Re: Indian Pacific hits freight train - Who's the Responsible Investigating body?

[mauried@commslab.gov.au (Maurie Daly)]

>Having an insulator between the wheel and the rail is usually fatal for any
>form of track circuit.

>Reading the above I think there is some confusion as to how a track circuit
>operates.  In it's most basic form (Steady DC), the sections are broken up
>by insulated joints.  At one end of the section, the voltage is fed to the
>track, and at the other end the track relay is between the rails.  So the
>circuit goes:
>B2 (positive side of battery), feed resistor, one leg of the track circuit
>(rail), relay, other leg of the track circuit (rail), N2 (negative side of
>battery).

>So the relay picks up when the track is unoccupied, and drops away when a
>train enters the section.  This is because the trains wheels act as a short
>circuit.  The relay will also drop, simulating the presence of a train, if
>you use a piece of wire to short the rails, (kids living near level
>crossings find this out annoyingly regularly), or if the bonding between
>consecutive rails breaks, or the rail itself breaks, or a track lead is
>broken, or if the ballast resistance falls due to heavy rain, or if the
>mains power fails and the track battery goes flat.

>So if a train is either lightweight (Victorian Sprinter), or is sanding
>VERY heavily (alleged to be 3801), the trainshunt (resistance of the axles
>of the train) can rise until its not enough to shunt enough current to drop
>the relay (Ohms law etc...).

>Therefore, a train drops the relay, and if it drops a lot of sand or goes
>high resistive, it 'disappears'.  This is what I understand to have
>happened on Cowan Bank, and with the Sprinters.

>European railways I believe have an arrangement where the track circuit
>will not pick up until the next track in sequence is down, apparently due
>to European ideas on proving relay operation.  This has the effect of
>ensuring that a train 'dissapearing' off a track circuit doesn't actually
>cause false operation of signals, points, level crossings, etc...  I
>remember reading that one of the recent resignallings somewhere in NSW will
>actually generate some sort of alarm whenever 'non sequentual' operation
>occurs, this may be what you are alluding to.

>Why can't this 'sequential testing' be used in Australia, I hear you ask?
>Well British, American and Australian railways have always used a different
>philosophy when designing anything vital within a signalling system, i.e.,
>if you open a contact with a vital relay in a vital circuit, you can
>usually assume that the relay will drop, whereas the Europeans almost
>always want to prove that the relay actually dropped.  When applied to
>track circuits, this philosophy goes that if you design a track circuit,
>then it must drop whenever a train enter the section.  If we started
>proving sequential operation, it would create problems if a track circuit
>drops, then it needs the next track down to pick up, which leaves problems
>if there is an intermittent fault, instead of self clearing, it would then
>get stuck down.  This then raises several problems, excessively unreliable
>tracks then mean more caution orders, hand throwing of points, continuously
>operating level crossings, all of which mean a reduction in safety anyway.
>Why bother, when you can put the effort into ensuring that the tracks work
>properly in the first place(i.e. avoid crummy low voltage circuits)?

>--
>Mr Notagunzel.
>Rail Transportation Connoisseur.
>notagunzel@bigfoot.com
>(Waiting for the next move at http://www.bigfoot.com/~notagunzel)


Fair enuf on all the above.
The thrust of my argument is that Railway Safeworking systems are supposed to 
be fail safe , ie if they fail its in the safe mode and not a dangerous mode 
where multiple trains are allowed to occupy a single track section.
The provision of sand across the rails is , if we beleive the 3801 explanation
a failure of the track ccts in the dangerous mode.
If the Rail Authority knows that sand can cause this effect then they have 
only a small number of choices.

1/ Hi voltage track ccts that can break down the insulating material.
2/ Entry / exit track ccts which are sequential , as you have explained above.
3/ Axle counting instead of track ccts altogether.
4/ Banning sanding wherever low voltage track ccts are used.
5/ Instituting absolute block working where locos that carry sand run over the 
track circuited terrain.
 
If however , the Rail Authority has done none of these, then they are 
expecting us to beleive that sanding from a steam loco can make the track ccts
fail in a dangerous mode , but not sand from diesels or electrics.
Hardly beleivable one would think.
Prior to the 3801 incident had the SRA , or indeed any other Railway in this 
country issued warnings about sanding in track circuited terrain?

MD