Of course I recall that the Paris carriages have three doors per side. I
spent a morning observing them from Châtelet les Halles platform, something
which Tony hasn't done (so don't make fake claims about struggling). They
are also longer carriages (27 m instead of 20 m). I have written all of
the details before. Unlike Tony, I have also had a session with the
designer of the Sydney carriages. They were intended to have a
double-leaf door plus a single-leaf door in each mezzanine. The operators
of the day deemed that unnecessary, and now we have the bleating about
'insufficient doors'.
Transport managers, politicians and (regrettably) transport hobbyists have
the feeling that a whole trainload is trying to alight at each stop, with a
fresh load boarding. Not so.
One of the differences between Australia and Europe is that our systems are
tidal flow. Europe (and Beijing) are far more short hops.
A Paris RER A train has 10 carriages x 27 m: 30 double doors for 270 m, so
9 m per double door.
Sydney has (and Melbourne could have had) 16 doubles for 160 m, so 10 m per
double. However, they could have had 12 singles too (and 14 if becoming
nondivisible, like Melbourne's new misnamed trains): 22 or 23 doubles for
160 m, 7 m per double door.
Hence, either could have beaten Paris for dwell time.
The reason for Sydney and Melbourne going the way that they have was the
importation of managers from UK: 'We don't have double deck trains in UK,
so you ruddy colonials shouldn't have them either'.

Dwell continues to be irrelevant for density: that is a function of
signalling.
Melbourne has three major sections signalled for 2 min headways. The
underground could have matched them, but that wasn't specified in the
original design. The government also underspecified portals.
The secret to success is speed-proving signalling. Drivers hate that, as
they are likely to spad more often, but that can be overcome with training
and experience.
Dumping the air at 60 km/h, 17 m/s, 1 m/s/s braking needs 17 s and 145 m to
halt. At 0.9 m/s/s service braking, it needs 18 s and 162 m to halt. With
160 m platforms, it is quite possible to have a second train entering a
platform as the preceding clears. That is what Sydney did at Circular Quay
in 1955 (possibly before Tony was around). It was removed because it was
deemed unnecessary, and now managers bleat 'we can't cope'. With service
acceleration and braking, those phases require 40 s of platform occupation,
allowing 80 s dwell time on 2 min headways, and I have never found Sydney
to need more than 70, and we could get down to 60 with ease with the extra
doors, hence the desired operational margin for whatever is going wrong
today.

Capacity of trains is rubbery, according to whether one counts six people
or four per m^2 for standees. Space is now consumed by wheelchairs and
bikes. Melbourne's useless designers have just failed to provide ceiling
hooks for bikes on its misnamed trains; Tait trains had them 90 years ago,
and parkiteer cages have them today. The secret to success today isn't
being single deck: it is removing seats so that people stand for an hour,
not for 10-20 minutes as on real metros.
The propaganda claims 1100 passengers for the new trains, 20% more than
others.
That counts the others at ~950 passengers (the contract figure for
overcrowding), when they can and did hold 1400 when built. What has been
gained by seat removal has been consumed by wheelchairs and bikes.
A Melbourne double-deck set would have started life with 740 seats and 1260
standees: 2000 passengers. Factor standees down to 4 instead of 6 for
comparability: 840 standees. Turn all 2+3 seating into 2+2: that is about
500 seats becoming 400, with standees going up by 200. We are now on 640
seats and 1040 standees: 1680 passengers.
The previous sock-puppet transport minister announced that the new tunnel
would be signalled for 23 tph. Assuming that she was wrong:
Single deck on 2 min headways, 30 x 1100 = 33 000 passengers per hour,
with 14 000 seated for a 60 min journey.
Double deck on 2.5 min headways: 24 x 1680 = 40 000 passengers per hour,
with 15 000 seated.
Double deck on 2 min headways: 30 x 1680 = 50 000 passengers per hour, with
19 000 seated.
At city-stations (about seven) dwell time might hit 60 s; at all others,
dwell time would be no worse than single deck (40-60 s). Years ago,
average dwell time was factored in at 20 s. That doesn't happen any more.
It is 30 s, but the extra performance of modern trains has absorbed that.
https://en.wikipedia.org/wiki/High_Capacity_Metro_Trains

Tony might well sneer at Melbourne, but we can laugh at NSW which couldn't
travel backwards: ' we will die of asphyxiation'. Now they are travelling
sideways.

There has been one useful post in the series: Town Hall as a station could
be solved by holding everyone on the concourse, and releasing them only
when their train is the next one. Increasing frequency would also help.
Sydney may well develop a few nodes (Parramatta has been one for years),
but the number will never be such that dwell time becomes an issue.

Speed is nice, but frequency, reliability and comfort count for more. In
Melbourne, the Belgrave line is timetabled for 65 min, but could be down to
~58 min if trains ran according to their designed performance.
There is no point waiting 30 min for a train which then is 10 min faster

Roderick

On Thursday, June 13, 2019 at 6:19:16 AM UTC+10, Prescott wrote:You might
recall Roderick that the Paris trains have three doors per side per car.
They struggle with that. Sydney has no chance with two doors.The planning
for new train lines in both Perth and Sydney is based on modal interchange,
not on walk-up. Speed of operation is essential for that, as well as for
traversing the long distances of very spread-out cities quickly.

>

>