Electrical properties of railway track....

[railproject]

Hi,
I am looking for the general electrical characteristics of mainline railway tracks. I am modelling the rails as a transmission line and need to have some rough figures (resistance, inductance, capacitance, cunductance) on the impedance and admittance of the rails.

I am a Uni student currently on a project looking at creating a more accurate train detection system. The basic concept of the idea would be to send a signal down the rails; the train acts as a short circuit which will reflect the signal bacl towards the source. By measuring the time taken for the reflected pusle to return it should be possible to calculate the trains exact distance.

obviously there are other signals on the rail, such as traction currents, electric motor noise.... are there any other signals which are going to pose an issue and will requrie some form of flitering? I have been trying to research the characteristics of the 'noise' but have yet to find anything.

If anyone could point me in the right direction or knows any of this info that would be awesome!

Sorry if this is not the type of forum where technical questions are asked.

Thanks for taking the time to read this!

 

[Nym]

Well, if I remember right in an RIAB report the shape of track is on there, it's steel, but then you have sandite, rust etc, to worry about, theres also the properties of the axles that change for conductance etc, then you have weather conditions changing the ballast conductivities and diferent track shapes, and then diferent types of welding and track connection overides...

Rather you than me hun

 

[Dennis]

I'd have thought that the rails themselves, if perfectly isolated from the ground would have much the same properties as steel (the grades which are used are listed http://www.iorw.org/jobknow4.html ) - datasheets for the individual grades should be readily available.

I don't know anything about electrical noise, you could try searching group standards .

Possibly a further consideration would be current loss through railpads (the rubber pad between the sleepers and rail) - most of the ones used by NR are filled with carbon black and have a small but finite degree of conductivity which will vary with both wear and ambient conditions (to name but two variables).


OOI, which uni are you at?

 

[Nym]

When I was looking at them, the main problem was the salt and ammonia in the ballast increasing the conductivity across the rails, unforunatly this cannot be predicted, unless you blank the rails on each empty pathway...? Btw, that ammount of the aboves may change every time a train runs over it...

 

[Dennis]

When I was looking at them, the main problem was the salt and ammonia in the ballast increasing the conductivity across the rails, unforunatly this cannot be predicted, unless you blank the rails on each empty pathway...?

You're correct there (not forgetting the influence of moisture) but ideally the railpads will electrically isolate the rails from the sleepers and ballast.

 

[Nym]

Yeah, but how much would all of them cost, compared to somthing like an 801.1 protocoll resistivity sensor multiplex for the length of the path...?

 

[rower40]

If I'm not mistaken, the Vaughn Harmon Level Crossing Predictor being trialled on the Cromer line works in this way; i.e. not treating a track circuit as digital (occupied/clear) but as analogue, determining where the train is and how fast it is approaching by measuring the rate-of-change of impedance as the short-circuiting axle approaches a fixed feed point.
I'm still gonna look both ways when I go across a level crossing...

 

[Up Carriage Liner]

Always worth consulting British Railway Track: Design, Construction and Maintenance, edited by G.H. Cope, 6th ed., The Permanent Way Institution, 1993.

 

[railproject]

If I'm not mistaken, the Vaughn Harmon Level Crossing Predictor being trialled on the Cromer line works in this way

Do you have a web address which has any information regarding this predictor?

 

[railproject]

so...

im wondering what the best option is in terms of detecting the reflected pulse.

I realse there are going to be all sorts of noise imparted onto the pulse signal and therefore it could be hard to detect when the pulse is returned.

Can anyone give me suggestions as to the best methods of filtering/detecting the returned signal?

 

[TI21]

Hi railproject,

If you're still looking at this then send me a private message and I may have some info I can send you (too much to post here!)

 

[railproject]

hey, i would like that info you have TI21. i have sent you a pm with my email address

 

[TI21]

Bother - just found I can't send or receive pms because I have less than 5 posts. I've put in a help request to try to resolve this.

 

[Nick W]

so...

im wondering what the best option is in terms of detecting the reflected pulse.

I realse there are going to be all sorts of noise imparted onto the pulse signal and therefore it could be hard to detect when the pulse is returned.

Can anyone give me suggestions as to the best methods of filtering/detecting the returned signal?

Fire a high-frequency pulse at least 10 times as frequent as any noise from currents induced in the rails and it should be clearly recognisable above any interference.

I have no experience with the electrical circuits required to filter out frequencies but a quick google search found this: http://www.swarthmore.edu/NatSci/echeeve1/Ref/FilterBkgrnd/Filters.html .

 

[TI21]

The trouble is that the track is so inductive that getting high frequency signals down the rails is a real challenge - anything above 10kHz-ish is a complete nightmare, and there's a fair bit of interference even up there! As railproject is discovering (we're in touch off-forum) this project is much harder than it looks - really good project though.

 

[Bill EWS]

None appears to have mentioned the track circuits. I believe this is 4 volts D.C. The locomotives, coaches and wagon wheels break the ciruit to place signals to danger behind the passing train. Where rail joints and plastic insulation is used wires are looped, usually in pairs, from one section of rail to the other. However, using electrical frequencies may be an answer but track ciruits fail for many reasons and may not be acceptable for a signal that has to remain constant such as being discussed I am not sure if lines that have Radio signaling such as the West Highland line would be a help or not.

I believe the railways would like to get rid of track circuiting but so far there is not one acceptable system that's been proved although trials are being carried out. Track circuit failure is quite regular.

 

[cc2]

The trouble is that the track is so inductive that getting high frequency signals down the rails is a real challenge - anything above 10kHz-ish is a complete nightmare, and there's a fair bit of interference even up there! As railproject is discovering (we're in touch off-forum) this project is much harder than it looks - really good project though.

cca 1 Mhz might work ,but only in a pulsed mode-modulated.in this case you should deal only with SURGE impedance of a rail.i don't know how much exactly it is ,but i guess something like 100 ohms.however,the problem with megahertz pulsed is high attenuation of signal ,due to skin effect in particular.good dection system you will need after few km of propagation!

as concerns low ,continious frequency , the inductance of single rail is about 2×10^-6 H per meter.
--- old post above --- --- new post below ---

I'd have thought that the rails themselves, if perfectly isolated from the ground would have much the same properties as steel (the grades which are used are listed http://www.iorw.org/jobknow4.html ) - datasheets for the individual grades should be readily available.

were the rails perfectly isolated that would be very dangerous.
they must be grounded.grounded in a special way,on several places, becouse they represent geometrically long circuits.traction currents can be very large,and passing them through rail just by the virtue selfinductance,can bring them up to dangerous potentials.if the potential rises above some 50-60 V anybody who touches them can be electrocuted.