Great Western Coffee Shop

Late yesterday evening a LNER Azuma (800109) and a LNER HST (43300) were involved in a moderately low speed collision outside Neville Hill Depot in Leeds. Several carriages of the Azuma were derailed in the collision. Neither train was in passenger service.

The location, I'm led to believe, is a stretch of track called the Down Hull Goods Loop which gives access to and from the depot. There is, I believe, permissive working on this stretch of line. The speed limit is 15mph.

(http://i598.photobucket.com/albums/tt68/bignosemac/EJSmojWXYAICOFI_zpsgml20ogw.jpeg)
https://twitter.com/stuartthomas/status/1194763023224918016?s=20

I think the HST cab came off better than the IET one although the latter is mostly 'for show' fibreglass......

https://m.facebook.com/groups/1295041497248086?view=permalink&id=2813644855387735

It's a bit of a mess for 15 mph.

If they were both doing 15, closing speed would have been 30 though.

The HST was, I believe, stationary.

Well, yes, but that's the same as hitting a wall at 15 mph. 'Closing speeds' have more appeal to journalists than to physicists...

I'm quite relieved there was a bit of a mess.  Better a mess at the front where the energy of a collision is designed to be absorbed, than that energy being felt further back where passengers might be sat.

I agree, but am slightly alarmed by the derailment. I am, however, no expert.

That Azuma is going to be out of action for quite some time, as for the HST it's time to say goodbye for good.

The 800 was still “snaking” across a couple of sets of points at the point of derailment.  I think the HST was basically on almost straight track.  I expect that affected the derailment?

Even on permissive-working stretches of track, you can't have trains travelling in opposite directions at the same time.  You can have train 1 moving, and approaching train 2, but train 2 has to be stationary.  If "Huddersfield Control" is implemented, train 2 mustn't have a movement authority AWAY from train 1 either.

The Down Hull Goods loop is used for stacking up trains to await their acceptance into Neville Hill depot.  It's only signalled for down-direction movements.  But from that picture, I would suggest that the coming-together occurred on the Neville Hill Depot Arrival line.  (The points visible next to the Azuma's 6th coach are where the Down Hull Goods Loop rejoins the Down Hull Main for trains not going into Neville Hill depot).  The track adjacent to the HST and Azuma is a depot Shunt Neck.

We can see the Azuma's rear lights in the broken fibreglass, so it seems as if the HST ran into the back of the Azuma.

After an accident, it's standard practice to switch on every red light available.  So I wouldn't rely on that to determine direction.

Also, on these Hitachi trains the Tail Lights are illuminated automatically whenever the driver keys out of the train - unless coupled, or overridden by the Emergency Headlight Switch.

Information posted elsewhere suggests both trains entered the depot line from the Leeds direction, first the HST as 5D24, followed by the IET as 5D29.
https://www.realtimetrains.co.uk/train/Y31299/2019-11-13/detailed
https://www.realtimetrains.co.uk/train/Y31303/2019-11-13/detailed

Ah, thank you. That's two railway lighting things I've learnt today!

When the cab is locked out on the Azumas the tail lights illuminate. The cab was locked out to preserve i would imagine.

The HST was stationary, the Azuma has effectively ran into a 410 Ton wall

There is still something rather alarming about such a low-speed collision causing several coaches to derail. I hope it's down to the somewhat ramshackle track of the depot arrival line. While the rear of the train was still over the crossover, the HST was also partly on a curve. Which was moving doesn't really matter in the impact, nor do brakes; it's all down to mass distribution.

While the HST power car is roughly twice as heavy as a Mk 3 (70 vs 36 t), the 801's driving car is not so much lighter (48 t, if first class was at the other end, otherwise 52 t). Not having engines, the intermediate coaches are fairly light, but still heavier than Mk 3s (ca. 40 t, I'd guess). It's light vehicles needing to convey a large impact force that are at risk of jumping, and it's not obvious whether that applies here.

Railway safety has been improved in several places, of which train protection and structural integrity of vehicle body shells are two major ones. Keeping trains on the track, and upright even off it, are important too, at least for low and medium speed collisions (at high speeds there is no overall "solution"). More rigid couplings help in that, though articulation (shared/Jacobs bogies) is the best. that's why this looks alarming.

The 26 m coaches of the 801s do have a longer overhang at the ends than 23 m stock. That does increase the leverage of end force (whatever its direction) at the bogies, so I wonder if that is relevant. Also, the gap between coaches is particularly long - does that mean the coupling is longer between its fixings into the body at each end?

I expect the RAIB will look into everything.

I wouldn't be surprised if they do!

9 car Azuma approx 431 tons, 2+9 HST approx 410 tons

I'm sure they will. It's an opportunity (and unplanned and unwelcome one, but one which must be taken) to look at the reaction in collision of an IET.  I'm sure it's all be simulated and worked through in theory but there's noting like the practical look too.   That look in addition to "how did they come to collide" in the first place!

This incident, along with the buffer-stop collision at Long Rock Depot, makes me think that these IEPs have some unexpected behaviours - or should that be vices - when it comes to braking.

Report summary from https://www.gov.uk/government/news/report-132020-collision-and-derailment-at-neville-hill


Full report (55 pages) at https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/935941/R132020_201118_Neville_Hill.pdf

That was an interesting read. I didn't see any convincing theory at the time as to why this train derailed so badly after a low-speed collision, so this was always a matter of waiting for the RAIB to tell us.

The basic dynamics of a train that has to stop very fast indeed, because it's hit something, is fairly simple. Each carriage body goes down at the front, and up at the back, and the high forces in the couplers then add to that effect and can lift the rear bogies off the track. It only takes a random initial yaw angle of any of the carriages to mean the same forces push those rear bogies sideways too. But that doesn't usually happen in low-speed collisions - but did this time - why?

What I hadn't realised was that the couplers have to collapse, and in most trains are designed to do so sacrificially (i.e. to protect the body shells). Hence in this case the RAIB identified that the failure of the couplers to collapse (shorten) was a primary cause of the derailment, and a design error. Exactly what should happen when the gap between the carriages has reduced to zero was not made clear (because it wasn't part of the accident). Presumably they should lock in some way, if only by being scrunched into each other.

The lengths of the body and the coupler also were not as significant as I imagined, though the gap between the carriage ends being so large did allow a lot more movement up and sideways than otherwise.

The cause of the accident was operational, though Hitachi's software played a part in the driver's errors, but how the train behaved is all down to Hitachi's design. And that does look embarrassing for them as a major train maker. While the standards and requirements may not have told them to worry about low-speed collisions, they should have know it was needed. The first technical director I had to present projects and proposals to would always say, if you told him your equipment was designed to meet the requirements, "but will it work?" This catchphrase became a bit of a standing joke, but it's a very good question.

So Hitachi have some serious thinking to do about how they ended up here, and what they can do about it (if anything). On top of that, the single thing I imagine is causing the biggest concern in Hitachi Towers (legal division) is this:


Some of those changes to the model are not related to the derailment, but one aspect of its oversimplification was:


Of course there are those (Stadler, Alstom, etc.) who will be quick to point out that even moderately high-speed trains ought to be using Jacobs Bogies in any case. That keeps the carriage bodies firmly joined to each other, though it must remove the scope for absorbing crash energy outside those body shells.

So the answer was right, but the wrong question was asked? The apparent damage caused at such a low speed does serve to demonstrate to the layman the difference between speed and momentum, and the wisdom of having crumple zones. If I have correctly understood the full import of the parts of the report that you have drawn attention to, Stuving, will there have to be a much more detailed examination of the design of the IET, or even a reconsideration of how safe it is?

I just wonder if the energy of the impact was below the give way point? There must be an engineering balance between being durable / reliable for every day duty and the point at which it needs to protect against impact.

Even at 15mph quite a bit of energy to redirect for a train that must be close on 300T for a 9 car hitting a stationary object

Whilst the slow speed derailment should not have happened I think a more significant part of the summary is:

"The collision occurred because the driver of the Intercity Express Train was focused on reinstating an on-board system which he had recently isolated, instead of focusing on the driving task. This was exacerbated by him unintentionally commanding too much acceleration due to his lack of familiarity with the train.

The driver had isolated the on-board system at Leeds station because he had been unable to correctly set up the train management system. He had been unable to do this because ambiguous documentation from Hitachi, the train manufacturer, had led to LNER misunderstanding the required process for setting up the train management system when developing the content of its driver training programme.

The driver?s lack of adequate familiarity with the train probably arose because LNER had not recognised that his training needs were greater than for his peers."

The first paragraph seems to go against all principles of railway safety that a driver should be able to command a train to STOP (even if not in time) rather than abitarily automatically accelerate.

As a friend said do drivers have to computer "Geeks"

Bit like on the GWR versions it was possible to when changing Pantograph ends to raise and hence the train VCB close while the other end was in the process of closing the earth switch on the lowering pantograph at the other end of the train, the trains have a 25kV bus from one end of a unit to the other.  couple of occasions of fire works in Paddington Stn

How do you get that form those words? The driver thought he was commanding a very low tractive effort, and he would have time to re-enable APCO defore getting to the HST. Actually the control was further forward than that, giving 20% of full power, so by the time he looked up he was very close to it. He did pull the controller right back to command the emergency braking, but the collision happened before any significant braking could happen.

Indeed, design for crashworthiness is a complicated business - so I'm afraid you'll have to read the report! That has a lot on the subject, but is still a simplified version to support their enquiry, rather than a design guide.

And I can recommend more reading, too. Hitachi have published several articles on the UK A-trains, which mention crashworthiness:
Railway-vehicle Technologies for European Railways (https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKEwj7k5XF4pHtAhWQecAKHbLnBrQQFjAAegQIBRAC&url=https%3A%2F%2Fwww.hitachi.com%2Frev%2Fpdf%2F2008%2Fr2008_01_010.pdf&usg=AOvVaw1MrpMuhd_orTB6XhMv2gEP)
Development and Maintenance of Class 395 High-speed Train for UK High Speed 1 (https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwiwhKi34pHtAhU7QkEAHbTYBWwQFjAAegQIBBAC&url=https%3A%2F%2Fwww.hitachi.com%2Frev%2Fpdf%2F2010%2Fr2010_01_109.pdf&usg=AOvVaw1eDnJ4gs6BRtghKBvs73cN)
Development of Class 800/801 High-speed Rolling Stock for UK Intercity Express Programme (https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKEwjMudSH4ZHtAhVKPcAKHfz8A3UQFjAAegQIBRAC&url=https%3A%2F%2Fwww.hitachi.com%2Frev%2Fpdf%2F2014%2Fr2014_10_105.pdf&usg=AOvVaw2jdGsC1_nl4Tg434oobCkN)

Between them they have several pictures of the front end of the body shell and its "crush me" sections - which are surprisingly small: about the size of traditional buffers. I've added a few of those pictures below.

One point they make several times is this:

I can't imagine what cultural factors have led the Japanese (people, government, and railways) to think that doesn't matter.

Roger Ford has an intersting article in Jan's Modern Railways on the interaction of the driver and the Train Management system. Basically he pressed the wrong button trying to set the correct headcode for the move. and the train partly because it wasn't mantioned in traning or hanbooks.

He also didn't realise with power setting he made the train would reach 14 mph whilst an HST would reach 7 in the time.