Paris CDG Roof Collapse (merged)

[radarcontrol]

Part of the roof at terminal 2E has collapsed killing six people.

http://news.bbc.co.uk/1/hi/world/europe/3739715.stm

sympathies to those involved.

[-<M4v3r1ck>-]

http://news.bbc.co.uk/1/hi/world/europe/3739715.stm

[car_owner]

I express my sympathy to the victims, the wounded and their family.

I cannot understand how this accident can happen in France TODAY.

[Dr Dave]

Picture at:

http://story.news.yahoo.com/news?tmpl=story&u=/040523/481/cdg10505231003

Looks absolutely horrific.

Dr Dave

[warkman]

I regulary use BHX-CDG with BA and that uses 2B. From the pictures, it looks like the part that collapsed is similar in design to the departure area gates at 2B, ceiling with curved glass from ceiling to floor, next to airbridge to plane or to the bus for the smaller aircraft.
Under the curved glass is seats and the air conditioning vebts next to the window, with further seats and the security scanners under the actual ceiling. Three police have died, which might have been the guards at the scanners.

Terrible.

[IRRenewal]

Dutch radio reports 6 have died, including a number of police oficers who were in the process of sealing of the area after a crack was discovered in the roof.

[alexmcfire]

Reminds me of the Transvaal accident in Moscow.
Swedish media claim that one Chinese and one from Ivory Coast where among the injured, may the unfortunate ones who lost their life today, rest in peace...:(

[seacue]

A great and tragic surprise.

I've thought that the French, above all, were masters of concrete construction since they do so much of it.

[Vick Van Guard]

I am sitting in my office looking out of the window at it, there is a large amount of damage to terminal 2E, one whole section has collapsed.

It’s occurred where the walkway from check in connects to the departures area. Part of the walkway has collapsed as I can see it hanging down at an angle and that appears to have taken the whole section of the departures area out with it.

So far our flights (from 2F) have not been affected, but if you are travelling allow extra time as you are faced with a long walk to the terminal as they have closed off the road to 2F & E.

Very sad day.

[Wino]

No earthquake and this time of year obviously no snow, what about rain water accumulating on the roof?

It defies belief that a roof could simply collapse of its own in this day and age. How about sabotage?

Sympathies for thei victims and their familes
Wino

[hart744]

My better half arrived back to CDG this morning from her AF flight and was informed by their flight coordinator of the accident. Other than that, she was not able to find any information. It was frustrating for her and she wanted to know if any of her colleagues were involved in the accident or not. Aeroport de Paris was not providing update information. The weather at the time the accident happened was fine and did not cause the collapse. There was no rain and no wind.

Video from France television http://videos.tf1.fr/video/news/lesjt/

[seacue]

Wino, you live on the west side of the Atlantic. We've had our share of concrete collapses during construction, though I don't know of a collapse in a new structure.

Some days I walk down the hall of the Building Research Division at NIST - they investigate construction failures- and see their gallery of photos of crumpled and twisted buildings, bridges, etc. They do happen here.

One of the press stories on the CDG collapse mentions "security" inspectors delaying the opening of the building. That is probably a mistranslation and should read "safety".

[WHBM]

Could be a whole range of structural reasons for this, but should be straightforwardly determined by building engineers on the spot.

Looks like the sort of collapse you get when prestressed or post-tensioned concrete fails (you put the reinforcing steel under stress, then pour the concrete over it, then let go. You can do lighter structures this way).

Also looks like trucks can be driven underneath. Any likelihood of them striking the support columns or the underside of the structure ?

How long has the structure been up (see it opened a year ago). With concrete construction there's always the chance of corrosion getting into the reinforcing steel inside the concrete during the construction stage.

Was it a particularly hot day in Paris yesterday?

[Wino]

There have been a few concrete collapses during construction over here as well (though not to recently, last one was in altantic city, a parking garage I think)

But not buildings that were competed as far as I know.

Strange strange....


Wino

[brockenspectre]

RIP all souls lost at CDG today.

Let the experts find the reason for the building failure and let other buildings of similar design and fabric have its weakness identified rapidly before something like this happens again in France or elsewhere in the world.

Were there any questions about the build? the contractors? doubts expressed?

:suspect:

[Kalium Chloride]

Were there any questions about the build? the contractors? doubts expressed?


Dunno. But the main contractor seems to have rapidly removed the pretty CDG terminal pics which were on its website earlier today.

[Evening Star]

With concrete construction there's always the chance of corrosion getting into the reinforcing steel inside the concrete during the construction stage.

It would be unusual for corrosion formed at the construction phase to cause a collapse within a year, although a lot would depend upon the detail.

News reports are vague as to technical details (what a surprise!). Was the structure reinforcement insitu, precast, pre or post stress? The stressed stuff allows for a very light weight structure but as a result will have less redundancy for errors in design or construction. Whichever type, such a collapse, like the tragedies more commonly reported here, would require a chain of errors and mistakes. Normal advice about waiting for investigation report applies.

[Dublinflyer]

PARIS AIRPORT EVACUATED


Charles De Gaulle airport in Paris has been evacuated after cracking noises were heard in the roof - a day after its partially collapse, killing four people.

The noises were heard on either side of the 30-metre section of futuristic terminal 2E which gave way on Sunday morning.

[Eboy]

Excerpt from New York Times article:

The concrete shell, built by the French construction firms Eiffel and Laubeuf, was constructed in rings "with three interlocking elements, positioned one next to the other," the news release said. To simulate the pressure on an underground tunnel that keeps the concrete tube intact, the news release explained, steel hoops were attached to the outside and carbon fiber reinforcements were glued onto the shell.


http://www.nytimes.com/2004/05/24/international/europe/24pari.html
(registration required)

[alexmcfire]

Same company was building a major building in Beijing for the Olympics, but now the building site is halted.

[Airbubba]

Looks like the rest of the terminal may be razed...

_____________________________________________

May 24, 2004

Cracking Sounds Heard at Paris Airport
By THE ASSOCIATED PRESS

Filed at 12:32 p.m. ET

ROISSY, France (AP) -- A day after a deadly roof collapse, new cracking sounds were heard Monday inside a terminal at Paris' Charles de Gaulle airport, and dozens of workers there were evacuated.

The scare came as officials said they were considering scrapping the entire Air France Terminal 2E after a 98-foot section of the steel, concrete and glass roof caved in at about 7 a.m. Sunday, killing four people. The terminal is closed indefinitely.

More than 30 people in airport offices and a nearby restaurant were evacuated Monday when the new cracking sounds were heard, airport director Rene Brun said. Airport staff will not be allowed inside until an analysis is completed, Brun said.

Cracking and puffs of dust minutes before Sunday's accident had proved a warning sign.

The area evacuated Monday is a separate structure from the tube-shaped boarding and waiting area where the roof partially collapsed. It is joined by small walkways. A third structure in the complex is reserved for checking in.

``We feared that there would be another problem,'' said Corinne Bokobza, spokeswoman for the airport authority, known as ADP. ``We're taking no risks.''

The $890 million terminal was touted as a jewel of design, safety and comfort when it opened 11 months ago after several delays that the French TV network LCI said were due to safety issues.

Officials initially believed five were killed in the collapse, but lowered the death toll after searching through rubble. Two of the dead were Chinese travelers on a stopover at Charles de Gaulle on route from Shanghai to Mexico via Paris, the Chinese government said.

Three people were slightly injured -- all police officers or security personnel called to the scene when cracking appeared minutes before the collapse of a portion of the terminal's roof. They began evacuating people from the area minutes before the accident, perhaps sparing lives.

The collapse has been an embarrassment for French authorities.

The building, the pride of the airport, could be razed if investigations prove it cannot be made safe, the head of the airport authority said.

``If all these rings that make up this terminal are irrecoverable, we'll tear it all down, of course,'' Pierre Graff was quoted in daily Le Parisien daily as saying, referring to steel rings that form a spine in the building. ``We will take no risks when it comes to security.''

The futuristic building is a tube-like structure of concrete, steel and glass, hundreds of feet long, that stands on pylons above the ground. The ceiling is honeycombed with hundreds of square windows that flood it with light. With slots for 17 aircraft, the terminal was designed to handle 10 million passengers a year.

The collapse involved a 100-foot section of the roof, preceded by several loud cracking sounds. As the roof fell onto a waiting area, it pulled down walls and crashed through a boarding ramp onto several parked cars below.

``We are very touched by the cruelty of this collapse,'' Prime Minister Jean-Pierre Raffarin said as he examined the site. Investigations ``will bring out the truth,'' he added.

Two investigations, administrative and judicial, were underway, and experts from some of the 400 companies that participated in the terminal's construction went to the scene.

One hypothesis being studied is whether the ground underneath the terminal had settled, LCI television reported.

Paul Andreu, the terminal's architect, was returning to Paris from Beijing, where he is working on a project to build the National Theater of China, said an assistant, Patricia Casse.

``I am completely stunned,'' Andreu was quoted as saying in Monday's edition of the French Communist daily L'Humanite. ``I cannot explain what happened. I don't understand.''

Francois Charrita, assistant director of operations at ADP, told France-Info radio, that ``without a doubt'' the investigations would focus on potential problems with construction.

A fault in design would, on the other hand, be unlikely because of multiple controls by experts, including independent ones, he said.

[BahrainLad]

In the same way as some doctors carry malpractice insurance, do architects/contractors carry the same?

Even if, would such a policy cover (gulp) €750 million??

[FOZ]

If you look closely at the pictures so far, it appears that the concrete elipses are still generally intact but now at ground level. It also appears that the concrete joist that ran between the pillars supporting the structure is no longer present. Could this be a catastrophic failure of that prestressed concrete joist?

[Flightmapping]

Yes, architects would carry insurance, as would all the other contractors. I have to admit to being shocked when I first heard about this.

Building collapses caused by structural defects, as opposed to earthquakes, fire or acts of terrorism are extremely rare.

I was just telling someone about Ronan Point, which was perhaps the last serious structural failure in the UK? (AFAIK)

Of course, lessons will be learnt from this tragic occurence, but to put it in perspective I was always taught (at Architecture school) that to make staying indoors as risky as going outside, you'd have to slow all traffic down to 3mph.

So, those unfortunate souls who were trapped by this collapse were indeed very, very unlucky.

[ATC Watcher]

Very sad indeed . As a regular ( and frequent ) user of Terminal 2, I have used the walkway many times and never thought about its possible collapse .
According one news speculator in French TV, the terminal was evacuated just prior the collapse, thus limited the number od casualties...who were mostly chinese apparently.
According the same news speculator , The pillars that support the structure are n likely to be the cause, as some of them were showing cracks , and were reinforced by carbon rings , during the construction .

Condoleances to the famillies.

[OFBSLF]

n the same way as some doctors carry malpractice insurance, do architects/contractors carry the same?Yes, architects, contractors, and engineers carry errors and omissions insurance. Their contracts often limit try to limit their liability, however. Here in the colonies, contract language typically limits their liability to their fee. Whether or not those limits hold in court is well beyond my expertise.

While press are talking about the architect, it is the engineering and construction that are more likely to be the issue. The structural and geotechnical engineers are responsible for ensuring that the structure will stand up (provided it is constructed as designed and with proper materials). The job of the engineers is often made more difficult due to the architect's whims, but it is still their job to ensure that the design is sound, no matter what monstrosity the architect sold to the client.

The cause of the failure could have been:

- bad structural design
- bad foundation design
- faulty construction (i.e., not built to the design)
- faulty materials

We'll have to wait for the investigation to determine just what the problem was.

OFBSLF (used to be a Geotechnical Engineer in a previous life)

[pdalla]

I have spent many an hour gazing out of the east (maybe SE) facing window of the AF FF lounge in the bowels of 2E looking at the construction work happening adjacent to the new terminal (no perverse pleasure here, just long transits due to missed connections).

Thought crosses my mind whether all that work digging away at nearby ground, and if I recall correctly some of it was concrete apron, affected the new structure. Bit like causing a local earthquake maybe. I have a friend here in ATL whose house was similarly affected by adjacent building work on a neighbour's foundations, causing damage to his house as if a major earthquake had rocked his house. He was lucky though, his house was mostly built of wood and it only got a little bent out of shape. Concrete however would not be so forgiving.. Didn't ADP also build a tunnel round there somewhere ?

[Evening Star]

I have a friend here in ATL whose house was similarly affected by adjacent building work on a neighbour's foundations, causing damage to his house

Back in the days when ES was a student surveyor learning all aspects of the science, one regular task was to carry out schedules of condition on buildings adjacent to proposed building work. Very boring work that was, but one became quite an expert at spotting even the smallest crack. Quite common in the UK, and if anything does move it provides evidence for any claim to reinstate the damaged property.

In the same way as some doctors carry malpractice insurance, do architects/contractors carry the same?

Even if, would such a policy cover (gulp) €750 million??

To reiterate the answer about insurance, yes. It is compulsory requirement by my professional organisation and bloody expensive it is as well. After the mortgage, it is the second most expensive outgoing in the ES family budget, and that is for my modest needs in the relatively safe area of hydrology.

As for how much cover, it is possible that the organisations involved in projects of this size will have that level of cover. In any case, any claim will be spread between the various parties (architect, engineer, contractor), so any individual organisation will not have to carry the full 750m.

Ronan Point, which was perhaps the last serious structural failure in the UK? (AFAIK)

There have been a few serious failures since then, although all that I can recall were the result of a fire or explosion rather than a simple structural failure. However, even Ronan Point was not a simple building collapse, but a gas explosion exposing substandard design and construction.

Point is that these things rarely have a single cause and it is worth waiting for the investigation.

[airhead10]

Architects, Engineers and Contractors have PI Insurance, (Professional Indemnity). Usually this is a statutory requirement built into their contract. this insurance will, (to certain limits) cover the cost of the damage arising from professional error/incompetence, however I can tell you from experience that the excess will likely be in the millions so whoever is ultimately found responsible will face a hefty bill.
Incidentally, it is the Architect, not the contractor, who is involved with the Olympic Stadium in Beijing.
Very tragic event, thoughts go out to all affected, some of them were from Hong Kong, so big news here.
:sad:

[Notso Fantastic]

Looking at the pictures, some of the concrete pillars are still standing. The whole roof section appears to have come down in one piece. From what I could see, the roof section, in side profile fairly circular concrete, is attached on either side of the walkway at a point below the widest horizontal diameter. It appears to have 'slipped' off the walkway and come crashing down. The pillars look substantial- I think the failure was at the attachment between the roof and the walkway. or the joist along the walkway between the pillars. Someone certainly got their stressing horrendously wrong- without wind or snow it overstressed.

[max6462]

I flew to dubai about 3 weeks ago.
There is a long terminal in a sausage shape that looks very similar to 2E in Paris.
Was it similary designed?
They look alike to me , though I don t know much about architecture .

[411A]

Many years ago the old Kenelworth Hotel (Arthur Godfrey used to broadcast from here in the winter) in Miami was demolished, to make way for a newer structure.
Engineers got quite a surprise when they examined the concrete and steel used.
The rebar was far too small and beach sand was extensively used in the concrete mix.
One mentioned at the time that if a force 5 hurricane had happened by, the whole building would likely have collapsed like a house of cards.

Godfrey had quite a few exploits as a pilot, perhaps the most notable was his buzzing of the control tower, with his DC3, in New York. His flying license was suspended for six months after the incident, and when it was restored, he sold the DC3 and bought a Convair 340...later on a Gulfstream one, with what was absolutely new to private flying at the time...a Collins PVD.

[BahrainLad]

I flew to dubai about 3 weeks ago.
There is a long terminal in a sausage shape that looks very similar to 2E in Paris.
Was it similary designed?
They look alike to me , though I don t know much about architecture .

They look similar, but Dubai is glass and steel whereas 2E used a lot more concrete. And I mean a lot more.....

[Lemurian]

Max6462,Bahrainlad,
Dubai airport and terminal 2E share the same design,the same architect (Andreu),the same design office (ADP).But not the same contractors -I think_.
In fact they're quite interchangeable.

[MarkD]

Obviously this is a tragedy for the families of those involved but the AP is reporting here that 2E represents 15pc of CDG capacity. Will CDG be able to accommodate all traffic in summer without 2E?

[pdalla]

Lemurian - FYI Dubai's terminal was built by Bechtel not ADP

[BahrainLad]

....that's what I thought, with ADP involved in the new, still under-construction, Terminal 3 and Concourse 2/3.

[Lemurian]

pdalla,
ADP 's role stops at the blue printing workshop.
The construction is the contractors'affair.
That's what I said.
P Andreu and the ADP design office have done the R&D work for more than 40 terminals worldwide.

[DeepC]

ADP 's role stops at the blue printing workshop.

Not if you have anything like CDM (Construction, Design and Management) regulations in France.

The designer has a duty of care to design for safe construction, use and maintenance of any structure he designs.

DeepC

[Lemurian]

Deep C,
Agree.
What I meant was that the design is ADP.the construction itself (the realisation) is awarded to contractors.
It didn't mean the designer had no responsibility whatsoever in the safety of the building,and I suspect a lot of questions are going to be asked about the design itself,the structure computations,the monitoring of the designers' specifications....the role of independant surveys during construction etc...etc.

[Stress Relief]

Yes my first post, please don't maul me!

Just to respond to Deep C, the CDM regs do empower designers to produce designs which can be constructed in a safe and expeditious manner, and to make the Contractor aware of any sequencing of operations that need to be followed in order to ensure the short term stability of the structure. They are in place in an attempt to reduce accidents during the construction phase, and as you say to ensure to that the building can be safely maintained.

The designers on a project can have little or even no involvement during construction, this being generally dependant on their appointment with the client. Some clients like to have their interests looked after during construction, but many don't want to pay for this service. In the UK , Design and Build contracts, are totally managed by Contractors, and many do not want a supervising presence on site.
It's along way removed from the times when Resident Engineers and Clerks of Works would be continually looking over the Contractors shoulder, to ensure that the quality of work was good, and that the designs were being correctly interpreted on site.

It'll be intersting to see what the Engineering press make of this collapse,later this week.



SR

[RatherBeFlying]

If Notso Fantastic's observation proves out, it won't be the first time that an attachment failure initiated a structural collapse.

The WTC collapse on 9/11 was put to attachment failures as well as the bridge collapse over the Hyatt lobby.

The stresses on the primary structures are generally well understood, but attachments are critical and an apparently minor change by the builder may have major consequences.

[OFBSLF]

well as the bridge collapse over the Hyatt lobbyThe Hyatt walkway collapse was not primarily caused by the design of the attachments. Yes, the attachments were poorly designed. But the primary cause was that the walkway as-built was significantly different than the proposed design.

http://www.glendale-h.schools.nsw.edu.au/faculty_pages/ind_arts_web/bridgeweb/Hyatt_page.htm

The key difference was that in the original design, there was one continuous support rod supporting both the fourth-floor and second-floor walkways. The beam for the fourth-floor walkway would only have been supporting the fourth-floor walkway itself -- the weight of the second-floor walkway would be supported by the rod.

When it was built, there were two rods -- one from the roof to the fourth-floor walkway, and a second from the fourth-floor walkway to the second-floor walkway. While this looks quite similar to the original design, it is not -- as built, the fourth-floor walkway support beam is now supporting both the weight of the fourth-floor walkway and the weight of the second-floor walkway. In other words, the fourth-floor walkway support beam has now had its load DOUBLED, but the size was not changed.

Yes, the attachment design also sucked. But the main cause of the failure was the doubling of the load on the fourth-floor walkway support beam.

[RatherBeFlying]

OBSLF, I can agree that it would be more precise to say that the attachment point failed in the Hyatt case.

[arcniz]

This unfortunate collapse does put one thing in perspective:

With 30,000 years experience, more or less, in putting up buildings, occasionally things still go wrong in unexpected ways.

Yet, with a mere 100 years of aviation history behind our modern turbine aircraft and the critical maintenance technology that keeps them safe, they show an incredibly tiny incidence of problems that actually result in injury to occupants.

And... for an added bonus, they move!

The airplane will get you to the destination considerably faster than will the airport terminal building.....In certain cases the airplane is perhaps also a safer place to be.

[DeepC]

Stress Relief,

I agree with everything you say regarding CDM. We are 'supervising' a D&B contract at the moment and there is no approving of technical details but just checking that the Contractor jumps through the correct procedural and legal hoops.

My email above was specifically referring to the quote at the top of the email from Lemurian.

Perhaps a PM to me if you want to continue to this discussion off the main board.

Cheers

DeepC

[OFBSLF]

OBSLF, I can agree that it would be more precise to say that the attachment point failed in the Hyatt case.Ratherbeflying:

I'm sorry, but I can't agree. When designing a beam (in this case, the 4th floor walkway support beam), you first determine the load on the beam, and then size it. Someone (and it is unclear who, whether fabricator, contractor, or engineer) changed the design to double the load on the beam, but did not change the size of the beam. Even if the attachment was better designed, the 4th floor beam would have failed because it was now undersized. The load on the 4th floor beam was too much for that size beam.

The problem with the attachment design was that it was prone to a ripping failure. The beam was made of two c-sections, welded top and bottom. The hole for the support rod was centered in the cross-section -- right through the weld. So the entire stress was applied at the weld. They should have reinforced the beam on the top and bottom with a flat plate at the attachment point. That way the stress from the nut holding the support rod would have been spread out over the surface of the beam and not concentrated at the weld. The plate would have prevented a ripping failure.

But as I stated above, even if the attachment point was properly designed, the walkway would have failed. The beam was 1/2 the required strength for the load. The reinforcement plates that I described above would have prevented a ripping failure but would not have prevented a bending failure -- the beam would simply have bent until it failed.

The problem was not just the design of the attachment point. The cross-section of the beam was insufficient throughout the length of the beam, not just at the attachment point.

[Stress Relief]

Ok, well the word on the street from our esteemed Professional rag, is that is a" failure of the column head supporting the vault roof or in the bolts connecting silicone bearings to the vault edge and column heads."

Have to say that the diagram they published shows this to be a rather unusual detail with the column head eccentrically loaded.

It is not thought that there was any failure of the concrete vault structure itself, which has generally retained it's shape.

In 2002 , all 170 no. columns were stiffened with carbon fibre bandages wrapped around the column heads, to relieve bursting stresses after cracks were discovered, soon after the roof was positioned. Instigated by ADP, who fulfilled both Architectural and Structural Engineering roles, it seems regretably that it wasn't enough.

[RatherBeFlying]

OBSLF, It sounds like your structural engineering expertise far eclipses mine -- all I have is a bit of physics.

That said, reinforcement of a poor attachment point might have produced a slower failure mode such that the deficiency might have been discovered before lives were lost.

[mini]

OBSLF,

Someone changed the design... ???

Beam size is not directly related to strength, you can double the load and yet retain the original beam size. The composite of the beam determines its yield strength, not its size.

C sections properly welded would not have had a stress raiser at the weld. If this was a fail point then the weld was at fault, not the design.

You state the beam was 1/2 the required strength. very strong allegation, any proof?

In a nutshell, Bullsh*t

[OFBSLF]

OBSLF, It sounds like your structural engineering expertise far eclipses mine -- all I have is a bit of physics.RatherBeFlying: I'm not a structural engineer, but I did take a number of structural engineering classes way back when. I was a geotechnical engineer (BS, MS, and ME in Civil Engineering).

That said, it looks like I misremembered. The problem was indeed the connection, not the beam cross-section itself. I was correct that the design change doubled the load that the 4th-floor beam carried. It also doubled the load on the nut supporting the 4th floor beam. It turns out that the attachment design at the nut was already 30% under-sized. The doubling of the load sealed its fate. See:

http://www.engineering.com/content/ContentDisplay?contentId=41009035
http://www.eng.uab.edu/cee/reu_nsf99/hyatt.htm#Causes
Originally, the 2nd and 4th floor walkways were to be suspended from the same rod (as shown in fig-1) and held in place by nuts. The preliminary design sketches contained a note specifying a strength of 413 MPa for the hanger rods which was omitted on the final structural drawings. Following the general notes in the absence of a specification on the drawing, the contractor used hanger rods with only 248 MPa of strength. This original design, however, was highly impractical because it called for a nut 6.1 meters up the hanger rod and did not use sleeve nuts. The contractor modified this detail to use 2 hanger rods instead of one (as shown in fig-2) and the engineer approved the design change without checking it. This design change doubled the stress exerted on the nut under the fourth floor beam. Now this nut supported the weight of 2 walkways instead of just one (Roddis, 1993). Analysis of these two details revealed that the original design of the rod hanger connection would have supported 90 kN, only 60% of the 151 kN required by the Kansas City building code. Even if the details had not been modified the rod hanger connection would have violated building standards. As-built, however, the connection only supported 30% of the minimum load which explains why the walkways collapsed well below maximum load (Feld and Carper, 1997).And this too:
The hanger rod detail actually used in the construction of the second and fourth floor walkways is a departure from the detail shown on the contract drawings. In the original arrangement each hanger rod was to be continuous from the second floor walkway to the hanger rod bracket attached to the atrium roof framing. The design load to be transferred to each hanger rod at the second floor walkway would have been 20.3 kips (90 kN). An essentially identical load would have been transferred to each hanger rod at the fourth floor walkway. Thus the design load acting on the upper portion of a continuous hanger rod would have been twice that acting on the lower portion, but the required design load for the box beam hanger rod connections would have been the same for both walkways (20.3 kips (90 kN)).

The hanger rod configuration actually used consisted of two hanger rods: the fourth floor to ceiling hanger rod segment as originally detailed on the second to fourth floor segment which was offset 4 in. (102 mm) inward along the axis of the box beam. With this modification the design load to be transferred by each second floor box beam-hanger rod connection was unchanged, as were the loads in the upper and lower hanger rod segments. However, the load to be transferred from the fourth floor box beam to the upper hanger rod under this arrangement was essentially doubled, thus compounding an already critical condition. The design load for a fourth floor box beam-hanger rod connection would be 40.7 kips (181 kN) for this configuration. ...

Had this change in hanger rod detail not been made, the ultimate capacity of the box beam-hanger rod connection still would have been far short of that expected of a connection designed in accordance with the Kansas City Building Code, which is based on the AISC Specification. In terms of ultimate load capacity of the connection, the minimum value should have been 1.67 times 20.3, or 33.9 kips (151 kN). Based on test results the mean ultimate capacity of a single-rod connection is approximately 20.5 kips (91 kN), depending on the weld area. Thus the ultimate capacity actually available using the original connection detail would have been approximately 60% of that expected of a connection designed in accordance with AISC Specifications.

Mini, just to be clear, RatherBeFlying and I were discussing the Hyatt walkway collapse in 1981 in Kansas City, Missouri. I don't know the cause of the Paris terminal collapse and can only speculate that it was caused by a problem or problems in the design, construction, and/or materials. We'll have to wait for the official investigation. Someone changed the design... ???Yup, the as built did not match the original design drawings. See figures 1 and 2 here:
http://www.eng.uab.edu/cee/reu_nsf99/hyatt.htm#Causes
Over here http://www.engineering.com/content/ContentDisplay?contentId=41009035 you will find the following text:
------------
"The hanger rod detail actually used in the construction of the second and fourth floor walkways is a departure from the detail shown on the contract drawings....During the 26-week administrative law trial that ensued, G.C.E. representatives denied ever receiving the call about the design change. Yet, Gillum affixed his seal of approval to the revised engineering design drawings.

Results of the hearing concluded that G.C.E., in preparation of their structural detail drawings, depicting the box beam hanger rod connection for the Hyatt atrium walkways, failed to conform to acceptable engineering practice. [This is based] upon evidence of a number of mistakes, errors, omissions and inadequacies contained on this section detail itself and of [G.C.E.'s] alleged failure to conform to the accepted custom and practice of engineering for proper communication of the engineer's design intent." Evidence showed that neither due care during the design phase, nor appropriate investigations following the atrium roof collapse were undertaken by G.C.E. In addition, G.C.E. was found responsible for the change from a one-rod to a two-rod system. Further, it was found that even if Havens failed to review the shop drawings or to specifically note the box beam hanger rod connections, the engineers were still responsible for the final check. Evidence showed that G.C.E. engineers did not "spot check" the connection or the atrium roof collapse, and that they placed too much reliance on Havens.

Due to evidence supplied at the Hearings, a number of principals involved lost their engineering licenses, a number of firms went bankrupt, and many expensive legal suits were settled out of court. In November, 1984, Duncan, Gillum, and G.C.E. International, Inc. were found guilty of gross negligence, misconduct, and unprofessional conduct in the practice of engineering. Subsequently, Duncan and Gillum lost their licenses to practice engineering in the State of Missouri (and later, Texas), and G.C.E. had its certificate of authority as an engineering firm revoked."
-------------------
I did not remember who made the design change. This report says it was G.C.E.

It is clear why the design change was made -- the original design was impractical to build. To build the original design, the 30' long hangar rods would have to be installed. Then the 4th floor support beams would have to threaded 20' up the hangar rod to the fourth floor. Major pain the rear. So the idea to change the design to two rods made sense -- they just should have redesigned the connection at the time, taking into account that the new design doubled the load on the fourth floor support nut.
Beam size is not directly related to strength, you can double the load and yet retain the original beam size. The composite of the beam determines its yield strength, not its size.Sorry for not being more accurate in my wording. When I said "sizing" the beam, that to me means choosing the cross-section. You are correct that there are typically several different standard cross-sections with the same extreme dimensions -- the heavier cross-sections carry more load. Back when I did design of determinate structures, we would typically determine the load on the member, then open up our AISC handbook and select the appropriate cross-section from among the standard shapes.
You state the beam was 1/2 the required strength. very strong allegation, any proof?I was wrong, it wasn't the beam itself, but rather the connection. Surf the links that I've provided. I think that's all the proof that you should require to agree that the design change doubled the load on the 4th floor connection and the 4th floor connection was not redesigned to support that load.
In a nutshell, Bullsh*tAnd a pleasant day to you, too.

[mini]

OBSLF

Firstly apologies, I assumed your post was relating to the CDG incident. My error for not reading the thread.

In light of the above, apologies also for the Bullsh*t comment.

mini (with tail between his legs...)

[maxalt]

Just been up the dome of St.Peters in Rome.

Built 1540 (approx)
Still standing.
Amazing.

[OFBSLF]

Mini: Thanks. Your apology is warmly accepted.

Built 1540 (approx)
Still standing.
Amazing.While we all rightly marvel at structures built long ago, we should also realize that the cr*ppy structures built a long time ago have either collapsed or been pulled down. The only ones left are the exceptional ones.

[Daysleeper]

just to drag this back to the point, what are the implications for terminal 2F which looks identical on the outside.

[mini]

OBSLF,

for the record...

You're a gentleman

[Vick Van Guard]

just to drag this back to the point, what are the implications for terminal 2F which looks identical on the outside.

The check in areas of 2E and 2F are the same, but the departure areas (the bit that collapsed) are different, so there should be no impications for 2F.

I transit though 2F quite a bit and the scale and design of the roof always amazes me!

[Airbubba]

Paris Airport Probe Publishes Findings

By THE ASSOCIATED PRESS

Published: July 6, 2004

Filed at 9:04 a.m. ET

PARIS (AP) -- The Paris airport collapse that killed four people in May was probably caused by metal struts piercing the building's concrete vaulted roof, an official report said Tuesday.

The Transport Ministry said preliminary findings suggested that the ``dominant cause'' of the partial collapse of terminal 2E was ``linked to the perforation of the vault by the struts'' and the use of poor quality concrete.

Four travelers -- two Chinese, one Czech and one Lebanese -- died in the May 23 disaster. Three people were injured when the roof gave way showering the terminal with falling glass, steel and masonry.

``It's probable that this perforation was made possible by the prior gradual deterioration of the concrete,'' the ministry's statement said.

Still to be explained is the reason the roof gave way exactly when it did, the ministry said.

The ministry, citing the findings, brushed aside speculation that the accident was caused by ground movement or problems during construction, which led to delays. Pylons holding up the section of the three-building terminal that collapsed had to be reinforced after cracks appeared.

An inquiry commission is using modeling techniques and laboratory analysis of the debris to establish the cause of the disaster at the futuristic terminal designed by architect Paul Andreu, 63, who is working on an opera house in Beijing, China.

Its final report, for which no date has yet been set, will determine whether the $920 million building will be repaired or torn down.

The terminal, which opened to great fanfare on June 25, 2003, was a key element in the development strategies of the airport's operator, Aeroports de Paris, and its main customer, Air France.

A judicial investigation into the accident also is proceeding, with experts trying to detect possible errors in the structure's conception or any failure in the building techniques.

http://www.nytimes.com/aponline/international/AP-France-Airport-Collapse.html

[Bre901]

According to Le Monde (http://www.lemonde.fr/web/article/0,1-0@2-3234,36-397322,0.html), the investigation report should be published in the coming days. Is is very severe for ADP and the architects, who are expected to be indicted very soon. The collapse is said to come from poor design and unsufficient concrete reinforcement as the the roof has been perforated by the pilars.

[edited as new articles are available]

[None]

http://www.usatoday.com/travel/news/2005-02-07-paris-airport_x.htm

[Bre901]

The Beeb (http://news.bbc.co.uk/1/hi/world/europe/4266591.stm)

[Bre901]

Le Monde (http://www.lemonde.fr/web/article/0,1-0@2-3234,36-402102,0.html)