• Welcome to the Internet Infidels Discussion Board.

Hyatt Regency Hotel Disaster

 
 
 
I seem to remember reading about that one. If I recall correctly, they had the walkways hanging from the ceiling at two levels, and the structural design called for a single, continuous hangar rod, passing through both walkways with a bolt and bracket supporting each walkway at that point.

The builders installed two hangar rods, one supporting the upper walkway from the ceiling, and the second supporting the lower walkway from the upper walkway.

They reasoned that the total load on the hangar rods was, of course, the same in both cases, so (as the rods were rated for that load) this change didn't require structural engineer approval.

However they failed to consider the load on the bolts. In the original design, each bolt and bracket carried the load of the one walkway it supported; But in the new design, the bolt supporting the upper walkway now carried the full combined load from both walkways. Now under up to twice its design loading, it inevitably failed when fully loaded.

Which is why we don't let unqualified people do structural engineering assessments. This has become a classic example of how easy it is to overlook something critical, and is now used for training builders, engineers and achitects.

As an engineer friend of mine remarked, engineers need to be more careful than doctors, because a doctor who makes a mistake might kill somebody, but an engineer who makes a mistake might kill hundreds.

(I don't watch online video, so my apologies if that's not the same event you are talking about).
 
Last edited:
Or look at the Challenger. The fundamental problem is the engineers assumed the joint would always be in compression. Yeah, there's a whole bunch of weight pushing down on it, it's obviously in compression.

Except at the moment the main engines lit. Then the whole rocket twanged (expected), the joints got momentary periods of tension. And the solids were lit before the twang died away. And one cold winter morning an O-ring failed to save the day that nobody knew it had been saving for many previous launches.
 
bilby said:
I seem to remember reading about that one. If I recall correctly, they had the walkways hanging from the ceiling at two levels, and the structural design called for a single, continuous hangar rod, passing through both walkways with a bolt and bracket supporting each walkway at that point.

The builders installed two hangar rods, one supporting the upper walkway from the ceiling, and the second supporting the lower walkway from the upper walkway.

They reasoned that the total load on the hangar rods was, of course, the same in both cases, so (as the rods were rated for that load) this change didn't require structural engineer approval.

However they failed to consider the load on the bolts. In the original design, each bolt and bracket carried the load of the one walkway it supported; But in the new design, the bolt supporting the upper walkway now carried the full combined load from both walkways. Now under up to twice its design loading, it inevitably failed when fully loaded.

Which is why we don't let unqualified people do structural engineering assessments. This has become a classic example of how easy it is to overlook something critical, and is now used for training builders, engineers and achitects.

As an engineer friend of mine remarked, engineers need to be more careful than doctors, because a doctor who makes a mistake might kill somebody, but an engineer who makes a mistake might kill hundreds.

(I don't watch online video, so my apologies if that's not the same event you are talking about).
Click to expand...
I remember that incident. At the time I was working as a structural steel detailer, which entails translating engineers’ plans into detailed drawings of each piece of steel. For instance, every steel beam in a high-rise building is detailed, showing how many and what size bolts to use and where to drill the holes in the beam, so the building goes together like an erector set. The engineers’ plans give the load and stress for each connection and the size of the beams, and the detailer calculates (using a large book of tables which I assume is digitized these days) the size and number of the bolts etc. Of course all the “drawing” is now computerized as well. But in the days of the above incident it was all still done by hand.

When the above incident happened my boss told me it was a detailing mistake; the detailer should have known better. Sometimes the detailers catch engineers’ mistakes, but in any case the engineers are supposed to review all the detailers’ drawings and approve them. Whoever signed off on the details in the above case was the person who took the blame.
 
For those interested in a entertaining book on this general subject, I recommend Why Buildings Fall Down: How Structures Fail, by Levy and Salvadori. It's non-technical but I believe accurate, and as I said, quite entertaining.
 
bilby said:
I seem to remember reading about that one. If I recall correctly, they had the walkways hanging from the ceiling at two levels, and the structural design called for a single, continuous hangar rod, passing through both walkways with a bolt and bracket supporting each walkway at that point.

The builders installed two hangar rods, one supporting the upper walkway from the ceiling, and the second supporting the lower walkway from the upper walkway.

They reasoned that the total load on the hangar rods was, of course, the same in both cases, so (as the rods were rated for that load) this change didn't require structural engineer approval.

However they failed to consider the load on the bolts. In the original design, each bolt and bracket carried the load of the one walkway it supported; But in the new design, the bolt supporting the upper walkway now carried the full combined load from both walkways. Now under up to twice its design loading, it inevitably failed when fully loaded.
Click to expand...
Reminds me of the Westgate Bridge Collapse. There was a royal commission into the collapse. One recommendation was that the review of all calculations must be more rigorous and use some engineers from outside the main architects and builders.
My father was a mechanical engineer and he was asked to check the calculations on bolt stresses that many others had already checked. All had to agree.
For a couple of decades after my father (and many others) would be checking of calculations on a contract basis. Tedious and time consuming but so important.

My father and a friend were actually at the public viewing site when the bridge collapsed. They did not see it fall, but were knocked over by the air blast as the bridge hit the ground. The friend was a keen photographer and some of his photos of the bridge on the ground and the rescue work appeared all over the world.
bilby said:
As an engineer friend of mine remarked, engineers need to be more careful than doctors, because a doctor who makes a mistake might kill somebody, but an engineer who makes a mistake might kill hundreds.
Click to expand...
Indeed they do.
 
I do not know if this is true or not but I did receive an opinion from an engineer who pointed out that the beams that held up the crosswalks were C beams that were faced towards each other in an open C configuration. He said that the strongest Part of the C beam is the back vertical section and if the C beams had been turned so that the open part of the C was facing away from each other and the rods aligned vertically between them with a steel plate underneath it his opinion is that the crosswalks would not have failed.

I do not remember the engineer's name and I do not know if this is true or not. I find it to be an interesting opinion that was not included in any of the reports that I read.
 
Last edited by a moderator:
Coleman Smith said:
I do not know if this is true or not but I did receive an opinion from an engineer who pointed out that the beams that held up the crosswalks were sea beams that were faced towards each other in an open C configuration. He said that the strongest Part of the sea beam is the back vertical section and if the C beams had been turned so that the open part of the sea was facing away from each other and the rods aligned vertically between them with a steel plate underneath it his opinion is that the crosswalks would not have failed.

I do not remember the engineer's name and I do not know if this is true or not. I find it to be an interesting opinion that was not included in any of the reports that I read.
Click to expand...
Was this written using speech-to-text?

There is no such thing as a "sea beam", as you seem to be aware, because you get "C beam" correct once, but (bizarrely) "sea" is written where "C" is meant three times.

I am not an engineer, but my feeling is that if a C-beam is close enough to failure under vertical load for the orientation to matter when mounted with the web (the back of the C) vertical, the structure probably lacks an adequate safety margin.

Mounting the web flush against a supporting structure might help with twisting resistance, but if that kind of torsion is a concern, a C-beam may be the wrong choice altogether.
 
I was living in Hartford Ct when the Hartford Civic center roof collapsed.

Luckily it happened after a sports game ended.

Mu8ltiple causes. A new design concept tested by early computer simulation. It was built on the ground and housed up while pillars were built under it. Underestimated weights and now and ice loads. Ice and snow built up the roof leading up to the collapse. Drains froze.

It was not built to specs, I remember something about a teacher was hired by the city as a building inspector.

People were finding sheared heads of bolts and rivets on the floor. The roof was slipping of the supports.

I got a look at it from a nearby building.
 
bilby said:
Coleman Smith said:
I do not know if this is true or not but I did receive an opinion from an engineer who pointed out that the beams that held up the crosswalks were sea beams that were faced towards each other in an open C configuration. He said that the strongest Part of the sea beam is the back vertical section and if the C beams had been turned so that the open part of the sea was facing away from each other and the rods aligned vertically between them with a steel plate underneath it his opinion is that the crosswalks would not have failed.

I do not remember the engineer's name and I do not know if this is true or not. I find it to be an interesting opinion that was not included in any of the reports that I read.
Click to expand...
Was this written using speech-to-text?

There is no such thing as a "sea beam", as you seem to be aware, because you get "C beam" correct once, but (bizarrely) "sea" is written where "C" is meant three times.

I am not an engineer, but my feeling is that if a C-beam is close enough to failure under vertical load for the orientation to matter when mounted with the web (the back of the C) vertical, the structure probably lacks an adequate safety margin.

Mounting the web flush against a supporting structure might help with twisting resistance, but if that kind of torsion is a concern, a C-beam may be the wrong choice altogether.
Click to expand...
I watched C-beams glitter in the dark near the Tannhauser Gate. So I got that.
 
I had to look up C beam

www.engineeringtoolbox.com

American Standard Steel Channels: Weight, Dimensions & Properties

Dimensions and static parameters of American Standard Steel C Channels
www.engineeringtoolbox.com www.engineeringtoolbox.com

It is not a solid beam, it is a channel.


Structural channel - Wikipedia

en.wikipedia.org en.wikipedia.org

The structural channel, C-channel or parallel flange channel (PFC), is a type of (usually structural steel) beam, used primarily in building construction and civil engineering. Its cross section consists of a wide "web", usually but not always oriented vertically, and two "flanges" at the top and bottom of the web, only sticking out on one side of the web. It is distinguished from I-beam or H-beam or W-beam type steel cross sections in that those have flanges on both sides of the web.[1]

The structural channel is not used as much in construction as symmetrical beams, in part because its bending axis is not centered on the width of the flanges. If a load is applied equally across its top, the beam will tend to twist away from the web. This may not be a weak point or problem for a particular design, but is a factor to be considered.[2]

Channels or C-beams are often used where the flat, back side of the web can be mounted to another flat surface for maximum contact area. They are also sometimes welded together back-to-back to form a non-standard I-beam.
Click to expand...
 
My father was a professor of Civil Engineering and taught a Structures class for 30 years. After the Hyatt disaster, he used it in his class as an object lesson about designing unique structures. There was no handbook of standards for suspended cat walks. The designer had to rely on the requirements for bridge building. The original design accounted for the maximum possible load, plus a safety factor. The problem was, it would be difficult to assemble and install, which is to say, expensive. The design change was made to make construction easier, and in other words, cheaper. If they had consulted the original designer, the crash could have been avoided. The point of the lesson was not only to design a structure that is safe, but to consider how it's going to assembled in the field.
In my neighborhood, they recently finished construction of a new library. The third floor extends out over the front of the building. Designing a cantilever structure is a compromise between weight and strength. Heavy is stronger, heavy requires more strength. As the interior was being finished, the cantilever began to sag. Everything was to spec and no welds failed. The steel simply couldn't hold the load and bent. No one got hurt and after a year of finger pointing, modifications were made and the library is open today. Engineering is serious business.
 
Just because a structure is strong enough to take a load without failing, that does not mean it won't deflect under load to an undesirable - even an alarming - degree.

Engineers sometimes forget this. People won't like or trust a steeply sloping floor in a cantelevered structure, even if all the calculations say that it will not fail catastrophically.

Another common instance of this error (designing to avoid failure, without considering deflection under load) is "bouncy" floors, where the flooring is strong enough to support the load between the joists, but sags noticably when loaded, and springs back when the load is removed.

And of course, resonance and damping are also often overlooked or miscalculated, sometimes with catastrophic failure as a consequence.

I strongly recommend J E Gordon's excellent works The New Science of Strong Materials, or, Why You Don't Fall Through the Floor (1975); and Structures, or, Why Things Don't Fall Down (1981) as an intelligent layman's guide to structural engineering basics.
 
Back in the 70s I knew a draftsman who worked at a structural engineering company. He talked about how serious they were. 24/7

There are engineering mistakes, but we only hear about the big ones.

In the 80s a large pipe was being installed in a building to connect with another pipe in another building. When it came time to connect a pipe between the two buildings it did not line up.

In the 80s I was in North Idaho. At a ceremony for a new trash transfer station in Coeur d'Alene when the first truck backed up to a door the door was too small.
 
As an aside, the largest property damage claim that I handled in my career was a structural failure of a different type. It involves a pipeline that extended from Tulsa to the Gulf of Mexico, pardon me Trump not the Gulf of America, that became corroded and leaked petroleum products.

The leaks did not extend the entire length but were substantial enough that the cost of remediation including replacing large sections of the pipeline and removing contaminated soil was in the millions of dollars.
 
You must log in or register to reply here.
Back
Top Bottom