Reinventing the REAL 1

This unusual engine is based on the Stuart REAL, designed in the 1970s. The engine used the cylinder and valve gear common to the Beam, Victoria, Grasshopper and James Coombes engines.

I bought Andrew Smith's book, Building a “REAL” Vertical Steam Engine, (cheapest source of drawings) on building the Real and started to look a pictures of them on the net, the more I looked the more I saw that I did not like and that included what Stuarts are asking for the kit. So over a few winter evenings I developed a different version that has gone by the working title "Unreal" which addresses some of these issues and also more closely follows the etching from Andrew's book that was his inspiration.

Although I could have made the flywheel and cylinder from scratch I decided to use just these two Stuart castings as I felt that others who may want to follow suit would find these the most difficult parts to make from scratch. I also went for a metric design based on nominal sizes rather than direct conversion of the old imperial sizes again with many beginners now of an age where they were taught metric at school that make sense and if using hand wheels on modern metric machine sit is easier to keep track of whole metric numbers rather than working to several significant figure after the decimal point. This results in a model with 25mm bore and 50mm stroke

I'll go into more detail during the build about the various modifications but a summary is as follows:

•Base - This is a very expensive casting at £45 for what is little more than a rectangle with tapered sides and rounded corners So something with a bit more shape and detail.
•Columns - Again something with more shape and detail than the plain tapered ones.
•Entablature - This is shaped the wrong way for any form of classic architecture being larger at the bottom than the top all for the sake of easy removal from the sand so that needs to be corrected and some additional detail added at the same time.
•Pulley - change for one closer to the one shown in Andrew's book not the standard Stuart one used on all their 2 x 1 models.
•Flywheel - remove the "lumps" they now cast on the rim, add some counterbalance weight and fit with a gib head key.
•Valve/eccentric - modify layout to do away with the dog legged valve rod and make an eccentric correctly shaped for vertical use.
•Cylinder - move exhaust position as it gets a bit tight between column and exhaust position.
•Crosshead and guides - change to rod guides which was another feature of the engraving in Andrew’s book.
•Conrod - change from the tuning fork design to one similar to the other 2 x 1 Stuarts and Andrew's etchings.

On the cost front this model worked out at about £200, half of what Stuarts are currently asking. That is based on buying most materials by the 300mm length so you will also have some over for the next engine; larger diameters by the 25mm length where available. The cost also includes enough for fixings etc as well as the price of the two castings.

A few more bits of tooling and cutters may be needed but those will last a long time and if used over several more models will not amount to much. I did make use of my CNC for some of the parts but just about all could be done with manual machines and only one part would need a bit more of a re-design to make it a bit easier to cut from solid.

I decided to make a start with the flywheel while I was waiting for a delivery of materials that would have allowed my to start more logically with base. Both the flywheel and cylinder castings were very good; no holes, no hard spots, mould halves lined up with no step and very little flash. The only real downside is that Stuarts seem to only supply castings from the pattern used for the twin Victoria which have the ugly bosses cast in where two flywheels are bolted together.

I like to fettle my castings before machining so there is no risk of spoiling any finished surface if you later slip with a file or grinder, so that was the first thing to do. As the inner rim is not machined I like to set that to run true and was able to do so by holding in the 3-jaw. This allowed me to turn the outer surface, one side, clean up the hub and also bore the hole all at one setting.

I like to bore the hole to finished size as a reamer is likely to follow a drill hole that may have wandered and I can also get a slightly closer fit than a typical reamer will give which reduces the chance of a wonky flywheel. Once I got close to size by measuring with the digital calipers I switched to using a piece of the crankshaft material as a plug gauge to get the final size.

I did the outside with the lathe running in reverse and used a boring bar mounted upside down which easily reached around the casting with minimal overhang.

I then changed to the 4-jaw and set the machined surfaces to run true using one indicator on the OD and another against the side.

It was then a simple job to turn the other face and clean up the hub.

You can see where I have turned away the bosses that there are some flattish areas these were blended in with a Dremel using an oval shaped grinding point.

Stuart show a grub screw to retain the flywheel but I used a broach to cut a tapered keyway and machined up a suitable gib head key from gauge plate to fit.

Jumping ahead to when I test ran the engine it was quite noticeable when running slowly with no load that the engine rotated faster when the crank was moving downwards than up which is due to the added weight of the conrod, crosshead, piston and piston rod all pulling downwards and then the engine having to work harder to lift that weight back up again. So some strips of lead flashing were cut and bent to the curve of the inner flywheel and held in place with masking take to experiment with how much counterweight needed to be added.

Once happy the two layers of Code 4 lead were bonded in place with JBWeld. Once hardened the excess JBW was cleaned up, a bit more U-Pol filler added to blend things in, and you would think it was all part of the casting.

Part two here