Magnetic Track Brakes for the British Main Line Network

[edwin_m]

That would be self protecting, even if the train subsequently left the rails, as a miscount leaves a train detection section in the occupied state.

There's a risk that a train (say a 4-car) uses the MTB and stops straddling a counter or a coupling breaks and the system reads eight axles and four MTBs as 12 axles, so the section is shown as clear when the last vehicle is still in it.

Perhaps cab car leading bogies would be the best minimum installation for an MU.

That increases the compressive forces within the train as noted in a previous post, as well as increasing the amount of energy each MTB has to dissipate.

 

[Falcon1200]

Somewhat surprised by Mr Keay's quote in the OP that the 1997 Southall collision could have been avoided had magnetic track brakes been fitted; Could they really have decelerated an HST from line speed quickly enough?

 

[HSTEd]

Somewhat surprised by Mr Keay's quote in the OP that the 1997 Southall collision could have been avoided had magnetic track brakes been fitted; Could they really have decelerated an HST from line speed quickly enough?

The RSSB specification limits maximum decelleration (from all operating braking systems) to 2.5m/s/s. A suitably specced MTB system should be able to attain this deceleration quite rapidly after the command to apply the brakes.

A simplistic constant deceleration curve would allow a train to stop from 125mph (~56m/s) in approximately 22 seconds. The train would travel approximately 630m in that time.

I don't know how that relates to the Southall accident, but it is quite fast.

 

[MarkyT]

There's a risk that a train (say a 4-car) uses the MTB and stops straddling a counter or a coupling breaks and the system reads eight axles and four MTBs as 12 axles, so the section is shown as clear when the last vehicle is still in it.

A possible failure mode I agree. Some extra safeguards might be worth investigating. E.g. setting up radio systems to issue emergency calls to other trains in the area automatically if the system is deployed, including closely following same direction ones that could be most at risk in the scenario. Might be enhanced where practical in CBI interlocking areas to automatically operate emergency replacement controls (or ETCS movement authority withdrawal equivalent) in the locality.

That increases the compressive forces within the train as noted in a previous post, as well as increasing the amount of energy each MTB has to dissipate.

If the MTB on the rear cab car also operated simultaneously that would help counter this. It might be easier in the FLIRT example to place the MTBs on a number of the articulated trailing bogies instead. Possibly more space with no traction equipment.

 

[northeast10]

Albeit not heavy rail but a track brake stops a Tyne and Wear metro in 150m from 50mph.

 

[MarkyT]

The BLS RABe 528, a recent FLIRT 4 design, is equipped with MTBs on the 3rd, 4th and 5th bogies (of 7 total on this 6-car train) The fitted bogies are all articulated and unpowered. I'd seen a comment on reddit about these trains having magnetic brakes in response to a query about the information panel. here:

https://www.reddit.com/r/trains/comments/19b8huk/what_do_all_these_writings_on_the_train_mean/?rdt=50777

MTBs are indicated by the code Mg. They're very difficult to see right under the bogie side frames from a normal viewing height, especially when at a station, so this slow mo low view video was useful to confirm:

 

[Technologist]

There is obviosly a potential to stop a train more quickly than is possible using convential braking.

So how are they deployed, does it form part of the normal brake application, would it be automatic when systems recognise that normal wheel/rail adhesion has been lost or is there an 'oh ****' lever or button in the cab which the driver operates when they realise that they are in trouble, or some combination of the above.

How do they cope with operation over points? Do the gaps cause a problem.

And another related thought, was adhesion as much of an issue when rolling stock was tread braked, could rolling stock have a proportion of tread brakes?

Obviously anything that improves safety is a good thing but if it then encourages riskier behaviour "because I always have the track brake" then the benefit is at least partly lost.

We've been using eddy current brakes for decades on roller coasters to allow trains to stop faster and more safely and enable new track designs. Why aren't we using eddy current braking to vastly decrease stopping distances and thus vastly increase the amount of trains we can run?

I'm expecting someone to say "requirement X says a train cannot decelerate at greater than X ms^2", which is hogwash plenty of other forms of transport subject their passengers to much higher accelerations and find ways around it.

 

[hexagon789]

We've been using eddy current brakes for decades on roller coasters to allow trains to stop faster and more safely and enable new track designs. Why aren't we using eddy current braking to vastly decrease stopping distances and thus vastly increase the amount of trains we can run?

I'm expecting someone to say "requirement X says a train cannot decelerate at greater than X ms^2", which is hogwash plenty of other forms of transport subject their passengers to much higher accelerations and find ways around it.

The only application of Eddy Current brakes on mainline trains I'm aware of are on the German ICE3 and certain Japanese Shinkansen high-speed trains, they are only permitted to be used on certain high-speed lines which have been specifically designed or modified for these brakes to be used.

Japan mostly switched back to regenerative braking on the trainsets which were equipped, as the ability to recoup some of the lost energy in braking was deemed more advantageous.

They are generally banned elsewgere due to either signalling interference or more commonly the creation of vortices under the train which disturb ballast and other items, increased rail temperatures or the potential to damage points and switches.

They also don't actually provide significantly better deceleration than regenerative/rheostatic brakes, the advantage is more in the reduced wear of braking components. On the ICE3, maximum Eddy Current braking is only 0.9m/s/s and in the upper speed range only 0.6.

 

[stuu]

We've been using eddy current brakes for decades on roller coasters to allow trains to stop faster and more safely and enable new track designs. Why aren't we using eddy current braking to vastly decrease stopping distances and thus vastly increase the amount of trains we can run?

People aren't stood around not even holding on to anything on rollercoasters, so the maximum deceleration rates have to be much, much lower

 

[MarkyT]

The only application of Eddy Current brakes on mainline trains I'm aware of are on the German ICE3 and certain Japanese Shinkansen high-speed trains, they are only permitted to be used on certain high-speed lines which have been specifically designed or modified for these brakes to be used.

DB only uses it on slab track equipped high speed lines I understand. The rails get very hot when the brakes are used and authorities were concerned about dissipation of that energy. I think slab must function better as a large heat sink than sleepered track on ballast. They had trouble initially with axle counter wheel sensors, the predominant method of train detection in Germany, but this has been solved.

 

[hexagon789]

DB only uses it on slab track equipped high speed lines I understand.

Yes, that's my understanding as well

 

[MarkyT]

We've been using eddy current brakes for decades on roller coasters to allow trains to stop faster and more safely and enable new track designs. Why aren't we using eddy current braking to vastly decrease stopping distances and thus vastly increase the amount of trains we can run?

Failsafe rollercoaster brakes using the eddy current principle can only operate at specific places on the track where additional equipment is placed. They typically rely on permanent magnets and a reactor plate or fin. Different designs place either the fins or the magnets on the train. The track mounted elements can often be moved out of the way by the control system to suppress braking where used for speed and block control. They return to the default brake position under spring pressure. Eddy current effects are proportional to speed with the highest force being generated at the highest speed, declining to nothing at rest. They must be supplemented by a friction brake for final stopping.

 

[Technologist]

People aren't stood around not even holding on to anything on rollercoasters, so the maximum deceleration rates have to be much, much lower

I wasn't thinking rollercoasters but buses, looking at standards and guidelines buses are expected to keep acceleration and deceleration below 1.5ms/s, however accelerations of 2.5ms/s still regularly occur. Accelerations of over 5ms/s are expected in the case of emergency braking.

If we look at the capabilities of battery trains (using the batteries to enhance acceleration on OHLE lines) to accelerate at basically the tractive limit we can potentially re-imagine how we use trains and stay competitive with exiting and emerging forms of transport. For example I scoped a battery electric Intercity train using off the shelf BEV electrics that cost a few % of a typical trains cost, this would be a 22MW train with 400 tonnes of mass.

This would be able to accelerate at 1.3ms/s until it hit ~95mph, there after it would be power limited rather than traction however the performance would be as follows:

100kph - 21s - 0.29km
160kph - 34s - 0.76km
200kph - 45s - 1.30km
250kph - 62s - 2.42km
300kph - 87s - 4.26km
360kph - 127s - 8.01km

With a vehicle of that level of performance it would be possible to run stopper services with the end to end timings of high speed trains. So let's say we ran stops 40km apart (top speed being 360kph) with the deceleration being the same as the acceleration and a 60 second dwell at the station, we'd average 260kph along the line and only have 4 minutes not accelerating or braking

Ergo you probably wouldn't need to run expresses at all and intercity lines could be re-designed as essentially tube lines with massive capacity and frequency. Instead of complex signalling and schedules as multiple routes join and leave main lines we'd just expect passengers to make connections as they do on metro systems.

If we change the paradigm to all trains a stopping every at least every 10 minutes, and spending most of their time accelerating or decelerating then passenger behaviour will change as it does on buses. People won't expect to leave food and drink on a table any more than you would on a bus or in a car, people standing will always be hanging on. Toilets will be for emergencies, given that a London to Liverpool high speed service stopping at Luton, Milton Keynes, Northampton, Leicester, East Mids, Sheffield, Manchester, Liverpool would only take 1:20 most people wouldn't use them or would just get off to use one in a station then re-board another train.

To circle back around to eddy current braking, we could potentially use it to allow trains to brake like rubber tyred vehicles whatever the weather, thus could allow us to operate them in a much higher performance manner with much fewer margins than we use today. It might also allow us to operate existing lines at much higher speeds without needing to expensively re-signal, to put this into context a train operating at 125mph stopping at 5ms/s stops (I have no idea if you can stop that hard just using a rail) in about 315m and puts about 50 degrees C into the rails doing so.

 

[eldomtom2]

If we look at the capabilities of battery trains (using the batteries to enhance acceleration on OHLE lines) to accelerate at basically the tractive limit we can potentially re-imagine how we use trains and stay competitive with exiting and emerging forms of transport. For example I scoped a battery electric Intercity train using off the shelf BEV electrics that cost a few % of a typical trains cost, this would be a 22MW train with 400 tonnes of mass.

This would be able to accelerate at 1.3ms/s until it hit ~95mph, there after it would be power limited rather than traction however the performance would be as follows:

But there's more to acceleration than just power and weight, and furthermore I'd be interested to know the details of how you think a train that weighs 400 tonnes and outputs 22MW could be built.

 

[supervc-10]

The impression I got from the original article was that this is a safety backup - not something for regular use. Think of it as being like a sprinkler system in a building - yes it ruins all your stuff with water damage, but it prevents people being killed and prevents fire from destroying the building. Track brakes could prevent crashes like Salisbury and Talerddig (off topic - but are 158/159s particular bad for slip?) - it wouldn't matter if some damage happens to the railhead or signalling equipment if it means that trains don't crash into each other.

I've been on trams where the track brakes were activated - once in Manchester and once in Toronto. In both cases it felt like being in a car when you slam the brakes on, the vehicles stopped very very quickly, and people did fall over. In both cases it was down to idiots on the line - in Toronto, a pedestrian, and in Manchester, idiot kids pretending to jump as we came into a station (watched them through the glass at the front). I'm not imagining they would need to be activated frequently on a mainline train. I certainly don't think it would be worthwhile to start retrofitting them - but maybe future designs could have them included as a way to reduce risk?

 

[MarkyT]

The impression I got from the original article was that this is a safety backup - not something for regular use. Think of it as being like a sprinkler system in a building - yes it ruins all your stuff with water damage, but it prevents people being killed and prevents fire from destroying the building. Track brakes could prevent crashes like Salisbury and Talerddig (off topic - but are 158/159s particular bad for slip?) - it wouldn't matter if some damage happens to the railhead or signalling equipment if it means that trains don't crash into each other.

A post-incident procedure could be to impose a low emergency speed limit after use of MTBs until the infrastructure involved can be properly inspected by PW & S&T engineers. Slowing down significantly before approaching a newly discovered low-adhesion site seems like a good policy anyway, regardless of infrastructure damage/malfunction.

I've been on trams where the track brakes were activated - once in Manchester and once in Toronto. In both cases it felt like being in a car when you slam the brakes on, the vehicles stopped very very quickly, and people did fall over. In both cases it was down to idiots on the line - in Toronto, a pedestrian, and in Manchester, idiot kids pretending to jump as we came into a station (watched them through the glass at the front). I'm not imagining they would need to be activated frequently on a mainline train. I certainly don't think it would be worthwhile to start retrofitting them - but maybe future designs could have them included as a way to reduce risk?

Large-scale fleet renewals for Wales and other UK regions are currently underway or recently complete. Without any retrofit program on these latest Bombardier/Alstom, CAF, etc. products, Wales and many parts of the UK couldn't benefit from MTBs for another three or four decades, apart from the tram-trains in the Cardiff valleys and the new trains for T&W Metro. I suggest it would at least be worthwhile looking at the handful of recent designs for feasibility. I wouldn't mind betting that identical or very similar bogies from the same manufacturers have MTBs already fitted somewhere in Europe, certainly the case with Stadler Flirts, for example. Probably not economic or practical to fit 30-40 year old 2nd gen units near end of operational life, unless a particular fleet was being substantially rebuilt for a significant life extension.

 

[supervc-10]

That's very true- if easily retrofitted to the FLIRTs, for example, then it would make sense to do so. Probably not going to be possible to do affordably for all recent units though, so probably needs assessment on a case-by-case basis.

 

[MarkyT]

That's very true- if easily retrofitted to the FLIRTs, for example, then it would make sense to do so. Probably not going to be possible to do affordably for all recent units though, so probably needs assessment on a case-by-case basis.

And risk assessment. On the Elizabeth line, I'd guess it would be less of a priority for their Aventras as they tread a heavily trafficked mainline route at both ends and are in tunnel through central London and at the airport. The same train family on SWR suburban will serve a wide network of leafy branches radiating from their mainline so they may be more deserving of fitment. Rural routes, in Wales for example, can face many unexpected obstructions for which assured and powerful stopping in an emergency could be very useful. It could help for tree falls, level crossing, trespass, livestock incidents, and low adhesion scenarios.

 

[Technologist]

But there's more to acceleration than just power and weight, and furthermore I'd be interested to know the details of how you think a train that weighs 400 tonnes and outputs 22MW could be built.

Battery assisted electric train: 22000/400=55kw/t
BEV car above: 745/1.2=620kw/t

Trains are not really pushing power to weight as far as passenger vehicles go!

Lets break it down, if we devote 5% of the mass of our 400t train to batteries we have a 20t battery. 5% is similar to the amount of weight that an 8** class train gains when you add the diesels and should be well within the capability to either be added to a design or for 5% of mass to be saved on an existing train given the fairly obvious weight reduction opportunities in most designs (Siemens claims to have saved more than 5% of the dry weight of their new Velaro). It should also be noted that in automotive usage as a rule of thumb if your battery pack is structural you can normally remove structure equating to around 15% of its mass.

A 20 tonne battery with a 150wh/kg (end of life state) LFP or LMFP battery pack would have 3MWh, this is enough to allow for ranges off wire of more than 160km even at intercity velocities. However in this context we'd be using it to allow for more rapid acceleration. A 3MWh battery would allow for 22MW peak energy usage, this would be a 7.1C discharge rate, the car above achieves 7.4C using a lithium NMC battery type (for energy density reasons), LFP or LMFP batteries are more tolerant of high discharge rates with 10C being possible. LFP batteries are also good for 5-10,000 discharge cycles so even operating on a high speed metro they would last for a good number of years. Energy density, cycle life and discharge rates (particularly at low charge states) are also only going up over time.

In terms of drive units a 22MW train is 30 Tesla Model S drive trains, that equates to about 6 tonnes in mass or probably less than the traction motors in a typical electric train. Motors in top of the line cars are several generations ahead of rail traction motors because of very rapid development cycles since 2012. The cost of 30 Model S drive trains is likely in the region of $300,000 so again cheap by rail standards. I did wonder whether they would need to be de-rated for rail use but actually I doubt it. As the train is so overpowered at full power it will hit 360km/h in 2 minutes dead, then cruise at around 1/4 power even at 360km/h. The biggest challenge is likely to be thermal management of the battery as it will always be charging or discharging at a decently high rate.

Regarding traction I've set the maximum acceleration in my model at 1.3ms/s which is the same as a tube train achieves, I have no idea if this is feasible for a mainline train in all weathers. Though I would have though that a Cross Rail type service that gets you from London to Liverpool in just over an hour and stops at every major town on the way might be worth some investment in cleaning or protecting the tracks.

 

[HSTEd]

But there's more to acceleration than just power and weight, and furthermore I'd be interested to know the details of how you think a train that weighs 400 tonnes and outputs 22MW could be built.

If you provide that power from Li-Ion batteries, its not really impractical.

The AGV uses 600kW permanent magnet motors on each axle that only weigh around 600kg. Motor technology has advanced further in the last decade and automotive motor power densities are pushing towards 4kW/kg.
A 12 car Class 700 (weighing ~410t) has 48 axles, which would be more than 22MW if all axles are motored at 600kW each.

Li-Ion battery technology is capable of enormous discharge rates, on the order 5-10kW/kg.
4kW/kg and 5kW/kg for traction motors and batteries respectively implies a total traction system weigh of ten to fifteen tonnes including controllers.


You would not be providing 22MW of transformer, obviously, indeed you would probably have less transformer weight than more traditional designs, once you utilise the battery to perform acceleration.

 

[Technologist]

You would not be providing 22MW of transformer, obviously, indeed you would probably have less transformer weight than more traditional designs, once you utilise the battery to perform acceleration.

Yes, good spot again if we look at high power power electronics from automotive you can save several tonnes of transformer weight by using DC-DC power converters to charge the battery. The basic principle being that the power electronics shift supply frequency to a much higher frequency which allows a much smaller transformer, a rectifier then converts this to DC to charge the battery or feed the drive train inverter.

 

[bahnause]

The impression I got from the original article was that this is a safety backup - not something for regular use. Think of it as being like a sprinkler system in a building - yes it ruins all your stuff with water damage, but it prevents people being killed and prevents fire from destroying the building. Track brakes could prevent crashes like Salisbury and Talerddig (off topic - but are 158/159s particular bad for slip?) - it wouldn't matter if some damage happens to the railhead or signalling equipment if it means that trains don't crash into each other.

Fun fact: A few days ago I[1] had an emergency brake application commanded by a safety system on a train with magnetic rail brakes. This actually triggered an axle counter fault.

[1] By I, of course, I mean my trainee, who had a rather optimistic assessment of the underlying ETCS braking curve. I have to admit I admired his braveness.

 

[edwin_m]

There's a good argument to have the emergency brake (as triggered by automatic systems) not apply magnetic brakes, because of the risk of injuries to passengers when it is applied, which would usually by a fault in a failsafe system rather than an actual emergency.

 

[MarkyT]

Fun fact: A few days ago I[1] had an emergency brake application commanded by a safety system on a train with magnetic rail brakes. This actually triggered an axle counter fault.

[1] By I, of course, I mean my trainee, who had a rather optimistic assessment of the underlying ETCS braking curve. I have to admit I admired his braveness.

Do you know the nature of the axle counter problem? Did it remain occupied due to miscount after the train left the section?

 

[bahnause]

Do you know the nature of the axle counter problem? Did it remain occupied due to miscount after the train left the section?

Not in this case, but back then when I was a young signaller it usually was a miscount. Track circuits somtimes were bridged by the metal abrasion of the brake.