Construction of a Test Track

[Le Corbusier]

Unfortunately Howard my gauges are the exacto scale gauges for this very reason ... and more curiously the area of track I am describing as having narrowed has been running absolutely fine over the last 4 months and only manifested problems following the recent hot dry spell. Colin's explanation is to date the only one which computes for me.

I was thinking about this in terms of B to B measurements. If there is a degree of shrinkage and expansion within the ply sleeper ... then perhaps this explains why a B to B at the wider limit in some cases improves reliability of running? ... the hypothesis - If the track is set to gauge at its driest humidity then as the humidity rises the gauge will widen. If the B to B is set at the widest limit then it will run ok when humidity is low and will cope with a degree of widening as humidity rises. The same would be equally true for a narrower B to B within the range, if the track were set to gauge with high humidity levels. So in an odd way the B to B perhaps needs to relate to the relative humidity at the time of laying or will require adjusting over time to achieve good running ... Just a thought

Tim

[Martin Wynne]

Interestingly I have found a couple of areas on my track where gauge narrowing has occurred over the last 2 or three weeks that were fine before I have separated the chairs from the sleeper and re bonded to gauge and all is fine. Slightly confused as to why this might have happened with the warmer weather? Any thoughts out there

Hi Tim,

There have been several posts on here and RMweb showing the poor fit of the current rail supplies in the chairs. This means that the chairs are under stress and will gradually deform over time to release the stress. This deformation will be accelerated in hot weather. Since there is a greater mass of plastic on the outside of the rail, the most likely area to deform is the inner chair jaw. The result is that the rail will move inwards, reducing the track gauge.

There is a lot of anecdotal evidence of all types of plastic-chaired track narrowing over time, on plastic or ply sleepers, and flexi-track.

The only way to prevent this would seem to be to have chairs which are an easy loose sliding fit on the rail. I believe the DCC Concepts 00 bullhead flexi complies with this, so much so that there are reports of the rail (stainless steel) falling over in the chairs. Likewise for the new Peco 00 bullhead flexi it is reported possible to push the rail into the chairs from above. The latter uses vertical rail (not canted).

For flat-bottom rail the fixings are the same on both sides of the rail, so any deformation of the fixings is unlikely to affect the track gauge.

Martin.

[Le Corbusier]

An interesting hypothesis Martin .... though it does beg the question as to why the narrowing is so localised?

[Martin Wynne]

An interesting hypothesis Martin .... though it does beg the question as to why the narrowing is so localised?

Hi Tim,

That question applies equally to any other explanation -- shrinkage of plywood, expansion of rail, etc. Perhaps a section of chairs from a different batch?

An interesting finding will be to see if the gauge is restored when the weather returns to cooler wet weather. If the chairs have deformed, it won't be. If the plywood has shrunk, it will be.

Martin.

[Le Corbusier]

An interesting hypothesis Martin .... though it does beg the question as to why the narrowing is so localised?

Hi Tim,

That question applies equally to any other explanation -- shrinkage of plywood, expansion of rail, etc. Perhaps a section of chairs from a different batch?

An interesting finding will be to see if the gauge is restored when the weather returns to cooler wet weather. If the chairs have deformed, it won't be. If the plywood has shrunk, it will be.

Martin.

I've already re-gauged so perhaps will never know.

edit ... not aware that the chairs were from a different batch as they were all mixed up in a box having been removed from the sprues - even if it is possible to tell at all. I provided decent rail gaps using 45 ft scale lengths and making sure the rail slid in the chairs so there is no issue as far as I can tell with rail expansion.

Plywood is a natural material and a laminate ... so i assume there would be a variation in any expansion and contraction .. not so with rail or plastic. Why does one piece of wood twist or split where another doesn't?

[Colin Parks]

Well, the hot weather has finally caused something akin to Tim's problem of gauge-narrowing on my test track!

It has occurred on one the stock rails of the tandem turnout, where the slide chairs are positioned. Having curved the [bullhead] rails to match the Templot template, I had not fixed the slide chairs to the rail web and all has been well until the past few days, when temperatures have reach 29.5C in the village.

Yesterday, I noticed that there was a smaller gap than there ought to have been between one of the reversed switches and its stock rail. The stock rail has expanded in the heat and bowed inwards at both switch positions. Loctite might hold the rail in place, but the stresses need to be relieved and this will be done by cutting through the stock rail in due course. This leaves me wondering if any other rails are being held under stress by the chairs.
In light of what I have observed, Martin Wynne must be correct in saying that the route of least resistance for an expanding rail held in functional plastic chairs would be inwards.

As with Tim's gauge-narrowing, my problem manifested in hot weather and as for being localised, once the stress is released, there would be no further bowing of the rail. Meanwhile, I wait with interest to see if the rail in question returns to normal alignment once the temperature reduces.

There have been no signs of lateral movement in the flat-bottom railed track - so far...

[Le Corbusier]

Hi Colin,

It would still be interesting to know from more experienced track 'experts' if gauge variations over time resulting from changing heat/humidity/environment is a fact of life with P4 and adjustments are necessary until everything settles down? Or if when such as Howard/Tony/Allan/Tim V/ .... etc build, the newly installed sections tend to be fine over time unless a mistake has been made at installation?

Martin's thoughts on functional chair quality would of course not effect ply and rivet ... but then ply and rivet will have expansion stresses that are not there using functional chairs. Both would be effected by sleeper expansion/contraction.

Equally my portion of track may well have been installed inaccurately and the heat only highlighted this, where the built in tolerances within the standards coupled to springing/compensation previously masked the issue. Although the track was clearly under gauge (I had to force the exacto gauge to fit between the rails), not withstanding this only one of my 4 pieces of stock actually de-railed (the clayton carriage) and this not every time - so in theory it should also have been possible to adjust the clayton to solve the problem (I did try this first and only checked the track when this seemed not to help - I still have to add some weight to the carriage).

[Colin Parks]

Hello Tim,

Your experiences with derailment of stock and gauge-narrowing are worth understanding before constructing the track work for your layout. Let us hope that with more experience, we will have no issues with under-gauge track!

I have also been checking my track with an Exactoscale gauge, and there are one or two spots that do need attention, but this is more likely down to poor laying than movement subsequently. The C&L roller gauges that I use grip only the rail head and even then, not tightly, so I doubt that the chairs have been forced apart and one slightly tight spot that I have detected is on a section of flat-bottom track with vertical rails.

There must be some movement in wooden track components, but like Howard says the linear expansion of a plywood sleeper or timber over a length of around 19 mm would be very small. However, when combined with other factors such as the baseboard material, there is certainly plenty of movement lengthways of both track and boards from what I have observed in the past at exhibitions on my old layout and again with this present test track. The gaps between the rail ends at the baseboard joint have gone from 0.5 - 0.7 mm to virtually closed up during the hottest weather at the weekend.

One concern that I had with plywood timbers when constructing pointwork was the number of twisted lengths which had to be discarded (5 or 6 out of 100). Unless quality control at the old C&L was non-existent, I doubt that the plywood was twisted when cut into timbering, so it mus have occurred after cutting to width. This leads me to believe that there must be unrelieved and hidden stresses in at least some of the plywood sleepers and timbers that I did use. How that will play out in the long term - who knows!

Colin

[Colin Parks]

A picture of the track magents which were mentioned in a recent post.

Seen here are three Geomag magnets, approximately 5 mm x 3 mm. They are press-fitted into holes drilled down into the baseboard, with the top of each magnet being flush to the underside of the sleepers, which are 1.6 mm thick.

There are black pen marks on all of the magnets to ensure that the same polarity is facing upwards. An experiment with a stack of three magnets in one position to see how that compares is next on the list of to-dos.

[JFS]

Just perhaps a couple of quick thoughts. Tim:- to answer your question about effects over time... generally speaking things very quickly settled down fine from the start and not got worse. SO FAR... However, I did have a number of chairs come un-stuck at Scale forum (which caused Colin's 4 SUB to derail - a previously unknown thing!) Fixed in seconds luckily. I don't know why they chose that moment to let go, but it was warm in there at that time.

The only thing which has got worse is the tendency of flexi track - used on the cassettes - to progressively tighten to guage. I have established that the stuff was OK to start with and I think this problem is due to me using "super-glue" to stick it down. I was with a mate this weekend and he uses double sided tape for the same job. I think that is both a cheaper and a less risky solution.

Given that your case was not related to the type of guage, and given Colin's observation, I am wondering if the gauge narrowing is actually a symptom of the longitudinal expansion of the rail resulting in a buckle - not enough to "burst" the chairs, but enough to distort them.

Here is an example:-

Expansion effect.jpg (44.44 KiB) Viewed 5863 times

I had this happen to me before I fully fixed all the chairs - it is easy to imagine that if the chairs had been fixed, there would have been a fair amount of stress involved - perhaps easily enough to push the gauge about!

In this case, I had broken my own rule about never having a single length of rail longer than about 420mm, and needless to say, rework was needed. Even if the rail is in short enough lengths, it is easy to allow insufficient expansion gap. I know Tony Wilkins follows a policy of never more than a scale 120ft length of rail for the same reason - he uses rivets of course and has found that the soldered joints tend to fail at the ends of longer lengths. Of course if you use soldered construction, by definition the rail is warm when it goes down. Not so chaired construction.

I wonder if it was a bit chilly when you cut your rails to length?

Just in this recent hot spell, I did have one expansion related issue - a length of conductor rail suddenly let go - it was held down by means of a soldered connection to a brass C&L "pot" and the soldered joint failed - obviously, I should have gapped the conductor rail also. Bear in mind Miories lives in a loft and it does get both hot and cold up there!

The more I think about it, I tend to think that expansion of the rail is a more likely cause than either the chairs or the ply, notwithstanding all our various and valid comments above.

Best wishes,

[Le Corbusier]

A picture of the track magents which were mentioned in a recent post.

Seen here are three Geomag magnets, approximately 5 mm x 3 mm. They are press-fitted into holes drilled down into the baseboard, with the top of each magnet being flush to the underside of the sleepers, which are 1.6 mm thick.

IMG_8927 (2).JPG

There are black pen marks on all of the magnets to ensure that the same polarity is facing upwards. An experiment with a stack of three magnets in one position to see how that compares is next on the list of to-dos.

Colin,

Is it just a misreading ... or do the rail joints in these pictures appear very tight ... particularly given Howard's thoughts

[Le Corbusier]

I wonder if it was a bit chilly when you cut your rails to length?

Just in this recent hot spell, I did have one expansion related issue - a length of conductor rail suddenly let go - it was held down by means of a soldered connection to a brass C&L "pot" and the soldered joint failed - obviously, I should have gapped the conductor rail also. Bear in mind Minories lives in a loft and it does get both hot and cold up there!

The more I think about it, I tend to think that expansion of the rail is a more likely cause than either the chairs or the ply, notwithstanding all our various and valid comments above.

Best wishes,

Thanks for taking the time to respond so fully Howard.

I certainly laid the track over the winter and the room temps were perhaps 12 - 15 degrees C cooler. I will check all the joint gaps but I haven't noticed them to be tight.

The rails were scale 45ft length in line with pre grouping Midland.

I would certainly not rule out a degree of incompetence on my part either.

[Tony Wilkins]

In this case, I had broken my own rule about never having a single length of rail longer than about 420mm, and needless to say, rework was needed. Even if the rail is in short enough lengths, it is easy to allow insufficient expansion gap. I know Tony Wilkins follows a policy of never more than a scale 120ft length of rail for the same reason - he uses rivets of course and has found that the soldered joints tend to fail at the ends of longer lengths. Of course if you use soldered construction, by definition the rail is warm when it goes down. Not so chaired construction.
Best wishes,

Hi Howard.
Largely correct, but the reason for this maximum was down to an experience I had with Heckmondwike, which I have already explained previously elsewhere.
In our naivety, we laid the ply and rivet track with metre lengths of rail and nicked the rail cosmetically where the joints were. One very warm evening, we started our running session, only for the first train out of the fiddleyard to pile up on entering the curve. On investigation, we discovered that almost an entire length of rail had parted company with the track rivets (only the droppers retaining it) and the gauge had spread several mm as a result, the loco falling into the four foot with a clatter. After this we cut all our rail lengths in half, which has been policy since. For myself, I now make my plain track in scale 60ft lengths from the off only having slightly longer rail lengths in some turnouts. None of my track is constructed in situ, so will be at room temperature when laid permanently.
Like Howard, I also have some spiral plywood from the same source. I was given to believe that plywood was stable, apparently not.
I came to the conclusion many years ago with my experiences with Green Street, that during the summer, the wood dried out and shrank whereas the rails expand compounding the effect. During the winter months the opposite happens. The difference in the rail gaps can be considerable, especially with Nickel Silver rail, which has almost twice the coefficient of expansion of steel. Half mm gaps close up producing shorts. Increasing the gaps can produce rather wide gaps in cold conditions. I have had some issues with gauge narrowing, but a soldering iron usually sorts these out.
Regards
Tony.

[Colin Parks]

To answer Tim's question about tight rail joints: the latest picture (with the magnets) was taken close to the end of the headshunt and a platform road, so what you can see are piercing saw cuts, about 200-240 mm from the buffers to the left of picture.

Today, the errant stock rail on the tandem turnout has had its expansion relieved by simply cutting through the functional plastic fish plate at the toe end of the stock rail. The rail then assumed its correct shape again - despite there being no reduction in ambient temperature. This has made me realise that unless a rail joint has to be insulated, such as crossings to closure rails, having functional fishplates, with the gap between the rails effectively filled with a plastic web, are not a good idea at all: what looks like a healthy gap is not anything of the sort. Oh well, it is a test track.

Having thought about this, I have decided to remove all the functional fishplates that I can and just have cosmetic ones where the rail gaps can close up if needs be.

Noting Howard's problem with a conductor rail in the heat, I am minded to stick my conductor rails to selected supports with Kwikfix adhesive which is quite elastic.

[Le Corbusier]

....This has made me realise that unless a rail joint has to be insulated, such as crossings to closure rails, having functional fishplates, with the gap between the rails effectively filled with a plastic web, are not a good idea at all: what looks like a healthy gap is not anything of the sort. Oh well, it is a test track.

hmmm! interesting . - lets say thanks to test track threads as well

[Colin Parks]

Yes Tim, test tracks and their threads have myriad uses and benefits!

On my only other attempt at track building of any note, there were no insulating fishplates, just gaps where needed. The rest of the joints were held with adapted Peco N gauge nickel silver joiners. Given that SMP flexitrack was used then and not pre curved, something robust was needed to hold the rails in alignment at the joints (and of course, whole 36" lengths were used, not knowing any better).

This time, I resisted the urge to save time and did not use the 36" lengths of flat-bottom rail where I could have so easily have done - glad I did not now! The C&L Hi-nickel bullhead rail, at 500 mm in length, is self-limiting to just over two scale 60ft track panels at most, but this fishplate issue has truly caught me out.

[bevis]

Hi Colin,
Another way round the fishplate issue is, where insulation is not needed, to file a small V in the rail ends so the cross piece doesn't get int he way of expansion. I can't remember where I got the idea from but it solved my issue of the rail ends chopping the fishplate in half and the halves falling off.
Bevis

[Philip Hall]

I think the best way to hold rails in alignment these days is to use the Peco bullhead fishplates. They have simulated bolt heads, are the right length, and also hold the rail in the web as well as the foot. For hidden trackage I use their N gauge rail joiners.

Re this expansion issue, of course right now we are experiencing some unusual conditions with the heatwave. I would normally have quite a gap between rails, themselves no more than 18” long. If they contract in the cold and we get a bigger gap, so be it, because that doesn’t seem to have caused a problem on my previous layout. I also chamfered the rail ends quite noticeably, which removed most possibilities of stock climbing rail joints.

The current plan for my main running lines is to have Exactoscale FastTrack for the plain track, which will allow for some movement in the chairs and is also very strong, with ply and rivet for the pointwork. Hidden trackage will be soldered PCB construction, so important to get the gaps right. I also quite like the idea of the track not being rigidly fixed to the baseboard to allow for some movement. At least the baseboards, having been built a year ago, have not moved at all, so I must have got that right at least. Or more correctly, my carpenter did!

Philip




Philip

[JFS]

This has made me realise that unless a rail joint has to be insulated, such as crossings to closure rails, having functional fishplates, with the gap between the rails effectively filled with a plastic web, are not a good idea at all: what looks like a healthy gap is not anything of the sort.

Ah, well, the rail web should be notched to allow the "bridge" piece between the fishplates some freedom to move - but not large enough for the rail ends to be able to touch. Somebody should have told you that ages ago.... oops...

I found out a while ago that if the rail web is not notched, then expansion tends to chop the fishplate in two anyway, so they drop off without you having to do the job!

Still, it does seem a bit drastic to get rid of them all - especially since it will raining again soon enough - especially where you live!

Also just to mention that for non-insulating joints, Exactoscale do a lost-wax cast brass fishplate which is rather beautiful - if not exactly cheap. Again, you need to notch the rail web...

Very Best Wishes,

[Colin Parks]

Re. the helpful responses to my fishplate crisis:

Bevis, your suggestion to notch the rail web is what I sould have done and will do in future on any track building where the joint does not need to be eletrically insulated.

Phillip, though I have seen Peco's bullhead turnouts, I have not seen their new design bullhead rail joiners as yet, but will look into the matter.

Howard, you did suggest notching the railweb, which I did not want to do for fear of expansion issues. However, if as from your experience, the fishplates might effectively be guillotined by the rail webs and drop off anyway, I shall wait for cooler weather when a saw can be inserted in the rail joints and remove those fishplates that do not have to keep the rail ends apart.

Colin

[Philip Hall]

Colin,

The fishplates have been on here:

viewtopic.php?f=5&t=5435&p=54000&hilit=Peco+bullhead#p54000

Philip

[Colin Parks]

Thanks for the link Phillip.

I must have missed Grovenor's post when I was in hospital last year. Our local model shop should be able to order me a packet of fishplates for appraisal. The bolt head detail looks very good on them.

All the best,

Colin

[Philip Hall]

I have been following the discussion on the fit of rails in chairs part of this thread, along with the results of the hot weather, and have begun to check out some of my FastTrack bases. Now most of these go back a long way, to the very first ones moulded in a browny orange colour. I have had some more recently from Phil Reid, although it is possible they too had been manufactured a fair while back. What prompted the experiment was that I have to hand rail from many sources, including some original Studiolith rail. I was recently able to obtain a large quantity of nickel silver rail, which I think might have come from the EMGS about thirty years ago, and I have some very recent EMGS rail, also N/S. I wanted to reassure myself that the bargain purchase rail would fit the track bases. The results have been interesting.

For each of the two rail sections I assembled a half metre length of gauge widened track, (grey bases) which I understood would be 0.2mm wider ie 19.03 mm. I also assembled a non widened length, using old (orange) bases for half the length and newer grey ones for the remainder. This of course should be 18.83 mm. All of these bases have been stored in boxes or drawers away from daylight and apart from a few odd ones are straight and true, as new.

The older rail was a reasonable fit in the orange bases, and quite loose (though not worryingly so) in the newer grey ones. Looking end on, one rail leaned inwards whilst the other was almost vertical, on both kinds. The gauge was 18.9 mm with very little variation throughout the length in both old and new bases, despite the varying lean in the rails. The gauge widened length had the same leaning (or vertical) tendencies, the fit of the rail was a touch loose as for the grey non widened bases and came out at 19.1 mm, with again, very little variation. The rail was able to waggle about a bit in the chairs but gauge was consistent throughout the lengths.

The lengths with the newer EMGS rail, which does look a little squarer in the rail head than the earlier secondhand stuff, was a much more snug fit in the chairs, and here the lean inwards was correct on both rails, and was the same for widened and non widened. The interesting bit was that the gauge was almost exactly the same as with the older rail, where I would have expected it to be a trifle less.

I’m not sure quite what all this is telling me, but probably that I should not worry too much about rail variations and just to use up what I have. I shall also not worry about the rail leaning inwards or not; after all, the pointwork will probably be ply and rivet, packed up to match, and to be honest, you can’t tell whether the rail is leaning inwards or not from virtually any angle. I will have a try with some of the Studiolith rail next but I suspect any variations in gauge will be insignificant. I think that once the track is painted and secured to the baseboard the rail will be able to move a bit with expansion but probably not too much. I should say that the temperature in the railway room is not too bad at the moment; the building is probably better insulated than the house and temperature variations quite reasonable. It’s tempting to make up a bed in there at the moment!

Philip

[Terry Bendall]

What prompted the experiment was that I have to hand rail from many sources, including some original Studiolith rail.

It might be interesting Philip to measure the width and height of all these different rail sections just to see what the variations in size are and even compare the results to what the prototype should be.

I’m not sure quite what all this is telling me, but probably that I should not worry too much about rail variations and just to use up what I have

It seems to me this is very reasonable and pragmatic way of approaching the matter. If it works, don't try and fix it! A gauge of 18.9 does not matter at all. 19.1mm might be a step too far but again if it works and stock stays on, why change it? This may be a heretical statement but I wonder if at times we get too hung up getting things exactly to tolerance? If stock falls off and the track is not within the tolerances then that may well be the reason but other factors may be the cause.

Terry Bendall

[Philip Hall]

I have measured the head width of both kinds of rail and they are to all intents and purposes identical, which surprised me. They look a little different, with the newer rail having a sharper squarer head, but the difference is very slight. The older rail looks a trifle finer, but that might be the ever so slightly more rounded edges to the head. Height is the same.

We have had this discussion about rails before on here, and as dies wear we do get these variations. I shall have to dig out the old Studiolith rail and have a look at that sometime. It was a trifle worrying to find that, with the older rail, one rail leaned inwards and the other was vertical, but it didn’t seem to affect the gauge that much! I think there is little point in trying to find some wider rails to reduce the gauge widening a bit. I could be doing that for ages instead of laying track.

I always used to check curved track by putting a rectangular Society gauge in it and making sure it could waggle about a bit, and this gauge widened FastTrack passes that test. I have a wagon fitted with some old Jackson aluminium wheels which I turned down, and I got a bit carried away with reducing the width and ended up with wheels 1.6 mm wide, and that doesn’t fall off. As you say, Terry, it doesn’t seem to be as critical as we might think.

Philip