Model Engineering / Track Construction

   

My approach to building a 5" gauge home railroad

This story appeared as What a Way to Build a Railroad! in the
Australian Model Engineering Magazine in November 1990 (Issue 33).
Unfortunately the issue is no longer available.
The following text has been updated and includes original photos with a few extra included.

Where it all began

Railways large and small have been a part of my life since childhood. The hobby grew with me and alternated between electric and live steam models. It was in the early 1970s when live steam took hold and construction of a locomotive commenced. As this was my first adventure into a project of this magnitude, joining a model engineering society was a high priority. Being a member of a society is a marvellous way to meet people with similar interests as well as a perfect source of motivation and all the help you need to complete a project.

During construction of the locomotive my thoughts turned towards building a ground level railway system at home, as my club track at the time was thirty kilometres away. After all, the locomotive will need to be tested from time to time! It is also a thrill to be able to steam up when you feel like it and not be confined to once or twice per month.

This is the story of how the track was built. The methods shown here are applicable to any outdoor model railway from 3-1/2 inch gauge upwards. My locomotive is a narrow gauge unit built to 2-1/2 inches per foot scale running on 5-inch gauge track, so sharp radius curves and steep grades are not really a problem for my railway at home. Short wheelbased standard gauge model locos would be able to run on this railway as well.

The layout of the track was not difficult to solve, as the location of the house on the land allowed a track to run around it in the form of an oval some 200 feet long. There is provision for a branch line to the wood stockpile at the back of the garage as well as a passing loop down the driveway with a spur into the garage.

There has been, and probably always will be, a great deal of discussion about the merits of "point to point" form of layout for "realistic" railway operation versus a circular pattern. However the circular pattern was selected simply because there are occasions when steaming out on long journeys are the order of the day. The fact that the scenery repeats itself every thirty seconds does not worry me in the least. Anyway, when you are concentrating on driving a steam locomotive you do not have a lot of time to admire the view. The family all agreed that it would be fun to have our own railway, so the proceedings began in earnest.

Making a start

The track alignment is pegged out around the front yard - January 1985.

The first step was to locate the centre of the curves at the front and rear of the house. From the centres a string "trammel" was used to measure out the twenty-one foot radius to peg the track centreline. After that the straight sections between the curves were pegged out. Having the oval pegged out; the next step was to establish the extent of the gradients. This was achieved by using a through-the-lens viewfinder movie camera set on a tripod and levelled. The staff was a white stick with measurements transferred from a tape measure. Yes it was rough, but quite adequate for the job. The end result was that an eighteen-inch high embankment would be required in the southwestern corner of the oval to reduce the gradient to 1 in 40 along the southern side of the house. This provided the reason for a timber trestle bridge at that location, as a solid embankment would create a dam every time it rained. The bridge is only six feet long and the remainder was built up with soil obtained from a nearby building site, with the blessing of the builder.

After allowing some time for the filling to settle, about twelve months as it turned out, the roadbed was prepared. In the backyard the roadbed was cleared to a depth of four inches and a width of eighteen inches, the extra width was required to allow the top of the concrete edging to sit flush with the lawn. This trench was filled with 3/16" (5mm) bluemetal until the track was laid so that the children could play safely without falling into the ditch. 3/16" bluemetal was chosen as ballast because of its scale appearance and its ability to compact tightly around the sleepers.

It was not that long after this section was prepared that the children were hinting about a cubby house. I agreed, on the condition that it took on the appearance of a station building and so the station/cubby house was born. The only addition to the basic plan was to extend the platform past the awning to allow tall people to use the train. At this stage the groundwork in the backyard was ready for tracklaying, but what style of track should be used?

The club track is made up of 20mm x 10mm black mild steel bar using the edge as the running rails. These are welded to sleepers made of 25mm x 6mm black mild steel. For the heavy-duty use at the club track this arrangement was fine, however I wanted something that looked realistic and not necessarily heavy duty. At the time, scale rail was not readily available but there was an alternative. During a visit to a fellow modeller's house I was shown a sample of track and pointwork that he had made. He had used 10mm square black steel as the running rails in 5 inch gauge and the result was quite good in realism and operation. Not long after that visit I saw a photograph of the North London Society track in a book by Martin Evans entitled "Outdoor Model Railways". This track was a happy compromise of scale appearance and strength.

The North London society used 20mm x 10mm steel used on its edge for the running rails with timber used as sleepers. The beauty of this design was the manner of attaching the rail to the sleepers. No welding, no screwing, merely a press fit of the rail into a slot machined into the sleeper. With the slot cut to a depth of 10mm it leaves 10mm of rail projecting above the sleepers; the remainder is nicely concealed below the ballast. Yes, this was the way to go, virtually a snap together kit. Well that is what I thought anyway, preparing everything to that stage was not as simple.

The sleeper factory

Son Jeremy wanted to help cut the 850 sleepers to length.

Because of the large number of sleepers required, I can understand why the North London Society devised a machine to process them. This allows a quantity of sleepers to be machined at the same time. The extra work of fabricating a machine is worth the trouble, especially when all of the sleepers need to have identical gauging. Here is a useful motto: Don't do anything by hand if there is a machine to do it for you! My postscript to that is: if there isn't a machine to do it, make one. The original idea was to use a pair of standard milling machine slitting saws or side and face cutters of 10mm width mounted on an arbour at the correct rail gauge distance apart. Simple enough, except slitting saws and side and face cutters are not available in metric sizes! Don't throw your imperial rules away yet because they are still not metric in 1990! Notch one up for imperial in the measurement war.

The sleeper cutting machine with a tray of sleepers in place.

This is the situation: The nearest imperial size to 10mm is 3/8ths of an inch which is 9.525mm in metric numbers. This results in a force fit of the rail into the sleeper which is excessive and causes the sleeper to split. Some experiments were carried out and there is a pile of split sleepers to prove it. To overcome the problem, two flycutters were made. These are in the form of a steel disc which holds one 1/2 inch square high speed steel tool bit in each disc. The tool bits were ground to allow a cut of 9.8mm width and 15mm depth; in practice they act like a pair of chisels gnawing away at the timber.

The machine was built with a sliding tray to hold ten sleepers at a fixed height under the cutters, the depth of cut is varied with the arbour if necessary. In my situation the slot depth is set at 10mm and the gauge is set at 5.0625 inches (to provide extra clearance at the 21 feet radius curves). The cutting tools are set at 180 degrees apart on rotation to balance the load on the motor (from an old washing machine). The operation of the machine is very easy: Ten sleeper blanks are clamped to the tray, the machine is switched on and the tray is drawn along the slide. The cutters mill both slots to the correct gauge and depth in about thirty seconds. You couldn't do that by hand! In fact the system is so efficient that the sleepers were milled in a couple of weekends, all 850 of them.

The tray of cut sleepers. You can see the cutter on the right.

The sleeper material was chosen after a few trials with various timbers of hard and softwoods. The best results were achieved with the common garden variety tomato stake which is a hardwood, more or less 25mm square and about 1.8 metres long. The price of these varied a great deal according to the supplier or the season. Mine were purchased by the bundle and each sleeper was cut to a length of ten inches for use in the slot milling machine. The cost averaged out to five cents per sleeper. While we are on prices, at the time of construction the whole railway was cheaper per metre than HO gauge flextrack. The method used for insect treating and weatherproofing the sleepers was to let them soak in creosote for a few days, after the machining process.

Sample of cut sleeper. The dark sleeper had been treated with creosote after slotting.

The only work required on the steel rail is in bending the curved sections. To achieve this a special type of curving roller was built. It was designed to roll along the six metre length of 20mm x 10mm black steel bar and set a curve in it. The finished radius is measured by comparing the curved bar to a chalk mark which has been scribed to the radius on a flat surface, e.g. A concrete path or patio. A curve template could also be used. Either way it only needs to be approximate as the final shape will set in the prepared roadbed by measuring from the centreline pegs.

Tracklaying

Having the machinery and tools makes track building and laying a very simple task indeed. Now track laying may begin. At this stage all the centreline pegs are in place with the ground suitably prepared. After trimming the grass and excess soil from the formation, a 20mm layer of 3/16" basalt (blue metal) ballast is spread onto the roadbed to form a base for the track. An advantage of assembling the track on site is that the rail ends can overlap each other by at least one metre. This helps to maintain a smooth curve at the join area, a feature not easily accomplished with the pre-welded "set track" arrangement.

The formation is trimmed and levelled. A 20mm thick layer of ballast is added over the track bed.

I mentioned previously that this form of track work was like a snap together kit, well, this is how it works: Lay the sleepers on the roadbed at three inches apart (a piece of scrap timber cut to three inches length can be used as a gauge). Next lay one length of rail onto the sleepers, beginning at one end press the sleepers onto the rail - checking the spacing with the gauge - continue until the first rail has all the sleepers in place.

At first a G cramp was used to press the sleepers onto the rail, however this proved to be very tiresome so another tool was born. This turned out to be a quick acting camlock type of pressing tool which not only eased the work but it sped up the process as well! With one rail mounted on the sleepers the track is flexible enough to check alignment with the centreline pegs. Once satisfied with the location, the second rail is placed onto the remaining slots. Be sure to offset the second rail end about a metre along from the adjacent rail. Press the sleepers into place, checking alignment as you go. With both rails firmly locked into the sleepers the whole assembly becomes quite rigid and will remain in place. The tracklaying procedure is more manageable when work is carried out on six metre lengths of rail at a time. Install one side then the other, move to the next section and repeat the process until the job is complete. Note: During the assembly it is important to continually check the alignment of the track.

The track laying under way and the sections have been welded together.

The rail ends were joined by welding, the ends being ground back to a 45 degree chamfer using a four inch angle grinder. After welding, the top and flange side of the rail joins were ground back flush with the running surface. A handy hint for joining, leave a few sleepers off each side of the join and clamp a solid bar about 300mm long to the underside of the rail with 150mm on each side of the join. This is to minimise distortion of the top and flange side during the welding operation. Because it was stated earlier that welding was not required, an alternative could be to bolt a 25mm x 5mm black steel plate 100mm long across the join on the outside of the rail. Use two 6mm bolts on each side of the join to restrict vertical movement of the rails. You could drill and tap the rail ends prior to assembly, or if nuts are used, ensure that they do not obstruct the flangeways.

Finishing touches

With the track in place, ballasting is simply a matter of pouring the 3/16" bluemetal over the track and tamping into place. Trim off the excess to sleeper height and shape the shoulders. Superelevation on the curved sections of the railway is just as impertant as the prototype. The correct elevation can be achieved during the assembly of the track or after ballasting. On my railway with sharp curves and slow speeds, an elevation of the outer rail by 1/8th of an inch (3mm) is used. The superelevation is measured by placing a 1/8" thick packing piece on top of the inside running rail. Lay a spirit level across the track from the top of the packing to the top of the outside running rail and adjust the track until the bubble shows level.


Front yard track complete with ballast and superelevation.
The cubby-house station - February 1985. I used a concrete block border in the back yard to assist with mowing the edges.

 

The chassis of the locomotive is sitting in the "service facility" in the back yard. An elevated area for servicing the loco before and after a run. The ramp swings on to the mainline for loco access, and swings away clear during the run.
With 3 tons of top soil and 30 cubic metres of pine bark, the front garden was built up over the soil dug out of the track formation. This reduced the amount of mowing in the front!


Over the passage of time, the garden has grown a bit!
A driver's-eye view of the track heading through the front garden. Native trees and ferns make it a peaceful scene.

Having completed the superelevating and ballasting, the great moment has arrived, the railway is operational. The methods described in this article can be applied equally well to a short straight test track or a complete railway system. As yet there is no pointwork on my railway, however due to the difficulty of machining slots in the timber sleepers for points, they could be constucted by using the traditional 20mm x 10mm steel rail welded to steel sleepers. Another method is screwing 10mm square section steel bar to timber sleepers. This method retains the appearance of the railway proper without the sleeper cutting problems.

There's no point having a bridge without a reason. This night view shows what I call Sandy River.
A night time view of the home-made waterfall feeding Rangeley Lake that overflows into the Sandy River. The Sandy River flows into an underground copper tank and is recycled back to the waterfall by an electric pumping system.

 

When the loco was finished we had a ribbon cutting ceremony in the front yard to officially open the track on evening of 6 October 1989. We had a neighbourhood party, it was a great night!

With the railway operational you will enjoy many happy hours of steaming with your family and friends. If you have any enquiries regarding this form of tracklaying then e-mail me and I will endevour to answer your questions. In the meantime, happy steaming!

 

Return to Home Page

Copyright 1999-2007, all rights reserved
Best viewed 800 x 600 with Javascript enabled