taken that the fit permitted the valve to slide freely from end to end of the valve spindle when set in the steam chest and before the locating bushes were installed. The face of the valve was rubbed on a sheet of medium wet-and- dry paper ( used dry ) to impart the minute scratches which, in service, retain a film of oil and assist in maintaining a tight joint under the steam pressure in the steam chest. it is of course important to ensure that the valve block is free to lift from the portface, it is held in contact during service by the valve chest steam pressure but has to be free to overcome any problems with condensate in the early stages of firing whilst the cylinders are cold.

The arrangement of threaded valve rods and nuts with locknuts employed for fixing the valve blocks to the rods on PIXIE is replaced on Pixie by 4 bronze bushes ( 2 per valve block ) machined from hexagonal bronze bar. These bushes were initially secured to the valve rod by mild steel grub screws which whilst proving sound enough in service, started to rust. These have been replaced by 5 BA set screws machined from stainless hexagonal bar. The hexagon head of the set screw permits greater fixing force to be applied once the correct setting has been achieved than was possible with the earlier grub screws. These stainless steel screws are secured in service using a retaining fluid.

A general point regarding grub screws which can save problems with head breakage in the case of smaller sizes is that the body and most of the head of the screw should enter the tapped hole. In this way, when effort is applied using a screwdriver, the lobes at each side of the slot in the head end are supported by the walls of the tapped hole in the bush, wheel or whatever the screw is intended to secure. Where possible, when set screws or grub screws are used, the spindle or axle should have a small flat or dimple on which the tip of the screw can bear when tightened. The use of two screws in a bush permits a small drill to be passed down one tapped hole to form such a location whilst the bush is secured by the other fixing screw.


As an aid to determining slide valve dimensions, valve setting and visualising the events of the slide valve assembly the modeller may care to make a pair of dummy slide valves. These can be simply prepared using 1 / 8 inch perspex sheet. The exhaust cavity position can be scribed on the base and further piece of perspex cemented vertically onto the ' valve '. The upright can be drilled for the valve rod and the valve spindle installed in the usual fashion, using grub screws or whatever fixing is favoured. Whatever the final arrangement for securing the actual valves to the spindle, for the dummy assembly two bushes fitted with grub screws allow trial settings. With the dummy valves set in place the engines can be taken through their cycle and, with the length of the valve scribed on the perspex, the valving checked out. This arrangement has proved particularly helpful in the case of Pixie where the motion was constructed to dimensions taken from survey and small innaccuracies arising in construction called for some adjustment of the valve lengths to achieve optimum performance. Where a model is being built to proven details this eventuality should not arise although it is still most interesting to see the valves in motion and to visualise the events.


The valve spindles are lengths of rustless steel, their ends terminate in bosses which are turned to shape, drilled and reamed for phosphor bronze bushes prior to being silver soldered into place. After the soldering process the bores are cleaned up by a pass of the reamer and a small bush pressed into place in each. When working on bushes of this size, the bore of the bush is often reduced by this squeezing action and require a pass of the reamer to bring them back to size. Small oil cups are mounted on the top of each bush, a fine hole being drilled to convey the 'amber liquid' to the bearing face.


These are from silver steel with mild steel forks soldered at each end. The forks are profiled by milling and filing. Milling the rounded ends is carried out on a pin mounted in the small 4 jaw chuck attached to a baseplate which can be bolted onto the table of the milling machine. Prior to the aquisition of the milling machine such work was carried out using a pin mounted in a machine vice secured to the vertical table. In each case care has to be taken to avoid climb milling with resulting snatch on the part. Prior to installing the valve rods onto the locomotive the silver steel pins were fitted into the forks and holes for split pins were drilled through fork and pin. A very small drill was used, run at the highest speed obtainable on the drilling machine.

Work of this nature is carried out on a sensitive drill constructed from a set of castings purchased from MESSRS REEVES of Birmingham. The drill formed an exercise in machining and boring and building the drill served to teach many a lesson. The result is a drill with which minute holes of considerable depth can be drilled with confidence, an extended arm controlling the movement of the quill giving a very direct ' feel ' right to that critical point where the drill breaks through on the underside of the material. One small amendment made to the design of the machine is the addition of a perspex guard around the twin pulleys driving the quill. One's face is often quite near to this area when concentrating on that last, critical hole in some assembly that represents some hours of work, and a broken drive belt could seriously damage a person's good looks ! Prior to building the sensitive drill I used an auxiliary drill chuck set in the lathe and used in conjunction with a drill-pad set in the tailstock, feed being achieved by the tailstock handwheel. This is quite an exciting operation, particularly with the lathe turning at top speed to suit the smaller number drills. Drama can result if, with the Super 7 and its self-ejecting mechanism, the drill-pad reaches the point of ejection before the bitt reaches the bottom of the hole !


Valve setting was carried out at the appropriate stage in the erection process. The best description that we have encountered in any book is that by LBSC in his account of building ' TICH '. As he always did, old LBSC describes the process in the simplest of terms, and there seems little point in repeating any but the key points here. these are basically that the 'crack ' of the port must occur on each dead centre In the case of the Hackforth valve gear, the inclination of the guides on the weighshaft determine the amount of the port opening. Unfortunately, compared with the Walschaerts' gear there is no combination lever to hasten the opening of the ports to exhaust, inlet and exhaust taking place at similar pace. This has not proved to be a problem in practice however and the little locomotive has ample pulling power.


The brake standard was turned from a piece of mild steel reinforcing rod, bored a clearance fit for the rod. The standard is supported by two angle steel straps bolted into the frame supporting the footplate. On Pixie it is set-up on the right of the foot plate although in some previous period it has been sited on the left. At footplate level the standard is based upon an oval plate and incorporates two angle steel supports between this plate and the standard proper. At the top a bracket takes support from the frame of the rear weatherboard. Near the top of the rod and within an enlarged section of the standard is a collar which, being pinned to the shaft, acts as a bearing resisting the thrust as braking effort is applied by rotating the brake handle clockwise. As the force set up by putting the brake off is relatively small, a keeper plate, retained by small round head screws tapped into the top face of the standard, is sufficiently substantial to provide a bearing surface. this is one of the few places in railway mechanical engineering that such screws are encountered.

The rod passes through the footplate into a substantial bush set in the fork of the actuating lever. the thread on PIXIE is of course a buttressed thread, that on Pixie, a ' common or garden ' Whitworth thread... we did say that Pixie was a near-scale model ! The brakeshaft sits in two bushes bolted to the frames. Actuating rods form a connection between cranks on the brakeshaft and transverse rods passing through clearance holes in the bottom of the brake hangers. The actuating rods are maintained in the correct transverse position by bushes and split pins, some allowance for adjustment to the brake set-up is provided by a series of holes in the back end of the actuating rods.


Smokebox saddle. Cylinders, ports, passages, covers and glands. Pistons and piston rods.

Smokebox door, coupling block and location of axle driven pump

Smokebox and and smokebox door. Hinges, darts and rings. Wet header, steam, blast and petticoat pipes. Blower ring.


The smokebox was turned from a piece of steam barrel, not on first sight a prepossessing piece of material. Turning in the four jaw revealed an immaculate piece of high quality steel however and this was soon brought to the correct diameter and length. Before removal from the lathe, lines were scribed longitudinally onto marking fluid on the outer surface at 90 degrees intervals as a guide for use later in setting out the position of the chimney, the openings for the steam and exhaust pipes, and importantly the holes for the bolts which are used to connect the smokebox to the saddle. This marking was aided by using the simple dividing attachment described later in the book, the lines being scribed using a point tool in the topslide tool holder and traversing the saddle.

The position of the hole for the chimney was next established and marked out, again using marking fluid. The tube was then set on end and clamped to the saddle with packing such that the centre of the hole was on the axis of the lathe. The hole was started using a centre bitt, followed by a succession of drill bitts up to the largest available, finally being brought to size using a boring tool mounted in the four-jaw chuck.

Two pieces of ply were bandsawn to the profile of the outer diameter of the smokebox tube. Spaced by a block of wood these permitted the location of the tube, on its side, on the table of the drilling machine. Where just a few operations are to be carried out, such an arrangement, whilst quickly produced is adequate in providing a sound support. If there were a number of similar operations something more substantial would be employed. Apertures for the steam and exhaust pipes were then set out. These openings were formed by chain drilling and some careful work with ' junior ' hacksaws. The ends of the saws being wrapped with insulating tape to avoid damage to the inside of the smokebox tube. Two saws were used, one the common ' junior ' saw with the blade parallel to the handle, the other with the frame cranked through 90 degrees enabling the cuts around the perimeter of the tube to be made. The edges of the openings were finished by filing. Holes for the bolts securing the smokebox to the saddle were then set out and drilled, still using the ply cradle to support the tube, these would be subsequently transferred to the top of the saddle casting.


A disk of 'mild steel' yielded the smokebox door, after a fight that was ! The blank proved to be extremely hard, possibly as a result of some long-past cropping operation. This is a problem sometimes encountered when ones material comes from scrapboxes, breakers yards and similar sources, although a problem which will not be encountered by modelmakers who patronise the trade supplying ' bona fide ' model engineering materials ! The result was that having taken the edge off, and the temper out of, several tools it became necessary to set up green-grit grinding wheel in a power drill and having protected the lathe bed with rags, applying this to the blank as it rotated in the lathe. Unorthodox as this may have been, it did the trick. Fortunately there was no hard spot in the centre of the piece, and after drilling the blank was mounted on a bolt whilst the inner face was relieved to leave just a small ' land ' which would ensure a good seal on the front smokebox ring. What LBSC would have described as a ' nobby ' handle, and what the writer sees as a Victorian cupboard door handle is screwed into the central hole to complete the door. Other prototypes employ darts and crossbars with perhaps keys and locking handles, these provide some exercise in watchmaking. Crossbars to retain the dart are seated into brackets rivetted to the front smokebox ring. These are not needed on Pixie as the door is retained by two angle steel brackets and ball handles


On PIXIE, as on a number of narrow gauge locomotives, the smokebox front is a piece of plate mounted on the firebox by a rolled steel angle. this arrangement is represented on Pixie by using a one-piece gunmetal casting turned to profile. Dummy bolts in the front face represent those connecting the plate and ring of the prototype. Such an arrangement often proves simpler in model form than would fabrication using separate components, in this case it makes the sealing of the firebox easier to achieve.

A plaster pattern had been prepared for the smokebox ring castings. The pattern was ' horsed ' as described earlier and yielded the two castings which, whilst crude, contained the necessary material. The rings were prepared by first machining the external form with the casting set-up on the reversed jaws of the four jaw chuck. The outer face of the front ring was turned to a snug fit within the smokebox tube, leaving the small projection which simulated the thickness of the front plate where it laps over the front edge of the tube. Set-ups of this type must be approached with care as thermal expansions tend to loosen the grip of the chuck. Frequent checking is essential. With the outer edges machined, the face and inner section of the casting could be brought to profile. This was carried out with the ring gripped within the jaws of the chuck, here again care is required to avoid distorting the now quite slender ring whilst tightening the chuck. The back ring was turned to a neat fit within the tube then set aside until the inside diameter could be ascertained from the boiler.

The front ring was next drilled for the bolts which on the prototype connect the front plate and the annular angle. The twenty bolts are installed nut outside, one shares the duty of supporting the lamp iron at the top in front of the chimney.


Mild steel plate was used for the hinge plate. In order that the back face could be turned to match the form of the door the blank was cut to rough outline and set up on a faceplate. A secure fixing was obtained by using bolts screwed into tapped holes in parts of the hinge plate which would later be machined away. With a good fit obtained between the hinge plate and the outer face of the door, the blank was filed to profile.

To ensure the airtight joint essential to good draughting under steam, care must be taken with the geometry of the hinge points. A few minutes work with pencil and card will reveal that the hinge point must be so placed that the door opens both outwards and away from the front ring of the smokebox. It may be necessary to slightly relieve the edge of the door adjacent to the hinge to ensure clearance. Once the geometry has been resolved the hinge plate can be milled or filed to final profile and rivetted onto the door. The hinge lugs are filed up to shape from square mild steel, set into the four jaw and turned to accept a thread. A bright steel hinge pin is lightly tapered at one end and fitted with a collar. The lugs can be slipped onto the pin which is passed through the hinge plate and the whole offered-up to the front ring. It may prove necessary at this stage to make some small adjustment to the shoulders on the lugs to bring the perimeter of the door into contact with the front plate to ensure the essential airtight joint when the door is closed.

Twin steel angle brackets secure Pixie's door in the closed position. These are mounted on studs set into the smokebox front ring and are tightened by two ' ball headed nuts ', which present a nice piece of ornamental turning. This is an example where a note on the survey sketch would have resulted in a more authentic model, these ball handles appear on later inspection to be of gunmetal and not mild steel as reproduced ! The manufacture of these handles was however most a enjoyable exercise. The profiles were generated offhand, using a bar clamped under the MYFORD tool clamp as a tool support. After parting off the blanks with a ball at each end were set into the drilling machine vice, sandwiched between two scaps of steel having countersunk holes to provide a locating grip for the blanks. With the blanks set to the correct angle the larger ball was then drilled ad tappped for the stud which is set into the front face of the smokebox.


The wet header was turned from a piece of gunmetal rod, it sits in a bush soldered into the smokebox tubeplate. Three stainless steel bolts pass through the header into holes tapped into the boiler bush. The steam pipe ( from the regulator ) enters the header and the joint is sealed by an ' O ' ring. A small diameter pipe from the lubricator delivers oil into the steam at this point. Whilst generally this is done by inserting tees into the steam pipes nearer the cylinders or by means of displacement lubricators mounted on the steam chest covers, the current arrangement appears to be satisfactory and does not lead to priming as the steam ports of the Stroudley type regulator are set high in the steam dome and thus considerably above the level of the outlet at the header.


After careful annealing, the steam pipes were bent to shape and fitted with the flanges to connect with the cylinder casting. The flanges were silver soldered to the pipes and care was required to ensure the correct orientation of these flanges to suit the location of the fixing studs in the cylinder block. The upper ends were then soldered into the wet header during the same heating as the lubricator pipe.

Exhaust steam pipes were formed from 1 / 2 inch diameter copper pipe which had to be annealed frequently in the course of producing the reverse bends between the cylinder block and the connection for the blower ring. A jig was prepared replicating the actual position of the holes in the cylinder casting for the blast pipes and the proposed location of the blower ring, determined by the type of ring selected. The connection between the twin exhaust pipes was filed up from two pieces of mild steel, the two halves were clamped together and the holes for connecting bolts drilled. The assembled joint was then drilled and tapped for the installation of the combined blast pipe and blower nozzle and a small slug of metal prepared to fair-in the joint and improve the flow of steam once the joint is assembled. The parts were all assembled into the jig and the flange and connecting block joints were all soldered at one heating. In the case of all flanged connections the flanges are prepared as a pair and then used to spot the holes for the fixing bolts. a centre popped mark ensures that the flanges are used where intended and thus minor discrepancies in hole centres are overcome.


As the ring on PIXIE is a loop of small diameter tubing and replication in model form would not perhaps have provided a satisfactory draught, a rather more sophisticated type of blast ring was incorporated. This took the form of of a combined blast nozzle and blower with an annular chamber around the nozzle pierced by very small holes to provide the blower jets. The arrangement has been detailed in several articles in the technical press over the years.


A piece of copper pipe from a recent domestic plumbing job was annealed ready for spinning. The pipe was plugged at one end to avoid distortion in the three jaw, the other end was then spun-up to a push fit in the chimney base using a polished piece of silver steel supported from a piece of stock bar clamped into the tool holder. The first plug was removed and the newly expanded end was plugged then spun. Frequent annealing permitted the eventual achievement of the required form. On erection the petticoat pipe would be fixed using a smear of instant gasket.


A boilerfrom a proprietary supplier, this one for the Allchin


Boilers. Home built vs. proprietary supply. Construction techniques. Location. Cleading.


Tackled systematically and following upon the advice of the masters such as L.B.S.C, Alec Farmer Martin Evans and Keith Wilson the task of boiler construction is within the capability of most model engineers. Where possible the modelmaker will be wise to enlist the assistance of an experienced person in constructing his first boiler, in the case of the writer this person was a dear old gentleman who had for many years been employed in all types of plate fabrication and jointing using welding, brazing and soldering and many different techniques for a range of materials. The following comment arises from experience in building that first boiler, the advice recieved from the aforementioned old gentleman who provided me with the basics of soldering practice. Other people have joined evening classes or attended short courses held during the vacation at various Universities and colleges. Usually these events are organised and staffed by experienced technicians prepared to guide participants through the tricky parts of the operation. They also provide access to that most important ingredient of good boilermaking, the substantial heat source.

The keys to successful boiler building are, a proven design, care in obtaining good fits between parts, absolute cleanliness, and an adequate heat source. Given these essentials together with space and time the beginner can produce what L.B.S.C described as the ' kettle ' for their locomotive. In the case of Pixie, experience gained in producing the boiler for an earlier 3 1 / 2" gauge locomotive had convinced the writer of the difficulty of providing sufficient heat using existing equipment, both propane and butane torches. To avoid problems, and economise on time available for model engineering work, the decision was made to sublet the boiler construction.

This proved to be a wise decision and should there be any doubt in the modellers mind as to whether he can afford the equipment or obtain the necessary support and assistance, he would do well to consider obtaining a ready-made boiler from one of the many experts advertising in the trade journals. In this case it may only be necessary to phone several suppliers to obtain a quotation for a boiler for a specific model to a published design. Various well known specialist suppliers spring to mind and one cannot go far wrong in purchasing from any of these. Although there are a few widely known suppliers, prior knowledge passed on by a fellow club member may direct ones choice to another fabricator who although perhaps not so widely known has proved his ability in a previous supply.

Non-standard boilers such as that required for Pixie call for a slightly different approach. Here, whilst the overall dimensions are known, the design of the boiler and its staying is best left to a builder with with a proven track record. After obtaining several ' budget ' figures it was decided to employ a boilermaker with an excellent record, although the budget price quoted was not the lowest. The fact that the builder had for many years manufactured a range of standard boilers, so much so that his firm had a ' household name ', generated confidence. As well as being approachable the proprietor was exceedingly knowledgeable and was most helpful in assisting with design input.

Armed with outline details, especially critical dimensions of items such as the location of the tank support brackets, firehole door, steam fountain and dome, arrangements were made for a meeting at the firms stand at the MODEL ENGINEERING EXHIBITION. Here the boilermaker advised upon such matters as the number and location of tubes, size and bore of bushes, desirability or otherwise, in fact otherwise, of superheater and so on. After the meeting these details were transferred to a finished drawing on which the final quotation was to be based. At this stage the builder knew exactly what was expected of him and as a result could provide an accurate quotation of what the cost would be, an essential to a satisfactory business deal. Here it should be bourn in mind that money that we spend on a boiler today would, a few years ago, have purchased quite a reasonable motor car ! Boilers from commercial suppliers come with a boiler test certificate which must not be confused with the certificate required by clubs where the locomotive is to be run. The certificate simply indicates the pressure contained during a raw test, probably before the bushes were drilled out to tapping size. Although the criteria vary from club to club, officials will require more stringent testing, including the efficiency of the safety valves, before they issue a club certificate.

In the event that the model engineer decides to go it alone, it makes sense, in the case of a standard model for which designs have been published, to make use of the ready flanged plate service offered by many suppliers. This will save a lot of time and, whilst costing more than obtaining the raw material and flanging up over hardwood or mild steel formers, can prove more economic in time, and material. The price of flanged plates can be maintained at an economic level because as well as re-using the formers many times in the course of production, the proprietary supplier has a better chance of planning his material usage in the course of a production run. He can work far more economically than does the individual attempting to cut a set of plates from his one piece of sheet copper.


A first class boiler, and we must not settle for anything less, demands excellent materials. These will be found in the catalogues of the proprietary suppliers to the hobby. With the materials to hand, the plates must be set out in the most economic fashion. Here paper templates are essential, these can be shuffled around on the raw material to provide a best fit and thus the utmost economy in material. The lines are best transferred to the material using a waterproof overhead projector pen. In this way scribing is avoided at this stage and the lines can be simply erased using a rag soaked in methylated spirits. Where essential the lines can be lightly scribed once the best layout has been achieved.

To cut the plating material obtain the largest pair of snips possible, grip one handle in the vice and slip a piece of pipe of suitable diameter over the other. With this arrangement, not only can considerable force be applied when necessary, but also very small cuts can be made slowly with great precision.

Whilst the platework can be prepared using the range of hand tools owned by most model engineers there are one or two additional tools which can make platework easier. a planishing hammer, a cross faced hammer and a large pair of tongs, mine are in fact medical forceps as used in delivering babies ! Plumbers gland or automotive pump pliers are useful for grasping hot metal. Also overlever wrenches such as the ' MOLE ' wrench make useful quick release clamps.

When flanging, anneal plate frequently by heating to a medium red then plunging it into cold water. The tendency to take a plate ' just a little bit further ' before annealing must be avoided as cracks will result due to work hardening of the metal. When preparing the tube holes in the tubeplates, countersink the holes, roughen the edges of the holes and file small nicks in the periphery of the hole to promote the flow of solder into the joint and around the tube.

Where it is necessary to drill copper plate for tubes, bushes and the like, Alec F. Farmer suggests a useful dodge which will be appreciated anyone having experienced the snatch of a drill as it breaks through the material. He recommends that drill bitts for copper plate should be ground in such a way that the tips cut the plate almost in the manner of a trepanning cut and so that the lips of the drill remove a curl of swarf from within this cut. The point of the drill seats within a pilot hole drilled after the centre of the hole has been set-out an positively spotted using a centre punch.


On the matter of soldering one can do no better than to read ( and view in the series of more than 300 photographs ) the advice of Alec F. Farmer. Not the least of the tips passed on in this treatise, 'Model Locomotive Boiler Making' is the mixture that Alec uses when making up flux. He uses 2 heaped TABLESPOONFULS of flux powder, a 1 / 4 TEASPOONFUL of household detergent and mixes these to a creamy consistency with water. Using this mixture we are told that the detergent breaks down the surface tension of the liquid, ensuring penetration into the joint as well as degreasing the components being soldered, the greases being replaced by flux and then burnt off. As well as adding the detergent to the flux paste, cleaning and degreasing of components, prior to painting on the flux, has been found to provide the best results. With the joint painted with flux and the metal quickly brought to red heat. the flux will be seen to glaze and melt. The solder stick or wire, applied with the metal at this temperature, will melt and flash into joints. If blobs appear these have to be dispersed using a scratch, a wire ground to a point and bent through a right angle at the end. The solder will flow to the hottest parts of the joint and can be drawn along by judicious heating. Once a bright line of solder is apparent along or around the joint the parts can be allowed to cool before immersion in pickle. Final cleaning of soldered joints is best carried out using steel wool or scouring powder, after 1 / 2 an hour or so in pickle.

Firebricks as a wall around the hearth which is banked up with crushed firebrick maintain all available heat in the soldering process around the parts being worked on. A useful source of firebrick is the redundant storage heater. If it should be necessary to purchase bricks then a local laboratory supply company will generally stock the genuine article.

It is advisable to have a trial assembly of as much of the boiler as possible ' dry '. Holes can then be drilled, in key positions, for small rivets to hold the parts firmly during the soldering operations that follow. The pundits have always advised the use of differing grades of solder for the ongoing soldering operations to avoid spoiling joints made earlier. Due to the way in which the solder amalgamates with the copper in the course of the soldering process whilst advisable this is however, not essential.


Arrangements had been made at the time of ordering Pixie's boiler for the provision of blind bushes for later attachment of the firehole door hinge plate ( Care being taken to ensure that these were set on the correct side of the firehole ring ), also a bush for the spur. The firehole door with its hinge incorporating two shaped gusset pieces was set in place and clamped in the closed position against the firehole ring. With the hinge plate, again fabricated from mild steel located within the projecting gussets the hinge pin position could be marked and the hole for the pin drilled in the door gussets. At that stage the fixing bush positions were transferred to the hinge plate, the plate drilled and the hole positions spotted through to the bushes which were in their turn drilled and tapped. The spur for the latch, profiled from a piece of mild steel, was set into the four jaw for the end to be turned and threaded for insertion into the bush on the boiler. At this stage the latch lever could be cut to shape and brought to thickness, ensuring a good fit in the spur with the door closed tightly. A baffle was then cut to size which allowed it to pass through the firehole ring without binding when mounted, using a small parallel collar to space it from the door proper. On Pixie a small length of chain is used to lift the latch. In practice it is more likely lifted by the fireman's shovel, however the chain is provided and as on the prototype it terminates in a ring through the small shelf above the firehole door


There has been considerable discussion in the model press regarding the insulation values of available materials. For Pixie the asbestos replacement material supplied by the trade has proved successful. The material was easy to cut to pattern, did not stretch or distort in the course of handling and installation and was sufficiently rigid to remain firmly in place until the cleading material was fitted and the boiler bands set in place. In shaping the material the use of patterns is essential, it takes but a few minutes with paper and scissors to produce quite accurate patterns, indeed the lady of the house may well find that this is an area where she can make a contribution ! The material is best cut on a sheet of hardboard using a sharp knife or a scalpel. Prior to installing the cleading the insulation can be held in place around the boiler using soft iron wire or even the plastic coated wire sold for use in the garden. These wires serve as an extra pair of hands and can be snipped away after the cleading is in place.

A further set of patterns are required for the cleading plates. It is important when cutting the plate material to remember to allow for the overlap at the joint. The time honoured method of using snips, with one of the handles secured in the vice and a piece of pipe on the other, makes short work of the straight cuts. A nibbler is fine for this job although it may be worthwhile protecting the metal from scoring by laying sellotape along the line of the cut. For curves and for the formation of the large diameter holes for boiler fittings, chain drilling and a sharp chisel will suffice or, rather more laboriously, a metal cutting blade in the saw frame. Precautions are required to avoid scoring or burring the metal and particularly, forming creases which will prove impossible to remove.

In the case of Pixie a supply of half-hard brass was available, this was used although the pundits advise using thin mild steel Although the material is a kindly one in use and tin snips can be used for cutting it to shape, care has to be taken in installation to avoid a local sharp bend which can produce an irremovable crease in the plate. Attempts to remove such a crease will only emphasize it so it is best left alone, that is if the builder can live with it, otherwise that piece should be set aside for a less visible use and a new plate prepared. In the absence of bending rolls, and few of us are fortunate enough to possess these, a variety of formers can be used to roll the plate, clean scaffold tube, preferably the lightweight variety, even a jam jar has been pressed ( with care ) into service, work being carried out on a piece of carpet to preserve the face of the material.

The boiler bands are fairly straightforward provided that the correct length is first established using a piece of wire wrapped around the cleading which is temporarily held in place by wires, as was the insulation. Small sections of angle steel, drilled for the fixing bolts are rivetted and soldered to the ends of the boiler bands, remembering to leave a tongue of the band projecting beyond one of the angles to maintain the line of the banding. With the bands assembled, the wires temporarily retaining the cleading plating can be snipped allowing the bands to do the job for which they are designed.




Plumbing. Gland and union nuts. Union liners and olives. Valve bodies, spindles and wheels. Pumps. Gauges. Sanding gear. Clacks and dummy injectors.


One of the attractions of locomotives large or small is the profusion of pipework, both in copper and brass. Much of this is visible and can be polished, adding to the aura of a model. When preparing runs of pipe, once their routing has been established it is advisable to replicate the pipe using a length, or lengths, of fairly stiff wire. Straightened out, the wire(s) will give an accurate length measurement which would otherwise be difficult to achieve. Many of the pipes on PIXIE have a joiner in mid-run, this is helpful when it comes to erection when the two lengths of a pipe can be installed, sometimes a fiddley job, the joiner inserted and the pipes coupled up.

L.B.S.C. had a lot to say about pipes and unions and there can be little doubt that he was right to ridicule the overlarge unions seen on many standard gauge models on 3 1/2 inch gauge track. His solution was to solder the cone directly onto the pipe thus avoiding union linings which have of necessity to be of such a diameter that the pipe can be entered into them. Size is not such a problem in 2 1/2 inch scale and on Pixie linings incorporating the cone are soldered onto the pipe ends. This soldering is best carried out by applying the heat to the inside of the lining whilst touching the fluxed outside of the pipe and lining with 1 / 16" diameter rod of silver solder.


Gland nuts for valves will be of standard sizes and a number should be fabricated for stock. A small recess cut with a minute boring tool will form the locating groove for an 'O' ring seal adjacent to the butress end of the nut, steam or water tightness can be achieved with just the slightest turn of the nut on the Model Engineer thread.

Gland and union nuts are made from appropriately sized hexagonal brass bar. The bar is first centered from the tailstock then drilled using the appropriate bitt for the thread employed. Model Engineer threads are used universally for fittings and a glance at any catalogue will indicate the appropriate size of threads for specific pipe sizes. It is worthwhile making up a ' D ' bitt to complete this stage of the work and provide a square face to the buttress. Next drill a few thousandths of an inch larger than the diameter of the pipe is taken just beyond the overall length of the nut. Tapping is carried out with the tap supported by the tailstock centre, finishing with a plug tap. A touch with a a chamfer tool or a smooth file ( with a good handle ) applies the chamfers and the piece is completed by parting-off.


Making liners is an enjoyable job, especially where a four tool turret is available ( and if it isn' t, now is a good time to make one! ) The angle setting of the chamfer tool forming the cone on the liner can be set from a centre-bitt mounted in the tailstock drill chuck and once set can remain in position until a batch of work is complete. A knife tool is set into another position in the holder. With a parting tool set in the rear tool post production can commence. A length of brass bar of diameter which will just enter the thread to be cut in the union nut is set in the 3 jaw, centred and drilled from the tailstock chuck, using firstly a centre-bitt then a drill of the diameter of the pipe bore. The hole is taken rather deeper than the turned and shouldered length of the lining plus the length of the cone. This is to ensure a through hole after parting-off. Next a drill of the a few thousandths of an inch greater than the outside diameter of the pipe is taken down to the line of the eventual shoulder, this provides a positive stop for the pipe end. The parallel portion of the liner is now turned to a diameter a fraction under that of the hole in the buttress face of the union nut. The chamfer tool is plunged in leaving a short length of the original diameter between the cone and the shoulder. The parting off tool completes the job, bringing the cone to length. If the dial readings are noted as the first component is made then manufacture of a batch of linings becomes almost automatic.


Small cones or olives can be used where it is required to reduce the overall size of a union or connector. From the range of copper tube available it is usually possible to select one that is a sliding fit on the size of pipe to be connected or secured. To obtain the correct cone angle a chamfer tool can be set to a centre bit in the headstock drill chuck. Several cones can be produced in quick succession from the copper tube held in the three jaw chuck. A sharp tool and high speed will be required to avoid distorting the tube, particularly when parting off which should be done leaving a small shoulder. These cones soldered into place on the tube, using the merest touch of solder require very little force on the union nut to achieve a tight joint.


Manufacture of valves is quite a time consuming operation and some builders may wish to buy theirs from the trade. It is always difficult to get exactly what is required for a specific prototype however and the following should prove helpful to those who wish, or are forced, to have a go for themselves. Trade suppliers market useful material for the body of valves in the form of tees and double-tee castings. These save the constructor the trouble of building-up bodies from bar or rod.

At this stage a set of tapped bushes will prove to be most useful. Produced in the three jaw, drilled, tapped and marked for number one jaw in a series of Model Engineer thread sizes, these bushes permit bodies to be mounted securely for turning and drilling operations without thread damage. The set of bushes, which should include 3 / 16, 1 / 4 and 5 / 16 sizes as a minimum, can be set aside and re-used, so that time spent in their manufacture will be amply repaid as work proceeds on further models . The valve body casting can mounted in the four jaw, the first end to be threaded centered and lined with the axis of the lathe. This can be achieved by centre punching the approximate centre of the spigot, lodging this on a centre set in the tailstock and then lining-up the body by adjusting the chuck jaws, the tailstock can then be withdrawn. With the body thus set the outside diameter of the body can be machined as far as the intersection of the limbs of the valve. The limb can be brought to length, centre drilled and a hole drilled a slight clearance over the pipe size. A suitable size of centre bit can then be used to form an internal chamfer to suit the liner. The external thread for the gland nut can now be cut, ideally using a tailstock die holder. In the absence of a tailstock die holder the nose of a tailstock drill chuck can be brought along to bear on the back of a standard diestock to maintain the correct line of thread. With the body still mounted in the mandrel the hole for the valve spindle can be drilled to the required depth and the body tapped to suit the spindle. Once one limb of a casting has been threaded externally it is a simple matter to mount it in one of the previously prepared mandrels, set in the three jaw chuck for further turning and completion of the other limb(s) ready for union nuts. With all limbs completed it becomes matter of a few minutes work with a succession of files to complete the body profile. For this operation the valve body, still in the mandrel can be transferred to a vice or the chuck set on the mounting plate described in the chapter on simple tools and attachments.


These are generally made from rustless steel and are best made several at a time. A length of material set in the three jaw and fine turned to the appropriate spigot diameter. this may if required be coned using a smooth, flat file. Threading is carried out using the tailstock die holder or a diestock supported from a tailstock drill chuck. The neck of the spindle can then be turned to diameter and, if required, the end filed square for attachment of the valve wheel. An accurate square section can be achieved using the simple indexing attachment described later. After parting off, if the valve wheel is to be fitted permanently, the gland nut should assembled onto the spindle before the end is swaged or punched. Where the wheel is to be threaded onto the spindle sufficient length of thread should be allowed to enable a lock-nut to be fitted.

It is important to ensure that valve spindles remain captive in the body of the valve whilst in their open position. Spindles which unscrew completely from the body of the valve are potentially hazardous in service. Spindles can be rendered captive by reducing the spindle diameter where it passes through the gland nut such that a shoulder is formed beneath the ' land ' of the nut. When the valve spindle is unscrewed this shoulder bears against the buttress so formed, preventing its complete withdrawal.


Those who manufacture their own valves will be faced with the production of valve wheels that convey the spirit of the prototype. There have been several methods suggested in the model press and the following are though to provide convincing wheels. A.G Neville in MODEL ENGINEER Vol 146, No 3644 described a method using six lengths of thick walled tube, silver soldered around the circumference of a central tube. These were held in place for the soldering operation by soft iron wire. Succeeding operations were to part off suitable lengths for individual wheels, lightly countersink the holes and file the rim to required final section.

LBSC describes a method of building ' cool ' wheels where a coiled wire is soldered around the periphery of a turned brass wheel centre. Requiring considerable care in production this does produce a good looking and effective wheel.

A person coming new to model engineering will do well to buy a few nuts and cones, or even a complete valve to serve as patterns for their plumbing work. From these the general proportions can be gauged rather more easily than from the many drawings that are published in the model press. Of course the fanatic can produce fully scaled models of fittings if they wish, but most of us are happy with conventional model fittings. as has been said before, we set our own standards to suit our purses and the time available.


The use of ' O ' rings for packing pistons and glands ensures that the very beginner can manufacture efficient pumps. In the days of graphited yarn and tallowed hemp, the amount of packing and the tightness of gland nuts was critical, too little and pressure was lost, too much and excessive friction resulted. With careful machining, in accordance with the manufacturers instructions, pistons and glands will be perfectly tight with the least possible amount of friction developing. In the case of piston packings some small pieces of tool steel ground to the correct width for the main sizes of ' O ' ring will save time in preparing the necessary grooves. For gland nuts, small diameter boring tools, again ground up to the correct profile can be set into the tool holder described later to machine the internal grooves in glands.

Pumps should be located such that the eccentric rod is as long as possible. This reduces the wear on glands as well as cutting out a certain amount of the friction in the system. The length of the ram is also critical in this respect.. In Pixie the pump is mounted behind the front beam, below the saddle, on a block profiled such that the axis of the pump lines with the centre of the front axle. This position was chosen as the available space was limited by the position of the main stretcher spanning between the motion plates. There would also have been insufficient clearance around the back axle for the sweep of the eccentric strap. The mounting block is secured to the front beam by countersunk screws hidden behind the front coupling, the pump in its turn being bolted to the block. This arrangement will permit the pump to be dropped down from between the frames in the event of maintenance being required.

Pump rams are best made of rustless steel, barrels are generally gunmetal castings incorporating the stand and valve box and are readily obtainable from the trade. Rustless steel balls, a suitable sized ' O ' ring and some hexagon brass bar complete the material requirements. Construction details have been spelled out many times in the model engineering press and there is little point in repeating them here. LBSC gave good coverage in his collection entitled ' LBSC's Shop Shed and Road ' which describes several types of pump and their construction. He notes the key points, for efficiency of eccentric driven pumps as being, ' fairly large bore and short stroke, ample sized ball valves with restricted lift, minimum clearance within the pump itself, and the water pipes ( especially the suction ) of sufficient diameter not to throttle the flow of water '.

A crosshead driven pump has a lot to offer, not the least being accessibility for adjustment and maintenance. Discussing the the advantages of a crosshead-driven feedwater pump in the case of ' Rob Roy ', Martin Evans rightly states that ' such a pump is readily accessible for servicing, if the pump doesn' t function for some reason such as balls sticking or becoming furred-up, it is only the work of minutes to take the whole thing off, strip it right down and clean the valves etc.'


There have been some magical articles in the model press, written by wizards who have made their own gauges complete with microscopic Bourdon tubes. The beginner will do well to buy-in his gauges from a specialist supplier. Gauges are not the thing to cut ones teeth on, particularly when dealing with the substantial pressures that our locomotives are capable of producing. Those who have an overwhelming desire to construct their own gauges should study the article in Model Engineer. This will probably prove sufficient to prompt them to start out for the nearest trade supplier of the finished article! The gauges on Pixie were purchased from a suppliers catalogue, one for the boiler pressure, taking its feed from the steam fountain, the other measuring valve chest pressure. The latter is fed from the right hand valve chest via a ' tee ' incorporating a drain valve complete with a tap reminiscent of that to be found on a antique Samovar, complete with ebony handle, replicated here by a turned, stained and varnished piece of hardwood on a mild steel spindle.


Sandboxes are a prominent feature on most narrow gauge locomotives, being especially necessary in the conditions in which narrow gauge locomotives operate. Making the sandboxes on any narrow gauge locomotive should give the builder great deal of enjoyment. Whether mounted on the running plates, on top the boiler or as in the case of the Wrenn Class on the front of the saddle tank, they probably include a degree of geometry. The many patterns can be produced by turning or using a combination of turned parts and barrel or formed sheet. Sandboxes are another of the features which generate atmosphere and care was taken to replicate those on the model. For Pixie the mild steel fronts and backs of the boxes were flanged up over a piece of 1/4 inch plate which was cut to the correct profile, with allowance for the plate thickness to be used, and rounded on one edge. As the flange was to be relatively narrow flanging could be done cold without too much upset to the metal. With the profiled plate sandwiched between the template and a slightly smaller piece of stock steel it was gradually flanged using hammer with a smooth face. Progressing around the plate about 1/8 inch of flange was formed without creases. After removal from the templates the flanges were brought to uniform thickness by filing, paired off and marked in two sets. With the plates set flange to flange, the outside edges were trued locally, again by filing.

A wire was bent around the perimeter of the flange to establish the length of the strips of mild steel infill which were to form the sides, top and bottom of the boxes. The strips were then bent to the required profile and small adjustments made to ensure a neat butt joint. At this stage a small hole was drilled centrally where the flange for the cover would be located. This was thought necessary to avoid any possibility of an explosion as each box was sealed during welding! Now the three pieces of each box were wired together and the welding operation carried out. Welding was only adopted for this assembly as access was available to TIG welding equipment, brazing would have been equally acceptable. After the assemblies had been cleaned-up the filler openings were drilled and bored and flanges for the covers were soldered into place.

With the boxes clamped to the topslide of the lathe the holes for bushes for the operating rod could be drilled. As the rod passes through four bushes, two in each box, the lathe set-up was used in order to ensure that the holes lined correctly, parallel to the faces of the boxes. The bushes for the operating rod were then soldered into place, location aided by turned shoulders, bearing against the sandbox sides and the insertion of a length of rod which had been thoroughly oxidised to prevent it becoming a permanent installation! Now the back of each box was drilled for the fixing bolts which were to engage in the blind, stepped bushes in the front face of the saddle tank. These holes were spotted from the template prepared at the time when the bushes were set-out on the tank. The cone shaped feeders at the bottom of the boxes were shaped and fly cut to radius then through-drilled for the sand flow. Pipe flanges were prepared and drilled at the same time as the feeder cones to ensure a good fit on final assembly.

Filler caps and flanges proved to be simple turning jobs and one of those little tasks that give so much enjoyment in obtaining the correct curves. These are achieved using a graver held on a bar secured under the clamping shoe of the top slide, and finished, dare I say, by filing in the lathe!

The operating lever with a boss at one end and at the other a tapped hole providing a fixing for the pivot bolt connection with the pull rod was prepared from mild steel. The pull rod itself is a length of 1 / 8 inch diameter mild steel rod terminating in an eye for the pivot bolt at one end and yet another ' Victorian cupboard handle ' at the other. The travel of this pull rod is limited by a strategically placed split pin which bears against the weather plate in the extreme rearward position. Small cranks and a dart valve inside each box control the flow of sand to the copper sandpipes which extend to the track level and are supported at one spot by pipe brackets bolted to the main frame members.


Feed water on Pixie is introduces via clack valves positioned where the prototype injectors are installed. Clacks are efficient and provide trouble free running. Lately time has permitted the production of dummy injectors which fit into the bosses on the boiler and replicate injectors for show purposes. These dummy injectors, whilst outwardly similar to the prototype, are in reality clack valves in disguise. they are soldered up assemblies of pieces filed from gunmetal offcuts from other operations. The screwed cap permits the installation of a spring and ball valve just as in standard valve. At a later date a working injector will be installed under the footplate.


An extra not to be found on the prototype is a feedwater by-pass valve. On Pixie this is mounted on the left hand frame member in the space between the back of the bunker and the boiler cleading. It is unobtrusive in this position, yet easily accessible. The valve itself is an essential brought about by the axle mounted water pump and the setting has to be monitored in steam to prevent over-feed and priming. That is if the driver wishes to avoid the ' hot shower ' which will otherwise result!


Below and at the rear of the saddle tank, feeds are taken to the axle pump located behind the front beam and the hand pump sited in the left hand bunker. From the pumps the feed goes to the boiler, that from the axle pump via the by-pass valve. Sited below the front of the saddle tank is a balance pipe which also serves for emptying the tank. On Pixie this is extended and the shut-off valve is, for convenience, accessed below the level of the main frame member.


Regulator and steam dome.


Regulator block, gland and lever. Chimney. Steam dome. Safety valves. Retro-fitted pump eccentric Cylinder drain cocks.


A search through the scrapbox revealed some short lengths of gunmetal, chucking pieces from a previous model, these were to form the body and gland. The machining involved is similar to that for any gland assembly, the body being a slide fit in the large bush on the backhead, secured in place using threaded and shouldered studs. Nuts on the threaded part of the stud tighten the gland into its recess. An ' O' ring was installed into a groove behind the flange of the regulator body so as to bear on the boiler bush and ensure a steam tight joint. ' O ' Rings, whether standard or purpose-made from a kit feature more and more in model engineering and, provided the manufacturer's limits on groove sizes and fits are observed give trouble-free service.

The regulator rod in rustless steel locates into a hole in the regulator within the dome and is maintained in position by a tiny bush of the same material which, pinned to the rod, bears against the back of the regulator body at the backhead. Some small allowance is necessary to avoid any tendency towards binding due to expansion in service. The gland slides within the block, guided by the threaded studs, here again an ' O ' ring forms an excellent steam tight joint with little or no effort applied to the nuts.

The regulator lever and handle, ( sometimes seen on the prototype with the handle facing rear-wards and sometimes facing forwards ) is from mild steel and the travel of the lever is limited by an angular shroud fitted with stops at the perimeter of the face of the block. Some experiment was required to ensure that the limits imposed by the shroud allowed the full travel of the actual regulator disc on the Stroudley type regulator within the steam dome


A number of reasons prompted the choice of a Stroudley type regulator, not the least was the straightforward construction and positive action. A further factor was that access for maintenance could be simply achieved via. the steam dome. These considerations have been bourn out in service although the fixing method would be revised in a further model. Being reluctant to drill the boiler to provide a fixing for the regulator the fastening is is made through the steam dome, using stainless steel bolts. This necessitates some juggling at the time of assembly and would, in a rebuild, be modified to the more conventional fixing using a strap screwed to the boiler shell.

Gunmetal was used for the body of the regulator, and the two valve discs. The body was drilled to provide the vertical steamway then sealed with screwed plugs. The fixed disc and two fixing brackets were silver soldered onto the body at one heating. The steamway was drilled through the disc and the hole for pivot stud drilled and tapped. Care was taken to remove the burrs formed in these drilling operations as they would otherwise prevent the valve disc from making a steam tight joint with the fixed disc. A ' vee ' notch was filed in the edge of the hole in the fixed disc to assist in providing gradual opening of the port as the regulator was opened in service. The hole for the steam pipe was next drilled and tapped. The loose disc was then prepared, being drilled a close clearance on the pivot pin. Holes were drilled and tapped for the linkage bolts. These were threaded such that when fully inserted into the holes in the loose disc they provided ' pinch-free fit ' when passed through the holes in the actuating rods. All screws and pins are stainless steel and all are drilled and secured using stainless wire to prevent loosening in service. The spring maintaining contact between the valve faces is of stainless wire wound-up around a pin in the 3 jaw, only a light spring is required here as steam pressure does the job in service. The actuating levers are from rustless steel as is the rocker bar. This fits onto a square formed on the end of the regulator rod, being retained by a nut threaded onto the rod. The rod terminates in the rounded spigot which engages in the regulator body.


The chimney was turned from a piece of reinforcing steel culled from the scrap heap on a building site. It was set in the 3 jaw, drilled using the largest available drill bit, then bored out. The boring tool used was an adaptation of a design published in MODEL ENGINEER many years ago and described later in this book. This has the advantage that it can be entered into small bores and, when modest cuts are taken, provides a a long reach without chatter. The tool uses small diameter cutters which can be produced from the remains of broken centre bitts. As with all long tools however it is essential that a number of fine finishing cuts are made to counteract inaccuracy resulting from the spring of the shank holding the tool.

The outer taper was turned in several stages, setting the top slide to the required angle and moving the saddle and locking it at each intermediate position. A small rebate was formed at the top of the piece to locate the cap, and, prior to parting-off to finished length, a similar rebate formed to seat into the chimney base. The cap and base castings were next machined to the required profile with the necessary rebate and projection to ensure a good register between the parts at the time of soldering the joints.

The chimney cap was turned from a piece of gunmetal, set in the 3 jaw. It was bored and the joggle formed for the joint with the chimney body. The curved surfaces were shaped offhand, using a tool supported by a piece of stock bar set in the tool post. Once bored and profiled the cap was then parted off. The bottom casting was turned, again with a joggle for locating the chimney body and the plain portion above the flair turned circular. The part was then set on the vertical slide for fly-cutting to match the radius of the smokebox. This was facilitated by using a washer and one of the medium Myford tee bolts set into the slot in the vertical slide. as in all such operations the slide was packed from the topslide to eliminate movement and avoid chatter. Finally the flair was ground and filed to shaped. Enthusiasts may wish to produce their parts by spinning or raising them from gilding metal. This is, it seems, an aquired art and presents a rather daunting prospect to the tyro.

The three parts were joined by silver soldering and the base drilled for the bolts to fix it in position onto the smokebox. As is the case when soldering boiler tubes, several small nicks in the joining surfaces will assist the flow of solder into the joint.

The petticoat pipe was spun-up from a piece of copper pipe, an offcut from the last plumbing job carried out in the house. Frequent annealing permitted the lower end to be expanded to a profile in line with the recommendations in 'The manual of Model Locomotive Construction' by Martin Evans. The upper end is a push-fit into the chimney and stays secure with the aid of a dab of instant gasket. Whilst on the face of things this may seem to be a pretty tenuous means of fitting it permits the petticoat pipe to be withdrawn when access is required to the otherwise congested smokebox.


The steam dome on any locomotive is a joy to produce. Whether turned from a casting, from solid material or spun to shape by the more adventurous and proficient model engineer the dome has to be a thing of beauty. The fortunate model engineer may find a suitable casting in the stocks of his local supplier. He may through perusing the details in articles in the model press find a dome casting that nearly matches his requirements and can be doctored to provide the required form. It was Henry Moore the celebrated sculptor who pointed to a massive piece of material and said that the form he wanted was ' in there somewhere' ! The dome on PIXIE has been fabricated from thin sheet metal and fortunately there is a joint in the material just below the convex curve at the top. In the case of Pixie this permits a turned dome being produced in two pieces, the lower from an offcut of steel of unknown parentage, probably a cropping from a stockists scrapbox, the upper a cast piece of gunmetal. Assembly from two pieces of material also facilitated the cutting of the opening for the base of the safety valve assembly.

A length of bar for the lower dome was cut by hand from the billet of steel.... blood and sweat ! The length required was determined bearing in mind the flare which fits over the cleading. This was then set in the 3 jaw chuck and a length slightly less than the dimension from the top of the flair to the top of the parallel portion of the dome was turned to finished diameter. A small rebate was formed in the end to provide a location for the dome top section. The blank was then turned in the chuck, gripped by the reduced section and the remainder turned to to the diameter of the finished flair. The body was then bored out to clear the inner dome.

Whilst means of producing the flair by milling alone have been described by Martin Evans, the tyro will find that the simplest solution to the somewhat knotty problem of the geometry involved will be to turn the upper, external concave shape which is viewed in the side elevation of the finished dome, fly cut the seating to match the boiler cleading ( NOT the boiler itself ) then mill away the evidently surplus material before finally finishing up with a succession of files. Ideally the edges of the flare should be very thin to replicate the prototype, this is easier said than done and the individual must work to HIS predetermined standard. Again a template will assist in achieving the correct profile.

Following the sequence described above, the base of the dome was fly-cut to the profile of the cleading. The part-machined blank was mounted on the vertical slide using bolts passed into the tee grooves via a washer fitted into the inner dome recess. In instances such as this where material is to be machined whilst attached to the vertical slide, particularly where there is any amount of overhang, some packing should be used to augment the support provided by the bolts or clamps. A piece of stock bar and some small blocks secured between the piece to be machined and the top slide will provide such support This precaution will be appreciated when for example the heavier of the intermittent cuts of a fly cutter are experienced. Once again, a reminder .... when fly cutting, first set the cutter to describe a radius greater than the required finished radius and then slowly reduce the radius of the cut, whilst advancing the piece, until the eventual diameter is achieved. Frequent checking against a template will ensure a good fit.

With the seating shape generated, the excess material was milled away and the final form achieved by filing, taking care not to remove too much material in any one place at a time. The filing operations were eased by mounting the dome blank onto a mandrel secured in the vice at the workbench. Final polishing produced an acceptable lower dome ready for drilling for the various valves which adorn the prototype.

Pixie's upper dome was produced by straightforward turning, finished freehand using a tool bearing upon a length of stock bar set in the toolpost. The part was then set, protected by paper at each jaw in the 3 jaw for the inner profile and the locating tongue to be turned. Later the plan profile of the the safety valve base was set out on the top of this piece using lines scribed in marking blue and the opening milled for final filing to fit at the time of erection.


The safety valves on PIXIE sit on top of the steam dome. This arrangement is replicated in Pixie by two valve bodies, containing the valve seatings, which pass through a shaped block into bushes in the top of the inner steam dome. A turned pillar which provides the fulcrum for the spring retaining beam is mounted on the block. When the valves lift, exhausting steam is directed away above the roof of the locomotive by two polished brass pipes slotted to fit over the retaining beam. These pipes seat upon a shoulder turned onto the valve seat bushes and are supported by a mild steel bracket bolted to the roof.

Where two safety valves are fitted, and this is recommended, they should be set to blow at the same pressure. Failing this the higher of the two may never be actuated and may seize-up.

In Pixie the valves are phosphor bronze balls drilled and tapped to accept stainless steel spindles. The seat for the balls was formed in the manner described by LBSC, a blow with a heavy hammer on a steel ball set onto the embryo seating. To avoid thread damage during this somewhat brutal procedure each valve bush in turn was screwed, as far as its shoulder, into a tapped hole in a substantial piece of bar. Valve springs were wound from stainless steel wire and were tested using a simple loading rig as described in an article by ' Laurie ' Lawrence on ' SAFETY VALVES AND A FEW OTHER SAFE ODDS AND ENDS ' in Model engineer Vol. 162 No.3847 of May 1989


The axle driven pump is, of course, quite unprototypical in the case of Pixie, although as with most things connected with locomotion, there are prototypical examples of the use of axle pumps. However, in view of often encountered comments regarding substandard performance of injectors with ' warm ' water an axle driven pump is mounted behind the front plate. Later experiments may result in this being discarded in favour of an injector. With subsequent removal of the pump in mind, and as the wheels are bonded onto the axles, the pump eccentric is clamped and pinned to the leading axle.

The standard pump is mounted on a pair of tapered blocks behind the front plate using countersunk headed screws. The inclination of the blocks lines the pump with the axle. A blank for the eccentric tumbler was first produced from a piece of stock bar, this was left oversize for eventual finish turning and the machining of the locating groove for the eccentric sheave. A length of bar of sufficient section to incorporate the clamping bolts was squared-up on all four faces. This block was drilled for clamping bolts. One of the faces was 'raised' by centre pop marks to ensure full penetration of the solder into the joint, and the block then silver soldered into position on the tumbler disc with regard to the approximate axle centreline. The assembly was then split by sawing, the sawn faces lightly cleaned up and the whole re-assembled to enable the axle hole to be drilled then bored. With the hole bored a thou or so undersize, the assembly could now be clamped onto a stub of material of the same diameter as the axle, mounted in the four jaw, offset by the throw of the eccentric. Now the eccentric was machined to final shape including the projecting guide. Later, during erection, the eccentric assembly was pinned to the axle using a short length of silver steel, this was inserted into a hole drilled through the clamping block. The eccentric thus produced works efficiently and can be removed when an injector is fitted in due course.

The eccentric strap was a standard casting intended for another locomotive. The lugs were drilled for the assembly bolts, the assembly sawn into two parts, re-assembled then faced and the inner radius and guide groove bored, to match the tumbler which was frequently offered up to ensure a nice fit. The strap was then clamped onto a stub of stock material for the other face to be machined. Provision of an oil cap on one lug and milling the halved joint for the eccentric rod completed the part. Binding between the tumbler and the completed strap was eased by some judicious application of grinding paste then metal polish followed by a good wash and scrub with scouring powder and detergent !

LBSC's words and music provided the basis of the method used to produce the drain cocks. The principle operation here was not that of machining the valve bodies but rather the manufacture of a tapered reamer for forming the valve seatings. The technique involved setting the top slide to the correct inclination and turning not only the tapered plugs but also a reamer in silver steel for forming the seating within the body of the valve. The reamer was then filed to shape and tempered. The cross hole in the body of the valve was drilled undersize and then reamed to what must then to be the correct taper for the plug. The ' thin ' end of the plug was then threaded for the retaining nut and washer and the ' thick ' end squared-up for the attachment of the operating lever. A little ' running-in ' with a small amount of grinding paste ensured a tight valve although all traces of the paste had to be removed using scouring powder then detergent. On Pixie very small diameter tubes are superglued into the valve outlets to direct the wet steam away from the loco with the drain cocks in the open position.


When first encountered PIXIE was fitted with steam actuated drain cocks. These are operated from the footplate by a valve similar to that found on the outlet of Victorian tea urns, a valve of character ! The actuating cylinder is mounted behind the front beam, the piston rod connecting to the cockshaft via a link and lever. The bar connecting the drain cocks on the right hand side of the locomotive has a tension spring which keeps the cocks open / closed according to the setting of the valve on the weather plate.

A short stub of gunmetal was turned then bored for the cylinder, quite fun as it is only just about an inch overall. A cylinder cover was turned to suit and bored for the piston rod. The piston and rod were prepared all as those for the engines, and the whole assembled using 8 BA bolts. The cylinder is mounted behind the front beam on an angle bracket, forked around the steam inlet union, again using small bolts. The crank, pivot and lever connecting to the cock provided some watchmaking experience, the lever being pinned to the cockshaft which is set in tiny bearing bosses on the frames.


Saddle tank cover with access panel on model

Platework, tanks and bunkers. Tank mock-up. Building the saddle tank. Tank mountings. Weather board / spectacle plate. Scale roof and back. Driving roof and back. Footplates and footplate supports. Ashpan. Cutting, bending and rivetting plates.


The saddle tank dimensions having been established from the survey of the prototype a start was made by constructing a card and plywood mock-up. Plywood end panels, 3 / 8 inch thick, generated the curves of the platework which in this mock-up was represented by thick cardboard. One of the plywood dummy ends was later used as a template when cutting a piece of 3 / 8 inch mild steel plate to be used as a flanging plate. Provision of the mock-up enabled an early trial fit to be made to confirm that the finished tank constructed to the measurements taken at site would seat properly on the components of the model erected so far

Blocks of wood set upon the frame angles to simulate the tank-support angle brackets enabled dimensions to be established for the location of angle brackets which would eventually be brazed to the boiler shell. The position of the curved steel angle support connecting the tank to the smokebox could likewise be established. Messrs, Reeves asbestos-free lagging material was to be used in conjunction with cleading and boiler bands of brass sheet. The mock up confirmed that there was sufficient space between the underside of the tank and the boiler for these items. With key dimensions of the model components established, work could be commenced on the tank proper.


It was decided to use mild steel plate for the tank rather than brass sheet which would have been the alternative. By making a few trial sections of flanges which included both convex and concave curves it was established that, with care, 16 gauge mild steel could be flanged cold without loss of flatness of the plane parts of the platework. This could be achieved provided that, whilst flanging, the plate was sandwiched between the 3 / 8 inch template and a stiff piece of plate slightly smaller than the flat part of the template. As the flanges were formed it became necessary to snip out excess material from the flange at the convex corners in order to avoid cockling and to provide a smooth land for the wrapper plate. A touch of weld-metal sealed the flange at these points. With care the metal could be spread sufficiently to accomodate the concave curves where the tank sits over the smokebox and boiler cleading.

In the prototype, due to the riveted construction and the lapping of plates, it was necessary to form a joggle in the flanges to accommodate the extra layer of plate at the lap joint adjacent to the bottom corner of each side. This was also necessary in the model, It was achieved by setting the plate into the twin vices , and using two substantial backing bars together with two offcut pieces of plate of thickness equivalent to that of the tank plating. With the offcut pieces of plate positioned either side of the tank plate, above and below the position of the joggle, the vices were tightened in unison and the offset or joggle in the tank plate thus formed as it would be in a press. A considerable amount of ' presswork ' can be carried out in the vice although of course the bigger the vice, the better for such applications. When folding or pressing, the workpiece should always be supported using quite heavy backing bars, in this way twist and distortion will be avoided.


At this stage fixings for the sandboxes were inserted into the front flanged plate. these consisted of brass bushes of 'top-hat' configuration in which were blind holes tapped to receive bolts which would be passed through the back of the sandbox. These bushes, as well as providing a secure fixing for the sandboxes served to space the sandbox from the front plate. The holes for the fixing sockets were drilled to template, the template then set aside for future use in drilling corresponding holes in the sand boxes, a small note was scribed on the templates as a ' reminder ' to reverse the templates when applied to the back of the boxes !


The tank wrapper plates were formed using suitably sized bars from stock set in the twin vices at the dirty bench. Packings sitting on the bed of the vice located two bars, one the radiussed bar which actually formed the bend, the other a square section bar to maintain the flatness of the plate. Timber drifts used driven by a builders ' lump ' hammer persuaded the plate to take up the right form. The certain way to generate blemishes in folded platework, when not using a folding tool, is to use too light a hammer when bending or flanging. Blemishes are also formed if the face or edge of a hammer is allowed to come into contact with the plate. It is essential that clean barstock, free from blemishes is used for the formers, otherwise blemishes will be transferred into the plate being folded. A hardwood block should always be interposed between the plate and the hammer as the metal plate is bent. The combination of hammer blows and pressure applied to the plate by the free hand ensures the desired curve being generated. With some steels, particularly stainless steel, it will be necessary to over-bend the plate slightly to allow for elastic recovery although care is needed to avoid forming a local kink beyond the freshly formed curve.

Difficulty can be experienced in correctly gauging the width of plates which are to be folded to fit into, or around, concave or convex flanges. It it is advisable to start with an overwide piece of plate which can be trimmed to size on completion of the folding operation. The approximate width of such plates can be established by bending a piece of fairly stiff wire around the flanged plate, straightening it, then making some small allowance for the effects of the bends to establish the cutting width. With the folding operation completed any excess can be cut away. In the course of forming wrapper plates frequent checking against the flanged plates is necessary to ensure a good fit. It is important to make index marks on the flanged plates and wrapper to ensure that the wrapper is located in the same place each time a trial fit is made. Temporary marks of this nature are best made with a ' Permanent ' grade of pen as used for overhead projectors. The marks can be removed using methylated spirits and a scrap of rag. Trial fits should be made against each end plate as, when using rolls, particularly when using makeshift ' bending rolls ' lacking control of parallelity, there may be some small difference between the two ends of the wrapper plates.


The tank currently mounted on PIXIE has been fabricated by welding, with dummy rivets, a grand job it is too! However, as locomotive was to be modelled as it was the first time it coasted into Pages Park Station on the Leighton Buzzard Narrow Gauge Railway. It had to be a fully riveted assembly on Pixie! In order that the rivetting process could be carried out on all visible joints it was decided that a closing plate would be introduced into the curved underside of the tank. When all had been riveted up this plate was to be inserted and secured by countersunk screws into tapped holes in butt straps along the joint. Being on the underside of the tank these would not be visible in the finished model.

Rivetting was carried out commencing with the installation of the tank front. Fortunately, in the prototype the flange on the backplate of the tank is located rearward-looking and these rivets can be inserted and formed last. Soft iron rivets were used, a dolly being set in the vice to enable the innermost rivets to be reached. As with all riveted joints it is essential that the plates being joined are held tightly together to avoid the rivet shanks distorting and actually spacing the plates apart. Rivet holes should be marked by a scribed line and set out along the line using dividers. Apart from satisfying the ' rivet counting brigade ', accuracy of line and spacing have a considerable effect on the appearance of the finished model, and any inaccuracy on the tanks or bunkers which form a prominent feature of many narrow gauge locomotives can detract from otherwise good quality workmanship. Once the holes in the wrapper are drilled the flanged plate can be set into position and two holes transfer-drilled from the wrapper. Two small bolts through these holes will then hold the plates in the correct location whilst the remainder of the drilling is carried out.

The reconstructed tank on the prototype has an access panel let into the tank top. In the model advantage was taken of this feature to enable access for lining the tank with Fiberglass for durability purposes. When laying up the fibreglass in the tank BONDAGLASS pre-wetted glassfibre tape was used to ensure a good fit into the corners Pre-wetting avoids a problem well known to laminators, that of air becoming trapped between the composite lining and the base material at angles. Entrapped air makes it difficult to fully ' wet out ' the glassfibre. At this stage a baffle was glassed into position across the tank below the filler opening to prevent water in the tank surging backwards and forwards when the model is started and stopped in service.


The tank filler cap flange was turned from gunmetal as was the cover. The cover was made a push-fit on the flange to avoid problems in service. These activities are great fun, especially the reproduction of detail such as the very domesticated knob on the filler cap ! After fixing the filler cap flange by rivetting the underside of the panel was primed and painted. The panel was then secured by screws into flanges fixed to the underside of the tank top.

Balance pipe and feed pipe connection flanges were fixed into position on the wrapper using countersunk screws into tapped holes, a seal being achieved using Araldite. Once the Araldite had cured, the holes in the flanges guided the drills used to penetrate the tank . Where a mating flange is required, for the feed pipes for example, good matching fits can be achieved if these items are prepared at the same time as those for the tank. Pairs of flanges drilled and, if necessary, tapped at the same time ensure an accurate fit when they are eventually fixed together when the plumbing is installed. Flanges pre-prepared in this fashion should be marked according to location ( and orientation ) and stored safely for later use.


The tank is located and fixed at the smokebox using a radiussed steel angle, connection being made by bolts passing through the tank front and through the top of the smokebox. Attempts to form this by heating a piece of angle then bending it around a former failed, due to distortion of the section. This problem was solved by producing a simple plaster pattern from which a local foundry produce a suitable casting. This casting was set on the lathe faceplate and turned to finished section and radius. The angle is quite prominent on the model and, whilst this solution amounted to the ' use of a sledge-hammer to crack a nut ', the correctly shaped mounting enhances the overall appearance of this part of the locomotive. An alternative would have been to rough-saw the shape from the solid and then turn it to final form. On the model only two bolts are actually used in fixing the angle to the tank front, the rest are simulated by nuts run onto studs set in the angle. Similarly only two bolts actually penetrate the smokebox, the remainder being dummies. In the case of this fixing, as in many instances encountered in model making, absolute replication of fixings and fastenings is unnecessary for the purposes of strength and may even complicate final assembly and later maintenance.

The rear of the tank is located and retained by studs set into mounting plates on the underside of the tank, these pass through holes in the brackets on the boiler shell, these holes are slotted to cater for movement as the boiler heats and cools. The plates on the underside of the tank were fixed using countersunk screws into the wrapper plate, the joint being anointed with Araldite before the mounting was fixed. The fixing studs project downwards through hardwood blocks, which exist for insulation purposes. Teak blocks are used on the model and these allow a small amount of adjustment to overcome inaccuracies in the level of the tank assembly measured at the time of erection.

Two holes were drilled in the upper flanged joint at the rear of the tank. These are required for bolts fixing the stay-rods to the weather board, or as it might be termed on other locomotives, the spectacle plate! These two holes are coincident with the pitch of the rivets at this point.


It is most enjoyable to see a model develop in gleaming raw metal, be it steel, cast iron or gunmetal, and there is always a reluctance to follow what is perhaps the more usual sequence of priming, undercoating and painting parts as the work of production and erection proceeds. It has to be admitted however that there is little worse than entering the workshop and finding that surfaces, which the previous evening had exhibited a satisfying shine, are marred by embryo rusty fingerprints ! It is difficult enough to prepare metal around rows of rivets for priming, without having to remove such evidence. In view of the mild steel construction it was decided to prime the finished tank before storage. Accordingly, the tank was spray coated with a rust preventative paint, ' BONDAPRIMER, this is a resin based paint containing zinc which has good covering powers and dries quickly to a uniform matt finish. This was a drastic step as painting a model is not something to be attempted lightly ! Experience has proved that this was the right course to adopt, the primer warded off the dreaded rust and provided a hard base for subsequent coatings, although they were applied many months later. More about the painting process anon!


The coal bunkers presented another folding job. This time a flanging plate was produced to the dimensions of the inside of the main face of the bunker. Two suitably sized pieces of stainless steel plate were prepared to the correct height of the bunker. The plate to be folded was sandwiched between the flanging plate and a substantial piece of plate from stock, this was cut just slightly smaller than the flat on the flanging plate. This ' sandwich ' was then gripped in the vice and force applied via. a hardwood block. Care was taken when applying the force to start close to the stock plate to achieve the necessary tight bend. To overcome difficulty experienced in gripping the small amount of plate projecting from the sandwich this was held between two further pieces of stock held by a ' mole ' wrench which allowed hand pressure to be applied. With the bends completed a check was made on a true surface to ensure that the whole assembly sat flat and that the back and front were perpendicular to the base. Any slight twist in the formed plate was adjusted at this stage before the back and front edges of each bunker was trimmed to the correct length.

The top edge of each bunker is trimmed by a beading. A length of 5 / 16 inch diameter rod was milled filed to the correct 'D' section ( filing would have sufficed ), bent to match the top edge of each bunker and then riveted into position. The openings at the bottom of each bunker were chain drilled and sawn to shape prior to fitting steel angles at the edges which would abut the back plates. Similar angle was fixed to the bottom edge of each folding and to the back plates for use in fixing the bunkers to the footplates. Before rivetting these joints, plates were cut to the correct profile to fit into the bunkers. It was decided to mount the boiler feedwater pump in the left hand bunker and that plate was slotted accordingly. The left hand bunker was chosen for the pump location as this would free the right hand for the ticklish work of stoking through the dimunitive firehole door ! Later a hole would need to be drilled for the feedwater pipe although at this stage in construction the exact position of the hole was not known. As stainless steel had been used in constructing the bunkers these could be left in their pristine, shiny state until the time came for the final paint job . A word here about working stainless steel. Many are the broken bitts that testify to difficulties encountered in drilling ( not to mention tapping! ) stainless steel. The secret appears to be that any tool used to fashion stainless must be kept cutting at all times it is in contact with the metal. Considerable heat is developed the moment the tool rubs rather than cuts, and the temper is lost or the tool jams with dire results in case of the small drills we use. Hundreds of rivet holes in stainless plate have proved the wisdom of using an anti-scuff paste as a lubricant. This combined with working the tool hard by applying plenty of force to the feed avoids tool wear, heating and damage


The platework forming the front of the rather spartan cab provides some interesting work in shaping the plate itself and fabricating the angle steel supporting frame. This is mostly straightforward, fashioning material prepared according to measurements from the survey into model form. The angle supports presented a problem as far as bending the metal without distorting the section. The difficulty was ovecome in this instance by cutting a disk of plate of the scale thickness of the angle and turning from this a complete circle of material of angle section. Segments of this section were then cut as required to meet the geometry of the support frame. The roof bearer was of sufficiently large radius to be bent cold in the vice. The curved segments, uprights and the roof bearer were clamped to a ply template and silver soldered. After this the plate could be clamped into position and the whole assembly drilled and bolted up.


Two roofs were made, one for exhibition purposes, the other for use when the locomotive is working. the latter has an opening which permits a good view of the 'works', the gauges and access to the firehole door and the regulator. Stainless steel plate was used here, with rectangular section bar stiffeners riveted at all edges, using the right number of rivets ! Correct curvature was achieved using blocks of wood, a clean length of scaffold tube and foot power. The plate was set upon the blocks cut to the correct rise on the chord and the scaffold tube applied ' rolling pin fashion ' by the sole of the shoe. Heath Robinson as this may seem, both roofs fit comfortably onto the roof bearer for bolting into position, and both roofs are completely interchangeable ! Provision was made in each case for the two bolts to which the support plate for the two brass columns of the safety valves was to be attached.


As with the roofs, two backs were fabricated, one for driving purposes and the other for exhibition. The scale back or rear spectacle plate ( which, due to in service modifications to the locomotive, may not replicate the back currently fitted to the prototype ), was built to the dimensions achieved from the survey soon after I first saw the locomotive. It comprises a framework of angles and a curved roof bearer, silver soldered together and clad with plate. Holes for the glazing were trepanned from the plates whilst they were mounted on the vertical slide, absolute care being taken to ensure that the face of the slide was set dead square to the lathe axis. This was ensured by placing one edge of a steel rule against the chuck face and bringing the face of the vertical slide tight against the other prior to tightening the bolts securing it to the topslide. A suitably shaped tool was ground-up for the fly cutting tool holder providing the nearest radius to the required opening. Bearing in mind the maxim that the tool must be kept cutting at all times the saddle was fed towards the cutter, propelled by the tail stock quill. In such an operation it is advisable to tighten the saddle locking bolt slightly, and essential to interpose a piece of hard plywood between the metal being worked and the topslide.

The glazing is secured within the platework of the scale back by two turned rings, one pair cut from brass bar and the other from steel. These are of course of such size that when fixed, using small screws, they provide a rebate into which perspex glazing material can be fitted. The driving back is a simple open structure, again silver soldered with just side plating added to aid with rigidity along the main axis of the locomotive. A small angle provides a fixing for the brake standard


On PIXIE the footplates are quite rudimentary. There is a plate which bears on angle steel brackets attached to the inside of the frames and two outer plates which are supported on brackets on the drag beam and cantilever brackets riveted to the outside of the frames below the bunkers. Stainless steel plate is used on the model, more drilling ! Angle steel stiffeners run around below the edge of the outer plates. The rather tight bend below the front of the bunkers can be formed by cutting a 90 degree ' vee ' from the flange and bending the material cold. Care is required here to ensure that bolting faces of the angle remain flat, this can be checked on a true surface as the bend is formed. The cantilever support brackets may be cut from angle steel or fabricated, I chose the latter course, adding the smaller angle for the footplate connection as on the prototype.

The plate between the frames has an opening for the damper lever and a small cranked plate that engages with this lever to locate the damper in the required position. It is advisable to leave the cranking process until the correct angle can be be gauged to the slots in the lever. Small checkouts will be required in each of the footplates to accomodate the angle brackets at the junction between the drag beam and the main frame members. In the central plate the drain pipe from the water gauge and fixings for the brake standard have also to be accommodated.

The ashpan and damper are important parts of the firing system. In Pixie they are constructed of stainless steel sheet and mild steel angle. The angles sit below the foundation ring of the boiler, located by stainless steel dowels set into tapped holes in the ring. The upper edge of the ashpan projects slightly into the firebox and thus provides support for the grate which is cut from a section of proprietary stainless steel grate material supplied by Messrs Spinks. A length of mild steel was milled, drilled and filed to shape and cut into four pieces to provide the hinge blocks into which a stainless steel hingepin is slipped.

The damper is controlled by a lever projecting through the central footplate and retained by a small, joggled piece of plate. Any of three checkouts in the lever locate on this to provide three settings, one fully closed and two sizes of opening...high tech ! In service the position of the damper does make quite a difference to the state of the fire, closed down it allows the engine to simmer comfortably.


At this stage a 'mini-erection' can be carried out ! The cantilever brackets supporting the outer footplates can be installed and the outer footplates set into position and bolted into place. Next the bunkers can be positioned and the fixing holes drilled through from the fixing angles. Then the weather board and the backplate can be set up. The weather board bolts to the bunker front and the back to the drag beam. Now the roof can be installed and it is time to step back and admire the fruits of our labours so far!



Erection, frames and motion, weighshaft and valve gear, reversing lever, valves and setting. A trial run on air. Smokebox, boiler, steam and exhaust pipes and blower, regulator, safety valves, brake gear, ashpan and grate, saddle tank, superstructure, makers plates.


As this chapter proceeds the text describes the work in erection. Although this book is not of the ' words and music ' type produced by illustrious writers such as LBSC, Martin Evans, Keith Wilson and other extremely experienced model designers and builders, it is written with an understanding of the problems that face the newcomer. It is hoped that the approach adopted will assist newcomers to model locomotives in their efforts.

Whilst the reader may be building a model of another prototype the principle procedures stand.


This is perhaps the most enjoyable part of the construction process, when at last the fruits of one's labours take finished form. The work of erection will be eased if the builder has access to both sides of the locomotive. Many builders employ an erection frame at this point in construction and there have been a number of examples in the various model publications. If the modeller intends to go along that route it is worthwhile making a frame that can be adjusted for use in subsequent maintenance operations, and indeed for further locomotive erection. End frames set upon an angle steel base can provide adjustment for length of locomotive, the ends being slid along the angles and clamped into position as required. A frame that permits the locomotive to be rotated and locked in any position is obviously going to be an asset although provided that the end frames are of square form and large enough to clear the projecting part of the locomotive this will permit access to the locomotive whilst the frame is set on its side or inverted.

Mr J Bowman described a useful building stand in MODEL ENGINEER Vol 159 No 3805, August 1987, this was built using two scrap car jacks obtained from a breakers yard. Pivoted clamps provided attachment to the front and back beams of the locomotive. The jacking screw arrangement can be used to elevate the model to the required working height, as the designer says, opening-up useful bench space beneath the model.


With the frames inverted the axle sets are be dropped into position, including any springs or rubber parts of the suspension. The hornplates are installed, their bolts being anointed with a spot of fixing fluid. At this stage it is convenient to install an axle pump if one is to be fitted, the bolts connecting the parts of the eccentric strap again being given a touch of securing fluid. The cylinder blocks can be attached to the frames, remembering to insert insulation between the back of the block and the frame, where required. The motion plates are next, the bolts, or dummy rivets in the case of Pixie being inserted and taken up tight. The slide bars can then be installed between the seatings on the piston rod glands and the lug on the motion plate, not forgetting to install the crossheads, the correct way round. At this stage the geometry of the engines should be checked out. Careful measurement of the distances between the piston rod and the slide bar ( s ) at the extreme ends revealing whether the two are parallel. This adjustment is carried out on a full size locomotive by sighting through the centre-line of the cylinder or establishing the centre line using a taut wire then taking measurements in the same fashion. Any minor adjustments can be made by insertion of shim washers where the slide bars are attached to the motion plates.

At this stage the cock shaft for the cylinder drain valves can be installed, as can the small steam actuating cylinder and cranks. The cylinder bolts behind the front plate on the right hand side of the locomotive. The straps linking the cocks can be installed. that on the right hand side of the locomotive having an extension for a return spring which is secured to an eye bolt through a plate projecting from below the front beam.

Coupling rods, with marine bearings in the case of Pixie, are next. Some anti-scuff paste assists in sliding them home onto the crankpins. In the case of crankpins constructed incorporating the retaining flange, scale marine bearings can be split on the centreline of the crankpin and bolted-up. With dummy marine bearings, the leading end of the coupling rod can be secured by a countersunk cap sized to replicate the head of the crankpin. A nice detail in Martin Evans book ' Model steam locomotive construction ' shows the use of either a cap threaded onto the crankpin, or the use of a cap secured by an ' 0 ' BA Allen screw. Whichever is chosen, securing fluid ensures that the caps do not come loose in service. In the marine bearing the upper bolt must be that having the groove for lubricant. Each bolt has a full nut, half nut as a locknut and a split cotter pin for safety. With the rear coupling rod bearing assembly completed, the crossheads can be installed on the slide bars, including slipper plates as appropriate, and the piston rod / crosshead joint made using a cotter or bolts.

The connecting rods can be positioned and gudgeon pins installed. On Pixie it is possible to drop the connecting rod by rotating the wheels to a position where the crosshead clears the wheel allowing the the gudgeon pin to be withdrawn from the back of the crosshead, this is helpful when it comes to routine maintenance. The gudgeon pins have single large nuts on the threaded portion, again safety split cotterpins are inserted.

With the coupling rods installed, connecting rods secured within the crossheads and the cylinder drain cocks open and plenty of machine oil on all the moving parts the motion should turn over sweetly. In the event of any binding the bearings can be slackened one by one to find the offender when a touch with a fine file or insertion of a piece of shim stock sorts the problem .


The weighshaft brackets on PIXIE are fixed by bolts through the stiffening angles on the main frame and through the frames themselves. In the case of Hackworth valve gear the position of the centre of the weighshaft relative to the remainder of the motion is critical to the valve events and these have to be checked out prior to finally fixing the position of the weighshaft brackets. A check was made prior to constructing the valve gear by making a twice scale sized model in card, set up on the drawing board. The location of the brackets on Pixie is exactly as established from the survey and fortunately the positions and sizes of components thus determined provided, in conjunction with scale sized components, sufficient valve travel.

The return is crank set on the rear crankpin, lined with the axle centre and both the bolts tightened. A spot of securing liquid assists in maintaining the correct position although with a well fitted, reamed hole in the return crank the locking bolts provide a remarkably strong connection. A variety of methods are employed to fix return cranks, ranging from pinned and bolted, slotted and bolted and simply bolted connections, each has its merits. That employed on Pixie is straightforward to make and effective in service.

With the die blocks installed in their guides the vibrating lever can be set into place, the top secured by the pivot bolt in the die-block and the bottom by a cap with a countersunk screw. Next comes the installation of the valve rod. Here it should be mentioned that wherever a connection is made with a simple pivot pin, the pin should be retained by a split cotterpin. Holes for minute split pins tend to wander in the course of drilling through the fork and the pin, and some time can be expended in attempting to reinsert split pins after a service or routine maintenance. A small punched mark made when holes for such minute split pins are drilled through the forks and pins will identify the resident pins with the forks and also the correct orientation of those pins within the forks.


On many locomotive the reversing lever and sector plate can be assembled independent of any part of the superstructure. On Pixie this is not the case as the sector plate depends for support at each end from bolts through the inner face of the right hand bunker. To facilitate access and valve setting a temporary bracket was fabricated to provide support at these points and the lever and reach rod installed. The front end of the rod pivots on a stud secured into a lug on the bottom of the slide. The rear end of the rod, forks over the reversing lever which in turn, pivots on a fulcrum bolt on the bunker side, replicated on the the temporary bracket. There are occasions such as this when a temporary arrangement allows access for adjustment which would be difficult to make with everything permanently in place. This time the use of the temporary bracket permitted the notches for the latch to be marked before the stage of final erection. The notches were actually cut after removal of the bracket and final erection of the sector plate and bunker assembly onto the bunker side.


The slide valves were dropped into place in the valve chest. The gland nuts were threaded onto the valve spindles which were inserted into the glands on the valve chest, not forgetting to insert the ' O ' ring packings. The rod was then passed through the first of the retaining collars, the slot in the valve, the second collar and then into the tail guide. Now the valve rod could be installed and the pins inserted. Valve setting followed the normal procedure of adjusting the position of the slide valve on the valve spindle, locking the collars when the optimum position had been established.

It has to be remembered that the collars must allow the spindle freedom in the vertical direction such that steam pressure in the valve chest forces the valve onto the port face. Now the weighshaft brackets can be fixed into their final position, also the position of the notches on the sector plate can be marked in full forward and full reverse.

In the event of a need to modify valve arrangements it is a great help if a set of perspex ' slide valves ' are produced. These can be simply made from 1 / 8" perspex sheet, joined by using plastic solvent. The exhaust recess can be scribed and coloured-in using marking blue. The dummy valves can be threaded onto the valve spindle and the events viewed as the locomotive is moved back and forth along a short length of test track.

With the valves set the valve chest cover can be installed, it is advisable that all the studs should be inserted and the cover checked for location prior to finally inserting the gasket and tightening down the nuts. This will avoid the situation sometimes encountered where, due to minor inaccuracies in hole location, the cover binds on one of the many studs and fails to seat down properly ! For the cover / valve chest joint, liquid gasket has proved satisfactory in service, and certainly saves a considerable amount of work otherwise required in the manufacture of traditional gaskets.


At this stage it is more than likely that, with the the engines attached to the frame and with the motion work installed, the builder will want to have a trial run. This can be carried out at the bench on compressed air with the frames set upon packing. The writer felt the urge at this stage to prove the work to date. In the absence of a length of track Pixie's first run was on the floor of a local garage. The frames complete with motion and a temporary steam inlet pipe were taken to the garage, the only source at that time of an adequate supply of air. To the amazement of two young mechanics, who whilst having intimate knowledge of the workings of motor car engines were apparently ignorant on the working of steam, the ' mechanism ' shot twenty feet across the workshop floor on the end of the air line, like a scalded cat. Remarkably too, on resetting the reversing lever and opening the regulator it shot back again. The power developed by the small engines had to be seen to be believed.

Such excitement! After more than seven years and much figuring and speculation as to the effects of scaling down parts of the valve gear, it's geometry, the steam chest dimensions and the dozen or so components of each engine, the thing worked, and powerfully too. In retrospect the whole proposition of a concrete engineer manufacturing such a complex mechanism in miniature, and one that worked as intended on first trial, seemed as unlikely as the prospect of the construction industry, as we know it, putting a man into space!


The smokebox can be secured onto the saddle, here stainless steel bolts are advisable. These can be turned up from stainless steel hexagon stock of the appropriate size. It is a good thing to hold a stock of the common sizes of hexagon material in stainless for bolts, and brass for unions. When there are a few minutes to spare between other jobs, or when only a limited time can be allocated for the workshop, a few unions can be turned up or the odd bolt prepared without too much upheaval and without too much effort in setting up machinery.


Boiler installation will render the space between the frames rather less than accessible so that it makes sense to insert any pipe runs and make-up any joints in those pipes before finally locating the boiler into the smokebox ring and seating it onto the bearing brackets on the frames. First attempts at boiler installation onto Pixies frames involved considerable sweat and no little amount of personal injury ! Despite the fact that the boiler matched the dimensions from the survey in every detail, ( indeed the accuracy achieved was a great credit to the builder ), the boiler just would not seat into position...... Enter Giant Despair ! ......Several cups of tea and a breather later however, on resuming what until then had been a struggle, the boiler suddenly slid into position as it has indeed on subsequent occasions. There may have been ' Gremlins ' in the workshop on the occasion of the first attempts at installation, although more likely the interference was caused by some rogue rivet head or overlong bolt, eased in in the heat of subsequent operations.

With the boiler installed a check should be made to see that the firebox end is free to move as expand everything heats up and the boiler expands. Holes in the bearing brackets need to be slotted to allow this movement as will those in brackets providing support for a saddle tank. The level of the boiler, relative to the frames, can also be checked and if necessary packings inserted at the bearing brackets. At this stage also, sufficient clearance can be established around blowdown valve (s) and if necessary any openings in stretchers adjusted as required.


Now the steam and exhaust pipes can be installed,. Some juggling is likely to be necessary to persuade these items into position through scale size apertures in the lower quarters of a smokebox ( Later thought was given to this and perhaps in a subsequent model, where some detail of the prototype is concealed in the model then some licence may be taken with sizes ! ) Eventually with such Chinese puzzles solved, the wet header can be bolted into place on the boiler bush ( not forgetting the 'O' ring seal ). The joints between the steam and exhaust pipe and the cylinder block can then made. Liquid gasket was used to seal these joints on Pixie and has proved perfectly effective in service.

With the blower ring installed the plumbing can be completed. The parts of the pipes between apertures in the smokebox and the connections with the cylinder blocks are enclosed in a sheet steel cover, secured by tiny screws into tapped holes in the smokebox tube. To render the apertures airtight small quantities of modelling clay can be smoothed around the pipes from inside the smokebox using one of Messrs Proops ' Doofers '. These can be obtained from a MODEL ENGINEER exhibition stand and although they convey a fearsome reminder of time spent in the dentists chair will prove immensely useful about the workshop!


The regulator can be mounted within the steam dome, using stainless steel screws into the lugs on the regulator body. A check should be made at this stage to ensure that the screws on the actuating rods are wired to prevent them coming loose in service. Stainless steel wire is a must.

With the regulator rod inserted into the regulator handle and the rod passed through the gland, the small bush can be set onto and secured to the rod by a stainless steel grub screw. This bears against the back of the body block, retaining the rod in its correct position within the regulator proper. The complete assembly can now be passed through the regulator block bush, the leading end being located in the back of the regulator body within the steam dome.


In the case of pixie, once the regulator and the regulator block and rod had been installed it was time to install the safety valve block and safety valve bodies. The block is in fact retained by the screwed portions of the valve bodies so that it became necessary to enlist the aid of the plastic gasket material once again. The fulcrum bracket which retains the safety valve balance beam screws into the safety valve block. The beam itself is mounted using a bolt as a pivot, the ends of the safety valve spindles move freely within bushes at the ends of the balance beam. Adjustment of the valves is achieved by screwed bushes set into those at the ends of the balance beam. Assembly of these items takes less time than is needed to write the description, although with such an important part of the ' works ' care has to be taken to ensure freedom of action of the parts.


There are three running plates on PIXIE. The two outer, low level plates take support from brackets rivetted to the main frame members and rails which are bolted to the drag beam. These plates provide seatings for the side bunkers, and importantly, the brake standard. These items are each in turn clamped into position and the location of the fixing screws transferred to the plates using a centre punch. If the platework is of stainless steel the holes for the fixing screws has to be drilled using a freshly sharpened bit with sufficient pressure to ensure that the bitt cuts metal all the time. As has been noted elsewhere, the slightest rubbing action will heat the bitt and spoil its temper.

Once the lower running plates and the left-hand bunker are erected, the positions of the holes for the feedwater pipes from the hand pump in the left hand bunker can be established. Here pieces of stiff wire to replicate the pipes facilitates the marking process. The holes drilled the the pump can be installed on a mild steel block of sufficient height to bring the operating handle clear of the bunker ends at the extremes of the pump stroke. The brake standard is positioned on the right hand, lower running plate ( where it was located at the time of the survey, although there is evidence in the form of a set of mounting holes, that it has been set on the opposite side at some time in the locomotive's history ) and the fixing holes and the hole for the brake spindle spotted and drilled.

The middle, upper footplate will require trimming to fit around the frame angles, also slotting for the damper lever. this is another situation where it is advisable to produce a strong card template and carry out a trial fit prior to cutting metal. Clearance for the angles joining the drag beam to the frame members is easy to arrange by pure measurement. To avoid wasteful errors in positioning the slot, the template can be slotted on the known line of the lever. Inserted into position, the card can be marked to indicate the position of the ends of the slot and the footplate marked and cut accordingly. A small metal catch is bolted to the plate to engage in, and retain, the damper lever. This was bolted into position on the footplate and with marking blue and a scriber the correct angle and depth of the three slots in the lever were established. Provision of one last slot for the drain pipe from the water gauge permitted the plate to be dropped into place.


The main brake shaft with just the fixing bushes in place, was offered up and clamped into position, square across the frames and level. This allowed the cut-outs necessary to allow the bushes to seat onto the frame members to be determined, marked out, then cut and filed. The brake nut was then threaded onto the brake spindle and the handle rotated until the brake shaft complete with actuating arm, two bushes and two drop arms could be bolted into place. With the hangers and blocks installed by bolting to the frames, the pull rods could be inserted and the rods which serve as brake beams with their retaining bushes threaded into position. With all assembled the holes for the split pins which secure the bushes in place were drilled and the minute split cotterpins set into place. This size of split pin has extremely sharp ends, so beware!

At the running plate two bolts pass through support angles on the standard downwards into the angles supporting the floor member. The screwed portion of the brake shaft can be engaged in the bush within the brake fork below the running board and the handle turned until the foot of the column contacts the floor when the fixing bolts can be installed. At the top, the brake standard takes support from the cab backplate support angle and is bolted into position using a shaped strap, reminiscent of a gutter bracket.


At the time of final erection the ashpan was set into position, clamped and blocked to prevent movement whilst the holes for retaining pins were drilled. After a lot of careful measurement, small diameter pilot holes were drilled through the mainframes and the ashpan sides. During the drilling of the holes a colleagues assistance was saught to sight the drill and ensure that the holes were put through horizontally. Transverse accuracy was obtained by lining the drill with a try-square set against the frames. The ashpan is located by two stainless pins, cross drilled for tommy bars at one end and tapered at the other to facilitate insertion. The pins themselves are retained by "O" Rings slipped into position once the pins are engaged. In the event of an emergency the grate can be dumped by withdrawing the pins when the ashcan will drop as far as the sleepers, spilling the fire.


With the boiler seated into position and the boiler / smokebox joint made, the fastening angle for the front of the saddle tank can be installed. Although there are a number of bolts through the angle and the smokebox tube, only two of these are in fact bolted connections, the rest are dummies, this to reduce the amount of work when dissembling for maintenance in future. On Pixie, the rear of the saddle tank rests on two brackets supported from the boiler casing and the front takes its support from the aforementioned bracket. Studs projecting from stiffening plates which also serve to reinforce the delivery pipe connection points from the tank, pass through timber blocks into the boiler brackets. the timber blocks are for insulation purposes. again the holes in the mounting brackets have to be slotted to permit expansive movement of the boiler.


After the arduous work of boiler and tank installation, the erection of the superstructure affords time to relax and enjoy ! The rather elementary weather plate is supported on angle steels which in their turn take support from the backs of the bunkers. The back of the cab, again platework on angle framing takes support from the drag beam. The weather board and cab back plating support the roof which is bolted to the curved angle steels forming the top of the superstructure framing.


At this stage it was time to mount the makers plates ( which had been prepared earlier ! ) Mounted on each of the bunker sides these proclaim ' Kerr Stuart etc......" Plates such as these are essential in conveying the atmosphere of the model and some thought was given to their production. The method eventually adopted was first to choose a photograph of one plate from the reference shots on transparency film. This slide was then projected onto a sheet of paper set on the workshop wall and the image drawn-over using indian ink to approximately five times the size of the required plates. These were photographed by a friend, to provide the required image size, using a film having high contrast negative emulsion. The negative was then sent to Messrs Reeves for the production of etched plates. The detail was etched onto brass plate of the required thickness. All that was then necessary was to saw and file these to profile, polish the edges and drill for the fixing bolts. A coat of enamel was applied overall, the plate was lightly filed to expose the letters then polished. With the availability of Digital Cameras this process can be simplified, and a dic with the appropriate photograph sent to the engraver the size being simply adjustable as required.

Ron Acock described the use of brass punched letters, obtainable commercially, for constructing nameplates. His article in MODEL ENGINEER Vol 161 No.3833 October 1988 will be found most helpful for the modeller taking that route.

Updated 25 February 2011

To read Chapter 17 and on........Click here bookonbuildingpart 3.htm

To return to Home Page click here index.htm