(08-05-2009, 07:18 AM)mirabadi Wrote: Dear PJW
many thanks for your explanation.
I think you are have commented on what I was asking, exactly.
I got the values of the margines, from an indian book. in setting the regulating resistance in DC TCs, it was mentioned that a 2.5 times PU voltage is considered. this is to make sure that the relay voltage does not go beyound that. (perhaps to prevent the relay from overvoltage)
VR <= PUV when Rb (the ballast resistance) is infinity and feed is in maximum value.
the checking for the reliability is considered to be:
VR >= 1.25 PUV , this should ensure that the relay is in pick up condition in worst case, when the feed is in minimum and Rb also in minimum.
and the last checkin, as you said for safety,
VR <= 0.85 DAV, this should ensure the safety, since ensures dropping of the relay when the track is occupied. in the maximum value of the feed and the ballasat resistance.
are these values true for the network rail as well or they depend on the design and situation?
and my main question was that do we need to consider similr values for the AC TCs as well?
If these calculations are not valid for the AC circuits, is there any reference to use?
I think it should be a bit more coplicated when impedance bond is added to the circuit,
Mirabadi
I suppose that the majority of NR's dc track circuits would be fed from a 904 (or nowadays 867) feedset- there is no feed resistance for the technician to adjust. They are self regulating, but there is an adjustment re the type of track relay and another if the rail leads are particularly long. Must admit I don't quite know how they work but I am sure there is nothing particularly sophisticated about them- they need a guaranteed 110V input, have a large choke and a bridge rectifier of zener diodes to give them ac immunity, but I think that the real reason why they don't need adjustment is that they are used as relatively short (c 600m) track circuits on good quality (from insulation perspective) trackwork and are also capable of putting a fair bit of power into the ballast without affecting the rail voltage too much. There is a disadvantage with them though- where the insulations are poor they charge up the ballast and then we get residual voltage problems with ballast battery effect.
Even where the more traditional (trickle charged lead acid cell and series feed resistance is used) the technicians don't do calculations, simply adjusting the resistance as necessary to achieve the necessary drop shunt. The issue is of course how they know what that actually should be, given that they cannot know what the ballast resistance is at that particular time. I think the answer is experience and comparison with previous readings on the track circuit record card for similar weather conditions, plus aiming to set up for a drop shunt no less than 0.8 ohm knowing that there is a bit of a safety margin before it would become non-compliant with the specified 0.5 ohm.
You are absolutely correct that the calculations for the varieties of ac tracks would differ- a common traditional variety used vane relays where the operation depends on the induced currents in the vane interacting with another magnetic field; phase angle is critical and thus to solve these equations requires the use of Maxwell's equations and complex numbers etc. I could just about have coped with these years ago when studying Physics at university, but I am glad to say that the IRSE won't expect students to do so in IRSE exam.
If an ac track calculation does come up, it'll be for something like the WR Quick Release where the feed end consists of an ac transformer with feed resistance and a tapping adjustment and the relay end consists of a bridge rectifier a series resitor feeding a normal track relay. Primary advantage of these I think was the higher rail voltage thus permitting tracks of 1000m or more, being ac slightly better at breaking down poor rail surface conditions, certainly no problem with charging up a ballast battery and the fact that the relay itself could be placed in a convenient relay room a long way from the track end (but a nightmare for doing a drop shunt before the days of mobile phones or back-back radios!) Fundamentally treat these as a dc track, but worth commenting that the assumption of zero volt drop down length of rail is an oversimplification.
Yes impedance bonds do make things more interesting; in theory they should provide a perfect zero resistance path for the large dc traction current to flow from either rail to the commoning plate and then through the next impedance bond back to the rails again thus bypassing the insulated block joints, yet presenting an infinite resistance for the ac track circuiting current, so that even those few milliamps is completely blocked. Of course this is not achieved in reality and there is some series resistance for the traction current and more importantly for the signal engineer some leakage of the ac track circuiting current to places we'd rather it didn't go. In simplistic terms, the main effect is that the track circuit appears to have a low ballast resistance permanently; indeed we often have to accept that the ideal minimum 0.5 ohm drop shunt cannot be achieved and sometimes have to accept only 0.3 ohms. There is however some compensation in that much of the variability due to weather conditions of the "real" ballast resistance is not obvious- the impedance bonds dominate the situation so the net effect is an effectively constant low value rather than one that may be infinite in good weather yet low in wet weather. Impedance bonds can be "tuned" for particular fequencies; indeed one preserved steam railway with no apparnet need for impedance bonds has had to install some to get their ac track circuits to operate properly!
----------------------------------------------------------------------
So on a related subject, here are some questions open to anyone thinking of attempting the track circuit calculation in the module 5 exam (assuming that it comes up of course) to see whether you have the relevant understanding:
Sometimes the question doesn't state a "percentage release" (ratio of Drop Away / Pick Up) for the relay and you have to choose one to assume; 68% is a favourite. The questions are:
a) if setting out to design a track relay, would the designer ideally want to achieve a higher or a lower figure?
b) what are the relevant factors in the design / construction of the relay which influence the figure achieved?
Don't forget that in most track circuit calculation questions there are certain marks reserved for the answering of such questions at the end of the calculations themselves .......
