In model engineering there is a cost scale that comes to mind before I continue with the technical bits. It’s simply, the more you make yourself, the cheaper the hobby, and I do this hobby on a boot strap budget. Most people get a set of castings and machine them to drawing, along with other bits and bobs, but these castings add up and before you know it you’re in trouble with the domestic CFO. Doing your own castings is an interesting exercise, one that I would not recommend unless you’ve done considerable reading and research on the topic. So before I continue, to keep the health and safety officer happy, don’t try this at home unless you have the necessary competency to do it safely.

Pattern making for casting

Simply put, casting is melting a metal and pouring it into a suitable mould where the imprint of a component has been left.  The metal fills the cavity and after solidification and cooling the sand is broken away and the cast component is left. The imprints in the sand are made using patterns, generally made from wood, but they can be made from practically anything that won’t be damaged by the ramming of the sand around the component. These patterns generally scare people away from casting due to the complexity of the geometry. If the geometry were easy why bother with casting at all; simple machining operations would suffice. In reality, any complex geometry can be broken down into simpler shapes and just glued together. I’ve found that printing a sketch of the geometry and gluing it to wood makes light work of pattern making. With a scroll saw complex patterns can be cut, then finished by sanding or machining depending on what the geometry looks like. Corners need to be filled to prevent the casting from cracking and this can be done using standard automotive body filler. Finally, the wooden pattern is sealed using a sanding sealer.

Pattern making for casting

The furnace

My furnace is an oil fired furnace I designed and built just for making the castings for the locos. It costs me practically nothing to run as the oil is the waste oil from my last car service. And yes, the furnace complies with the environmental protection act; I designed it to have primary and secondary combustion to prevent damage to the environment. If you’re still not convinced, the furnace uses less oil per kg melted than my car uses petrol getting to town (around 15km), and nether have any visible smoke from the exhaust during normal running conditions.

Home Made Furnace

Making the moulds

The moulds are made from sand, bentonite and water. The sand is rammed around the wooden pattern, the mould is opened, the pattern is removed and then the metal is cast into the cavity that remains.

Most of the metals I use come from scrap. Copper which is the base metal for brass, gunmetal, etc. typically comes from discarded motors I find lying around. I think my wife gets a little annoyed when we’re on our way to a social event and I stop to save an appliance left on the side of the road destined for the city dump. I’ve stopped telling people I make my trains from old vacuum cleaners etc. because they think I’m being facetious, but there’s actually a lot of truth in that statement. Below is a picture showing the casting and the final (machined and polished) safety valve cover I made using plumbing fittings from when our geyser burst. It was interesting explaining to the plumber why I wanted all the spare parts…

Moulds for Casting

Finishing off the castings

When the moulds are cool enough to handle, the sand is broken away, the risers, etc. are cut off and the castings are sandblasted. I have a bench type sand blasting cabinet which works really well. A tip on sandblasting; it’s important to buy the correct compressor for the type of blasting you want to do, both in terms of pressure as well as air flow rate. The compressor needs to keep up with the consumption rate which is typically high with sand blasting.

Below are a few castings I did in a morning, with the materials ranging from cast iron to phosphor bronze, gunmetal and one brass casting.  All risers were removed and the castings sand blasted clean.

Machining some of the castings

As you can imagine there are quite a few components that are cast and need to be machined. I’ve decided to focus on the main drive wheel and the cylinders (next time). I think those two assemblies illustrate a number of machining techniques that might be useful to the reader. If there’s any particular component readers are interested in you can leave a comment and I can add it to a future article.

The main drive wheels are machined on a faceplate, due to the sheer size of the wheel (almost 300mm including the tyre and weighing in excess of 4kg). I generally use the cross slide to position the casting by moving the tool (with the machine off!) until it touches the casting; I then zero the cross slide dial and turn the casting 180°. I again bring the tool to touch the casting and the amount on the dial is the offset of the component. The casting needs to be moved by half that amount. This is repeated at 90°, and after some iterating you’ll have the casting close enough to the position you’re looking for. You can see from the picture I had to machine from the back end of the tool post to clean the edge of the rim, I have the 550mm geared head bench lathe, and I just managed to machine the wheel in my lathe. The centre hole was machined for a press fit to a 25mm shaft.

For those starting out with machining the speed and feed rate are important, and I get loads of questions during courses I present around what speeds etc. to use. Unfortunately, speed and feed rate change based on the material, but as the diameter increases you need to decrease the cutting speed. If you get tool chatter the first thing to look at is the cutting tip, thereafter the feed and speed. Generally if you decrease the speed and increase the feed it will solve the problem. If you get a burr in front of the cutting tool while cutting again the first thing to check is if the cutting tip is suitable for the material you’re cutting. Thereafter,  it helps to increase the feed. The same applies to milling processes. I’m hesitant to give fixed speed and feed rates but I’ve found that when facing I start with a cutting speed twice that I would use for the outside diameter. When cutting stainless I start with a cutting speed half of that I would use with normal mild steel. If the chips come off blue you either need to decrease your cutting speed or use cutting fluid. With high carbon steels even using cutting fluid the chips come off blue, but specialty steels are not that common in the home workshop.

Machining the castings

Below is a collage of the various stages in making the main drive wheel, from the pattern to the sand mould, the wheel broken out the sand and finally the machined product with the tyre heat shrunk into place.

Various stages of finishing the casting

The wheels are then pressed onto the axil assembly which includes the eccentrics and cranks. I have a 20 Ton hydraulic press with gauge which is perfect for the job. I’ve found the gauge very handy and if any of my press fits exceed around three tons I start to get worried and I know the metals have started to bind and my tolerances weren’t up to scratch. It’s also worth while taking some time to set the press up properly. During assembly the columns need to be square and the cylinder in the centre, this will result in balanced forces making future work much easier. I often wonder how workshops manage press fits without knowing how much force they’re applying. Contrary to popular belief if too much force is applied it can actually weaken the fit. The materials can crack in operation due to the excessive forces; you can get binding between the materials, etc. I normally put some light oil between the surfaces to prevent binding, and this helps with disassembly should it be needed at a later stage.

Wheels are then pressed onto the axil assembly by a Adendorff 20 ton Press

The next article will take a look at the cylinder assembly and some other interesting machined components…

Adendorff Machinery Products used in A TRAIN IS BORN PART 2

R 1,695.00
R 2,095.00

16 thoughts on “A Train is Born, Part 2

  1. Craig says:

    So have you ever tried using K-BOND home made alternative to Petrobond for casting. You have to some how get hold of Bentone which is a modified Bentonite apparently. By the way where do you get the blueprints from?

  2. Luke says:

    Hi Craig, No I haven’t. The only commercial binder I have used is a phenolic based system, but I stopped using it because you can’t reuse the sand. The correct bentonite is the one the farmers use for their dams. Its readily available, and I paid just over R150 for a 40kg bag. And I’ll probably never have to buy again, because you use such small amounts. Not sure how much it costs now it has been a while! For the cores I make up a linseed oil binder, using stuff from the local grocer. These are baked cores and can be cured in a commercial oven when the boss is out for the day.
    I unfortunately don’t have blueprints, but plans for a similar loco was published in model engineering in the 80’s and my design is loosely based on that. I did manage to find a book on these locos, by chance, that has a general arrangement drawing which has been very helpful. According to the book most of the drawings for the Stirling singles were destroyed during the war. You can apparently import a set of blue prints from some of the model engineering shops in the UK, but its very expensive.
    Thanks for the posts!

  3. Clive says:

    An excellent article thus far – I really trust that part 3 is on the way.
    I am so very interested in the wonders of steam and would very much like to build such an engine.
    Would you consider commercially supplying the moulded bits and pieces required ?

    • Luke says:

      Hi Clive, thanks I’m glad you enjoying the series, part 3 should be loaded. It will be difficult to commercially supply a set of castings for you, due to the time it takes to make them. These castings took me the better part of a year to complete. What I did with my previous loco is I sold all the spare castings, but in this case I didn’t do to many spares due to the size of the loco. Regardless, the spares I do have probably won’t help you as I’m not sure too many people have tackled this model, making any commercial castings hard to come by.

    • Luke says:

      Hey Craig,

      I’m busy designing my next loco, and I would suggest looking at adapting life size plans for a model. I’ve found a number of books on loco design, combined with existing plans from some of the greats like LSBC you’ll be able to design a unique loco that will definitely do the job. I’ve actually just finished optimizing the valve gear using simulators available online, something the men of old didn’t have access to, which makes the job considerably easier.

      I’m hoping to start the patterns and castings in the next couple of weeks!

  4. Johannes Robbertse says:

    Good day Luke
    I do engineering myself and sometimes really have the need to do castings
    is there any way that you could help me with plans and design for your Oil fired furnace
    I would really appreciate any help or ideas on that

    Regards Johannes

  5. luke says:

    Hi Johannes I can do one better, if you look at there’s plans for an oil fired furnace and burner that will work for most metals. Unfortunately there is more to casting than just having a furnace and you might need to do some reading up on metal chemistry and dosing to get good quality castings. I’ve found a wealth of information on university online archive platforms were books out of copy-write are available for free download and they actually better than some of the modern books on foundry techniques.

    • Luke says:

      Hi Tiaan,

      As nice as all the commercial products are, it is possible to get very good quality castings without using all the fancy foundry supplies. With the exception of the crucible (and cast iron additives) I don’t purchase any consumables from foundry suppliers. Even my sand is from the local river (washed free of charge). My dosing spoons and plungers are normal stainless cutlery, I make my own mould wash, and core sand, again with items from the local hardware shop and grocer. I have successfully cast a number of bronzes including AlBr and the tin copper alloy range for bearings. In terms of cast iron, generally I stick to grade 17 but I have melted disk breaks for the firebox grate.

      In short foundry suppliers weren’t around during the bronze age and some of those castings look better than what I’ve seen in foundries…


  6. Luke says:

    Hi Guys,

    I’m pleasantly surprised by the feedback and interest in the casting section of this series. I thought you might be interested in some of my new ventures in casting. I’m tackling the patterns for the new loco a little different. All my previous pattern work was done using wood, which required various cutting and woodworking techniques and loads of sanding. For the new loco I purchased the cheapest 3D printer on the market, and with a little engineering improvement I’m able to print first rate patterns. The nice thing is; I can now print risers specific for the mould which has worked very well so far. All the wheels have been cast, and not like the stuff you get from overseas, it has the same detail as the originals with correct spoke drafting and the rounded rim. The axle boxes and pedestals have been cast, basically finishing the casting for a rolling frame. Much quicker than the Stirling!


  7. Mike says:

    Hi Luke

    Congratulations on you inspiring series. I enjoyed them tremendously.

    I looked at the website. Which model of furnace did you build?

    Is it possible to obtain the MEW articles from you?


    • Luke says:

      Hi Mike,

      Sorry for the belated reply, I’ve been busy with the new loco! Thanks, I’m glad you enjoyed the series!

      My furnace is my own design with a oil fired injector that works on a similar principle to the ones on the web. Essentially I just optimized what they had done allowing me to melt a little quicker using less oil. I melt 4kg at a time, this gives an idea of the size.

      You welcome, I have scanned them in but my loco is far from the original design. The boiler was resigned to code and the whole back-head including the valves etc. were designed as working valves and to resemble the original back-head. I don’t want to give my details on an open forum but if you give Adendorff a shout they will pass on my details. I suggest you give me a call if you want to build this loco as some of the dimensions and drawings have minor issues and won’t give you an easy working loco.


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.