LOCOMOTIVE KITS – REBUILDING A WINSON
Sam, a 16-year-old member of the Sussex Miniature Locomotive Society, describes how the used bargain of an unfinished kit from a defunct manufacturer provided his introduction to the model engineering hobby.
Ever since I was a young child I have enjoyed watching and riding on trains, old or new, big or small. I also remember riding at the Beech Hurst Railway run by the Sussex Miniature Locomotive Society (SMLS) in Haywards Heath, West Sussex.
In the spring of 2017, I was given the chance to help run and maintain the miniature railway at Beech Hurst. I was very lucky to be able to drive locomotives with supervision round the track but also to meet very helpful and knowledgeable people at the club. It was clear to me that I really enjoyed this hobby and I wanted to try something new. I had shown interest in owning my very own locomotive but didn’t really know where to start...
Introducing the 14xx
It was around November of 2018 when fellow club member and a good friend of mine, Andrew Brock, just happened to come across a 1999 Winson Engineering kit for a Great Western Railway 14xx 0-4-2T (Photos 1-3). I was not specifically looking for a loco at the time, but it just happened to suit me so well – it was almost as if the project found me!
As I was just a beginner I didn’t really want to start a scratchbuilt loco as such projects take lots of time and experience. The 14xx was going for a very low price and everything appeared to be in the kit including the boiler. I bought it for the low price advertised because I wanted to learn with this project.
The loco seemed to be a simple build, with its running chassis already made, but later on we found out that this was not going to be as simple as we thought, with the loco having several underlying problems that went unaccounted for when being sold.
The issues included the wheel quartering, the slide valves not
seating, the eccentric pin heads fouling on the eccentrics and connecting rods that were too large.
It became apparent that this was going to be a larger project than I had expected but luckily with excellent help from other members of the club the build soon gained traction and I was to learn a great deal from the whole experience.
The prototype
Designed by GWR chief mechanical engineer Charles Collett and built in Swindon works from 1932 to 1936, the 1400 class locos were built for branch line work in the GWR and later BR (W) regions of the UK.
Originally classified as the ‘48xx’ class, the locos were designed to run using the auto-coach system where the driver could drive from a cab in the leading coach. However, in 1946 these locos (4800-4874) were re-numbered to 14xx to make way for the 12 ‘28xx’ class engines being converted to oil-burning so they could use the 48xx class number.
The 14xx locos could run on any of the GWR and BR(W) branch lines and were quite successful. But around this time diesel locos had started proving themselves and the lines on which
14xx locos ran were being closed by
Dr. Beeching. As a result of this, the locos started to be withdrawn from service in 1956 and by the November of 1964 the final four had been retired.
There are currently four of the locomotives in preservation:
No.1420 is stored at the South Devon
l Railway and as of 2021 not in service
No.1442 is a static display loco at the
l Tiverton Museum in mid Devon
No.1450 is now at the end of its
l 10-year boiler ticket and awaits overhaul at the Severn Valley Railway
No.1466 is under overhaul at the
l Didcot Railway Centre in readiness for the Great Western Society 60th anniversary in 2021.
The beginning…
Very soon after purchasing the engine, it was decided to hydraulically test the boiler, however I needed some boiler bungs as the loco did not come with any. I was allowed to borrow some from a friend, but I was still three short. They all needed to be 3/8-inch diameter with a 32tpi thread and making these provided my first machining job on the lathe. With help, I got there.
I used a brass hexagon bar and mounted it in the lathe. I learnt how to ‘zero’ the dials and then turn the material until it was 3/8-inch diameter. Having turned each bung, I then set up a 3/8-inch x 32tpi die in a die stock – this was also my first time using a die properly on my own. I used the tailstock of the lathe to apply a small amount of pressure onto the die and bungs so it would cut a good, straight thread. Once all was set up, I was ready to start cutting the thread.
I used the lathe bed to stop the die holder from spinning and manually turned the lathe chuck to cut the thread (make sure the lathe is out of gear when you do this!). After a couple of full turns I released the pressure from the tailstock and unscrewed the die to break the bur and brush it off, I then applied pressure again and started to turn further up the bung.
I repeated the process on the other two bungs and I now had all that was needed for a hydraulic test and before the test commenced, I wrapped the bungs in PTFE tape to stop any leaks that might occur through them.
With the bungs in place the next thing to do was to measure the boiler capacity. The boiler ticket from the factory did not state a boiler capacity, only boiler pressure. So, with a measuring jug and a funnel we managed to measure the capacity – it was only a small boiler at 2.05 litres.
Now to commence the test! The boiler had been tested from the factory at 180psi hydraulic pressure (90psi steam pressure), so we replicated the test to that pressure – I didn’t think such a smaller boiler would ever have to go to that pressure!
The boiler passed with flying colours with only a pinprick leak that let the smallest amount of water out. It was a very well-made boiler with
nothing to worry about – my project had made a good start…
However, this was where the restoration began. After some initial tests were carried out on the ‘running chassis’ it became apparent that the valve gear was out of time, numerous air leaks were discovered and the coupled wheel sets were not fitted properly to the axles.
Degrees of difficulty
To start with we had to check that the cranks and both wheel sets were quartered to 90 degrees. This was all checked on a Myford Super 7 lathe by holding the axles between dead centres and using the lathe bed, a ground vertical fixture and slip gauges to determine the size of the error, if there was one.
Unfortunately there was a problem because the original builder had keyed the front wheel set with an error of 0.008-inch. Because the left-hand side wheel was also loose, this measurement of 0.008-inch was at the ‘good’ end of the error and at this stage we could only assume this was where the left-hand wheel should be set. Before making any hasty decisions, we checked the rear (crank) wheel set and found this to have an error of 0.016-inch but with both wheels loose!
With this information known, the front right-hand wheel was removed, thoroughly cleaned, set up in a friend’s quartering jig and fixed using Loctite 638. Once the Loctite was fully set, a 5BA stud was drilled into the axle and wheel to fully secure the two together. The process was then repeated on the left-hand wheel, so the front axle was complete.
Afterwards, the wheels on the crank axle (rear driving axle) were quartered using the same technique. Both wheels were then re-attached, again using Loctite 638 but were not pinned at this stage.
We then needed to make new bronze bushes for the coupling rods; this was because the original bushes had been elongated by 0.020-inch just to make the wheels turn!
With the wheels now set properly, quartered and back in the chassis and the new bushes turned, drilled and reamed to fit the crank pins, an initial test proved the wheels could now rotate 360 degrees with no compromises and only a very minor tight spot.
The valve gear timing could now be checked. This showed the eccentrics to be at the correct position on the axle, however the pistons were just kissing the front cylinder covers as the connecting rods were too long. There was also a considerable air leak through the slide valves.
Making two new connecting rods would be too big of a task at this stage of the build, so a secondary solution was needed to solve the over-length issue. This solution involved cutting the existing rods, shortening them by 1/16-inch and brazing them back together. This proved 100 per cent successful– the pistons now cleared the cylinder covers with no issue.
The cylinder covers were also relieved to allow for the bolt which was secured to the front of the piston rod. Furthermore, new gaskets were also made to seal the covers at both ends. However, there was still a significant air leak from the slide valves and investigating further, it was clear that the faces of both the valve chest and the valves themselves were unsatisfyingly rough.
To clean these faces we would need to remove the cylinder block from the chassis once more – one step
forwards, two steps back! While they were rough, the faces were good enough to not require any machining, we hand lapped them on a surface plate using various grades of emery paper. This process started with 120 grit, through 400 grit, 800 grit and finally finished using a 3000 grit paper for a smooth and mirror like finish, but essentially to stop any steam leaks coming through the crevasses in the material (Photos 4-7).
Once finished, the cylinder block was returned to the chassis and two temporary gudgeon pins were made from mild steel to secure the little end of the connecting rods to the crossheads. We planned to re-make these using either case-hardened steel or silver steel at a later date.
With these tasks now finished, the locomotive could now be tested on air again. The main cylinder block, gaskets and valve faces sealed well, however there were still minor leaks from the pistons and valve glands. To fix this, the old gland packing was removed and temporary split viton O-rings were added. This largely sealed the leaks.
Valve adjustment
Further air tests were successful but proved the valves themselves could not be adjusted equally about their mid-point, because the valve spindles were too long – another setback but we were still making progress!
With another job added to the ‘to do’ list the valve spindles were checked, removed from their buckles and shortened by ⅜-inch to allow the more or less equal adjustment about their mid-point. The thread on the buckle ends of the valve spindle was also reduced in length by ⅓2-inch as it protruded beyond the thickness of the buckle and risked jamming the valve in the buckle.
Another task on the to-do list was to turn up, and thread, some draincock blanks. These would replace the auto draincocks that were supplied with the loco and which could be individually removed to help with testing.
Finally, with the to-do list finished and the valves re-assembled, a fully successful air test on the rebuilt chassis took place in late September of 2019, which coincided with the running season drawing to a close at Beech Hurst.
The chassis was now largely complete and running well in forwards and reverse. This meant it was now time for stripping, painting and the replacement of the temporary valve gear pins used during testing…
■ Sam will continue the description of his rebuild in coming issues of