I believe you were misinformed. The rails have not a real freedom to slide, they have two different forms of resistance, the one provided by the ties and the o

I believe you were misinformed. The rails have not a real freedom to slide, they have two different forms of resistance, the one provided by the ties and the one provided by the junction, there is only a given temperature range where they can slide. The rail goes into a cycle, you need some 3-4 C° to win the resistance of the joint (in this part the rail is compressed by some 6000 kg of resistance) and some 6-7 degrees more to win the resistance of the ties (some 10-12000 kg more of compression), then it becomes "free sliding" until the gap in the joint closes, this needs 12-13 C° more and the rail is again compressed by the ends. When the temperatures lowers, the same cycle is reversed, the rail first contracts until the gap starts to appear and widens, then you have again a range where it is free sliding, then the two resistances come into play again, when the temperature lowers further the excursion is limited by the joints (usually they allow 14 mm travel) and the rail is subject to traction (as opposed to compression). The only point that never moves is the tie at the center of the rail length, you have to imagine a spring that you try to expand by pulling the two sides, it will return to its original shape, all forces are relative to the center, so there is not any reason why the rail should not come back to its original position. Check this: https://pwayblog.com/2017/03/06/jointed-track-response-to-rail-temperature-variations/ https://pwayblog.com/2017/03/06/jointed-track-response-to-ra... In practice - strangely enough - rails behave very similar to the theory.