You are here
Results of our academic investigation of the I-35W bridge collapse
I attach an essay we wrote for a book that will be published by University of Minnesota Press titled "The city, the river, the bridge." The essay is a transcription of part of a public lecture I gave on infrastructure and on the bridge collapse.
After the bridge collapse there were several posts on Imechanica that included speculation about the cause of the collapse, including fatigue crack growth, lack of redundancy, etc.. Our investigation determined the collapse was a result of an undersized gusset plate that reached its plastic limit load.
| Attachment | Size |
|---|---|
 ballarini and okazaki the city the river the bridge essay.pdf | 561.17 KB |
»
- Roberto Ballarini's blog
- Log in or register to post comments
- 7399 reads
Comments
Minnesota bridge
Dear Roberto,
Thanks for the nice report. I start this term with delivering it to my undergrad students in Elasticity and grad students in Continuum Mechanics to motivate them to go beyond simple analyses of rod tension and beam bending...
-Kosta
Minnesota bridge; thanks Kosta
Dear Kosta:
I hope all is well. Thanks for reading the essay and your feedback. I agree that introducing undergraduate students to the concept that stresses and strains vary from point to point and calculating them requires solution of a boundary value problem is very important. I am glad that you believe the gusset plate represents a good illustrative example.
Regards, Roberto
Time-dependent failure?
Dear Roberto,
Thanks for your helpful article. I watched your video
from your department website, and read the NTSB's report before.
I feel confused by one important issue: you used elastic-plastics
analysis or elastic-plastics fracture mechanics, and they're all quasi-static mechanical behaviors.
Therefore, if "concrete added in later years increased the weight of
the bridge by more than 20% and thus represented a significant increase
in demand on the structural components," failure should initiate in that
year, not in 2007. I believe time-dependent
failure such as creep or fatigue crack propagation was involved but I
couldn't find this issue from the NTSB report.
Roy
The additional concrete did not fail the bridge
Roy:
the wight of the additional concrete increased the extent of plasticity in the gusset plate that failed. Nowhere in the article is it stated that this additional weight rendered the demand on the gusset plate equal to the capacity of the gusset plate. In fact this is not the case, and therefore the bridge did not fail at that time.
The demand on the gusset plate reached the capacity of the gusset plate when a very large weight, in the form of sand, gravel and water was placed on the bridge, in the vicinity of the failed gusset plate, a few hours before the collapse. As a result the bridge collapsed.
Creep in concrete is irrelevant in this failure.