Hitachi trains grounded

[Liam]

LNER have temporarily reintroduced a class 91 and Mk4 diagram. Also a significant 91 to bring back. 
 

 

[Reorte]

44 minutes ago, uax6 said:

It would be easier than you think....

The worrying thing was that there was an incident where ESR’s disappeared from the signalling system during a data re-boot... and therefore the drivers didn’t get told.

That sounds... disturbing, to say the least.

[Reorte]

26 minutes ago, corneliuslundie said:

And if it is a physical network forget about sabotage, all it needs is a careless JCB driver. I must admit though that that American pipeline business had me wondering how safe such systems are.

 

Depends on how it's organised, whether some damage would cause only local disruption or wider. I suppose it's the balance between robustness to damage and the efficiency of putting all your eggs in one basket. The basket can be reinforced so there's also the choice between frequent minor disruption or occasional major disruption.

Edited May 11, 2021 by Reorte

[corneliuslundie]

Re the Cambrian ERTMS problem, there is a thread on here somewhere about it with the report.

Jonathan

[Coryton]

4 minutes ago, corneliuslundie said:

Re the Cambrian ERTMS problem, there is a thread on here somewhere about it with the report.

Jonathan

 

I think this is the RAIB report.

[Afroal05]

1 hour ago, DY444 said:

 

Any 125mph capable replacement without ATP can't run east of Reading on the Mains at 125mph so the reason may be as simple as that especially if they hope to maintain some semblance of a high speed timetable in the Thames Valley.  If however the GW IET fleet looks like it might be stopped for an extended period then it is possible that a 100/110 mph interim emergency timetable might be required to accommodate whatever non-ATP fitted traction they can find to supplement the 387s.  

 

The bottom line is we're all busking at the moment and until an engineering assessment is available of what is required for a return to service and thus a timescale then everything is going to be up in the air.  As I said before if this is going to be prolonged then all bets are off and we could see any one of a range of interim scenarios most of which would have seemed ridiculous a week or so ago.

 

The last advice was 4 days per unit and up to 12 months to return the full fleet.

 

With 387s and turbos currently providing the only GWR traction between Paddington and Reading there isn't currently a 125mph timetable that can be run.

 

With 3 TPE 802s now anticipated to arrive this week they won't be going beyond Reading until there is some official ruling or guidance that supersedes the current rules that I have to adhere to!

[Gibbo675]

On 10/05/2021 at 10:26, APOLLO said:

Fabricate some big thick steel plates  and steel castings for awkward shapes, all bolted to the aluminium body at multiple points to spread the load / stress. Job sorted.

 

Jacking point failure point area I presume. Looks to be  Just not "beefy" enough (if as welded below).

Bogies swivel, move up and down, sometimes quite violently. look at the weld surface area second picture - not much. Yes this is not the jacking point of an 800 (see above), but the fixing area etc look the same.

 

 

 

Trains were once built like brick sh*t houses and lasted for ever structurally - now built to a cost, weight, etc.

 

Brit15

 

 

The above photographs shew just why the premium apprenticeship system worked, all modern stuff is now scoped, specified. and designed by people that, through no fault of their own, think that sitting in an office and doing sums using a box of magic switches is engineering.

 

23 hours ago, black and decker boy said:

AFAIK The aluminium stir-friction-welding is not yet up & running at NA. The bodies for the First Group OAO services to Edinburgh were built abroad and shipped for final assembly as they have been pictured being unloaded and stored in an industrial estate awaiting the call to NA. Similarly, the bodies for Avanti fleet are arriving from Japan. It’s been published that Hitachi hope to have the new equipment installed & producing bodyshells for the last56 vehicles of the Avanti order see here

https://www.railinsider.co.uk/2020/07/04/first-bodyshells-for-avanti-west-coasts-new-trains-arrive-in-the-uk/

The welding technique used in the video of Newton Aycliffe was of the MAG process in which the filler material is used as the conductor for the welding current and is fed into the weld deposition area shrouded by a gas that contains both anti oxidising agents and/or alloying agents depending upon the grade of alloy used.

 

22 hours ago, billbedford said:

 

Don't they use ultrasonics to test for cracks?

Ultra sonic testing is used for crack detection although the precise nature of the weld preparation edges and weld penetrations along with component shapes will have to be studied before any reliable trace may be used to indicate a structurally sound weld. Ultra sonics will indicate areas of metal fatigue in that the trace will become rather diffuse and not send back reflections of the back side of the component due to micro cracking scattering the signal.

 

12 hours ago, MarkC said:

Correct re the dye pen, SS.

 

End drilling the cracks to stop the crack extending is indeed a 'quick fix' - it stops the end of the crack being a stress concentration - but you still have the problem of stripping the interior, disconnecting electrics etc before welding.

 

This is not going to end soon - or end well for some bank account(s)...

This course of action may not be possible due to the likely axis of the cracks and the orientation and access issues with regard the sections and shapes involved. Frilling holes at the end of cracks was also tried on the class 40 bogies and didn't work overly well with them either !

 

7 hours ago, david.hill64 said:

My recollection is that the first class 158's also suffered from early failures of the yaw damper brackets due to fatigue cracking. Weld repair of cracks in aluminium is a fruitless exercise. Aluminium, unlike steel, doesn't have a fatigue limit: every cycle of stress causes fatigue damage. In steel, below a certain level of stress, the material can withstand an infinite number of stress cycles. In aluminium you have to make sure that the fatigue stresses are so low that cracking cannot develop. Welding (almost) always introduces small defects that act as stress raisers. If you weld repair an area that is already cracking you will achieve at best a temporary solution.

 

IIRC the affected 158s had to be withdrawn, stripped down, the affected area cut out and a new, stronger, section welded in. A lot of computer modelling was done by BR Research to ensure the repair would be robust (my bridge partner at the time was the head of that section and used to regale me of the progress between hands of cards).

 

So I think that Hitachi is in for a very expensive time to do a redesign and replacement exercise on all affected fleets. I would guess that some clever people are going to do some complex stress analysis and fracture mechanics analysis to determine how big the cracks can be before they represent a real risk of failure. We may well find that an enhanced monitoring regime will enable affected units to return to traffic safely until modifications can be carried out.

 

 I was confused by the references to fatigue cracking of the jacking points. There shouldn't be any stress in these except when in use. Looking at the photos it seems that the jacking point and yaw damper attachment are in the same piece. For the reasons given by Jim Snowdon early in this thread there will be a lot of high frequency low amplitude forces transmitted from the bogie through the yaw damper. I strongly suspect these will be the issue. I can understand that the general public can understand a jacking point but will not have the foggiest idea about yaw damper fixings.

 

As to the question of whether the design should have been better, obviously it should have been. But I have a lot of sympathy with Hitachi. Under the current approval regime they must, by law, comply with the EU Technical Standards for Interoperability and subordinate euronorms called up by these standards.  If the design complies with these standards (and it must), the assessment body is not allowed to question the design. If we go back to BR days enforcement of standards was subject to contract law and failure to comply could only result in worst case as a civil matter. Nowadays failure to comply is a criminal offence. If BR didn't like an aspect of the design then this would have been picked up in design scrutiny and a redesign mandated. Provided BR had a genuine reason the contractor would always comply. Whilst the current suite of euronorms provides a coherent and usually sufficient suite of standards I think we have lost that level of group knowledge that BR had and the industry is poorer for that. If the current suite of standards fails to account for Network Rail's crappy track quality then an issue is inevitable.

I would whole heartedly agree with your entire post only to add that the welding of aluminium is made even more difficult by way of contamination due to any dissolved salts carried into the cracked areas by rainwater washed mud into the cracks. No doubt one of the reasons the 158 had large sections cut out and replaced was to get to areas that had not suffered micro-cracking an such contamination.

 

 

What I would also add is that should anyone read a little more deeply into fracture mechanics you will find that alloys of aluminium are especially prone to brittle fracture. Even more so once they have been subject to work hardening stresses such as have the weld areas about which the great big lever that is the yaw damper bracket is attached to. The length of that bracket as a force multiplier is quite something as far a piece of poor design is concerned, as is the area to which it is attached which ought to have either a casting or better yet a forging of greater proportion welded into the body side to bolster the forces.

 

The method of building trains in the way the 800's have been constructed is to me inherently dangerous as the longitudinal welds offer a perfect "Pea Pod" stress raise in which the entire structure could easily unzip in a crash. Worse still the brittle fracture characteristics of aluminium as were shewn in the method of failure of the structure of the DVT involved in the Great Heck incident. From what I can see of the construction process in the various videos posted, the sections that are welded seem to have a single pass 12-15mm single vee butt weld holding them together, even with a similar weld on the back side of the extruded section, the cross section of weld material will be less than the section of the extrusions joined.

 

I would guess that the grade of aluminium used is 5083 as it is suitable for both welding and extrusion and does not require heat treatment after welding.

 

https://www.aalco.co.uk/datasheets/Aluminium-Alloy-5083-0-H111-Sheet-and-Plate_149.ashx

 

Gibbo.

[Peter Kazmierczak]

Was up in GW land this afternoon and saw two IETs in service on the Reading - Newport route; 800006 and 800009. So a few are still about.

 

[Mike_Walker]

6 minutes ago, Peter Kazmierczak said:

Was up in GW land this afternoon and saw two IETs in service on the Reading - Newport route; 800006 and 800009. So a few are still about.

 

You saw the entire operational fleet!

[phil-b259]

1 hour ago, uax6 said:

Ah yes, but it shows that the modern systems aren’t completely fool proof... 

And l hear that NR are looking to sell off its telecoms network again, which potentially adds in other risks to data transmission.

 

Andy G

Correction!

 

The Government are trying to FORCE Network Rail to flog off its Telecoms assets to generate some short term cash and satisfy the Free marketers within the Conservative party.

 

The same stunt was tried a couple of years ago when NR was told to look at selling off its major station portfolio relying on complex leases to actually have a railway running through them.

 

Yes the country has a gaping whole in the finances - but have we learn’t nothing from Privatisation? It would seem not so as from a Whitehall perspective....

Edited May 11, 2021 by phil-b259

[newbryford]

1 hour ago, Liam said:

LNER have temporarily reintroduced a class 91 and Mk4 diagram. Also a significant 91 to bring back. 
 

 

LNER were already planning to bring back some (4?) 91 sets for summer services - the 80x saga has just speeded it up a bit

[Peter Kazmierczak]

8 minutes ago, Mike_Walker said:

You saw the entire operational fleet!

Indeed Mike. That's what a Train Manager on another GW service told me when I said I'd seen those two running!

[5944]

[Coryton]

28 minutes ago, Gibbo675 said:

I would guess that the grade of aluminium used is 5083 as it is suitable for both welding and extrusion and does not require heat treatment after welding.

 

https://www.aalco.co.uk/datasheets/Aluminium-Alloy-5083-0-H111-Sheet-and-Plate_149.ashx

 

Gibbo.

 

I'm pretty sure I read somewhere that they are 7000 series. Which may or not be actually true.

[MarshLane]

 

5 hours ago, DY444 said:

 

That's defective ATP on stock fitted with ATP.  Different situation for stock which doesn't have ATP to start with.  For example 387s don't have ATP and they are permitted to do 110mph on the GWML.  Voyagers don't have ATP and they are permitted to do 125mph on the Up and Dn Main between Didcot East and Reading.  

 

As for what NR or anyone else will allow and what they won't then I think that depends on how long this situation persists.  I don't think it's entirely out of the question that this could go on for months given they have to figure out the cause, figure out a solution which will solve the problem permanently, and shop each unit to apply that solution.  If that turns out to be the case then plenty of out of the question right now options might come into play. 

 

I am sure that none ATP fitted stock is permitted at a lower speed - GBRf took 87002 into Paddington on their charter a couple of years back and that wasn't ATP fitted.

 

4 hours ago, Afroal05 said:

Fair point! The 57s on the sleeper don't either. However any rolling stock that is due imminently to bolster the fleet (and I'm not talking about the C2C 387s) isn't likely to go beyond Reading and the reason stated has been because it isn't ATP fitted.

I'm now trying to find what guidance I have for non ATP fitted stock on ATP fitted lines.

 

I am told the C2C 387s will basically drop into the GWR fleet as a complete set not to be split up - although I am sure somebody told me earlier today they need a software update for some reason before they can enter service.  I haven't confirmed that or asked why however.

 

3 hours ago, Flittersnoop said:

So these trains are actually in danger of disintegrating? I thought the cracks were only millimetres long? From your comment it sounds like they were stopped just in time!

 

No, this is purely external parts (lifting beam brackets) that may detach at speed. No danger of an 80x disintegrating and no effect or damage to the structural integrity of the units.

 

7 minutes ago, newbryford said:

LNER were already planning to bring back some (4?) 91 sets for summer services - the 80x saga has just speeded it up a bit

 

Yes. Should have been June - second 91 set expected to come back into traffic on Thursday.  Working off Leeds again.  1A04, 1N80, 1Y82, 1N83, 1Y86, 1N87, 1Y90, 1D29 is the diagram I have been given.

 

I am told that Hitachi has completed the risk assessment on returning services to traffic, it is now with the ORR / Safety bodies to assess the viability as I understand things.  If they say no, its back to the drawing board.  If they say yes, then we may have an option for units back in service later in the week, depending on what action needs to be taken first.

 

The suggestion about TPE units coming south, I am told is an option, but it sounds very much like it is a last ditch one for a whole host of reasons.

 

As ever, its not wise to assume or try and predict things.

[MarshLane]

8 minutes ago, Coryton said:

 

I'm pretty sure I read somewhere that they are 7000 series. Which may or not be actually true.

 

Yes the are 7000 series aluminium.

[Liam]

19 minutes ago, newbryford said:

LNER were already planning to bring back some (4?) 91 sets for summer services - the 80x saga has just speeded it up a bit

Wasn’t aware of that, but thanks for the heads up - will try and incorporate a ride on a set during my planned All Line Rover in the summer. 

[Liam]

6 minutes ago, MarshLane said:

Yes. Should have been June - second 91 set expected to come back into traffic on Thursday.  Working off Leeds again.  1A04, 1N80, 1Y82, 1N83, 1Y86, 1N87, 1Y90, 1D29 is the diagram I have been given.

Out of interest, do you know where they were before being ‘called up’? I’m guessing Ely? 

[ess1uk]

2 hours ago, uax6 said:

Ah yes, but it shows that the modern systems aren’t completely fool proof... 

And l hear that NR are looking to sell off its telecoms network again, which potentially adds in other risks to data transmission.

 

Andy G

Thread drift

not sell off wholesale

but get someone else to pay for greater bandwidth and let them use any spare the railway doesn’t use.

[MarshLane]

13 minutes ago, Liam said:

Out of interest, do you know where they were before being ‘called up’? I’m guessing Ely? 

 

The Mk4s are sat in Doncaster Belmont Yard, possibly a couple of sets at Neville Hill I think.  The set that came into traffic today was at Neville Hill as its been doing crew refresh runs once or twice a week for the past few weeks with 91110.  I think the 91s are at Neville Hill, but two or three are (or should be) in Wabtec Doncaster Works for G Exam - all 12 will go through this process before next January.

 

LNER are keeping 12 Class 91s in total and and seven Mk4 sets until summer 2023 (possibly 2024), which should have come back into service in June when the Kings Cross remodelling project was completed.  Bringing them back earlier has just helped the current situation.

Edited May 11, 2021 by MarshLane
Updated info

[Peter Kazmierczak]

GWRs current entire operational fleet of IETs in two pictures; taken within 16 mins of each other this afternoon...

[johnofwessex]

If we had the 'worse case scenario' does Hitachi have enough money to cope with it or could the company go bust?

[phil-b259]

2 minutes ago, johnofwessex said:

If we had the 'worse case scenario' does Hitachi have enough money to cope with it or could the company go bust?

 

Yup, its not exactly small.....

 

https://en.wikipedia.org/wiki/Hitachi

 

Allowing Hitachi to go under would be pretty much unthinkable to the Japanese.

[RJS1977]

Hitachi have assets of 9.93 trillion yen, I don't think the compensation and cost of remedial work is going to be of that magnitude....

 

(Though of course that figure is spread over any number of subsidiary companies like Hitachi Europe).

 

[D1072]

3 hours ago, Flittersnoop said:

Photo of detached bracket was of a different class of train, wasn't it? Not sure about the 28cm cracks, either. I just wondered how the risk of failure was affected by train speed - the stresses experienced at 125mph must be many times greater than at 70mph. The only train that can be 100% guaranteed not to shed bits as it goes along is one that is in a museum.

The source of the 285mm crack reference is post 1394 on the railforums "Entire 800 fleet stood down" thread from someone inside the industry; it states that is the longest crack found so far in the lifting pocket area.    What we don't know is if the cracks propagate in a direction which would mean material could fall off or not from here.  That message also said that the yaw damper bracket cracks were considered less serious.    I think the detached yaw bracket photo under a green train  has been identified as Irish stock, not GW, but any chance of material detaching will need to be assessed.