|
|
| Re: W train returning when Phase I of the 2nd Av subway opens in December 2016?. (1386432) | |||
 |
|||
| Home > SubChat | |||
|
[ Read Responses | Post a New Response | Return to the Index ] |
|
||
|
Re: W train returning when Phase I of the 2nd Av subway opens in December 2016?. |
|
|
Posted by randyo on Sun Feb 21 18:27:31 2016, in response to Re: W train returning when Phase I of the 2nd Av subway opens in December 2016?., posted by Henry R32 #3730 on Sun Feb 21 15:09:11 2016. The number of merges among the several lines does cause schedule making to be a rather complicated process. The problem with paper railroading vs the real world is that while running times between locations are often in odd amounts such as 2 min and 15 sec, scheduling even computer generated can only be done to the nearest half minute. Thus the 2 min 15 sec runtime would have to be bumped up to 2 min 30 sec in order to avoid the train ending up late based on the written timetable. Also while 2 min clearance between trains is the ideal, trains are often secluded 1 1/2 min apart which is the absolute minimum clearance between trains. Back in the days of manual scheduling, when a train had to clear 2 services as is the case, for example a 2 having to clear both a 5 at 149 St and a 3 at 135 St, if the clearance were tight, we would merely increase the running time for one train so that it would clear both services. With computer generated timetables, such a procedure would be cumbersome so that the solution became adding junction points like 142 St Jct on the IRT or 11 St Cut on the BMT into the internal timetable so that trains could be held at those junction points without interfering with trains at stations where such complex mergers occur. The auction points are usually omitted from the timetable that’s sent to the road so that they become in effect “hidden holds” performing essentially the same function as the variable runtime changes did under the manual system. I also used to be SOP in scheduling that the timetable for the route with the most merges got generated first and the other lines built around it. On the IRT it used to be the 2, but over the years it has now become the 5 since it has to merge with the 2 in the Bx, the 4 in Manhattan and Bkln and with both the 2 and 3 at Nostrand Jct. On the BMT/IND, the R has to deal with the M in Qns and the N and Q in Manhattan but now both the F and the M have merges with 3 other lines so which one gets done first is anyones guess although I’m sure a current SOP is in place. Internally, a conflict where 2 trains arrive at the same merge point at the same time is referred to as a “bang” and we used to generate manual bang sheets with all the possible conflicting merge points and train intervals written on them, Although computer generated bang sheets are available, When I was working there, we still found the hand written ones more useful and a bit less confusing. The problem came in when in moving one train out of the way of a conflict, we would sometimes create a conflict with another line which had to be resolved yet somehow we managed to get it done. |
|
|
|