Re: Modelling stub switches

Posted by Jeff Young on Sep 08, 2015; 8:11am
URL: http://c-sng-discussion-forum.254.s1.nabble.com/Modelling-stub-switches-tp2408p3406.html

Some larger spring wire arrived in the post.  (I ordered both 0.032" and 0.039" just to be safe.)

The 0.032" will throw the switch.  The shiny copper under the moving rails at the headblock was bothering me, so I hit it with a bit of Neolube to darken it.  This improved the throw of the switch as well, leading me to believe there wasn't a large margin of error in the 0.032" rod.

Making the lead rails longer will also decrease the torque required by the switch machine.  The prototype used a particular length to get a spiral easement into the switch.  For a #6, the lead rails should be a bit over 13'; for a #8 about 18'.  That works out to about 8 ties and 10 ties (assuming 22" tie spacing).

However, I don't think the spiral easement needs to be exact.  FastTracks, for instance, in their stub switch builder's guide, solders a #6 switch's lead rails a full 16 ties back.  While I think that's perhaps taking things too far away from the prototype, I experimented a bit with removing the last few spikes in my #6 turnout, and the switch machine clearly has an easier time with lead rails 9 or 10 ties long than with lead rails 8 ties long.  So I'm going to start spiking both #6 and #8 turnouts 10 ties back.

Cheers,
Jeff.