A few weeks ago, after connecting the newly laid track at Kankool to the DCC bus, I ran a test loco through the turnouts and was happy with the result.
Here is a short video of the movement.
I have made more progress on the Kankool turnouts. I have now fitted tiebars.
I have always wanted to have tiebars that look as realistic as possible, within the bounds of practicality for HO scale and reliable operation.
Some time ago, Rene kindly sent me one of his 3D printed tiebars. I had planned to look at his design and see if it could be adapted to my new track standards.
I put the whole thing on the back burner for a while, but resurrected it nearly 12 months ago when it was getting to the point that I was going to need something to be able to complete the turnouts.
I put the idea to a mate who is a bit of an expert on 3D design. He thought the idea had merit, so I left it with him to come up with a design.
Over the last few months however, I began to have reservations about the 3D print idea. Because the plastic tiebars would have to be glued to the points, I was worried about the mechanical strength of such a bond. Also, as the 3D printing thing is still relatively new, the materials are untested over time. Will they deteriorate over the next twenty years?
When Andrew and I built the P87 turnouts for Bowen Creek, we used thin pieces of printed circuit board that would allow for a soldered joint to the points. So began the manufacture of a prototype.
My design uses brass strip soldered together in a similar design to Rene’s. It still allows the tiebar to slide underneath the second sleeper back from the toe. It also has a drive tube underneath into which the wire-in-tube turnout activator will fit.
I had to have a method of holding the tiebar underneath the turnout whilst soldering to the points. So, using a piece of old melamine shelving, I marked out and routed a recess for the tiebar to sit in.
The turnout was then located over the tiebar and clamped in position.
Soldering the points to the tiebars requires a great deal of care. It is very easy to solder the points to the stockrails! To reduce the risk of this, and also to reinforce the soldered joint, small pieces of brass strip are soldered to the tiebars then to the points. The point is clamped against the stockrail then the brass strip carefully soldered in place.
Once the first point is soldered in place, the turnout is repositioned so the other point can be fixed in position. Again, the point is clamped to the stockrail, and ensuring the opposite point is the correct distance from the stockrail to allow for wheel clearance, the second point is soldered in place.
The finished tiebar! It is then tested for binding and to ensure both points sit snugly against each stockrail. The turnout can then be removed from the melamine base. The plan is to temporarily clamp the turnout in position on the roadbed and connect the actuator and test the operation using the wire-in-tube. Once it is proven to work effectively, the turnout will be painted and prepared for final installation on the roadbed.
A similar recess for the tiebar was routed out in the cork roadbed at each location.
Well, after a lot of procrastinating over the last few months, I have finally started laying down the first sections of Central Valley Model Works (CVMW) tie (sleeper) bases in the Temple Court section.
First, a bit of background as to why I chose the CV tie strips over traditional flex track.
After I had finalised the track plan way back in 2009/10, I pretty much knew how much track I would need for the scenicked sections. There is a total mainline length of 458 feet (139 metres) between staging yard exits, including the helix. The helix and short section from the bottom of the helix to the start of the Temple Court scene is 109ft (33m). This leaves 349ft (106m) of scenicked run. Add to this the trackage in the loops at Kankool, Ardglen and Pangela of 51ft (15m), this takes the total scenicked trackage to 400ft (122m).
Traditional flex track is easy to install but is expensive. For example, if I were to use Micro Engineering Code 70 flex track, I would need around 130 lengths (3ft), which at $7.25 (from The Railcar) would cost me $942.
For the equivalent three foot length of Central Valley ties plus code 70 rail, at the time (2010), cost me around $3.00. The CV ties came in a bulk pack of 50ft for around $45 and code 70 rail was around $40 for 33 lengths (3ft lengths). So you can see the savings I’ve made. The CV ties and rail were purchased from Proto87 Stores.
The other downside of flex track I have found, especially Micro Engineering, is it is very hard to lay out in curves, as the rail is quite tight in the moulded ‘spikes’. The sleepers also are a bit on the thin side.
So, the CV track I eventually went with was their #2002 ‘Branchline’ type. These had 8 foot long sleepers, which according to the Trackwork Manual from Greg Edwards, is what NSWR plain track was laid on. The sleepers also came out very close to correct width (9”) and height (4.5”). Even though CV label this track as ‘Branchline’, the sleeper spacing came out closer than their ‘mainline’ track, which looked a little too close for me.
Anyway, I decided to give the stuff a go.
So, the idea of the CV ties is that they are glued down to the roadbed and the rail is then glued to the ties. The CV team have always recommended a product called Barge Cement, which is essentially a flexible contact cement.
Over about the last 12 months or so, I have been trying to source this Barge Cement here in Australia but to no avail. It is available overseas, but I was always wary about getting it shipped here as it is a flammable product.
I eventually came across an equivalent product called Bostik Contact Bond. This also proved difficult to obtain, in Port Macquarie at least, but I eventually found some at the local Officeworks store.
Now, I have done a lot of research on this gluing of rail to plastic, and there seems to be division among modellers. Some users swear by it, others not so much Some say that both the base of the rail and the ties must be coated in the glue, others say only the base of the rail is required to be coated.
The idea behind the gluing method with the Barge Cement, is that it can be activated/reactivated with MEK. This means that once the rail is in position on the sleepers, a small amount of MEK is applied with a brush which reactivates the glue and makes the bond. This essentially bonds the glue on the rail with the plastic sleepers.
A thinned down mixture of the glue and MEK has to be made. I followed an article I found on Model Railroad Hobbyist. The Bostik product remains mixed over time and does not separate from the MEK.
The photo above shows (L to R) the Bostik glue, the glue/MEK mixture and the MEK. The MEK is Plumbers Priming Fluid which can be obtained from hardware stores or plumbing supplies.
I made a 2 to 1 mixture (MEK to glue) as a test first, and tried it on a short section of the CV track. The thinned down mixture is applied using a cotton bud. A good, thin and even coating is applied to the underside of the rail, then the same done on the tieplates on the sleepers. Initially I only coated the base of the rail, but this did not produce a strong bond. The rail is then held in place and a small amount of MEK is applied to the joint. Slight pressure is required for a few seconds to make sure the bond has occurred.
This first test seemed to go OK, but was a bit hard to judge how it would go with long lengths of rail. I may end up using some Proto87 Stores scale spikes in strategic places just to be safe.
Oh, the rail must first be cleaned with methylated spirits prior to applying the glue.
Before laying of the CV track could start, some preparation work had to be done at the exit of the staging yards.
The thickness of the CV ties are slightly thicker than the PCB strips I used for the storage yard track, so this necessitated a slight recess to be routed out at the transition point. I simple jig was setup up for the router and about 0.7mm was removed from the top.
Next job was to mark a centreline on the foam so I had a guide where to lay the track. As the foam is black, I needed something that would leave a line that was visible. I bought a white paint marker pen from a local art supplies shop and fashioned a timber jig to hold it vertically and whilst held against the spline, I could move it along and leave a line.
The grade from the Down staging yard through Temple Court and to the helix is steeper at 1.8% (1 in 55) than the rest of the layout at 1.19% (1 in 84). This meant I had to try and ease the transition from flat to grade at the exit from the staging as smooth as possible.
I carried out some work at the location to allow for a vertical easement. See the pictures below. I hope they are self explanatory as it’s difficult to put into words what I have done.
Once this was completed, the spline was fixed to the staging yard benchwork.
As the CV track sections are approximately one foot long, I decided to glue three sections together into three foot lengths to make laying a bit easier. At each end of the track sections, there is a half sleeper that when glued together at opposing ends of the track sections, a full sleeper is made thus making long smooth runs of track base. Have another read here for a full explanation.
Through most of the research I have done on the use of CV track, many users recommend gluing the ties to the roadbed using a flexible sealant eg. caulk or No More Gaps. This allows for some working time after laying the track in place. I chose to use a coloured acrylic caulk as I wanted to be able to see it when I applied it to the black foam roadbed. I applied a thin bead of caulk either side of the track centreline which was then smoothed out to a thin, even layer using a metal spatula.
The track sections were then placed in position on the centreline and checked for alignment. This was easily done by getting down low to the track and looking along it, using the raised tieplates as a guide, which were quite visible in checking for rail alignment.
When I was satisfied everything was OK, some weights were placed on top until the caulk cured sufficiently, which was generally a minimum of 12 hours. There is sand in the containers.
I did experiment with hand painting the track with water based paints prior to laying, but this proved time consuming, so the plan is once the track is down, to use the airbrush to paint the sleepers a base colour, maybe a grey, then mask off the sleepers and spray the tieplates a rust colour, then paint the rail the same colour, then glue the rail in position. Then, prior to ballasting, some extra painting will be done, maybe some thin washes to add some randomness to the sleeper colours. This may all prove a waste of time, as once ballast is applied, it will probably cover the sleepers anyway. Maybe just the base grey colour will suffice. We’ll see how all that goes!!
That’s where things are up to at the moment. I have got track sections down to a bit over halfway to where the Peco track currently ends.
Once this is all down, then I will start painting.
Things have been pretty quiet since the last post, but there are always things happening behind the scenes and sometimes aren’t worth posting about.
I have been itching to get back into trackwork and have started building the first set of turnouts at the Werris Creek end of Kankool.
The Kankool (1941) signal diagram is shown below. Click to open a larger version.
The first set of turnouts to be started are numbers 7 and 9 as shown above. No.7 is the Up Main to Loop and No.9 is a catchpoint in the Main facing Down trains. As can be seen above, there is a short runaway siding that extends from the loop through another turnout that is also designated No.7. These two are worked from the same lever. The runaway acts as a catchpoint for the loop facing Down trains as it is thrown for the runaway when lever 7 is normal in the frame (ie set for the Main).
In looking at photos I have that show snippets of this arrangement of turnouts at Kankool and a similar arrangement at Ardglen, I was intrigued to find that No.9 catchpoint is not a normal single blade type, but virtually a full turnout without the vee crossing.
The photo above shows a similar catchpoint at Ardglen. Note the short run-off rail and the timber block where the inner rail ends at the stockrail. The run-off rail extends under the point rodding. There also appears to be some sort of ‘checkrail’ as well.
I decided to build the arrangement on the workbench, so out came the Greg Edwards Trackwork Handbook, and some photocopies of the 1:6 and 1:8 plain turnouts and 1:6 symmetrical turnout (wye) were made. I had previously used some templates to get a rough idea on how the two would marry together.
My main workbench was found to be not perfectly flat, so I dragged out an old table I had and checked it – perfect! I knew I had a sheet of glass somewhere that I could use to stick the paper templates to. I started to lay out the templates, but soon found they had not copied accurately enough. For some reason, and I’m guessing it was the photocopier, there was a slight kink in both templates. Now, this error wasn’t helping me line up centrelines etc.
I decided to send an email to Greg Edwards explaining my problem and if he would be prepared to send me his CAD files of the turnouts. Well Greg replied very promptly with attachments of the turnout drawings I required. Greg’s only proviso in giving me the drawings was that I did not distribute them and that they were for my personal use only. Thanks Greg!
I then proceeded to manipulate the drawings in TurboCAD to combine a 1:8 plain, 1:6 symmetrical and 1:6 plain together to produce a new template. See below.
The above result was a much more accurate template I could use. As mentioned above, No.9 catchpoint arrangement can now be seen. The road off the runaway at the top of the drawing will be extended more when in situ.
The template was printed out over three sheets of A4 size paper, cut and joined, then taped down to the glass.
Once the template was in place, PCB sleepers had to be glued down. I started marking out where to place PCB in strategic locations but then changed tack to make every sleeper PCB. This may seem like overkill, but it makes it easier than trying to position timber ones in place later. Anyway, I thought I’d give it a go.
Clover House PCB strips were used. Turnouts timbers are generally 10” x 6”, whilst general track sleepers are 9” x 4.5”. Clover House #1266 scale out to approx. 10” x 5” and #1267 to approx. 11” x 5”. They are a tad wider than they should be, and it’s difficult to see the difference, but the turnout timbers need to look ‘beefy’ compared to general sleepers.
Once all the PCB’s were glued down, I started to think about laying the first piece of rail. During these thoughts, I decided to have a go at laying the rails on etched tieplates for that extra bit of detail. The tieplates were something I got etched as a detail item under the IR Models brand many years ago. I’m not sure even if I ever ended up officially having them for sale.
As can be seen from the image above, the tieplates came with a convenient centreline to assist in placement on the sleepers. Firstly. the backs of the tieplates were pre-tinned with solder. I then started by placing the etch in position over the drawn railhead on the template and with minimal solder, fixed them to the PCB. This was repeated for the length of the straight stockrail. In places where the curved stockrail converges on the straight stockrail, ‘half’ tieplates were used. See below.
The straight stockrail was then soldered in place to the tieplates, again with minimal solder from the ‘rear’ of the rail (ie, the non viewing side). Once this was complete, a 1:8 vee crossing was assembled and soldered in position, gauged from the stockrail.
The close-up shot of the crossing shows the support ‘plates’ the crossing is attached to. These were made from strips of 0.1mm thick brass. On the prototype, these ‘plates’ perform the same job as the tieplates under the rail. The vee crossing was spiked in place to these. On the model, they served the purpose of lifting the crossing up by 0.1mm which is the thickness of the tieplates.
The rest of the turnout complex was completed using the same techniques I used when building the storage yard turnouts. Prior to fixing the tieplates in place for a second rail, individual plates were removed from the fret and placed at strategic locations, and with rail held temporarily in place with gauges, these initial plates were soldered in position. The rail was then removed and the alignment of the plates checked against the template. They were generally pretty spot on, so the remainder of the plates for a particular section were then fixed in place using the template as the guide, and the process of fixing the rail to these was again repeated using gauges.
The following images show some shots taken during the construction process. Another vee crossing was also assembled, this time a 1:6 for the runaway turnout.
The final shot above shows the arrangement at the stage where all rails are in place. I’m pretty happy with how it has all progressed. Whether I continue with installing tieplates on future turnouts remains to be seen, as looking at a lot of photos, a lot of this detail is covered by ballast, and it is very time consuming. I might wait until the Kankool turnouts and track are painted, weathered and ballasted before I make that decision. After all, it will be a while before I need to worry about the Ardglen turnouts.
The next job is to fit the rail brace chairs, point blades and checkrails. I also plan to fit the tiebars and associated apparatus, ready for the ‘wire-in-tube’ connection from the lever frame, all whilst it is on the workbench.