TGV derailed Paris - Strasburg (test train not in passenger service)

[jon0844]

If so, that's an episode of Seconds from Disaster that will be easy to produce.

 

[Crossover]

I did read earlier on BBC that speeds of in excess of 200mph may have been involved (217mph) - one hell of a speed to come off the track!

 

[Tim R-T-C]

Hopefully not going to turn out to be an Aeroflot Flight 593 type incident.

 

[MarkyT]

The curve measures at less than 1000m radius on Google Earth. There's no way that could be negotiated at the claimed speed of 300+kph and the site of the curve on the canal bridge clearly contributed to the severity of the consequences of derailment. With safety systems apparently not activated, there was nothing to intervene and control the speed correctly on approach in the event of driver error. Without the full protection of signalling and ATP, clearly a test train should only be carrying essential staff, not the families and children alleged to be on board.

A very sad and disturbing incident with echoes of Santiago de Compostela and a number of subsequent derailments of over-speeding passenger trains on curves in the USA. From the SNCF publicity point of view they were 'lucky' it occurred on the same day as the Paris terrorist atrocity, as news scrutiny is be focused elsewhere and the incident may not receive the same attention it would have at other times.

 

[Groningen]

I think that that curve is the slowest part of the new highspeedline. On Google Maps Streetview one can the start of the building the line from August 2008. 48.685558, 7.712084

 

[Murph]

I did read earlier on BBC that speeds of in excess of 200mph may have been involved (217mph) - one hell of a speed to come off the track!

I would take any speed being reported by the media at this early stage with a large bucket of salt! Journalists have a terrible habit of translating maximum possible speed and/or maximum theoretical speed into the actual speed of the incident. Sometimes it's a form of Chinese whispers, sometimes it's pure sloppy journalism, sometimes it's a lack of understanding and/or asking imprecise questions when interviewing a witness or expert.

 

[Zoidberg]

I would take any speed being reported by the media at this early stage with a large bucket of salt! Journalists have a terrible habit of translating maximum possible speed and/or maximum theoretical speed into the actual speed of the incident. Sometimes it's a form of Chinese whispers, sometimes it's pure sloppy journalism, sometimes it's a lack of understanding and/or asking imprecise questions when interviewing a witness or expert.

And, sometimes, such reports are accurate. Let's see what results from the investigation.

 

[Murph]

And, sometimes, it's accurate. Let's see what results from the investigation.

Yeah, occasionally the journalists do get it right. I'm not saying they did get it wrong, just that any numbers being thrown around at this early stage have to be taken with a good quantity of salt.

I find it rather hard to believe that the sort of railway professionals that would be involved with a test train would cause such a horrific incident through careless and extreme excess speed alone. That makes me doubt the numbers being reported in the media, and suspect that it may well be a more complex issue, or that it's unrelated to speed and instead something like a track/train defect/failure.

The investigation hopefully will reveal the true causal factors, and all conclusions thrown around by media prior to the results of the investigation should be salted to taste.

 

[Tim R-T-C]

That makes me doubt the numbers being reported in the media, and suspect that it may well be a more complex issue, or that it's unrelated to speed and instead something like a track/train defect/failure.

Given the colossal damage caused and the fact that the train managed to jump a canal and continue a few hundred metres up the other side of the water course, it is indisputable that speed is a major factor, even if the numbers doing the rounds at present are just conjecture - a train doing the required speed for such a curve could not have crashed in such fashion were the accident caused by a mechanical fault.

The only train failure that might have caused such an accident is the equivalent of a stuck throttle, I've seen it in motorsports accidents, but I'm not sure how possible it would be on a TGV. Presumably they have some sort of 'kill switch' that would cut power to the engines?

There is the possibility the driver was incapacitated due to sudden illness, but again
the trains are I presume fitted with dead man's handles or similar - unless these were not functional or deactivated for some reason.

I find it rather hard to believe that the sort of railway professionals that would be involved with a test train would cause such a horrific incident through careless and extreme excess speed alone.

Unfortunately it is not impossible - the Santiago de Compostela crash was afterall caused by driver error and excess speed. Particularly if there is the possibility that extra people were in the cab, distraction is all too easy.

 

[Gordon]

Bear in mind my earlier post that they were supposedly planning to test at up to 352 km/h. This equates to the 217 mph being quoted.


.

 

[notadriver]

Looks like it was coming to the end of the LGV where it meets the classic line. With no warning of where to start slowing down I fear the worst.

 

[Murph]

Given the colossal damage caused and the fact that the train managed to jump a canal and continue a few hundred metres up the other side of the water course, it is indisputable that speed is a major factor, even if the numbers doing the rounds at present are just conjecture - a train doing the required speed for such a curve could not have crashed in such fashion were the accident caused by a mechanical fault.

Well, look at the Greyrigg derailment at 95mph. You don't need 200mph to scatter a high speed train across the scenery. Also, look at the photograph of the power car in the canal — aside from being off the rails and in the canal, it's relatively upright and intact. Speed may well be a factor, but there's simply not enough solid evidence available to us to reliably conclude that it was the root cause, or even a causal factor.

If something about the train or track caused it to derail at safe line speed for the curve (or safe + 10% for testing), we could easily see pictures of the entire train scattered around chaotically.

Bear in mind my earlier post that they were supposedly planning to test at up to 352 km/h. This equates to the 217 mph being quoted.

Yes, and that's exactly the classic scenario where the media get it wrong. They take the published max speed for the test or service and assume that was the speed at the time of the incident.

 

[SpacePhoenix]

Presumably they have some sort of 'kill switch' that would cut power to the engines?

? TGVs don't have engines. They have an emergency stop button, which presumably applies the brakes at maximum, with perhaps the traction motors being placed in reverse or in full regenerative braking mode, and sand being applied.

Do all TGVs carry any sort of data logger/black box?

 

[4jochen]

I like to add some tech. ideas here:
a) it's a test train => it records much more and much faster a lot of sensor signal info compared to any Black-box, there or not. If it was the Train the investigators will know within a few days. There is enough recorded data.
b) the curve radius (at the bridge location) is 800m to 1000m by Google Maps measurements.
c) to pass such a curve with 350km/h means (some calculation here) => 10m/s or another 1 g horizontal centrifugal acceleration.
I don't know trains, and their rail physics, but in such a situation you can stand 45° inside the train. 1 g by Earth vertical and 1g horizontal. I never experienced that in real live.
=> I think a train will derail at 1g vertical.
d) with "only" 50m/s or 180km/h the centrifugal acceleration is "only" 2.5 m/s that's a quarter g. That feels all right to me.

Formulas here:
https://www.easycalculation.com/physics/classical-physics/centripetal-acceleration.php

e) the aerial pictures show scratches to the grass next to the track already very early. For me that's in the very beginning of that curve. Maybe the track has a problem, not the train.

I don't believe they go max. speed into that sharp curve.
The derailment happen too soon to be "only" excess speed.
Also at 180km/h the picture of deconstruction look quite same. At such high speed we can not picture that our selves. Don't try to compare accident pictures to estimate speed.
For me this could be an 180km/h accident.

 

[Emblematic]

Do all TGVs carry any sort of data logger/black box?

Yes they do, and it has been secured, mentioned in the press briefing. We also have a large number of witnesses onboard, presumably most with significant railway expertise, and possibly monitoring and recording equipment has survived with data intact. Plenty for investigators to work with.

 

[Tim R-T-C]

Well, look at the Greyrigg derailment at 95mph. You don't need 200mph to scatter a high speed train across the scenery. Also, look at the photograph of the power car in the canal — aside from being off the rails and in the canal, it's relatively upright and intact.

You do know that was the rear powercar?

The front car had the speed to completely vault that canal and run up the other side.

 

[edwin_m]

Also, look at the photograph of the power car in the canal — aside from being off the rails and in the canal, it's relatively upright and intact.

Pretty sure this is the rear power car. The rest of the train appears to have jumped the canal and ended up somewhere on the other side.

 

[jopsuk]

I like to add some tech. ideas here:
a) it's a test train => it records much more and much faster a lot of sensor signal info compared to any Black-box, there or not. If it was the Train the investigators will know within a few days. There is enough recorded data.
.

Depends if the data storage survived the crash. The standard recorder will be crash proof, the test instrumentation storage won't be

 

[LNW-GW Joint]

One of the SNCF messages says, if my French is accurate enough, that they "do not know the speed of the train before the black box is recovered".
https://twitter.com/SNCF_infopresse

 

[Groningen]

Wiki says that the train travelled eastbound. Entering a curve with 350 kilometers; that must be a hell of a curve. I am no expert, but think that at the most 150 is possible. The picture below is the "best" picture i could find. It is clear to see that most carriages are separated from each other and some way left of the bridge.

 

[LNW-GW Joint]

Wiki says that the train travelled eastbound. Entering a curve with 350 kilometers; that must be a hell of a curve. I am no expert, but think that at the most 150 is possible.

The design speed limit at the junction with the classic line is said to be 160km/h.
So it's quite likely the approaches from the LGV are the same speed.

 

[4jochen]

that 160 km/h is the range of speed I calculated in my post #74.
The Wiki just give us the max. speed (as many journalists) of the train for the tests, not at that part of the track - not at the time of accident.

 

[Redonian]

An up to date report albeit in French.
This confirms the line speed at the point of derailment as 176kph and says that the driver has survived and that he says he was travelling within the speed limit.

http://www.dna.fr/actualite/2015/11...mes-la-motrice-a-deraille-et-entraine-la-rame

 

[Murph]

An up to date report albeit in French.
This confirms the line speed at the point of derailment as 176kph and says that the driver has survived and that he says he was travelling within the speed limit.

http://www.dna.fr/actualite/2015/11...mes-la-motrice-a-deraille-et-entraine-la-rame

Interesting. So, while not yet conclusive, that seems to shine the spotlight on defect/failure in track or train.

176kph is approx 110mph. So, a little faster than Greyrigg, but possibly interesting to compare those two. Without digging through the history, I'd guess we may be able to find other UK incidents at around 110mph for additional comparison.

 

[Crisso]

Can somebody kindly provide a detailed track plan of the immediate area of phase 2 of the LGV Est, where the accident happened, including the connections to the Classic Route?

Looking at the photo supplied by 'Groningen' it appears that the train was proceeding eastwards on the right hand track of a single line connection.
(I do realise that right hand running applies in this region but am personally not sure where it starts and finishes.)

It also appears from the photo that the northerly connection line in the photo is in fact double track? Is that so or, is the photo slightly blurred?

Thanks for any clarifications in advance.

 

[Murph]

Can somebody kindly provide a detailed track plan of the immediate area of phase 2 of the LGV Est, where the accident happened, including the connections to the Classic Route?

Looking at the photo supplied by 'Groningen' it appears that the train was proceeding eastwards on the right hand track of a single line connection.
(I do realise that right hand running applies in this region but am personally not sure where it starts and finishes.)

It also appears from the photo that the northerly connection line in the photo is in fact double track? Is that so or, is the photo slightly blurred?

I'm less than 100% on this, but I've formed the impression that it's the end of a normal double track which splits and diverges into 2 singles for a flying junction with the old line it connects to, with the northern (westbound) of the 2 crossing above the old line before diving down to connect to it. The incident is on the southern (eastbound) track.

 

[richw]

An up to date report albeit in French.
This confirms the line speed at the point of derailment as 176kph and says that the driver has survived and that he says he was travelling within the speed limit.

http://www.dna.fr/actualite/2015/11...mes-la-motrice-a-deraille-et-entraine-la-rame

Does the black box back him up though?

 

[Zoidberg]

Does the black box back him up though?

The statement linked to finishes by saying that analysis of the black box should allow, more precisely, the cause of the incident to be known.

So, no-one can, yet, answer your question.

 

[LNW-GW Joint]

Can somebody kindly provide a detailed track plan of the immediate area of phase 2 of the LGV Est, where the accident happened, including the connections to the Classic Route?
Looking at the photo supplied by 'Groningen' it appears that the train was proceeding eastwards on the right hand track of a single line connection.
(I do realise that right hand running applies in this region but am personally not sure where it starts and finishes.)

Looking at the various maps, the LGV makes the classic connection not with the main line via Saverne, but with the branch towards Haguenau, just north of the junction between the two lines at Vendenheim.
There are then 4 lines towards Strasbourg.
The transition from left to right hand running seems to take place at this junction, with the Paris-bound line rising over both Haguenau and Saverne lines and heading over the canal as the southerly track.
Meanwhile the northerly Strasbourg-bound LGV line crosses the Saverne line on a bridge and then joins the Haguenau line at grade for right hand running towards Strasbourg.

This map which I posted further up, shows the LGV line in construction over the canal, and the classic junctions.
If you zoom in you can see bridge piers in construction, and even a bridge span waiting to go up at Vendenheim.
http://wikimapia.org/#lang=en&lat=48.681611&lon=7.715127&z=14&m=b

I think both tracks over the canal are single.
All routes will probably be bi-directional.
For anyone with the Schweers & Wall Germany map, the area (without the LGV) is shown on p92.

 

[33Hz]

Google has useful pictures looking back up the line at the crash site.

From the aerial photos it looks like the right-left crossover is after this point (at the bottom of this picture by the classic line), so the TGV was probably wrong-line running.

There appears to be catch rails going back to the start of the curve and interesting that there is road access up to the small building just before the curve.




https://www.google.co.uk/maps/@48.6...AAAAFv64/xoHNvonSglI/s203-k-no/!7i1536!8i1024

https://picasaweb.google.com/Luftbildarchiv.FlyFoto.DE/VendenheimElsaLuftbilder#6102642788003967106