Metro Dome collaspe

[NITTANYRAY]

If I remember correctly, this is not the first time they have had trouble with snow on the Metro Dome roof.
I am sure the building code in Minn requires roofs to support more than 17" of snow.
Did the owners get a special variance because the roof is supported with internal air pressure?

It appears that when the roof deflated the snow on the roof all went to the center and caused a tear in the fabric.
I hope when they repair the fabric that they reinforce it in the center in case it deflates again.

 

[MikeHalloran]

Maybe they should repair it with a Velcro strip so it will re-tear in a predictable manner.



Mike Halloran
Pembroke Pines, FL, USA

 

[concretemasonry]

The roof deflated after there was a fabric failure the allowed the deflation. The snow removal was stopped because of safety and weather conditions a few hours before the failure. Actual snowfall, which is misleading was somewhat more and the changing wind and drifting caused localized loading.

The roof structure is supposedly stronger when deflated, but does not provide the space between the layers for air and heat circulation. The sagged/failed roof surface and cables still support the lights, speakers, scoreboards and other loads even after the failure. - The question is how fast it can be repaired since the materials are available. Three large cranes (interior and exterior locations will be used for detailed inspection and for lifting the structure to allow remaining snow removal. The practical/economic life of the 30+ year old structure is questionable because of other teams and development, but the infrastructure is in place for something different. It was intended to be a multi-purpose building with some practical use disadvantages now that other new specialized stadiums (baseball and football) are now available.

There were also 3 other air supported structures that collapsed during to 17" to 20" snow and winds. - No injuries and no major property damage.

Dick

Engineer and international traveler interested in construction techniques, problems and proper design.

 

[NITTANYRAY]

According to ASCE 7 Minneapolis has a uniform ground snow load of 50 PSF. ASCE 7 also requires domes to be designed for drift loads. While the cables held the fabric is part of the roof structure and it should have been able to support the drifted snow. Either the fabric was worn out and should have been replaced or it was never designed properly.
The collapse should have never happened. It casts a black-eye on the design profession much like the Kansas City Hyatt collapse. When I hear that this was the fifth time the roof has deflated I can only shake my head. How many warnings do they need. It is fortunate that no one was injured.

It reminds me of an experience I had with some builders and building officials in Gulfport, Miss.
We had designed an elderly housing building for a 140 MPH wind load. The developer complained that the local code only required 120 MPH. The building official agreed saying that the 140 MPH would not take effect until 2006 since they were still using an old code locally. I did some research and found out that Hurricane Camille hit Gulfport in 1969 with winds up to 145 MPH. I explained that it would not be ethical for me to lower the design wind load because public safety was more important than saving a couple thousand dollars. For some reason the locals did not think a hurricane like Camille could happen again.
Had the Church sponsoring the building not run into financial problems the building would have been under construction in 2005.
Knowing what I know about the local attitude about building codes, I was not at all surprised when I saw the damage Katrina did in Gulfport and the surrounding areas.

The apartment building was one of the first new structures built after Katrina and it was built to meet ASCE 7 design criteria with no objection from the developer.

 

[hawkaz]

I need to disagree about this being a black eye on our community. This was a freak storm and I don't believe people criticize engineers for something like this. The bridge collapse in Minneapolis is a different story. It came down without warning or an obvious outside stimulus.

Every year a heavy snow storm hits somewhere, causing roofs to collapse. Should we increase snow loading to prevent this- Perhaps, but where do we draw the line. We could increase the snow load in Minneapolis to 100 psf, but is that realistic? We base our loading on statistics and a rational approach. There will always be anomalies.

We don't design for tornado level wind gusts. We could, but no one could afford to build our buildings anymore. The auto industry could design cars that are 100% safe, but no one could afford to buy one… Whether we like it or not, we need to draw a realistic line somewhere. 50 psf ground snow load in Minneapolis sounds reasonable. If something changes to cause this to be re-evaluated, then it will be done.

By the way, NITTANYRAY, I applaud you for sticking to your guns on the wind speed. As soon as is was deemed inappropriate, it probably should have been addressed in an emergency addendum for that locality rather than waiting for the next code cycle. This is how moment welds were addressed after Northridge.

 

[a2mfk]

Not exactly an exhaustive investigation yet, but this is interesting:

http://www.treehugger.com/files/2010/12/metrodome-roof-failure-symptom-of-larger-failure.php

"The membrane was weathering as anticipated and had exceeded its service life of 20 years; it recommended planning for replacement of the roof fabric, and noted that planning and implementation would take an additional five years and cost $12-$15 million."

Sounds like it comes down to economics at the end of the day, unfortunately, like everything else it seems...

Nittany- I wouldn't say it casts a black eye on the profession unless they do an extensive investigation and determine that it was a design error. Like you said, and so does this article, it looks like it is a maintenance-related failure of the fiberglass roof membrane material. This is a lot different than an original design error. A roof of a building you designed could collapse because the owner reroofed over the secondary scuppers and let leaves clog up the primary drains...

Like in your case study, which is very interesting (I had some unusual experiences in Biloxi post-Katrina with building construction attitudes), sometimes you have to out-fox the owners to keep them or future owners from hurting themselves. Sometimes designing to the letter of code maybe is not as prudent as realizing what that building may eventually be asked to do, regardless of the owner's original intent. Usually they are not wise enough to know that you may have cost them a few extra bucks for the greater good, and you have to fight them, which you wisely did.

When utilizing a material or system that is state of the art, there will be failures along the way, hopefully none that are catastrophic or due to negligence or poor design. There have been monumental dam, bridge, aerospace, and building failures that have advanced the state of the art by leaps and bounds (or made the systems obsolete or proved their lack of functionality).

I wouldn't say fabric roofs themselves are even the problem, I did a little research with the Denver airport roof and only found one failure in a huge snowstorm in 2006. It said there was a small tear in the roof that allowed some snow to fall in. Maybe tensioned fabric roofs are more dependable than those requiring constant internal pressure to support them.

Maybe the savings up front in the dome roof at the Metrodome proves to be worth these "inconveniences" every once-in-a-northern clipper. Knowing so little about how these roof systems work, almost as a layperson it seems that its a bad idea in a heavy snow area...

Also, I have not read if the air pressure is kept constant or do they increase it during snow events to balance out the snow loads?

 

[NITTANYRAY]

Hawkaz,

The average person doesn't know who is at fault when something collapses. All they know is that it collapsed.
The politicians that made the decision to put off the repairs are going to need scape goats and the designers are considered to be low hanging fruit.

When you talk to a layman about the WTC, they want to know why it collapsed. The real question should be how did it stand up so long because it was never designed for the impact of a 737 or a 2000+ degree fire.

People expect buildings to withstand extreme events because so many do that it has become the expected norm. When collapses do occur, they start to raise questions in the minds of the public which is not good for the profession.

Raising the snow load to 100 PSF is not the answer especially since the 17" snow fall is only about 20 PSF +/-
depending on the density of the snow. The winds were less than 90 MPH and the snow was less than 50 PSF and the code requires that drift loads be accounted for. The "freak storm" as you described it, was within the design requirements of the code.

It is possible that it was a maintenance problem, but who ever made the decision to not replace the roof fabric that had exceeded its useful life was just rolling the dice and they came up "snake-eyes".

As far as tornadoes, there are buildings designed for tornadoes. We designed a jail that could withstand a F-3 tornado at no extra cost to the owner. Because of security reasons, jails already have considerably more reinforcing than other buildings. With carefully detailing we were able to build a structure capable of surviving a mid level tornado.

 

[Qshake]

Regarding the codes and statistical analysis. You're correct in that the codes are based on this...and more.

But ours is a changing environment and we need to evaluate the code through statistics and other tools to see if the loads are, in fact, sound now as they once were. This works both ways as I'm sure there are certain loads that become obsolete.

So in effect I do beleive that should an analysis show that Minneapolis or some other large region need a 100 psf snow load then that's what it should be.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[structuresguy]

I have a hard time understanding how anyone can call this a freak snow storm. The amount of snow deposited was much less than the snow loads required by code. So this roof failed at less than code level loading. This is a major problem. This is also not the first time it has failed, far from it. This should be screaming "Bad Design" or more simply bad choice of structure (or lack thereof) for the given environment. There should not be any requirement for snow removal to keep the roof from collapsing.

 

[MikeHalloran]

The fabric roof is a temporary structure, used well beyond its stated design life, apparently without plan or budget for its eventual INEVITABLE failure/removal/replacement.

That's not an engineering problem.

That's not a code problem.

That's a management problem.



Mike Halloran
Pembroke Pines, FL, USA

 

[hawkaz]

Per the Minneapolis news- This was the third worst storm in their recorded history. I'm not there- so I can't speak to it personally and my info (as is most everyone else’s) is hearsay and conjecture.

I have always had an issue with calculating density of snow. Anyone from snow country can tell you, there is wet snow and there is dry snow. Dry snow can have a density as low as 6.2 lbs/ft^3, wet snow can get to 31 lbs/ft^3. This can be 50% higher than the code required density calc for a 50 psf area.

I don't know which type of snow this was. If they has 17"-20" of wet snow, they could have been very close to or even exceeded the design snow load. If there was already old, heavy snow on the roof, this would further increase the loading.

I don't know enough about the geometry/slipperiness of the roof and the pockets created by the tension wires, so if we theoretically use flat roof design and ignore drift for these purposes:
Ce=.9, Ct=1 (debatable), I=1.2, pg= 50 psf
Pf=.7x.9x1x1.2x50=38 psf or 15" of wet snow.


Wet snow doesn't accumulate as much as dry snow, so 15" would be a lot- most likely the density was somewhere in between. I wouldn't be so quick to say that the snow load didn't exceed design loads without more info.

 

[NITTANYRAY]

I have personally measured fresh fallen wet snow and it weighs about 12 PSF. Over time the snow will consolidate especially if there is a freeze and thaw cycle or rain on snow. The higher densities you mention only occur if the snow has been on the roof for an extended period of time or from the build up of several snows. In a 12" deep cube of snow the bottom half is generally denser than the upper portion due to the weight of snow above. This is why the density of snow as specified in ASCE 7 increases with the ground snow load.
There was probably drifting because of the high winds but the code requires domes to be designed for unbalanced snow loads.

 

[concretemasonry]

There were actually 2 snow events. The first started late Friday was an "Alberta Clipper" followed by cold air and wind), which is a light dry snow. On Saturday morning we had 4-7" of light, dry snow and still had to wait in line to get into a local pancake restaurant for breakfast because there was no travel inconveniences. Flurries continued lightly during the day and then a southern flow of moisture hit the colder air air and snow was the result. From what I understand, the dome roof had excess snow removed by shovels and it is melted by hoses and drained off. This procedure has worked.

The second phase changed the snow moisture content, wind direction and the drifting effects. The actual snow (at the airport) was about 17", but varied widely depending on exposure and measurement location. At my home in a NE suburb, the accumulation was somewhat less. Southwest of me the amounts were less and southeast of me they were somewhat greater (over 20").

The snow removal was abandoned because of conditions and safety at about 1:00 AM on Sunday and the dome was monitored and a problem was detected about then. At about 5:00 AM, the video cameras were turned on to tape the failure. It was not an immediate failure or an immediate collapse and everyone was gone and there were no fatalities.

Since U.S. codes are generally written for life safety and not absolute, immediate "collapse" (TV term) was not the problem as many codes in some other countries.

The weather situation really was not the 3rd worst, since the second and third previous storms (20") were only several days apart and it only was the 5th worst.

When designing, it is important to separate the immediate collapse (seismic, etc such as the the Albanian earthquakes where about 20,000 died) and the gradual failure that allows evacuation due to the time involved.

Fortunately, the codes allow the engineers to design according the the conditions possible and the failure modes to over-ride the provincial standards.

Engineer and international traveler interested in construction techniques, problems and proper design.

 

[dhengr]

We got about 18-20" of snow in about 18 hours, approx. 11:00p FRI. through 6-7:00p SAT., 11DEC10 and had very strong winds from the north, thus lots of serious drifting. That’s not a wimpy storm around here, we were pretty well shut down on SAT. & SUN., airport, transit, and all but the really stupid people who go out and try to get stuck, and then prevent street plowing. It was the 5th largest snow fall depth and blizzard in recorded weather keeping, largest snow fall ever in DEC. Concretemasonry, I live in WBL, hide your edress, something at something dot something so we can talk. I had 2 or 3' deep drifts in my driveway. Several of the neighboring houses have approx. 4' deep drifts up on the south planes of the roofs. I weighed several blocks of snow and it came in at about 9#/cf on SUN morning when the snow stopped.

If you saw the video: at the underside of the roof and then later from above too, you see what is almost a quilted effect in the fabric, from the tension strands. There is an inner layer of fabric which you see, and a second outer layer of fabric, but they both show some form of this quilting effect from the tension cables. Thus, you get a fabric panel which is supported on three or four sides by cables, which can deflect down from snow load. Obviously, you will get drifting on the leeward side of the roof, and drift depth can increase as the fabric and cables settle, the arched shape of the entire roof changes under this loading, almost a ponding effect, but worse than would occur on a stiffer roof structure. Furthermore, the roof is warm enough so they do melt snow up there, adding to the ponding effect, and I think the video shows both snow and the water. They would have been warming the building for the game on SUN, adding to this melting action. Finally, they were up on the roof shoveling and melting snow, with warn water, to try to get it off the roof, when an edge (triangular) panel broke and cause the roof to deflate; then all the snow and water tended to slide to the center and caused a couple of square fabric panels to fail, and dump on the field. The roof, particularly the cable system, was still basically structurally sound, and intact. But, at this point individual square fabric panels become the weakest like, and are obviously overloaded.

Now, last night, while they were working to clear the field and get ready for the roof repair, another square fabric panel, near the center of the roof failed, with a large dump of snow near an end zone. I agree with MikeHalloran, I don’t think this is really an engineering or code problem, or the fault of bad engineering. It’s a building maintenance and management problem, they got what they asked for when it was built, and must then learn and understand what is needed to keep the roof functioning under these rare conditions. We’ve had a number of other smaller air supported structures fail also.

 

[a2mfk]

As a kid in Wisconsin I remember how crappy a job snow shoveling was, I wouldn't want to be the poor saps up there, doing it on top of a fabric dome supported by air pressure. Whatever you do, DO NOT poke the blade of your shovel in the roof!! Hope they had some good tie off system

Just thankful to live in Florida when I hear your blizzard stories...

 

[NITTANYRAY]

Here is another opinion from the Architecture and Design web site:

"The Metrodome was not lovable; it was brutal and ugly and cheap. It was not durable; one sixteenth of an inch of fiberglass is not going to last. It was not particularly flexible, incapable of adapting to the seasons and it certainly wasn't frugal, needing continuous supply of electricity just to stay up.

They will stick on a few patches, re-inflate it and be playing inside again soon. But it will remain an object lesson in how not to build."


They should have never built an inflatable structure in an area with high snow loads. A structure similar to Madison Square garden would have been a better choice.

Part of good design is choosing a structural system that can function in the environment it is placed with only a minimum amount of maintenance required for it to function properly. Value engineering should include the cost of maintaining it. I am sure that what ever they saved buy using this cheap roof has been spent several times over just to pump the air to pressurize the building

 

[a2mfk]

I'm a huge fan of retractable roof stadiums, especially in northern climates where it is awesome to be outside until November or so... And in really hot climates like down here in Florida, which is why the new Marlins field has that type of roof (and because of severe afternoon thunderstorms all summer).

But there is also something special about Lambeau in December or better January!

 

[concretemasonry]

The Vikings do actually like an outdoor field.

The next game will be moved to the new state of the art University of Minnesota stadium (50,000 capacity) and played outside on national TV Monday night. Because the stadium was winterized/shut down the cost (including snow removal from the field and seats) will be $400,000 to $800,000 plus the cost of erecting additional temporary stands since the sold tickets exceed the university capacity. The activation of the water and sewer system may be a bigger challenge than the snow removal (24"), because there are many students on break that want to get paid for working before Christmas and there are plenty of loaders and other equipment and there is land for snow dumping.

During the inspection, using 3 cranes they discovered some additional weak panels/spots in the roof surface, but the cable structure and ring were apparently satisfactory. The relocation and the scheduled use of the dome will allow more time to install additional panels before tractor pulls and the basketball games start and buy time to decide on a replacement stadium if and when it is necessary.



Dick

Engineer and international traveler interested in construction techniques, problems and proper design.

 

[gobsmacked]

There used to be a code in the UK for farm buildings, for which the snow load depended on the number of hours per week of human occupancy. As if the snow 'knew' what kind of building it was falling on!!!! I see the point of it, but it was the wrong approach to safety.

 

[anchorengineer]

Signs of the youtube age... When something is caught on video now it goes around the world in seconds! Convenient that Fox had their cameras running all night. If TV and video were as crazy during WWII, would FDR have ever been elected? People will forget about this quickly because no one was hurt and there will be a million other things that their brains get inundated with soon enough. And it was a maintenance thing.

Clarke Engineering Services, PC
Construction Consulting & Anchor Testing

http://www.anchorengineer.com