I started this project coming up for nine years ago, when the wheel castings were machined. For most of that time the engine units sat as rolling chassis, taking up bench space and slowly developing the patina of age, which is one of the drawbacks of living in a coastal rust belt.
Work finally restarted just after the 2023 GSSU about 14 months ago. I didn't have a new project to start and I decided to sort out the engine units and at least get them running on air. After a while I got stubborn and dug in for the long haul. While it took many more months than planned/expected, the construction phase was completed a couple of days ago. The light at the end of the tunnel turned out not to be an oncoming truck.
Project drift set in early and the final model is slightly different to the original intent all those years ago.
Prototype Story
The project as originally planned was to build a model of the proposed Beyer-Garratt that the Victorian Railways ordered from B-P in the mid '20s. The VR wanted a 2-6-2+2-6-2 but at the end of the process they got a 2-6-0+0-6-2 that became the VR G class. One of which is preserved on the Puffing Billy Railway near Melbourne.
By the time I restarted the project I had built my backyard track and found that the 2-6-2+2-6-2 would struggle with the curves in the wye, so the project was cut down to a 2-6-0+0-6-2. I didn't want a VR G class, so the project became a model of an APC Garratt.
APC was the Australian Portland Cement company, which had a quarry and cement plant near Geelong in Victoria. During the '30s they purchased two medium sized Garratts to handle the increase in limestone traffic between the quarry and the kilns as business recovered from the Depression.
The APC Garratts were an updated version of the 3'6" gauge WAGR M/Ms class Garratts, which were the third Garratt design, entering service in 1912/13, just before WW1. The VR G class Garratts were similarly derived from the M/Ms classes, but more heavily modified for the narrower 2'6" gauge.
The quarry line was 3'6" gauge and had a 1:37 grade up from the quarry floor, as well as the longest railway tunnel in Victoria. The two Garratts spent the next 25 years or so blasting their way over this line with loads of limestone until they were replaced by a diesel and then finally the railway itself was replaced with a conveyor belt in the late '60s.
One of the APC Garratts was preserved, as were the rest of the remaining APC steam fleet and is now located on the Bellarine Railway near Geelong.
The APC Garratt was chosen as a prototype for the model build purely because it had run in Victoria and I liked its appearance. When I was doing the original research, I was unable to find a source for the Beyer Peacock drawings, so the model was built from photos and relevant info from a set of VR G class drawings.
Model Story
There are 50 odd progress photos of the model build available, but a Garratt is basically just two conventional models slung together. I built this one the same way I build a conventional model, it just took longer because there are two of everything.
The cylinders and valve gear are scratchbuilt, based on the design in Brian Wilson's book. The wheels were machined from Walsall iron castings. The rest was made from steel or brass sheet and bar.
The photos below are composite photos of the specialist parts specific to a Garratt model, the flexible steam lines and the large boiler.
The first photo shows the engine units and the boiler cradle at the point where the steam lines and flexible joints had been installed and were about to be tested on air. The third part of the pic shows the chassis/cradle running on air.
The flexible joints are the black rubber sections of steam piping over the engine unit pivots. The rubber is a grade that is resistant to steam, oil and hot water. The connector design came from a friend who uses them for water lines in G1 tender engines which can be under boiler pressure when pumping water. I made up a couple of connections for testing and found they stay tight up to about 300 psi. At pressures higher than that, water is forced between the brass and rubber parts, acting as a lubricant and the rubber is pushed out of the fitting. No rupture, the rubber just quietly pushes out of the fitting and starts to leak water. At 40 psi boiler pressure, that is a 7:1 safety factor, which is more than adequate for purpose.
The copper and rubber tubing used for the steam lines is sized so that there is minimal restriction at each tee junction and the internal cross section is adequate for free flow of the steam. The extra care taken over the sizing seems to have worked as planned, as each engine unit starts at about the same pressure and runs at about the same speed.
The second composite photo shows the boiler after assembly and during testing.
As mentioned a Garratt has two of just about everything. With two engine units, that means four cylinders, so if you want something like prototype performance, you'll need twice the heating area in the boiler and twice the output from whatever source of heat is boiling the water.
This model is gas fired, so the boiler has been designed with two flues and two burners. Scale size for this model boiler works out to 75mm OD by 220mm length. That's a big boiler.........
The first part of the pic shows the boiler after being silver soldered and cleaned up for hydraulic testing. The two burner flues are apparent, as are the nuts for the longitudinal stays, as well as the bushes for various fittings. For comparison, the piece of copper tube beside it is the shell for a boiler the size of a standard Roundhouse Type 1 boiler, as fitted to a Lady Anne, etc.
The second part of the pic shows the boiler getting a pressure test, holding test pressure, with no leaks, or distortion.
The boiler has been steamed on the bench, but the engine units have not run under steam and won't until the model is painted. The gas tank is 80ml capacity and has been fitted in the rear bunker/water tank. With two Roundhouse burners to feed, this should give at least a 30 minute run. We will see......
The last photo shows both sides of the assembled model sitting in the afternoon sun yesterday, with the old engineer giving some idea of scale. The loco is about 800 mm over the couple pull lines and with a boiler full of water ready for a run it weighs a bit over 8kg.
I still have to build a carry case for the loco that will take the weight and also work up enough enthusiasm to paint it when the weather cools off. Don't hold your breath.
Learnings
The major learning is that a Garratt isn't a quickie project...........
I put 6 months into this model back in 2016 just getting it to the rolling chassis stage, then after I restarted work it took another 2 months to run it on air. Another 2 months to install the boiler, then the last 9 months was spent making all the bits that make it look like a model of a real Garratt. The materials weren't cheap either.
A bit of time would be saved by using a couple of Roundhouse chassis kits to make the engine units, but the cash outlay would increase by a factor of about three.

I don't see another Garratt in my future though. There are a couple of Roundhouse chassis kits under the bench and I haven't yet decided what sort of locos they'll be used for. One thing is for sure, the next steam loco project won't be a big loco and it won't be a bendy loco.
I think I'll go back to building wagons for a while, or maybe something small and battery powered. There were times during this build when going back to HO started looking like a good idea.

Regards,
Graeme