In article , Jon Crowcroft
writes
I thouht the points were a big limiting factor between
Hitchin and Cambridge and at Hitchin too...but I am not
a Jarvis employee of course:-)

For a non-stop class 365 EMU, the bottlenecks are as follows:

King's Cross approaches start at 15mph and ramp up to 100mph by, IIRC,
Finsbury Park (my sources for this are a bit out of date).
Limit is then never below 100mph to Hitchin.

Hitchin junction is 40mph for 34c (just under 700m) northbound, 26c
(just over 500m) southbound. Southbound trains are also limited to 75mph
through Hitchin and for about 600m south, then IIRC 70mph over the
connection to the fast line.

Limit is then 80mph to Royston (but this may be out of date).
North of Royston my sources predate the major works of a few years ago,
but show limits varying between 50 and 90mph; the major restrictions are
at the north end of Royston and at Shepreth.

Shepreth Branch Junction (on to the Liverpool Street line) is 30mph.
Then 90mph to Long Road, then 80mph northbound, 90mph southbound, to
Hills Road, then 35mph into the station.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

So just for fun I did this trip on the 10.45 cruiser kings Cross-Cambridge this
morning -

By chance, it was held up behind a slow train for 5 minutes, and yet arrived on
time, this _despite_ a stop outside kings cross (bottleneck: simply
because of 4 tracks from 11 platforms to get out) and coming into cambridge
(bottleneck: due to only 1 stupid platform for 3 trains requiring extreme caution
in correct use of semaphors for 2 sets of points), AND a stop at hitchin (as
northbound trains to cambridge cross the fast line, there's always a good chance
that there's some southbound GNER in the way...)

On the topic of ENgland's longest platform - someone tried to sing its praises
as oh so friendly to disabled people (no bridge or underpass or lift to
negotiate to get to the other platform) - well as a person still on crutches
after my bike accident, I can tell you that if you have to navigate from
platform 4 to 1 on the uncertain surface (and lets say that for some unusual
reason it is slightly damp too), it is not at all friendly to disabled people
at all.

Anyhow:

I would say that if the slow train (which the driver told us was not according
to schedule) had not been there, the cruiser could have been in cambridge in 40
minutes, (I checked the times on each section, but cannot speak for the speed of
the train). If the other bottlenecks were not there, the journey would have taken
approx 38 minutes. Admittedly, the 2 or 3 other bottlenecks wouldbe expensive to
remove and make safe alternatives.

I have no idea about the theory, but this was an empirical result

(actually I've been on the cruiser one or two times when its taken 42 minutes
platform to platform outside of busy times....)

oh, I dont suppose now is the time to complain about the lack of safe places to
put bikes (someones bike fell over as we went around the curve after hitchin - i
--
Jon Crowcroft

In message , at 13:30:34 on Tue, 4
Jan 2005, Jon Crowcroft remarked:

On the topic of ENgland's longest platform

That's at Gloucester, Manchester or Colchester, depending on who you
ask...

--
Roland Perry

In article , Jon Crowcroft
writes
By chance, it was held up behind a slow train for 5 minutes, and yet arrived on
time, this _despite_ a stop outside kings cross (bottleneck: simply
because of 4 tracks from 11 platforms to get out)

4 from 11 is about the right ratio (design standards suggest at least 2
platforms for each line entering a terminus).

Trains should be timed not to intersect; the Rules of the Plan specify
minimum margins for crossing movements in the timetable.

It's unusual to be stopped on departure; most trains depart on B or D
roads (which become the Down Fast and Slow respectively after Gasworks
Tunnel).

and coming into cambridge
(bottleneck: due to only 1 stupid platform for 3 trains requiring
extreme caution
in correct use of semaphors for 2 sets of points),

For some reason I first read that as "metaphors".

I'm not sure of your point here. If there were 3 trains in the platform,
you'd have been coupling up to one of the others, hence the dead slow
running. If not, do you mean that you had to wait for one to depart
before you could enter?

They're not semaphores in either the programming or railway senses.

Each set of points can be in one of three states: needed normal (one of
the positions), needed reverse (the other), or free (not currently
needed by anything). The signaller at Cambridge selects a route (by
pressing two buttons). The signalling logic has a list of points and
positions that that route requires, plus a list of routes that are
incompatible even though they don't conflict in their points
requirements (usually routes in the opposite direction over the same
track). If any of the points are already needed in the other position,
the route is left unset. Otherwise all those points are set to "needed"
(causing them to move if they're currently wrong). One the train passes
over each set they move back to "free" (unless another route is also
holding them).

[Clearing the signal requires other tests to be met as well; for
example, all the points must be detected as locked in the correct
position.]

At Cambridge this is done by a network of relays; current to energise
one relay passes through contacts on all those defining the appropriate
conditions.

AND a stop at hitchin (as
northbound trains to cambridge cross the fast line, there's always a
good chance
that there's some southbound GNER in the way...)

When trains are running to time, there won't be. Either the GNER was
late, you were early, or you arrived during the "pathing allowance" -
basically a scheduled delay to let another train through.

I would say that if the slow train (which the driver told us was not according
to schedule) had not been there, the cruiser could have been in cambridge in 40
minutes,

It's allowed 43 for the journey, of which 30 seconds at each end - IIRC
- is for station duties. So 42.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

In message , Clive D. W. Feather
writes
Otherwise all those points are set to "needed" (causing them to move if
they're currently wrong). One the train passes over each set they move
back to "free" (unless another route is also holding them)
What's the chances that the points freed at Poters Bar, whist the train
was going over them, allowing the first part of the train in the right
direct and the second part by the point being able to move?
--
Clive.

In article , Clive Coleman
writes
Otherwise all those points are set to "needed" (causing them to move
if they're currently wrong). One the train passes over each set they
move back to "free" (unless another route is also holding them)
What's the chances that the points freed at Poters Bar, whist the train
was going over them, allowing the first part of the train in the right
direct and the second part by the point being able to move?

None at all, the points "fell apart" because there were strategic bolts
missing.
--
"now, the thing you type on and the window you stare out of are the same thing"

In article , Clive Coleman
wrote:
In message , Clive D. W. Feather
writes
Otherwise all those points are set to "needed" (causing them to move if
they're currently wrong). One the train passes over each set they move
back to "free" (unless another route is also holding them)
What's the chances that the points freed at Poters Bar, whist the train
was going over them, allowing the first part of the train in the right
direct and the second part by the point being able to move?

--

In article , Clive Coleman
wrote:
In message , Clive D. W. Feather
writes
Otherwise all those points are set to "needed" (causing them to move if
they're currently wrong). One the train passes over each set they move
back to "free" (unless another route is also holding them)
What's the chances that the points freed at Poters Bar, whist the train
was going over them, allowing the first part of the train in the right
direct and the second part by the point being able to move?

Where a branch line diverges from a main line, as at Hitchin, when idle
and no train is immediately expectd. the points normally return ("default" it
would be called in computer terms) the straight ahead main-line setting.

Is this a question about the Potters Bar crash? I don't think it happend like
this.

Michael Bell
--

In article , Michael Bell
writes
Where a branch line diverges from a main line, as at Hitchin, when idle
and no train is immediately expectd. the points normally return ("default" it
would be called in computer terms) the straight ahead main-line setting.

That's called "self-restoring" and it is *NOT* the normal arrangement.
Indeed, if you watch at Hitchin you can see it not happen.

There are a few places where points are made self-restoring. There are
rather more cases where points may move for a not-immediately-obvious
reason. For example, at crossovers - the links between adjacent tracks -
it is normal for both ends to move at the same time. So the points at
the end of the northbound platform at Hitchin have to be set for
Peterborough before a northbound fast train can go through, even though
it doesn't cross them. There are more complex layouts where the
relationship is less obvious, but it's there nevertheless.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is:

In article , Clive Coleman
writes
In message , Clive D. W. Feather
writes
Otherwise all those points are set to "needed" (causing them to move
if they're currently wrong). One the train passes over each set they
move back to "free" (unless another route is also holding them)
What's the chances that the points freed at Poters Bar, whist the train
was going over them, allowing the first part of the train in the right
direct and the second part by the point being able to move?

Zero.

As well as various controls in the signalling logic (note that I said
"once the train passed, modulo misspelling), there is a separate "direct
track locking". If the track circuit covering the points is not clear,
the points are not allowed to move. The relay concerned is "slow to
rise" to allow for a momentary failure to detect the train.

This was allowed for in the investigation. There was no train movement
in the near future requiring those points to be swung, so there would
have been no reason for them to try to move.

In any case, the points were found to be correctly set and locked. The
accident happened because one of the blades broke free from the locking
mechanism underneath the train. The processes involved - and the errors
in assembly - are well understood by now; the question is *why* the
points were wrongly assembled.

--
Clive D.W. Feather | Home:
Tel: +44 20 8495 6138 (work) | Web: http://www.davros.org
Fax: +44 870 051 9937 | Work:
Please reply to the Reply-To address, which is: