Welcome on Henk's website

find all about fiNe-scale modelling
in a scale of 1 : 160


Walschaerts valve gear


Egide Walschaerts was the chief engineer of the main workshop of the Belgian State Railways in Brussels-Midi. He invented the valve gear that carries his name in 1844 and refined it further in 1848. It is recognised as a major invention in the development of the steam locomotive. Until that time all kinds of varieties for reversing of steam locomotives were incorporated and not all of them were a success. The Stephenson gear was introduced only a few months earlier. The importance of the Walschaerts gear stems from the fact that all parts are on the outside of the loco and easy to reach for maintenance, there is no need for expensive driving axles with eccenters and above the control of the valve inlets can be equilised on both sides of the piston smoothing the unbalancing effect of the moving weight of the connecting rods. No wonder that this became the most widely used mechanism for control of steam locomotives and saw use until end of steam.

In the whole world? ..eh  No there was a small place called Brittanica that withstood all foreign influences for a long time. It was not until the first World War, with its enforced grouping, that enlightenment could gain foothold and this superior valve gear was introduced on new machines.


drawing showing named components

Figure 1: This drawing shows the names of the main components of the Walschaerts valve gear.

As you may expect from things developed over a century there may be more than one variety to this theme. The drawing gives the most common form of Walschaerts gear as found on continental locomotives. There are however certain other forms with modified components that were not rare. One of these is the construction of the expansion link that can be of opposite construction. In the drawing the radius rod is contained in the expansion link, another common form is a flat expansion link contained in the radius rod. This variety is often observed on locomotives of American origin. Quite often there is a lifting link between arm and radius rod, replacing the sliding hole in the end of the radius rod. On older locomotives with flat sliding valves and outer admission the link between combination lever and radius rod sits below the plane of the valve. The slidebars can be of a double type with the crosshead in between, necessitating another vertical arm between anchor link and crosshead called crosshead arm or drop link.
Not shown are the various brackets that carry the whole lot. These are very dependent on the prototype and study of photos or a general arrangement drawing is necessary for modelling these.

Modelling options

Most of the components are just thin levers, a flat piece of metal with holes in both ends or sometimes 2 small holes with some metal around them to keep them apart. That cannot be difficult to make. Thus take the challenge to make this entire lot into a moving set of model levers. However for starters there are options for taking a shortcut instead of directly taking the plunge into the deepest of swimming waters. The easy way out is making not all components moving, notably simplifying the expansion link into a flat piece of metal only swinging by movement of the eccenter rod and forget a moving valve piston. In that case the radius rod can be fixed to the end of the valve piston. The combination lever is also flat and driven by the crosshead. But the challenge is of course making your model with working valve piston rod, radius rod, expansion link and lifting arm.
Independent of the type we choose we are faced with a moving crosshead that is very visible. The double sliding bar crosshead is just a form piece with two grooves for the sliding bars. Modelling the single sliding bar crosshead however is not that easy. This requires a lengthwise rectangular hole for the sliding bar. The easy way out is ordering a suitable replacement R-T-R item, Arnold for instance has dimensionally correct plastic crossheads on its Pruessian locos. However getting these as replacement components is complex and very unreliable nowadays. The direct approach is spending some of your free time with file and other small tools and shape them from metal. The slidebar itself is not difficult, a piece of (steel) bullhead rail filed down in height to 0.8 mm will do nicely. Particularly on the larger locos these slidebars are heavy chunks of metal, think of cross dimensions of 12cm square or larger. The bullhead rail not being rectangular is fortunate as this creates some extra clearance.
A fully working expansion link requires careful planning as the curve of the elements depends on the length of the radius rod. But apart from getting the dimensions correct there is not much difficulty to it. To arrive at a working version of any Walschaerts gear it is important to observe the necessary clearances. Particularly the offset between the holes for the valve rod and that for the piston rod so that the combination lever runs free. But also observe the clearance between the pins of the coupling rods on the first set of driving wheels and the inside face of the crosshead. On the other hand try to keep the total width over the cylinders as close as possible to the prototype as any overwidth here directly spoils the head-on view of your loco.

Filing the crossheads

The drawing shows a general representation of crosshead, a flat body with some decorations in the form of openings, oilbox on top and below that a means of fixing the piston rod, the connecting rod and anchor link. This is a 3-dimensional part that works best if the piston rod, connecting rod and slidebar are lying in the same vertical plane. Typical dimensions are a little over 3 mm long, 3mm high and 1 mm thick, but refer to your prototype drawing for exact dimensions. Because of the colour and durability I find steel a good choice as base material. As with most other components on a steamloco the crossheads come as a handed pair. Thus if we choose a piece of scrap steel of 10mm width x 1.5mm thick that would suffice, ideally this is a several centimeters long for ease of handling. Steel plates from N-gauge tankwagens turn out to form a good source. If we make them as pair besides each other then we are sure not to end up with two left ones or two right ones!

initial stages of crosshead

A photo clarifies more than words can the initial setting out and cutting away of superfluous metal.

The main difficulty is of course how to create a rectangular hole of 0.5 width x 0.8 mm high for the slidebar? Lengthwise drilling and filing is not a smart idea as it probably ends with broken drills and holes that are either too large or not rectangular and a waste of time. The fast way out is to create a slot at the not visible back and close it later with a lid soldered on. The slot can be cut with a large hacksaw in a minute. The only requirement is to cut the slot parallel with the surface of your steel plate. Instead of trying to make the slot parallel to the top of your piece of steel it is much easier to use the slot as basis for your dimensions. Therefor cut the slot at least 3 mm below the top of your steel plate and about a mm deep. This leaves about a 0.5mm thickness on the frontside which later can be filed thinner.

Once you have the slot, set out the basic dimensions by using your digital caliper as accurate scratching tool (grind the ends a bit sharper for this). The piece of steel is wider than we need, thus leave a slot in the center to partition into a left and right hand. This part can be cut out with a normal jewellers saw.
The next step is setting out the openings for the piston rods. Hardened steel pins are ideal for that, nip off the head with good quality cutters otherwise a dent in the cutters will the result. Standard pins have a dimension of .6 mm diameter which is fine. Drill lengthwise from both sides parallel to the slidebar slot and parallel to the front.

For fitting the connecting rod in the same plane you can cut a slot parallel to the front using a jewellers saw. The width you need depends on the head of your connecting rod but these can be thinned to 0.3 mm. A coarser saw say no. 1 will do first time out, otherwise use a barfile to open the slot to width.

The photo also shows a trace of other drilled holes which are used to create a circular cutout in the lower edge. The next step is fitting the piston rods that are secured with solder. Roughen the end of the pins on an oilstone and use an aggressive flux such as S90 to create a good bond preferably with 240 degree solder.

Now with the rough work out of the way what is left is the finer sculpting work and drilling the smaller holes for the pins. We start with the latter. Holes of .2 mm thereabouts are ideal. You can create such drills from a hardened steel pin and mini drilling machine with slitting disc. Best results are created with two mini drilling machines one for the pin the other for the disc but a pinvice and steady hand suffices in most cases. First make a thin cylinder on the pin end, flatten one side and create a 45 degree cutting edge on top. A length of 1.5 mm is enough for this job. Use this to drill the 4 holes for the pins of the connecting rod and anchor link. This can be done best slowly by hand using an archimedean drill. The sculpting work can be done with jewellers saw and files, fine set of files nr. 4 cut are a good investment for jobs like this.

near finished

The slot can be closed with a lid from thin brass taking care that solder does not run into your slot. A wooden splint or such helps to keep the solder out of the sliding bar slot. If you are not successful, unsolder, clean up with fiberglass pencil and try again. A test on the bench is necessary with the connecting rod and a loose pin to see if it can move freely in the slot. The anchor link comes on the front and is less critical. Any smucking-up with outer details can also be done with overlays of thin plastic card welded on top of the metal.

If you rather prefer to do it straight up that is is also possible of course, but less accessible for hacking slots in the lower part.

A working expansion link

The expansion link is the most complex part of the Walschaerts gear. This part can be viewed upon as set of 4 curved parallel bars linked at top and bottom. A reliable way to produce these is from .8 mm brass tube. First bend around a suitable object, a lid with the right diameter will do for this. Cut two pieces with a bit extra length for handling and make sure that they have the same curve. The radius is determined by the effective length of the radius rod. Also here we need handed pieces for each side.
Both tube sections are filed to a square cross section that fit on top of each other. We now have to cut the slot for the radius rod by filing each part so that it opens along on the center line. The open ends of the tubes are closed by soldering a small piece of .5 mm wire into them. Of course this should leave the freshly cut slot open. We now drill the pivot holes so that this can be used for a thin wire to locate the 2 parts in the same position during soldering. Position the link parts with the handles on the lower end. For this locating a stainless steel wire is suitable as it won't take any solder. The holes are later used for short pins to locate the expansion link into the bracket that is attached to the frame. However as there is hardly any flesh for the pins it is easier to solder first a square millimeter reinforcement on the outside. Alternativily use the holes with the pins in the bracket. Around these reinforcements the link is filed until it opens out at all sides leaving the 4 bars. To arrive at the right cross dimensions some careful filing is necessary. The end with the handle is cut last. The end of the pin is filed into the lever end with the hole for the eccentric rod. The radius rod moves up and down in the expansion link. This requires a pin in the radius rod. In order not to destroy our expansion link when inserting this, it suffices to fit the pin on 1 side only and make angled face to it, so that it slides in easy but not out.

Other parts and pinning together

The only further item that needs more attention are the combination levers. This is a forked item and can be produced by soldering parts together but carving it from the same piece of 1.5 mm steel as the crossheads exercising the jeweller saw is an alternative. By making it assymetric instead of the real symmetric item you can gain some clearance space. Using a flat piece is only advisible if the valve rod is stationary and not driven by the levers. Drilling the holes first takes care that these sit neatly opposed in both arms of the fork.
The valve rod itself is tricky. It requires drilling a tiny hole in .3 mm wire for a pin. Several trials are necessary before you are left with 2 working versions. Producing a flat first helps, which can later be filed circular.
Pinning can be done with soft annealed brass wire but using angling wire turns out easier. When treated with a hot knife it forms a convenient mushroom head, the free end is threaded through the holes and when these are pressed together with a pincet the wire can be treated again with the hot knife. The durability of this wire turns out well.


NMBS type 97 loco with only flat components
The flat solution on a loco, scratchbuild with heavily upgraded top.

NMBS type 81 loco with fully working Walschaerts

Under merciless enlargement: this is the high end working solution, but on a still it is not impressive at all. The carpetrider's amount of dust however is, still it is a full scratchbuild loco except for the cylinders and cabin.

Folding it up from sheet

Using the computer on hand it is also feasible to fold the whole lot from sheet metal. The crosshead can built thought as layers or slices and these can be drawn using a CAD. Draw an edge around it and print it on transparent and carry it over to NS sheet by light, drop it into the etching tank and ready is your crosshead. Then you only have to fold it together and use some solder for fixing. If you want more than a single set this will be of course faster than filing each time a new set from massive metal, note however that this neat solder job is not a thing for a beginner.

finished product using 6 layers of 0.2 mm NS

copyright: Henk Oversloot
date: 14 december 2004
extended: august 2006