On the real railway, they tend to lay the long-welded or continuously welded track in relatively warm weather, so that the usual problem is one of contraction rather than expansion; this puts the rails under considerable tension, but steel has a very high tensile strength, which it is why it is used in pre-stressed concrete structures. The heavy concrete sleepers (“Ties” in North America) and high ballast shoulders keep the track in place (like the concrete in the pre-stressed beams just mentioned). Where significant temperature variation is known to be a problem, we can ensure track is laid at the normal summer temperature and anchor it well with glue & ballast or pins/screws, and solder any joints (e.g. with fishplates). Unfortunately, if you have already laid you track in relatively cold weather, the best you can probably do if you’re experiencing problems is to cut the track into lengths representing the normal fishplated track, so that the effective length of the whole thing is reduced and the changes in length due to expansion & contraction are divided up amongst a large number of small gaps instead of just a couple that open out in cold weather to an extent that interferes with good running. Naturally, you would have to do something to ensure continuity of electrical supply along the tracks by soldering a feed to each rail or by soldering a connecting wire across each gap – one that will flex with any movement in the track. Air conditioning may now begin to sound the more practical solution!

If it is humidity there’s probably not much you can do with a finished layout. I build a lot of pine cabinets and have found that by sealing each piece before assembly I eliminate the problem. I’d just fix the track and put in some rail joiners hopefully the problem won’t come back.

Thanks for the thought. Will look at a way to fix it. If you know of a way I would gladly like to know. Thanks.

Hope this helps you
Don

A single loop of steel rail around a 6×4 ft plan might be around 15 ft long. 15 percent of 13/16 inch is about 1/8 inch. You might expect that much expansion if you get a 100 degree change in temperature (not too difficult to imagine). I would guess brass or nickle-silver would have a larger coefficient of expansion and thus more expansion. I would not expect your rail to straighten itself once it is bent.

We resolve the problem by anchoring pipes at intervals (every 200 ft on long runs) and putting expansion loops in the middle of each set of anchors. We may guide the pipe in the direction of expected expansion but do not otherwise restrict its movement. We consider both changes in ambient temperature and the change in the temperature of the commodity inside the pipe from a cold state to operating temperature. Steam piping temps may be 600 degrees F or higher and these pipes really do grow. Failure to resolve the issue adequately can, and has, resulted in dramatic failures of structures and equipment.

For an imperfect solution in an imperfect situation you might want to consider anchoring your tracks in the middle of the long runs and allowing them to expand into the curves. A dislocation of 1/64 inch in a curve might be manageable. Best to get the expansion rate of your particular track and do the math first. Good luck.

I have a figure of 19 ppm for nickel -silver, so if you built the layout in your garage in winter at zero degrees C, and summer temperature got to 40 degrees C, a two metre length of track would expand 1.6 mm, enough to cause problems if the rails were tightly end-to-end.

So as Robert says, leave a small gap between the ends of the rails when connecting track sections so it can slide a little in the fish-plate..

But as a really neat bonus, this will also help to provide a more realistic “clickety-clack” sound effect if the rail gaps are prototypically spaced.

Just one more thought:
Are you sure the problem is not also due to the wooden base drying out and shrinking ?

Correction, my layout is 4.5m x 1.7m. Hit the wrong key.