August 12, 2007

Engineers: Weight On Bridge Far Below Design Capacity

It takes a while for the Star Tribune to get to its buried lede in today's update on the St. Anthony Bridge collapse, but engineers have thrown more cold water on the theory that construction work caused the disaster. The story focuses on the fact that MnDOT and construction crews did not consider the effect of the extra weight on the bridge, but the story answers the question itself:

As federal investigators continued to pursue evidence Thursday in the Interstate 35W bridge collapse, state officials said they had no reason to analyze the potential impact of resurfacing the bridge before authorizing that work.

The Minnesota Department of Transportation (MnDOT) said it didn't need an analysis beforehand because the resurfacing work wouldn't add weight to the bridge.

Construction officials and some engineers deny or doubt that the resurfacing contributed to the collapse of the steel deck truss bridge. However, an unusual federal alert was put out Wednesday to other states that have similar bridge repair projects underway.

Why didn't they consider the weight distribution of the heavy equipment? As it turns out, the bridge design supposedly covered a much heavier load than anything a construction crew could bring. MnDOT chief bridge engineer Dan Dorgan told the Star Tribune that the bridge met military design specifications:

Kent Harries, an assistant professor of structural engineering and mechanics at the University of Pittsburgh, said that means the bridge should have been able to support a traffic jam of flatbed trucks loaded with M1 Abrams tanks.

"The loading that that bridge was seeing at the time it collapsed was considerably lower than its design loads," Harries said.

That doesn't mean that the weight distribution may not have played a part in the collapse. However, given the lack of any warning about the bridge's safety, no one would have thought it dangerous to bring that equipment on the bridge, considering the reduced load from the already-milled deck and the restricted amount of traffic on the bridge. A few construction trucks and gravel loads obviously pales in comparison to a bridge full of M1A1 Abrams tanks.

Clearly the bridge failed to support the weight of the load on it at the time. The question remains as to why the bridge could not support its design load, not whether the design load got exceeded. One problem in analyzing this kind of catastrophic failure is that such failures are so rare. All factors have to be considered carefully, but it's important not to assume anything from any one factor out of the ordinary. Correlation is not necessarily causation, and in this case it seems unlikely that a few construction trucks created the problem, although they could well have been the straw that broke the St. Anthony Bridge's back.

Meanwhile, Minnesotans continue to look at other bridges around the state -- and in Stillwater, one overpass labeled as "basically intolerable" has travelers wondering whether the state should shut it down:

Corroded strands of rebar jut from the sides and pillars of the cracked Hwy. 36 bridge near Stillwater, while jagged pieces of fallen concrete litter the ground below.

Every day, nearly 10,000 vehicles travel eastbound over the crumbling structure. Most of the people in the huge trucks, cars and school buses on the bridge are unaware that it has been listed federally as "basically intolerable."

Its last inspection report gave the bridge a rating of 28.3, significantly below the 50 given to the St. Anthony Bridge before its collapse. This overpass runs over a highway with some significant traffic, so it could present a danger to those on it and below it. The crumbling underside of the bridge prompted a local resident to send pictures of it to his state representative. However, the good news is that MnDOT had already planned to start work on repairs next month, which will take a little over a month to complete and will require a shutdown during the construction.

This highlights the fact that MnDOT has worked on prioritizing bridge repairs and replacements for quite some time. They estimate the minimum costs at $1.4 billion over the next twenty years, which sounds like a lot of money. However, if the state maintains the spending level for MnDOT at what it has been for the last few years, MnDOT will have approximately $44 billion in funding in which to make those repairs ($2.2 billion per year). That amounts to 3.2% of MnDOT funding in 2007 dollars for that period.

It doesn't sound like funding will be a problem, nor does that require tax hikes to acquire.


TrackBack URL for this entry:

Comments (23)

Posted by RBMN | August 12, 2007 10:30 AM

True, the Hwy 36 bridge is worse, but there's also a world of difference between a car falling 20-feet onto dry ground, and a car falling 50-feet into a river.

Posted by Dusty | August 12, 2007 10:58 AM

"... said that means the bridge should have been able to support a traffic jam of flatbed trucks loaded with M1 Abrams tanks."

Just to nitpick, I suppose Harries was speaking for lay accuracy and not relating engineering precision. The M1, at a spec weight of 61.4 tonnes, was not put into service until 1980, 13 years after the bridge was constructed, whereas the tanks in service at the time, M60 and M48, weighed approximately 46 and 45 tonnes, respectively.

Of course, the engineers could have been instructed to plan for the heavier, future M1 if it was in the planning stage, but I'm pretty sure they weren't instructed so. In any event, such a lesser load would be about the same as, guestimating, an equivalent bumper to bumper dump truck lined bridge.

Posted by cirby | August 12, 2007 11:23 AM

A fully-loaded dump truck should top out at 80,000 pounds or so (the current GVW for interstate highways without an "overweight" permit), or 40 tons.

So no, a handful of dump trucks loaded with gravel wouldn't be heavier than the same number of trucks loaded with tanks. And from the photos we've seen, there were no more than two or three dump trucks on the center bridge truss at the time of the collapse (they wouldn't need more than two or three because of the pace of resurfacing and the room available). From the article above, the worst load was the cement truck, which would have had the most concentrated heavy weight on the least number of tires.

Posted by Dusty | August 12, 2007 12:29 PM

Agreed, Cirby, with regard to the actual case loads. Live loads at the time are nowhere near the uniform loading case noted in the article.

As for your former argument, your use of the "same number of trucks" is not an approach an engineer would use. It might be correct if a tank carrying flatbed with cab was equivalent in length to the fully loaded, longer combination category of dump truck you cite. If not, then the shorter would have to calculated to determine the controlling case.

Then again, an alternate worse case loading might be two standard dump trucks owing to the differences in length, too. Surely, though, length is controlling, not equivalent number of vehicles. And, that said, since I only guesstimated, the 1967 tank carrying design criteria might still be the higher, but it is much closer than the lay accuracy implied.

In any event, I originally pointed this out only as an aside in reminding that the proper comparison is original 1967 design load criteria versus possible loadings at the time of collapse and this mention of M1 tank could make one overlook that. I did not intend to suggest the bridge was subjected to a maximum uniform live load at the time of the collapse and am pretty sure that's clear. It's certainly clear now.

Posted by Conrad | August 12, 2007 12:36 PM

The bridge was built to military specifications. So we can assume that if corrosion reduced the thickness of good metal in the structures, bad bearings and non-functional expansion joints had no effect on the integrity of the bridge over the years, because the bridge was built to military specifications.
So what do we need inspections for? Do inspection reports carry any weight in determining if a bridge is too unsafe and should be closed to trafic? And if so who is responsible to make that decision?

Posted by agesilaus | August 12, 2007 12:52 PM

So far as I know the entire Interstate system of highways, including bridges was originally designed to military spec. Eisenhower intended that this system be a defense highway system intended to move troops across the country.

The problem with this particular bridge is that it was designed without the 'fudge factor' normally included in bridges before and after this period. Calculations would be done based on the maximum expected load multiplied by this factor which ranged from 3X to 10X.

If a fudge factor was included those suspect gusset plates might have been made out of 1 inch steel rather than 1/2 inch. And might have used a stronger alloy.

Engineers of this period seem to have suffered a bout of extreme hubris and built bridges to the exact maximum calculated load. That leaves no room for design calculation failures, unexpected increases in load, or loss of strength because of damage and corrosion.

Posted by Dusty | August 12, 2007 1:16 PM

[agesilaus at August 12, 2007 12:52 PM]

Do you have a reference you are working from in making statements about loading design criteria and "fudge factors" for this "particular bridge"?

And when you say, "Calculations would be done based on the maximum expected load multiplied by this factor which ranged from 3X to 10X", are you speaking to what the before and after standard was/is or with regard to this "particular bridge". I infer that it is the "before and after" criteria because the "3X to 10X" is a fudge factor, which you say wasn't used, but I cannot be sure.

Thanks in advance for any response.

Posted by MO | August 12, 2007 1:41 PM

Note that the Star Tribune article makes no mention of the fact that the new St. Croix bridge was delayed (and the old bridge still in place) becuase of an ENVIRONMENTAL lawsuit. Once again, the liberals are putting the lives of birds and fishes ahed of human beings.

Posted by JB | August 12, 2007 1:46 PM

A few observations about the collapse. Below is pure anecdotal speculation.

This web page from the MN DOT has all kinds of interesting information on the I35 bridge including inspection reports and hundreds of sheets of bridge drawings.

From the most recent inspection report (June 2006) it seems abundantly clear that the bridge was in terrible shape. The report is filled with observations about fatigue cracks, rusting steel, frozen hinge bearings, cracked welds, water saturated concrete, etc. Money quote, “The long term plans for this river crossing need to be defined with replacement, redecking, etc. Due to the “Fracture Critical” configuration of the main river spans and the problematic “crossbeam” details, and fatigue cracking in the approach spans, eventual replacement of the entire structure would be preferable.” It is unclear from the news reports if the ongoing bridge repairs were resurfacing or redecking or if those terms are being used interchangeably. Were the repair recommendations enumerated in the June 2006 report being implemented? The bridge was clearly not being replaced.

Additionally, in March of 2006 the bridge was given an “Operating Rating” of 53.9 (out of 100, I think). Not sure how low the Operating Rating needs to be before a bridge is deemed unsafe for use. See

The bridge contains “fracture critical” structural members. As defined by the Army Corps of Engineers, “fracture critical members or member components are tension members or tension components of members whose failure would be expected to result in collapse of a bridge.” Maybe there was a fracture critical structural member that was somehow compromised as a result of the ongoing repairs or maybe the inspection teams missed it in their inspections (the report mentions that some of the areas of the bridge could not be inspected because of pigeon screens.) This point leads me to believe that maybe it is possible that there was no systemic structural degradation or universal congenital defect (like compromised or inferior gusset plates). Could the failure of a single structural member collapse the entire structure?

In the security camera video of the collapse it looks like the collapse initiates at the south end of the span. The main span between piers 6 (south end) and 7 (north end) falls as a single piece. It therefore looks like something south of pier 6 (the connection at pier 5?) gives way as the south end of the bridge drops first. (Having never been to MN I might have north south reversed. In the video I am assuming that the end of the bridge closest to the camera is the south end.) I would guess that the engineers who are looking into the collapse are going to focus on the areas of the bridge leading to the main span. Has anyone seen any video footage of the collapse from the other side of the river?

Again, this is all pure speculation.

Posted by Dusty | August 12, 2007 2:09 PM

MN local news is not my bag of tea, MO, but you might be talking the lift bridge here:

not the bridge in the story CQ links to, at least wrt the photos. I could be wrong. Your point is still appropriate, though, in the overall problem of maintenance/replacement delays v. safety.

Posted by Eric | August 12, 2007 3:21 PM

The St. Croix bridge wasn't delayed so much for birds and fishes as it was for the scenic "value" of a new modern construction bridge slashing across the national scenic riverway. The Sierra Club additionally railed against it because they are against sprawl at all costs and such a new bridge, their argument goes, makes redevelopment of fertile farmland in Wisconsin into impermeable driveways, roads, and roofs a net negative to the watershed of the river and urban heat island.

Again, not so much birds and fishes as other intangibles. Still, as a regional geographer, I disagree with the Sierra club's delaying tactics. They instead could have rallied for a beautiful bridge across the National Scenic Riverway, and could rally for more intelligent development that compensates for impermeable surface gain.

Posted by Eric | August 12, 2007 3:24 PM

Dusty, MO is talking about both. There is a proposed new bridge crossing the St. Croix running through where this pohotographed decaying bridge is because the lift bridge in Stillwater is outdated.

Posted by pk | August 12, 2007 3:58 PM

about a century ago the railroads got tired of their bridges and trestles falling down under heavier and heavier trains.

they developed a loading specification, it is called off as "coopers loading". e-72 was the most robust loading when they made up the system and it was used for steam locomotives pulling ore trains in the misabe range.

has anyone heard these terms used in conjunction with this bridge.


Posted by Dusty | August 12, 2007 5:10 PM

OIC. Thanks, Eric.


pk, not yet, though I have heard of it. It seems to me that railroads encountered, big time, the problem of fatigue failure in the switch to the revolutionary iron bridges. IIRC, bridges in service after a while reached a 25% fatigue related failure rate -- too brittle and no plastic area on the yield curve, thus the invention of steel.

In looking through google a bit, I see a lot of reference to fatigue in connection to railroad bridges and aircraft, areas where high repetitive stresses is the primary concern. Off hand, I 'd say that is not the primary design issue with road bridges.

I wouldn't be surprised though if there are tables of some kind of cross tabulation of Cooper to something road bridge related. Don't rely on my word, however; I'm just throwing it up before finding the definitions. (Now where is my little black book of every scientific table and measurement system in existence ....)

Posted by agesilaus | August 12, 2007 5:40 PM

My primary source is Engineering School and some Civil Engineering courses I took, I was not a CE major tho. I was told in those classes that the normal overdesign was 3X and up in those days. The Brooklyn bridge apparently had an overdesign ratio of 6X:

"In comparison, the Brooklyn Bridge, with wire cable suspension, was designed with an excess strength factor of six, which proved fortunate owing to a contractor's substitution of wire weaker than that specified. (This was discovered before completion and additional strands were placed in the bundles.)"

"At the time the bridge was built, the aerodynamics of bridge building had not been worked out. Bridges were not tested in wind tunnels until the 1950s - well after the collapse of the Tacoma Narrows Bridge in 1940. It is therefore fortunate that the open truss structure supporting the deck is by its nature less subject to aerodynamic problems. Roebling designed a bridge and truss system that was six times as strong as he thought it needed to be. Because of this, the Brooklyn Bridge is still standing when many of the bridges built around the same time have vanished into history and have been replaced. This is also in spite of the substitution of inferior quality wire in the cabling supplied by the contractor J. Lloyd Haigh - by the time it was discovered, it was too late to replace the cabling that had already been constructed. Roebling determined that the poorer wire would leave the bridge four rather than six times as strong as necessary, so it was eventually allowed to stand, with the addition of 250 cables. Diagonal cables were installed from the towers to the deck, intended to stiffen the bridge. This turned out unnecessary, but they are kept for their distinctive beauty."

Here is an interesting entry on general engineering design:

“… the amount of redundancy the designer puts into a structure to avoid total failure in case of local failures … varies with the type of structure. Structural redundancy essentially allows the loads to be carried in more than one way–i.e. through more than one path through the structure–and must be considered a needed characteristic in any large structure or any structure whose failure may cause extensive damage or loss of life.”

Here is a study on Bridge redundancy:

Now I read an article on the I-35 bridge failure that made the point that prior to the 1960's bridges were usually designed with conciderable built in redundancy, up to 10X. But with the flood of articles I haven't been able to google my way back to it. And then for a period starting in the 1960's and running for 10 or 15 years, engineers thought that they could exactly calculate the loads on bridges and design to that load. These are the non-redundant bridges that were built then. Including the I-35 bridge. After that period they came to their senses and started adding redudunancy back into bridge design. Of course it is cheaper not to over design a bridge, until it falls down anyway.

Posted by James I. Hymas | August 12, 2007 6:25 PM

Cap'n Ed However, if the state maintains the spending level for MnDOT at what it has been for the last few years, MnDOT will have approximately $44 billion in funding in which to make those repairs ($2.2 billion per year). That amounts to 3.2% of MnDOT funding in 2007 dollars for that period.

It doesn't sound like funding will be a problem, nor does that require tax hikes to acquire.

It seems to me that this point is too facile - just as facile as the automatic assumption that increases in taxes are required.

I'd defer conclusions until results of the inquiry - I'm assuming there will be an inquiry - are received. Then I would look at the MnDOT budget on a zero-based basis.

How much money is truly required to do all necessary and prudent resurfacing? I think that's something that needs to be done every 10 years or so. How much to rebuild the road-beds every 50 years or so? How much to inspect bridges and how much to replace them at the proper time?

I'll bet a nickel that most, if not all, of this information is available in MnDOT budget papers. And I'll bet another nickel that if resurfacing has to be done every ten years to avoid the need for road-bed reconstruction twenty-five years early, there will be a politician - supported by his constituents - who will say that a bunch of money can be saved by deferring the maintenance for an extra five years. That's what happens in Toronto, anyway.

Posted by PETNSandwich | August 12, 2007 6:44 PM

Embarrassing – being “designed to meet military specifications” does not have anything to do with loading the bridge up with M-1s bumper to bumper.

Standard AASHTO bridge loading for design is HS20-44 which is basically an axle load of 26,000 lbs on top of a lane load of 640 lbs/lin.ft.

What they surely mean is by “military specifications” is the AASHTO “alternative military load”, which in lieu of the HS20-44 is two 24,000 lb axles spaced at 4 ft apart.

Ain’t no military commander going to order all his tanks to bunch up on bridge, “That’s it boys, get nice and tight together” even in peacetime at home. [Rosie says, google it - bridge FM5 “military load”]

Posted by Dusty | August 12, 2007 9:37 PM

PETNSandwich, thanks for pointing out that some mil spec is not the appropriate load criteria.

There's a nice summary of the evolution of the live load models in this 154 page pdf (warning: 9meg file alert). It starts on pg 7 of 154.

I am a little surprised the 640 lb/lf/lane uniform load dates at least as far back as the 1931 AASHO 1st edition and the H20-44 came in in 1944. I wouldn't have expected it to be that consistent over 70 to 80 years.

I'd be interested in the difference in treatment of impact factor from the 60's, which looks to based on the 1956 AASHO edition, and when changes started in the 70's. Was it of any great significance and was it primarily to address fatigue?

Posted by Dusty | August 12, 2007 10:28 PM

[agesilaus at August 12, 2007 5:40 PM]

I overlooked your response, agesilaus. Thanks for the feedback.

Posted by notanengineer | August 12, 2007 10:44 PM

"The question remains as to why the bridge could not support its design load, not whether the design load got exceeded."

No ... that's not right.

The question remains as to why the bridge could not support substantially LESS than its design load.

Posted by AnyMouse | August 12, 2007 11:04 PM

The Hwy 36 bridge over by Stillwater which was in the STrib on Sunday included a photo of exposed rebar. The caption mentioned that until a few days ago that part of the bridge had been covered by fine mesh, but it was removed. The article did not mention why the mesh was removed. I suspect it was removed...because someone was inspecting the situation.

The STrib did not mention the St. Croix bridge being delayed by environmental complaints. The STrib also has avoided pointing out that a large chunk of what-used-to-be road funding has been reserved for public transit projects.

Posted by Alan Kellogg | August 13, 2007 5:47 AM

What was the bridge capacity after 40 years of wear and tear and the repair work being done at the time of the collapse?

Posted by Carol Herman | August 13, 2007 6:12 AM

They had to bring 10,000 pounds of gravel to the bridge? Why? They had these heavy trucks there, just waiting to roll the cement around? Why? You can see the damage. Even where the trucks and roadway materials went down.

Gosh, it's like watching kids jumping up and down on a bed. Until the adults call out "ENOUGH." "STOP." "HALT."

Some kids? They'll keep doing it.

HOWEVER, the lawsuits will bring reason to this mess. I doubt another contractor is gonna pile on weight. "Because it's easier." No. It. Ain't.

Post a comment