Le Corbusier wrote:Thanks Tony for the explanation.
Strange and confusing in equal measure. I would have thought that a pivoting rivet because of the movement would have been under less stress than the last of the soldered rivets and therefore that this would have been even more likely to fail ... followed by the next etc ... but this self evidently is not the case.
I have certainly known this sequential failure mode to occur where the switch rail is short and relatively stiff. For some years, I made the mistake of only allowing the switch blade to move only where there were slide chairs. All the turnouts on Green Street were made that way and suffered from this. One reason why I prefer to make my switch and closure rails all one piece. The explanation of Block chairs in the P way handbook is not that clear and it took me some years to realise that there were in fact two types, some that allow movement and some that don't. Interestingly this failure mode only seemed to affect the first two soldered joints where the moving block chairs should have been.
It was only when I was asked to build some EM turnouts that I had to increase the moving length and add extra slide chairs to allow for the wider 1mm opening that I began to dig a little deeper and realised the error of my ways.
I suspect that the pivoting rivet failure mode is down to the twisting moment repeatedly applied to the solder joint. Typical electrical solders are not that strong.
Curiouser and curiouser said Alice!
It does beg the question ... is there any point in trying to incorporate rotation at the heel at all, even if you do have separate switch and closure rails?
I would be interested in your thoughts.
There are engineering solutions for most problems. It is an issue we will have to solve with the current NAG layout as we have bravely / foolishly decided to incorporate a turnout with pivoting switchblades in one of the sidings. Several approaches have been tried, but the current favourite seems to be to extend the foot of the rail beyond the main body of the blade and bend it down 90 degrees to act as a pivot. i.e. file away the head and web for about 10mm to leave just the foot remaining.