More about Linespeed

[JC]

Does anyone know what the safety margin is assigned to linespeed?

I found a report from 2000 by IMechE that stated tilt trains could run at 25% less safety margin than not tilting trains but it didn't say or qualify what the normal safety margin would be.
I know linespeed is derived based on many things but this safety margin seems to be arbitrarily assigned.

Any clues?

 

[asylumxl]

I'm not sure myself, but I was under the impression the adhesion on a curve was proportional to the speed, and that linespeed is dependent mostly on reduction of wear,track design/quality and passenger comfort. The counteracting of centrifugal force in the curve makes it more comfortable, and hence tilting trains can go fasTer.

 

[anonymous0101]

Testing is normally to +10 percent of line speed.

 

[Peter Mugridge]

Someone once told me that the margin varied according to the type of track, and he said that for example over pointwork there could be a 50% margin but a bit lower on plain straight track.

No idea what the current margin figures are - this was many years ago!

 

[Old Timer]

The margin of safety is typically in the range of 10/15%.

However this is based upon all the track components being at minmum tolerance (i. e. worn), as well as the suspension of the rolling stock, wear on the wheels, etc being at the minimum acceptable limits.

As a rough guide you can get through S&C at the following speeds.

15 = 20
20 = 25
25 = 30
30 = 35/40 depending on switch type
40 = 50
50 = 60/65 depending on switch length
60 = 70/75 depending on switch length
70 = 80/85 depending on switch length
100 = 115
125 = 135

 

[JC]

Thanks for the responses all,
My understanding is trains to exceed linespeed by 35% or more is not unusual so I'm assuming the safety margin must refer to passengers falling inside the train etc rather than derailment speeds.
Which leads me to the further notion that derailment speeds are not often calculated as the linespeed is based on comfort and infrastructure wear and tear and these are assumed to be far lower than a derailment speed. I'm speculating becasue I can't find any real evidence in RSSB or the ORR about overspeeding except in respect against linespeeds already set.
There is no coverage of how linespeed is determined if a new line were to be created which seems very odd.

 

[Peter Mugridge]

That sounds right because the conversation I had all those years ago related to derailing speed rather than comfort speed; would that 50% margin figure sound about right for that?

I'm pretty sure ( from a rather frightening experience a few years ago! ) that it's not a case that it's the same everywhere; I can be pretty sure that for Mk1 stock the derailment speed margin on a particular mild curve near London is 30%...

 

[Ploughman]

Permissable speed will be determined by the curveature and transition lengths.
Gradients and changes in gradient also play a part.
Straight track is theoretically infinate speed but then signal limitations come in to play.

 

[Old Timer]

To add to Ploughmans excellent response.

Linespeed for passenger trains is determined by the ability of the train to run with minimal dynamic response. As the response of the vehicle increases then comfort falls off. A vehicle will become alarmingly uncomfortable immediately prior to the point at which it will derail.

Linespeed is not determined by the speed of the trains but rather as part of a package that involves the track geometry, the types of train which will operate, the spread of speeds, and the various different types of rolling stock.

In building a new railway all these factors will be considered, and the track designed to the best parameters to suit all planned vehicle types.

The build of the vehicle will be determined by the planned operating speed, its loaded, and tare weight.

These are all assessed to minimise wear and tear on the track infrastructure and the rolling stock. Derailment speeds are never considered as they are far higher than the speed at which trains are planned to run.

The above is a very much simplified response of a very complex subject.

 

[JC]

Everyone, thanks again for the input it's certainly been useful for clarifying my thoughts ready for debate with colleagues and will enable me to target my questions appropriately.

 

[coral reef]

My bible on this subject is O. S. Nock's "Two Miles A Minute" book I got as a kid, charting the development of the HST and the APT.

On curved track the weight of the vehicle goes downwards, the centrifugal force goes outwards and the combination of the two is a line that points down and outwards from the curve. For example, if the the weight is 10000N and the centrifugal force is 10000N then the the angle of the resulting force will be 45 degrees from straight down.

Obviously 45 degrees is a bit high. However, say for example the speed is much lower, resulting in an angle of 4 degrees, this is the sort of figure we get on the railways. The track can be tilted (cant) by 4 degrees to match, resulting in the force being balanced across the two rails and the train experiences no lateral acceleration with respect to the track.

However, the train and its passengers can cope with pushing it a little bit further. Going faster than the balanced speed of a track causes whats called a cant deficiency, in that to balance any lateral acceleration the track would need more cant, and the cant deficiency is measured in degrees.

The rub is that the cant deficiency points in one direction when going faster than the cant-equilibrium speed, but then if the vehicle stops on the same canted track the cant deficiency then points in the other direction. This means that you can't keep tilting the track forever, because it becomes equally uncomfortable to have cups of tea sliding off tables if the train stops on the corner.

A can't deficiency of 4 degrees is about what conventional trains allow. With this you can work out how fast a train can go around a corner just by knowing the curve radius and the weight of the train.

A tilting train such as the APT was designed to allow a cant deficiency of 9 degrees. I believe the class 390s have a lower max cant deficiency.

However also in general there seems to be more of a safety margin built in to the designs of track and train the faster you go, which makes sense. Interestingly they tested the APT locked in full tilt the wrong way around the curvenear Dover at 90mph (30mph limit or something) and it didn't derail, so I think the safety margin with respect to the train staying on the track is huge.

Curve radius, conventional speed, tilting speed;

1km, 100km/h, 135km/h
1.5km, 160km/h, 180km/h
2.0km, 190km/h, 240km/h

HSLs seem to have more actual track cant than normal lines (freight?) allowing slightly higher non-tilt linespeeds.

 

[Jan]

HSLs seem to have more actual track cant than normal lines (freight?) allowing slightly higher non-tilt linespeeds.

Yes, as freight trains usually travel at slower speeds, they will experience a cant surplus (is that the correct term?), which causes lateral forces to act on the track bed and rails, all the more as freight trains usually are both longer and heavier than passenger trains, which in turn increases the necessary maintenance. On a HSL (or a suburban rail network for that matter) all trains travel at the same speed, so you can simply build the optimal cant.
Maximum cant also seems to depend on the country, in Germany for example it's 150 mm (= 6 degrees) plus another 100 mm (4 degrees) of allowed cant deficiency, and there's a curve on my local tram network where this is fully utilised.

 

[TGV]

Cant excess or cant deficiency are the terms.

10% overspeed testing is usual. For both infrastructure and rolling stock. But this depends on rolling stock condition, wheel profile, track condition and conditions of running.

The mechanisms of failure would be derailment caused by flange climbing (depending on the curve radius and wheel angle of attack) or track failure due to excessive forces on a curve (could lead to rail-rolling or gauge spreading). It's worth noting that a common misconception is that tilting trains reduce track forces. In fact they are usually MORE harsh on track than a non-tilting train because the extra tilt induces a larger lateral force at the wheel/rail interface. It's purely a comfort mechanism.

On straight track, the limitations are down to the vehicle suspension and condition of the wheelsets and design of the suspension. A test in France in the 1960's when a locomotive reached 331km/h on a straight section uttlery destroyed the track because the hunting of the bogies caused such massive forces it deformed the rails. You may note that without Yaw Dampers fitted, rolling stock in the UK can't travel above 75mph, and it is for that same reason - controlling of the bogies and wheel sets at higher speeds. Some EMU's in Scotland are having yaw dampers fitted and once that's done, without any other physical modifications, their service speed limit has been raised from 75mph to 90mph.

In cab signalling allows overspeed before the system cuts the driver out. Typically in the region of 10-15km/h.

 

[coral reef]

Are yaw dampers part of the reason why the class 365s are 100mph but the class 465/466s are 75mph? I've always wondered that, as they easily have the power to get to the 90mph linespeed of the area. I can understand why they were never built for higher speeds, as they were intended solely for the suburbian lines that rarely are 90mph and the train would be stopping at some point anyway. But now they are regularly seen on the Tonbridge mainline which is significantly 90mph. Or have the 465/9s been sped up?

 

[Peter Mugridge]

As far as I know they are still listed as 75mph?

 

[chuffchuff]

Yup 465/9's are still 75mph Max