I was recently asked about the benefits of riveting or bolting joints in valve gear. This has prompted me to set down my method for jointing rods.

I have chosen to illustrate this with Walschaerts’ valve gear, partly because that is the motion I seem mainly work to with and so have examples to hand for illustration, but also because it is potentially the most complicated working set of joints a modeller will encounter. I discount the special inside valve gear sets that can be assembled as working or cosmetic as a positive choice will have been made to fit these. If your loco of choice has Walschaerts’ you must get it right as it is a fundamental feature of the locomotive.

Before describing the method, it may be wise to consider the positions in valve gear where you may consider making break points to ease the process of taking the motion apart. In general I have found that you can limit these to the components around the expansion link: the joint between the radius rod and the combination lever; the pivot of the expansion link; or the return crank – either the joint with the eccentric rod or by removing the crank pin complete with the return crank.

Breaking the joint between the radius rod and combination lever allows the con rod, crosshead, union link and combination lever to be removed as an assembly. Removing the pivot from the expansion link allows the radius rod; expansion link, eccentric link and return crank to be removed as an assembly. These are detailed later.

In addition to removing the motion links, separate cylinders and motion brackets can be an advantage. Some kits provide for this wholly or in part. Sometimes you just have to be inventive. Whatever is done here the main working part of these are the crossheads and slidebars.

When an investment cast set of slidebars complete with a rear cylinder cover is provided, they have a decided advantage. They make good structural unit. Even so they benefit from some attention to ensure smooth operation.

The running surfaces of the slidebars should be smoothed with a small file. Treat etched slidebars the same way by removing etched cusps on the edges. Similarly the running faces of the crosshead and piston rod should be cleaned up. The faces are straightforward, but a cast piston rod poses some difficulty. I use the modelling drill; insert the piston rod into the chuck with a fraction over the width of a flat needle file showing. Start the drill and apply the file. Pull the rod out of the chuck another file width and repeat the exercise. Carry on down the length of the rod in steps. It is important that the file is only applied to the rod next to the chuck to stop a bending load being applied. When an appropriate length has been cleaned up, I trim the rod and then further polish it with emery paper. This time I work the polishing along the length of the rod.

When all parts are cleaned up, check the fit of the rod into the crosshead. It may be necessary to gently bend the slidebars to ensure they are parallel to the axis of the piston rod in both planes as in Figure 1. A good check for free running is that the crosshead should gently fall out of the bars under gravity.

In some cases where the slidebars are separate from the rear cylinder cover, and in the case of a three-bar slidebar arrangement, I have found it beneficial to add a tube for the piston rod to run in. The two-bar slidebar gives three points of contact: the upper and lower bars and the piston rod gland. The tube takes over this role and guides the crosshead relative to the slidebars.(Figure 2)

And so to the joints

I usually use milled rods when they are available, but the method described below works with laminated etched rods too.

I use 3/64 in diameter brass rod soldered into one half of the joint. If you drill a hole the same size squarely into a piece of mdf, (Figure 3) this will support the rod whilst soldering. The holes in the rods usually have to be opened out to fit the rod. I use a cutting broach rather than a drill. It is then possible to get a relatively tight fit for the soldered side of the joint.

I use a paste flux for this job in preference to any other; it only goes where you want it. Just a touch of the iron with a very small quantity of solder is enough to make the joint. If you feel there is too large a fillet of solder around the pin, put a pin vice over the pin, nearly gripping it, and rotate it down on to the solder. The jaws then mill away the fillet. The pin is then reduced to be just proud of the surface of the rod.

The hole in the rod on the other half of the joint I make very slightly larger than the rod to ensure it is a free fit. Again it is broached rather than drilled. The hole in the rod on the other half of the joint I make very slightly larger than the rod to ensure it is a free fit. Again it is broached rather than drilled. (Figure 4)

The pin is then given a good coating of Carr’s Solder Mask. (Figure 5). I have found that if the lid is left off the pot for too long it dries out. To resuscitate it, I use isopropyl alcohol; when stirred in it becomes a paste again. If over-thinned it can be painted on. This is a benefit for these joints. By gently applying a hot soldering iron it can then be encouraged to dry rapidly onto the components.

Assemble the parts together with more solder mask. Slip a 14BA washer over the pin and solder. Again paste flux is best here. To make really sure the solder takes to the washer I brighten one face by polishing it on some emery paper. The soldering only takes a moment and the joint can be made without appreciably heating the earlier soldered joint. Trim off excess pin and file nearly flush to the washer. Now you have a very low profile joint. (Figure 6 )

Check the joint for free movement. Initially it may feel tight but as long as it does move this means that you haven’t soldered it solid and a few flexes of the joint will displace some of the solder mask and it will become free. A final oiling once it is on the chassis will complete the job. There is no need to remove the solder mask. It is a carbon compound (though it smells like shoe polish to me) and will help lubricate the joint. (Figure 7)

Break points

There are three main potential break points. (Click for picture). On the motion bracket of this model I have soldered a 12BA nut onto the inner part. The LMS style of motion bracket particulary lends itself to this approach. A 12BA screw passes through this and tightens on to the other side of the bracket where the nut is fixed. It doesn’t quite touch the outside part of the bracket. This serves to make the pivot for the expansion link. It may or may not support the radius rod too. However it is often possible, where you are presesented with a fair representatation of the expansion link, to sandwich it around the radius rod and locate the the rod in forward gear at a small degree of cut-off. The reverser lifting arm serves to hold it in place.

Another break point is at the end of the radius rod. By simply fitting a pin in the end slightly longer than for a fixed joint, it can engage with the combination lever with no washer. The slidebar arrangement of the valve chest and valve spindle serves to constrain the combination lever and they do not come apart.

The next picture illustrates the long pin and the arrangement of the expansion link pivot. Here it is a short length of the thin-wall brass tube provided by Eileen’s Emporium. You will see that it extends towards the inside of the motion bracket. It is spacing the link to the outside of the bracket.

For a time return cranks have been my bête noire as they have had a tendency to slip in the wheel. If the joints in the motion are free running, then no significant load should be applied to the crank, but even so I have had them move. However, making the crank as in Figure 8 has proved to be very successful.

It assumes a Slater’s crankpin and a etched laminated crank. Shorten the bush to just over the thickness of the connecting rod, too late once assembled. Drill and tap the bush and crank to 10BA.

If the crank is made from two half etch components it may be better to make a new inner one in full-thickness material. Fit the bush and crank onto a 10BA bolt, tighten the two together and then solder. (Figure 9). The outer half etch part of the crank can then be laminated over the inner on the assembly hiding the end of the screw. Tap the wheel 10BA. The plastic centre of a Slater’s wheel does not need to have a tapping drill to open the crankpin hole provided a first or second tap is used. The joint between the return crank and the eccentric rod can either be a soldered pin or a bolted joint.

The return crank is fitted to the wheel as shown in Figure 10. The geometry reflects outside admission. Sometimes I have reduced the offset as there seems to be too much swing in the expansion link – possibly a kit error? It is simply screwed into the wheel and tightened to give the required alignment. This may mean a degree of over-tightening but the plastic will be compliant enough to accept it. Complete the installation with a nut tightened down on the back of the wheel and trim any excess thread flush.

A crankpin bush tapped 10BA can also be used in the usual coupling rod positions, a screw passed through from the back of the wheel and tightened into the bush. It is essential there is a washer between the inside of the bush and the wheel, otherwise the rod can catch on the edge of the wheel fixing screw. This style of fixing is particularly beneficial where there is limited clearance behind the crosshead, a particular feature of LMS locomotives. Additional clearance can be generated by counterboring the coupling rod to accept the head of the bush. This is easier with milled than laminated etched rods, as the frictional heat can unsolder laminated ones.

In conclusion I will touch on a riveted joint. To successfully rivet a ball pein hammer is essential. I illustrate one in Figure 11. The ball end is used to set the rivet.

When a joint is assembled with the rivet, it is probably a moot point whether the rivet head is on the inside or outside of the joint.

In addition you will need to make a ‘dolly’ or receptacle for the rivet head. A block of steel drilled to accept just the head is suitable. Rounding the hole can be done with a dental burr if needed. The rivet is trimmed so that about one-and-a-half times the diameter protrudes beyond the components. Sit the head in the dolly and support the components. Strike the rivet with the ball end of the hammer. The edges of the rivet end will deform and start to make a dome. It is easy to steer the head around the rivet to form the head.

Stop once there is sufficient deformation to prevent the components separating. The parts should move freely afterwards. Of course, you can rivet the joint tightly if required to suit the application.

I hope that these notes provide some help with an issue that can be covered as a one-liner in kit instructions.