bolier testing
loco designs
scale calculations
pattern making
lost wax casting
lost wax examples

Lost Wax Casting - Examples

Lost Wax - Case Study         - words by Rob.

Jump to:   Introduction     Making the Patterns     Results     Conclusion

During the on-going construction of my 5" gauge locomotive, I have strived to create a good likeness of the original. However, I was puzzled over how to make certain small detail items, which add to it's realism, particularly where several identical copies are required. Some people may wish to make every piece, but my time in the workshop is limited, and I'm not too fond of repetitive tasks !

The solution in my case, 'Lost Wax Casting' described on the preceding page, was written shortly after the process was introduced to me. This second article describes my experience and results.

The parts selected to be cast reflected the advanced stage of my project. These included dummy boiler wash-out plugs - 12 required; 4 larger rings that also fit into the boiler casing; and the awkwardly shaped sander valves - 6 required, which are also bronze colour in full size. The photo shows the finished patterns used to make the lost wax castings described.

Later, some unique 16 sided handwheels were cast for use on boiler controls, featured on the prototype's turret valves.

Making the Patterns.

The adoining photograph shows the cladding of the full size locomotive, with holes cut out to allow access to the wash-out plugs. These square headed, taper threaded plugs, are removed once a month or so, to allow a hose pipe to be inserted to "wash-out" accumulated impurities from inside the locomotive boiler. This is not necessary in our much smaller copper boilers, but we can still represent them to achieve the right look.

Working from an outline drawing, and a photograph of the boiler cladding showing rings welded around the hole edges, I decided that the 'copy' should have a finite depth and show a 'plug', rather than just be a hole in the casing. So, scaling the dimensions up slightly to allow for contraction in the casting process, I turned a piece of scrap brass on the lathe, making a short tube with a small flange at one end to locate it in the cladding, rounded to appear as in the photograph. The tube was then gently squeezed in the vice to give the correct oval shape.

To complete the pattern, a flat piece of brass was attached to the other end, and edges finished by filing smooth as shown. I decided not to include the square plug at this stage as I didn't want them all to be in exactly the same position, so just a centre punch mark was included for locating a drill later. Actually either a hole, or a representation of the plug could have been formed in the pattern at this stage.

Similarly for the sander valve, I worked from an outline drawing and a couple of good photos. The pattern was built up to produce a single piece, including dummy pipe flanges held by 12 b.a. Screws and nuts, with a 1/4 " hole drilled in the bottom, to take the sand pipe. A short length of rod of appropriate diameter, (in this case a broken drill), was passed to the caster with the pattern to allow the hole for the sand pipe to be replicated in the casting.

As mentioned previously, a mould is first produced, from silicone rubber, allowing the use of many materials in pattern making. Because the mould is jigged, once cut open, it can be used many times and allows multiple wax copies to be produced, then all cast together.

The Results.

Here was the beauty of this process, I only had to make one good pattern; then a few weeks later, I received a set of castings ready to use. The finished castings, had been sawn off the cast 'tree', and still had short stems attached. Otherwise, they were very true to the original.

The real advantage was spending say just two minutes cleaning up each washout plug with a small file, and it was ready for fitting. All I had to do then was carefully prepare a hole in the brass cladding sheet, drop it in and retain with a small fillet of soft solder. The 4 larger rings were lightly skimmed in the lathe, then secured to the cladding sheet using 12 b.a. Screws as shown in the photo; very straigtforward.

The surface finish was good, although there were very small ribs where the silicone mould had been split. The silicone mould is opened using a very sharp knife, which can leave a small join line between the two halves, although negligable in real terms.

Shrinkage for the components described in this item, which were broadly less than one inch long, was hardly discernable, say 1.5% in total, (although this would increase with size, see previous article).

Similarly with the sand valves, I was particularly impressed at the fine detail, even the 12 b.a. Nuts and threads were well represented in the copy. The attached photograph illustrates that quite complex shapes can be easily accomodated. Illustrated are the original pattern for the sander valve, a couple of interim stage wax copies, and a completed lost wax casting.

Threading the top them to be attached to the base of the sand-boxes, and the photo shows a valve ready for it's pipe to be soldered or glued into the hole cast in the end. As these were brass castings, they are maleable, and readily take threading (and soldering).

It is worth re-stating that for many components, the result is so close to the pattern that holes, bolts, details, and even certain threads will reproduce. The process is also ideal for shapes that are pretty impossible to create in metal in the first place, as the original can be built from virtually any material. Although my patterns were made from metal, as it was convenient to do so, this was certainly not a necessity.


The critical thing is to weigh up the initial cost of producing a silicon mould, as this is the most expensive component, and where possible consider having several items produced in the same mould. It may be possible to share the mould cost with other people's components.

Ideally, if producing lost wax castings for your project, it is most economical to produce all the wax patterns in one moulding block. Then, if at some future date, should you wish to produce some additional castings, the caster will only have to locate 'your' block to produce a batch of individual waxes.

I found this a rewarding exercise, and one I would recommend; the parts enhance the look of my model, and only required me to put my energies into producing one prototype for each shape. Because my project was already a long way through when I came across this facility, the items selected are somewhat random.

However, if embarking on a new project it would certainly be worth investing in the production of all the required patterns in one go, producing a master mould for the project.

Although several model engineering suppliers offer lost wax castings for various detailed parts, these can be very expensive. So if you have a number of similar or complex parts to make, why not consider commissioning your own castings ?