Sydney Skytrain (Sydney Metro North West) - Concrete spans buckle and crack

[Takahama Trainwatcher]

For anyone interested in engineering failures, the linked article may entertain. I don't think I would like to stand under any of these spans.

 

http://www.smh.com.au/nsw/email-trail-reveals-highlevel-warnings-about-skytrain-safety-threat-20170111-gtpc6u.html

 

In a nutshell, there are concerns over the construction methods of the elevated part of the line. The government authority in charge of roads is not allowing the builders to build over the roads until they prove it will be done safely. The builders are under pressure to avoid this delay because they will be penalised for not meeting deadlines. Competing demands?

[Ochanomizu]

Hello,

 

Oh, so, if I understand it correctly, one government department is delaying the contractor so another government department can penalize the contractor financially ... right?  Good to know Australia is corruption free !!!

[katoftw]

No TT.  One department wants their project to move forward regardless of concerns raised or previous failings.  The other is asking questions and blocking construction unless safeguards can be put in place to protect its customers.

 

Personally I see no issue with the roads department blocking construction.  With chips and falling concrete everywhere.  It is only a matter of time before a real accident happens.  I wouldn't want concrete hitting my windscreen and myself turning into oncoming traffic.  big injuries happen with two cars travelling in opposite directions at 60kph hit each other headlight to headlight.

Edited January 19, 2017 by katoftw

[kvp]

I think it's more of a quality issue during assembly than a design problem. The buckled spans were assembled without proper dilatation gaps required for compensating temperature changes. The only design fault is using an elevated track system that allows improper positioning during assembly. Imho the proper way to proceed would be to remove the damaged sections, check the already built sections for similar, but unsurfaced errors and replace them and then keep building.

[velotrain]

How do you know the improper positioning happened during assembly and not design?

[kvp]

If the design was faulty, then the first two and all subsequent spans would have buckled too and not just some of the ones installed lately.

[velotrain]

How do you know the order of installation / failure for the spans?  

 

The design doesn't have to be universally faulty.

 

Poor quality materials, workmanship, and/or insufficient curing of the concrete are also possibly involved.

 

In any event, it looks like a true project from hell that will come in way late and well over budget, and may not even be safe then.  It also sounds like there has been an inadequate program for testing and QC from the start.

 

 

Regarding "installed lately", the cracking has been reported for at least one full year - quote from the article linked in the initial post:

 

 

"An Italian consortium, Salini Impregilo, won the $340 million bid to build the viaduct. In 2014 the then prime minister of Italy, Matteo Renzi, even appeared at a ceremony in Kellyville to inspect early works.

 

But the documents obtained by the Herald show that, less than a year later, RMS started to seek assurances from Transport for NSW that cracking in the overhead spans would not place motorists at risk.

 

In November 2015 RMS's manager on the project, Anthony McMahon, warned in an email that "the critical path for the entire project is at risk" if a squabble between the Italian contractors and Transport for NSW was not resolved.

 

The Herald then revealed in January last year that two of the viaduct's spans had suffered cracking.

 

By May 2016, concrete slurry had started to fall from a section of the viaduct over Memorial Drive in Rouse Hill. The slurry fell onto four cars after one span suffered cracking."

 

 

 

[kvp]

I've checked the photos and it had the cracked sections at the end of the construction on the last two installed sections. It seems like they kept on building without going back and fixing what was broken. That's actually really bad...

[velotrain]

It seems they're stuck in the first stage of grief . . . .

 

 

"without going back and fixing what was broken. That's actually really bad... "

 

What's even more bad is not trying to analyze the cause.

 

 

I'm also wondering if perhaps there is insufficient steel structure to support the concrete, which could cause cracking.

 

I can foresee a lot of time in the courts.

[velotrain]

An article from October has some additional info.

 

I had assumed a structural steel frame under the concrete, but apparently that's not the case.

 

"The spans between concrete piers have an average length of 40 metres and are held in place by steel cords that run though the segments.

 

Sources say the proof engineer for the viaduct had refused more than a year ago to sign off on the design because it was concerned stresses within the spans were too great.

 

The sources also blamed the cracking of the spans on quality and construction issues such as the use of poor concrete.

 

Australian engineering standards for a viaduct and cable bridge of the complexities of the Skytrain project are less stringent than in the US or Europe, where there are limits on how much stress can be placed."

 

http://www.smh.com.au/nsw/span-cracks-buckling-risks-causing-major-delay-to-skytrain-for-northwest-metro-line-20161018-gs5g50.html

 

 

This article has some specific info on the buckling problem.

 

http://www.smh.com.au/nsw/investigation-reveals-disappointing-failures-in-skytrain-construction-20161222-gtggmt.html

[cteno4]

cable stressed concrete design requires high quality very strong concrete or you get collapse like this with the stressed cables buckling the structure.

 

jeff

[velotrain]

Based on the second linked article, the root cause was a pseudo ad lib join between the two sections involved - it was the one case where both sections were not cast while adjacent to the other.

 

Jeff - do you know just how the sections are supported until the cables are attached?

 

I found a photo of the inside of the tube, and take it that the numbers are the order of tensioning - about halfway down.

 

http://www.dailytelegraph.com.au/newslocal/rouse-hill-times/sydney-metro-north-west-rail-link-to-fast-track-replacement-for-damaged-skytrain-span/news-story/31da8cbac1ead05bc3ed8fda366157bf

[cteno4]

Charles,

 

When they do cable stressed bridges they usually have a traveling framework that hangs off the end of the growing end that lifts the next section and its bolted on and suspension cables added, repeat. Then cable tension thru the structure.

 

Jeff

[kvp]

So it was an improperly fitted expansion joint. The problem of this suspension bridge technique is that the cables are the only thing that holds up each section of each span. It's like using several sections of Kato elevated tracks, then suspending the chain on two piers only and hoping the whole things hold up if we push the two ends strong enough.

 

The proper way would be to add a slight arch to this construction or to use form locking pieces like the Tomix elevated tracks or just cast each section as a monolithic block like everyone else.

 

For the damaged join, the correct way to handle it would have been to put each end on its own pillar by using a single cast double pillar and leave an expansion gap.

 

Now they can't go back and recast everything as they have to remove the pieces until a good one is found and then recast everything from there. An alternative would be adding two double pillars, one to each ends and running a new batch of sections for the fix. This must be done each time one section needs to be replaced, which is imho insane. A traditional loose fitting solid span would be drop in replacable without such hacks.

Edited January 21, 2017 by kvp

[velotrain]

So it was an improperly fitted expansion joint.

 

 

No - you seem hung-up on expansion joints, but they don't appear to be involved in this situation.

 

Did you read this article, which explains what happened?

 

http://www.smh.com.au/nsw/investigation-reveals-disappointing-failures-in-skytrain-construction-20161222-gtggmt.html

 

 

Have you ever created a two-part plaster mold?  You first pour the plaster to make a mold of half of the pattern / master.  You then use this - covered with a resist layer, and the pattern / master to make the second half of the mold.  This creates a very tight seam between the two halves of the mold - this is quite similar to the process by which almost all of the segment joins were created, save the failing one.

 

" To achieve a secure join between the concrete segments, the majority of segments were cast against those that they adjoin."

 

Using this method, friction between the flush-fitting ends of two adjacent sections prevents slippage as the cables are tightened. 

 

"But where that has not occurred, the segments are stitched together with concrete and clamping.  Mr Staples, said the method used to "stitch join" the two segments that buckled "had not been done anywhere else on the program.  The report commissioned by Transport for NSW, conducted by engineering consultant Aurecon, showed the failure occurred due to insufficient clamping, the lack of strength of the concrete used, and an "inappropriate" process of injecting epoxy to fill the gaps in the concrete."

 

Due to the lack of concrete friction, and the relative plasticity of the filler materials, when the cables were tightened, the two spans buckled at the point of least resistance.  Somehow the phrase "stitch join" doesn't seem like an acceptable process in the precise field of enginering ;-)

 

On the surface, there seem to be multiple methods of correcting this, so we'll need to see what they come up with.  I haven't seen any info on just why this joint was treated differently in the first place, and it does seem like a possible "what were they thinking" (or not) situation.

 

"the lack of strength of the concrete used" gives cause for concern regarding everything that has been built so far.  Adding to this, "Australian engineering standards for a viaduct and cable bridge of the complexities of the Skytrain project are less stringent than in the US or Europe."

[kvp]

This was not a matched join, but a plain one as it seems they restarted the casting process. This speeds up the construction as you could potentially start from two places. Another reason could be that they used the wrong element to cast to. Anyway, using epoxy as a filler was a bad idea. This mismatched join could have been installed without problems, by leaving a small gap, adding spacers (and bracing) and supporting both elements on a dedicated pillar. The process above leaves a small gap, turning the continous bridge into two smaller bridges, but without recasting anything. Now that this was missed, the builders actually face two mismatched joins at the two ends of the damaged section. Leaving gaps and adding extra support underneath makes it easier and faster to fix the problem.

[velotrain]

I just had a thought.  After they get the concrete situation figured out, but before they start laying any track, it would be good to do a stress test on one or more of the Skytrain spans - preferably not one over a road ;-)

 

This would entail determining the total weight of the track and control/signal equipment, and however many loaded railcars would fill a typical 40 meter long span - to simulate two trains passing.  The weight could either be averaged over the span length, or averaged for the track, etc. and calculated and divided for/by the bogie positions on the cars. 

 

Of course, this would only be a static test, without the additional vibration stress from moving trains.  However, if there are potential problems, it would be far better to discover them before the track is laid and all the equipment installed - and trains test run with (non-operating) crews.

 

I'm hopeful that the master plan for the project already includes something like this, and if not the current situation might inspire them to add it.  I also found a most interesting and insightful quote in one of the other SMH articles referenced in the initial link posted by TT.

 

The Greens transport spokeswoman, Mehreen Faruqi, who has a doctorate in engineering, said problems in the project were "hallmarks of an infrastructure program where the public sector has been hollowed out of engineering expertise, and is unable to effectively oversee and supervise project delivery".

 

My thought of a static load test was inspired by a famous similar test done by the great architect Frank Lloyd Wright - at the behest of local building inspectors, when the strength of his radical column design for the Johnson Wax Building was questioned.  This structure was built over 75 years ago, but still has a modern look even today - not to mention the adjacent tower completed in 1950.

 

Image - http://evstudio.com/wp-content/uploads/2015/11/Loading-of-Johnson-Wax-Columns.jpg

 

Text - https://en.wikipedia.org/wiki/Johnson_Wax_Headquarters

 

See the Construction section

[medusa]

The Greens transport spokeswoman, Mehreen Faruqi, who has a doctorate in engineering, said problems in the project were "hallmarks of an infrastructure program where the public sector has been hollowed out of engineering expertise, and is unable to effectively oversee and supervise project delivery".

 

Hmm that's like it is here. On construction of a new subway line in Cologne a sinkhole opened which swallowed the entire city archive building and took two lifes (the construction workers fortunately noticed early that something is really going bad and had enough time to warn most of the people in the surrounding houses to run away).

Supervision of the project was done by the construction company itself. Also due to the fact that the city government itself was unable to effectively oversee and supervise due to staff cutdown.

[velotrain]

http://www.spiegel.de/international/germany/cologne-archive-catastrophe-were-subway-builders-cautious-enough-a-612129.html

[medusa]

Yes that's the report about it.

 

Worse is, they should have known. Just 5 years before about 200m from the sinkhole there were massive problems with a church tower from the same subway construction. Link is German but the photos tell the story.

https://www.technodoctor.de/baptist.htm

Edited January 23, 2017 by medusa