I think you might be confusing slightly different things-
a) approach release of a signal for a slow speed diverging route,
b) modified aspect sequence utilised where signals on a line have to be positioned closer than considerations of braking distance would dictate and thus a "modified aspect sequence" is provided- which does involve approach release
c) the "delayed yellow" of a warning class route due to the Restricted Overlap beyond the following signal- which again involves approach release.
Raynes Park control is generally thought of as applying to the first of these but actually is applicable to any. Indeed one could say that it is a special case of the requirement always to utilise the correct version of a function, when in fact there are multiple copies (i.e. generations of repeat relays at the same site; or at one site repeating that at another; or held in multiple sub-system's memories). It is all about which is the FUNDAMENTAL and which is the DERIVATIVE from that fundamental.
In normal cases, most of the time, "copies" will be just that, the same as the original. However they will always lag a little in time at every change of state. and this may cause difficulties if the overall design doesn't make allowances for this. In failure conditions, they may become out of step permanently and there is then the problem that "A thinks something and B thinks something else"; if A and B then both act according to what they think, then it can be most unsatisfactory as they think different things are true!
Consider a track circuit which is designed as far as possible to fail right-side which means-
1. TR up means that we can safely assume that the track really is clear
2. TR down means either:
i) the track is truly occupied, or
ii) the track has failed and may show occupied but really there is no train.
Hence any time that we use a back contact of the TR we need to be careful; we may carelessly think that it means that there is a train detected on the section and perhaps 999/1000 it does mean that, but there is a non-negligible chance that there is no train but just a fault.
Now if the fundament track relay (generally taken actually to be the TPR) is down and it is used both for the aspect of the signal in rear (front contact) and to give approach release for the signal of which it is the berth TC (back contact) then if it happens to be down because of a failure, then
a) we do get the A/R of the forward signal which is a bit undesirable, but
b) it isn't too bad because by the same relay the signal in rear is held to red.
Therefore whereas the false A/R is in fact a Wrong Side Failure it is actually a Protected WSF.
The only train that could see the signal clear invalidly would be one that had already passed the signal in rear when the failure occurred-
a) it firstly is a low likelihood as the "window of opportunity" in which such a failure could occur and give the bad result is strictly limited, and
b) secondly we do know that train had only passed that outer signal at yellow and therefore should be expecting the next signal at red so should be going slowly and thus the consequences of the false A/R are also minimised.
HOWEVER a much worse problem arises if utilise a
different relay to give the A/R than the controlling the aspect of the the signal in rear.
In a failure condition where the relay controlling the aspect of the rear signal is up, but its repeat relay controlling the A/R is down (perhaps a fuse has blown, cable become disconnected etc), then we could have the scenario of the forward signal which should be at red as routed over slow speed points is actually showing Green + route indicator, whilst the signal in rear shows Green.
The result can be train going at maximum speed suddenly encountering the slow speed divergence through the pointwork and derailing through overspeeding.
All Raynes Park control entails is ensuring that [i] the relay controlling the aspect of the signal in rear is either
a) on the same relay as the one whose back contact is used to implement the A/R, or
b) a later generation repeat- i.e. if the first relay is down when it should be up then the repeat relay will also be down and thus protect the failure.
By extension, Raynes Park also means the careful use of duplicate track ids within SSI interlockings (it used to be permissible to copy across both "positive" and "negative" to prove definitely up / definitely down from one part of the electronic brain to the other, but even that has changed a bit recently) and indeed any other case of needing to thin carefully about system behaviour during foreseeable failure scenarios.
Like so may signaling controls, "Raynes Park" is named after the place (suburb of south London) where the control was NOT provided and the incident occurred!
Also see
Saraswathi's post
If you have the IRSE Study DVD ROM then under module 2 you'll find a Pwwerpoint presentation about Braking calcs, headway and aspect sequence which icludes an animated sequence relating to Raynes Park Control.
I think the thing that lies at the heart of your confusion is that you are thinking only of how the system SHOULD work in normal conditions; you are not thinking about what would happen when faults arise. A signal engineer has to make sure that the signaling at least remains safe (or as safe as possible) in failure conditions.
{Also be aware that when there is a modified 3-aspect sequence, the whole point is that there are NOT consecutive Yellows- the driver may well SEE one signal at yellow and then SEE the next signal also at yellow, but the signals themselves are never both showing yellow simultaneously. his is really no different to train12 following train 1 along standard 3 aspect signaling and receiving yellow at signal after signal because the trains are close and travelling at the same speed.}
In tyour case you have correctly drawn the modified 3 aspect sequence which means tyhat when 5 is at red then initially 3 is held to red and so the train is given warning to brake at signal1. This is necessary since signal 3 is not at braking distance from 5 at maximum permissible speed, whereas signal3 is at full braking from signal . Once the train has been braking so that it could stop at signal3, then it is correct that this signal can then clear as the concern due to insufficinent braking has been overcome.
You have indicated that the "release" is by some form of timed occupation of CT track. the Raynes park issue is that the timer which achieves that release must be "down-proved" within the controls for siganl1. Hence if it should fail in the "timed" state then although there is a false release of signal3, it is sufficiently mitigated since signal 1 will be held to red.
More generally any device that on the one hand gives a release, should be proved not in such a state initially. hence if a treadle were used, then the QRR contact would give the release of 3 once it was depressed by the train's wheels, but its QNR contact would be included in sgnal1 aspect level. Hence if it is jammed down for any reason once train 1 operates it, then train 2 will stop at signal 1 and be talked past that signal at danger by the signaller- the driver will then see the false release of signal3 but it isn't dangerous because already going at slow speed.
(21-08-2013, 07:37 AM)NJK Wrote: Dear Members,
I have few doubts in Rayne's park control.
Normally if we provide approach control for one signal means, then the aspect sequence for that line will be modified aspect sequence.
In the example of 3 aspect case, Consecutive yellow (Signal in rear to the approach controlled signal will show yellow, Approached controlled signal will show RED, once the train occupies the berth track of approach controlled signal, then it will go to yellow ) will be provided.
Here my doubt is how the failure of approach controlled signal's TPR will result in signal in rear show higher aspect then the permissible single yellow (Green in case of 3 aspect).
The signal in rear will show green only when the approach controlled signal is green.
Or else the modified aspect sequence is introduced only after the incident in Rayne's park ?
Please find the attachment also. Kindly clarify my doubt. If I am wrong kindly correct me
Regards,
NJK
