Does taper or flange handle train curves, Or both?

[hexagon789]

Manchester Piccadilly platform 14 appears to have acquired a check rail on the outer rail???

I think they are required on curves of less than 10 chains radius.

 

[507 001]

The terrible hunting you used to get on Metrolink was similarly caused by such a compromise.

Used to?!

The M5000s as delivered had a terrible low frequency hunt at all speeds on segregated lines, so much so that in certain places on the Altrincham line (where the problem has always been worst) an M5000 only speed restriction was put in place. Indeed if you look at the track in places you can see the change in wear pattern where the flange has struck the rail.

Some work was done a couple of years ago to investigate the problem, and a new wheel profile and damper set up trialed. Vehicle 3024 was the trial vehicle and 3115 onwards were built with the new set up with the rest due to be done at their first major mechanical overhaul.

However... the hunting didn’t go away, it just changed to a much more disconcerting high frequency hunting motion, this time mostly at speeds above 30mph. At times it felt as if the flanges were slamming against the rails.
Again, the problem seemed to be confined mostly to the Altrincham line, although 3116, 3117 and 3118 spent some time stopped as they were really bad between Derker and Shaw.
The problem has mostly gone away over the last year or so, there’s been an increased rail grinding program which has probably helped. There’s still a couple of older vehicles that have this high frequency hunt, but it’s not as bad as it was.

Apologies for rambling about Metrolink in a heavy rail thread.....

 

[edwin_m]

I think they are required on curves of less than 10 chains radius.

Indeed, but physics and common practice says they need to be on the inner rail. This one is on the outer.

 

[hexagon789]

Indeed, but physics and common practice says they need to be on the inner rail. This one is on the outer.

Didn't realise that, not sure what reason there is for that.

 

[Faulty]

just a bit odd that i still get people telling me that the flange does play a large role in curving, im still not sure what to believe though

 

[furnessvale]

just a bit odd that i still get people telling me that the flange does play a large role in curving, im still not sure what to believe though

As soon as you are in a situation of cant deficiency or cant excess, the flange is essential or the wheel will simply drop off the rail. Trains are very often in a cant deficiency/excess situation.

Any slight imperfection in track geometry would also throw a flangeless wheel quickly off the railhead.

 

[edwin_m]

As soon as you are in a situation of cant deficiency or cant excess, the flange is essential or the wheel will simply drop off the rail. Trains are very often in a cant deficiency/excess situation.

Not sure about that. The coning also creates a "steering" effect that keeps the wheel close to the optimum path, even if there is a certain amount of lateral force.

 

[furnessvale]

Not sure about that. The coning also creates a "steering" effect that keeps the wheel close to the optimum path, even if there is a certain amount of lateral force.

I'll let you try the first train to test that theory.

Let me know how you get on!

 

[snowball]

I think edwin_m is right here. I have three arguments.

1) Train propulsion and braking depend on what railway people call adhesion, which is just friction between wheel and rail acting in a longitudinal direction. If this is possible, then friction must also be possible in a lateral direction.

2) For any given curve there is only one speed at which a train can traverse it without cant deficiency or excess. So if furnessvale were right, the only way for a train to avoid flange contact would be to traverse each curve at exactly the correct speed.

3) If furnessvale is right, what was the point of all that research into wheel/rail contact?

 

[furnessvale]

I think edwin_m is right here. I have three arguments.

1) Train propulsion and braking depend on what railway people call adhesion, which is just friction between wheel and rail acting in a longitudinal direction. If this is possible, then friction must also be possible in a lateral direction.

2) For any given curve there is only one speed at which a train can traverse it without cant deficiency or excess. So if furnessvale were right, the only way for a train to avoid flange contact would be to traverse each curve at exactly the correct speed.

3) If furnessvale is right, what was the point of all that research into wheel/rail contact?

So all those sidecut rails I transposed or turned were a figment of my imagination.

 

[edwin_m]

So all those sidecut rails I transposed or turned were a figment of my imagination.

I never said that it was always the case. A defect in the vehicle or track can cause flange contact, and hunting is likely still to occur sometimes, so the flange is still essential even on straights or gentle curves. For example the lateral ride on Metrolink has been mentioned many times on here, including observation of sidecut rails even on straight track. I haven't been down that way for a while but someone posted on here that changes to the wheel profile have improved things.

But I suggest that in the absence of any of these problems it's possible for a wheel to ride a curve with a certain amount of cant excess or deficiency without flange contact. I remember visiting Heaton Lodge with the local civil engineer when on graduate training in 1988, and he pointed out the single narrow shiny strip indicating that the wheels of the trains coming from Huddersfield were following the exact same path along the rails of the junction, with rust on the gauge face too. As it was the curved leg of the junction there would have had to be some cant deficiency. The trains were almost all nearly-new Sprinters and the junction had been remodelled not long previously, so this was close to ideal in vehicle dynamic terms.

 

[snowball]

Once again I agree with edwin. I don't say flange contact never happens. I do say it doesn't start the moment there's the slightest difference between the train speed and the neutral speed for the curve, which is what I believe furnessvale said in #36.

As I understand it, it's a correct use of terminology to say that, the moment there's any difference between the two speeds, then you have cant deficiency or cant excess.

I believe the coning can handle a certain amount of cant deficiency or excess before flange contact happens, for the three reasons given in #39.

 

[furnessvale]

I think we are all dancing on a pinhead here.

Yes, I could have worded #36 better. I certainly gave the impression that the slightest cant deficiency would be enough to derail, perhaps that is an overstatement. I have certainly seen miles of straight track where the gauge face is rusty, giving testimony to the steering effect of "worn wheel" profiles.

However, I am still of the mind that the lateral forces involved with the weight of a train would require very little cant deficiency or excess before a flange became an essential requirement to avoid derailment on curved track.

 

[DaveTM]

One thing that I haven't seen mentioned here yet, which has to be a significant design aspect, is what happens when an axle locks up or slips. When the wheels stop but the train slides onwards you can forget about the conicity; the only thing keeping the train on the tracks is the flanges. If the average traveller knew what percentage of station stops in autumn involve a significant amount of WSP activity (railway term for ABS), they would be a little more appreciative of the job of the driver...

From my priviledged seat in the crumple zone I get a good look at the greasers which (are supposed to) reduce flange/rail wear and squeal. There is always a good coating of mucky grease for a few cm after the greaser on the side of the rail. Then a wider patch about one wheel rotation later, and then one wider still two rotations later, and so on until the different wheel sizes of different stocks results in the grease patches on the side of the rails merging. What i've never understood (but am thankful for) is that the grease doesn't seem to migrate to the top of the railhead where it would make my life very interesting indeed. Can anyone give a good explanation of why?

 

[MarkyT]

One thing that I haven't seen mentioned here yet, which has to be a significant design aspect, is what happens when an axle locks up or slips. When the wheels stop but the train slides onwards you can forget about the conicity; the only thing keeping the train on the tracks is the flanges. If the average traveller knew what percentage of station stops in autumn involve a significant amount of WSP activity (railway term for ABS), they would be a little more appreciative of the job of the driver...

From my priviledged seat in the crumple zone I get a good look at the greasers which (are supposed to) reduce flange/rail wear and squeal. There is always a good coating of mucky grease for a few cm after the greaser on the side of the rail. Then a wider patch about one wheel rotation later, and then one wider still two rotations later, and so on until the different wheel sizes of different stocks results in the grease patches on the side of the rails merging. What i've never understood (but am thankful for) is that the grease doesn't seem to migrate to the top of the railhead where it would make my life very interesting indeed. Can anyone give a good explanation of why?

I could be wrong but I guess if any of the grease gets on the railhead it may have gone in the contact area after the first few axles have passed over due to the high forces involved, as HSTed stated for water contamination here: https://www.railforums.co.uk/threads/lines-where-ato-could-work.162659/page-2#post-3409405