|
|
| Re: TA is obsessed with CBTC, and ''New'' tech for no reason. (579848) | |||
 |
|||
| Home > SubChat | |||
|
[ Read Responses | Post a New Response | Return to the Index ] |
|
||
|
Re: TA is obsessed with CBTC, and ''New'' tech for no reason. |
|
|
Posted by Stephen Bauman on Mon Mar 3 09:38:24 2008, in response to Re: TA is obsessed with CBTC, and ''New'' tech for no reason., posted by Jeff H. on Mon Mar 3 03:58:47 2008. Your thought experiment is too long-winded for me...I would have to draw it out to follow what you are saying.Your last paragraph indicates that you accept my conclusion so you may be spared that agony. However, I think there is benefit in examining it thoroughly. I have some documents that the BOT used to lay out the IND signal system. They start with a required service level... What an arcane design approach! If you assume a block length of 600', there is a red at the station entrance, a red 600' back, and a yellow 1200' back. The follower would need to be 1800' to see all greens. I'm assuming a block length less than 600 feet. First off, the average distance between signals on the system is 350 feet (from one of the TA's presentation slides before the start of CBTC). Secondly, empirical inspections indicate that signals in and around stations are generally 300 feet apart or less. Thus, I assumed that if the follower were 1200 feet back, it would see all greens. However, if you want 1800 foot separation, then that translates to an additional 13.3 seconds for the follower to come to a complete stop in the station. That would bring the minimum headway to 99.1 seconds for 36.3 tph. One reason I went into detail was to present an analytical framework for determining headways for various signaling strategies. One merely has to place the follower at the appropriate distance from the leader and calculate the time for the follower to come to a complete stop. As for keying-by, of course that was not intended to be required to meet service under ordinary conditions. However, when there was a variation in dwell time causing bunchng (often beyond the crew's control, e.g. door holding) then keying-by allowed better recovery from the disturbance. Eureka! There are variations! Their cause is immaterial. How to deal with them? There are two approaches: reduce them to insignificance or introduce a mechanisms to live with them. I'd assume those BOT design documents date from the 1920's. Reducing the variations to insignificance requires time standard and clocks throughout the system that are synchronized to that standard time. That technology wasn't available back then. The technology, in the form of keying by and station timers, were available. Their introduction could reduce the effect of the variations, which could not be reduced with the technology then available. Technology changes. Time synchronization is child's play, unless you are trying to set up, operate or use a Windows based time server. |
|
|
|