The Roy Link kit I am using is for 7mm scale, but I am building this small 18" gauge Bagnall to 1:32 scale, a third as big again. Because Bagnall's seemed to scale the functional parts of their locomotives to the gauge they were going to be used on, most things below the footplate don't require modification, and nor do the functional bits, such as boiler, firebox and water tank. The human bits do, so the cab has to be higher, some of the controls - handbrake in particular - need to be upsized, and because humans don't like coal smoke in their faces, the 1:32 model requires a much taller funnel. This picture shows the funnel that had to be made next to the cast funnel supplied in the kit

Now I have a Peatol lathe (now sold as Taig lathes) so the obvious route was to turn the funnel from brass rod. Thanks to the kind offices of Nearholmer of this parish I now had a GA of a Bagnall Sipat loco and was able to use that to make a dimensioned drawing of the funnel to use as my guide.

The first step was to mark off the positions where the different features would be on the brass rod. At this stage accuracy to within 0.5mm is all that is required, in fact that level of accuracy is plenty along the length of the funnel. It's the cross-sectional dimensions that require the digital vernier's attention.

The next step was to drill a ⅛" (3.2mm) diameter hole the length of the funnel. The purpose of this is to provide a means of alignment later on.

I decided to work from the base end of the funnel so that the top was at the chuck end. I have a compound slide so I was able to set it to cut a taper of 1.5º and shape the main part of the funnel. Now the important part is not the actual angle of cut but the cross-section measurements at the top and bottom, so about halfway some adjustment was needed to the compound slide setting. Making that adjustment while there is still a fair bit of metal to remove means that further adjustments can still be made if necessary.

I did cheat a little in that I used the radius of the cutting tool as a form tool for the curvature at the cap end, as well as putting some rounding in at the bottom end.

Creating the flaring at the base of the funnel is one of the more difficult tasks in loco building. Some advocate filing to shape, others creating a huge blob of solder to carve, or doing the same with filler. I thought I'd try an old John Aherne method, page 82 of "Miniature Locomotive Construction". That required creating a cone shape and afterwards boring it out so it only had a thin wall. The cone shape was made first. The compound slide was set by eye to create an approximate cone and then refined as the last cuts were taken. With the slide set to the angle to the last cut on the outside of the cone the tool could be changed and a boring tool fitted to bore out the inside to a depth of 3.6mm

The funnel cap could now be shaped and the funnel parted off. The funnel was then turned around and fitted in the chuck - a bit of masking tape wrapped around it to protect it from the chuck jaws - and the end faced off flat. The final step was to reset the compound slide to 1.5º and bore out the funnel pipe for some distance.

This is not the first time I have tried this classical method of forming the base flare of a funnel, but the method as described by Ahern does require three hands and a flexible spine. One hand to hold the funnel upright, another to hold a small block of wood on top of it as protection and a third to hold the hammer used to persuade it into shape. Plus that flexible spine so you can see what you are doing while tapping with the hammer. As I don't have that third hand nor that flexible a spine any more I thought I'd use a vice and hold things with a jig.

The jig was a piece of steel turned to the diameter of the model's smoke box. It then had a ⅛" hole drilled in the side. The four jaw chuck allows this to be done accurately. A piece of ⅛" silver steel rod, which I have to make axles from, was then used as a locating pin. The funnel could be mounted on that and it would be accurately held at a perpendicular to the former.

Before attempting to apply any pressure however, the base of the funnel needs to be annealed. That means heating it to a dull red heat and either quenching it or allowing it to cool naturally. This softens the brass and makes it more amenable to taking up the desired shape. Then jig and funnel were put it the vice and the whole lot gently squeezed.

The funnel needed a couple of trips back to the blowtorch for further annealing. I didn't do this the first time and the flaring cracked.

Photographs of Brede and its sister locomotives do not show the funnel with its flaring directly on the smokebox. Instead the funnel is mounted on a thick plate. From the drawings it looked as if that could be made from 1/16" thick brass, the sort of stuff we used to make loco side frames from. So I had a bit lying in the drawer, unused since c1985. a hole was drilled and a piece cut off and then bent to shape over the former. The funnel was then soldered to this, again held upright by a ⅛" rod inside.

The final step was to return the funnel to the lathe and very carefully, using the lightest of cuts, trim the base to be a perfect circle.

And that gave me the funnel shown in the first picture. Next time I'll report on progress with the cab.