Ing. Chiara Crosti: "La lezione di Minneapolis", Ispezione dei ponti IABMAS ITALIA 2015

MILANO, ITALY, NOVEMBER 20TH , 2015
LA LEZIONE DI MINNEAPOLIS
Chiara Crosti
“Sapienza” University of Rome,
chiara.crosti@uniroma1.it

COLLAPSE OF THE BRIDGE ON I 35-W MINNESOTA, AUGUST 1ST 2007
CASE STUDY

BEFORE
CASE STUDY

CASE STUDY

The fixed bearing assemblies were located at piers 1, 3, 7, 9, 12, and 13.
Expansion (sliding) bearings were used at the south and north abutments and at
piers 2, 4, 10, and 11.
Expansion roller bearings were used at piers 5, 6, and 8.
Bridge Scheme
CASE STUDY

AUGUST 1ST, 2007
CASE STUDY

27/10/2015 17franco.bontempi@uniroma1.it
CASE STUDY

27/10/2015 18franco.bontempi@uniroma1.it
CASE STUDY
Scheduled for reconstruction in 2020-25

After this tragedy, the Federal Highway Administration (FHWA) focused its attention on all the
465 steel deck truss bridges present in the National Bridge Inventory [NTSB, 2008].
“The term “fracture critical” indicates that if one main component of a bridge fails, the entire
structure could collapse. Therefore, a fracture critical bridge is a steel structure that is designed
with little or no load path redundancy. Load path redundancy is a characteristic of the design that
allows the bridge to redistribute load to other structural members on the bridge if any one member
loses capacity. “
CASE STUDY

FORENSIC INVESTIGATION
National Transportation Safety Board,
NTSB, 2008
“Collapse of I-35 W Highway Bridge,
Minneapolis, Minnesota, August 1, 2007”
Accident Report, NTSB/HAR 08/03 PB 2008-
916213, Washington D.C. 20594..

THE MAIN CAUSE:
The primary cause was the under-sized gusset plates, at 0.5 inches (13 mm) thick;
U10-W
5/67
[*] National Transportation Safety Board, “Collapse of I-35 W Highway Bridge, Minneapolis, Minnesota, August
1, 2007” Accident Report, NTSB/HAR 08/03 PB 2008-916213, Washington D.C. 20594. 2008.

FINITE ELEMENT MODEL FOR OUTSIDE WEST GUSSET PLATE AT U10W
1977-1998
Stress contours for outside (west) gusset plate at U10W at time of bridge opening in 1967
Yield stress
of 51.5 ksi
South North

2 inches (51 mm) of concrete were added to the road surface over the years,
increasing the dead load by 20%;
6/67
THE ADDITIONARY CAUSE:
1977, Renovation:
Increased Deck Thickness
1998, Renovation:
Median Barrier, Traffic Railings,
and Anti-Icing System
2007, Repair and Renovation:
Repaving

Stress contours for outside (west) gusset plate at U10W after 1977 and 1998 renovation projects
Yield stress
of 51.5 ksi
South North

The extraordinary weight of construction equipment and material resting on the
bridge just above its weakest point at the time of the collapse
7/67
[*]
U10-WNorth South
184 380 lbf (820 kN) of gravel
198 820 lbf (884 kN) of sand
195 535 lbf (870 kN) of parked construction vehicles and personnel

The extraordinary weight of construction equipment and material resting on the
bridge just above its weakest point at the time of the collapse
7/67
[*]
U10-WNorth South
Pier 6
184 380 lbf (820 kN) of gravel
198 820 lbf (884 kN) of sand
195 535 lbf (870 kN) of parked construction vehicles and personnel

Stress contours for outside (west) gusset plate at U10W on August 1, 2007
Yield stress
of 51.5 ksi
South North

Stress contours for outside (west) gusset plate at U10W on August 1, 2007
Yield stress
of 51.5 ksi

8/31
12/44
BRIDGE COLLAPSE

8/31
14/44
BRIDGE COLLAPSE

8/31
15/44
BRIDGE COLLAPSE

8/31
16/44
BRIDGE COLLAPSE

8/31
17/44
BRIDGE COLLAPSE

8/31
18/44
BRIDGE COLLAPSE

8/31
19/44
BRIDGE COLLAPSE

8/31
20/44
BRIDGE COLLAPSE

8/31
21/44
BRIDGE COLLAPSE

8/31
22/44
BRIDGE COLLAPSE

8/31
23/44
BRIDGE COLLAPSE

8/31
24/44
BRIDGE COLLAPSE

/31
25/44
BRIDGE COLLAPSE

8/31
13/44
BRIDGE COLLAPSE

THE MAIN CAUSE:

Fixed Bearing
Expansion Bearing
Expansion Bearing Expansion Bearing
THE MAIN CAUSE:

Fixed Bearing
Expansion Bearing
Expansion Bearing
THE MAIN CAUSE:
Fixed Bearing

Temperature Effects
The deck truss portion of the bridge was designed and constructed with a fixed bearing at pier 7 (on
the north side of the river) and roller bearings at the remaining three piers (5, 6, and 8).
Postcollapse examination of the roller bearing components showed the presence of roller wear
marks, indicating that the deck truss was moving relative to piers 5, 6, and 8 in response to thermal
contraction and expansion. This movement limited the amount of longitudinal force applied to the
top of any pier
On the day of the bridge collapse, the temperature had increased approximately 20º F from
morning to early evening, the time of the accident. The finite element analysis incorporated this
temperature increase with fixed bearings (but allowing for pier flexibility) to evaluate worst
case temperature effects. The effects of a difference in temperature on the east and west trusses
arising from the position of the sun were also evaluated. This analysis showed that the loads
necessary to initiate the lateral shifting instability of diagonal L9/U10W increased with increasing
temperature, and this result did not change when differential temperature was included, indicating
that the change in temperature on the day of the accident did not play a role in initiation of the
collapse.
OTHER POSSIBLE COLLAPSE SCENARIOS CONSIDERED FROM NTSB *

Corrosion Damage in Gusset Plates at L11 Nodes
None of the available evidence (video recording, evaluation of the fracture and deformation
patterns, and finite element analysis) indicated a possible initial failure associated with the L11
nodes
Fracture of Floor Truss
Safety Board investigators considered the possibility that the collapse of the deck truss could have
resulted from an initial pure tension or pure compression failure of a member within a floor truss.
However, examination of the floor trusses reconstructed at Bohemian Flats revealed none of these
types of fractures. Although a portion of a fracture in the upper chord of floor truss 10 contained
brittle fracture characteristics, this fracture was consistent with bending loads applied when
diagonal L9/U10E struck the lower surface of the upper chord of the floor truss, clearly a secondary
event.
Preexisting Cracking
Neither of these cracks, however, was in an area associated with the identified area of the collapse
initiation, and no evidence was found that preexisting cracking contributed to the collapse or
significantly affected the collapse sequence
Pier Movement
Postcollapse survey measurements indicated that piers 5 and 6 exhibited no settlement or
displacement, but that piers 7 and 8 were tilting about 9º southward toward the river. The Safety
Board evaluated the possibility that movement of one or both of these tilted piers initiated or
affected the collapse.
OTHER POSSIBLE COLLAPSE SCENARIOS CONSIDERED FROM NTSB *

INSPECTION REPORTING FOR I-35W BRIDGE, 1983-2007

INSPECTION REPORTING FOR I-35W BRIDGE, 2006
GUSSET PLATE???

BOWED GUSSET PLATE AT NODE U10

INSPECTION REPORTING FOR I-35W BRIDGE

Continued gusset plate load ratings on 25 Trunk Hwy specified bridges that began back
in October 2007
February 2, 2008: FHWA Issued 1st gusset plate analysis method draft
March 2008: Gusset inspections began on the 25 specified bridges
March – August 2008: Gusset plate load rating analyses refined based on inspection
findings
February 2009: FHWA published final draft of gusset plate analysis method titled -
"Load Rating Guidance and Examples For Bolted and Riveted Gusset Plates in Truss
Bridges"

St. Cloud Bridge
closed and replacement accelerated
Blatnik Bridge in Duluth
gusset stiffened to restore full safety factor
Winona Bridge
reinforced gusset due to corrosion
Hastings Bridge
unsupported edge stiffened to meet code requirements

Jennifer L. Zink, P.E. Bridge Inspection Mn/DOT Bridge Office

RESISTANCE OF GUSSET PLATES:
GUSSET PLATES IN TENSION
GUSSET PLATES SUBJECT TO SHEAR
GUSSET PLATES IN COMPRESSION
FHWA GUIDELINES, (February, 2009)
26/67
RESISTANCE OF FASTENERS
SHEAR RESISTANCE OF FASTENERS
PLATE BEARING RESISTANCE AT FASTENERS
http://bridges.transportation.org/Documents/FHWA-IF-09
014LoadRatingGuidanceandExamplesforGussetsFebruary2009rev3.pdf

CRITICAL REVIEW OF THE FHWA GUIDELINES:
• Stiffness of framing members, that increase the ultimate compression capacity of the gusset
plate;
• Influence of the initial imperfections, that decrease the ultimate compression capacity of the
gusset plate;
• Edge buckling vs. Gusset plates buckling, from that the importance of making consideration
not only on the length of the free edge, but also length of equivalent column is important for
buckling
40/67
Framing member stiffness
Gusset Plates
For LRFR and λ ≤ 2.25
(assumes δ ≤ L /1500)
What if δ > L /1500) ?

NIST PHYSICAL INFRASTRUCTURE PROGRAM
NIST Physical Infrastructure Program

FHWA SETUP**
[**] Iadicola M., Ocel J., Zobel R., “Quantitative Evaluation of Digital Image Correlation for Large-Scale Gusset
Plate Experiments”, IABMAS2012, Stresa, Lake Maggiore, Italy, July 8-12.

FHWA TEST, VIRGINIA (2010)

Advanced computing modeling
Hand calculation
7/28
FHWA, 2009
SIMPLIFIED CONNECTION MODEL

Connection element 1 Connection element 3
Connection element 4
n. connection elements: 5
Each connection element has a
6x6 stiffness matrix
N. Nodes: 28330
n. Dof : 169980
n. Elements S4R and S3R: 27670
45/6
7
MODELING OF GUSSET PLATE CONNECTIONS
SUB-STRUCTURING ANALYSIS – SIMPLIFIED LINEAR CONNECTION MODEL

59/67
SUB-STRUCTURING ANALYSIS – SIMPLIFIED NONLINEAR CONNECTION MODEL
MODELING OF GUSSET PLATE CONNECTIONS

ALL RIGID JOINT
U10 W
ALL RIGID JOINT + 1 SEMI-RIGID JOINT
NorthSouth
3D MODEL OF THE I35-W BRIDGE
3D FINITE ELEMENT MODEL
Nodes: 1172
Beam elements: 1849

56/67
NorthSouth
0
1
2
3
4
5
6
7
-0,7 -0,6 -0,5 -0,4 -0,3 -0,2 -0,1 0,0
LoadFactor
Dz (m)
RIGID JOINTS
SEMI-RIGID JOINT
Node at midspan
18/28
NONLINEAR ANALYSES RESULTS

0,0
0,2
0,4
0,6
0,8
1,0
1,2
1,4
1,6
1,8
-2,0E+07 -1,0E+07 0,0E+00 1,0E+07 2,0E+07 3,0E+07
LoadFactor
Axial Forces (N)
CONNECTION 1 CONNECTION 2 CONNECTION 3
CONNECTION 4 CONNECTION 5 AXIAL CAPACITY CONNECTION 1
AXIAL CAPACITY CONNECTION 2 AXIAL CAPACITY CONNECTION 3 AXIAL CAPACITY CONNECTION 4
AXIAL CAPACITY CONNECTION 5
57/6319/28
Compression Tension
What is important to underline is not only the possibility to catch the collapse due to the failure of
the connection, but moreover to classify the cause of the collapse which, in this case, happened
because of the achievement for one of the connection elements of the maximum capacity in
compression.

Deformed shape (scale displacement of 10)
at the ultimate load (Pu) of 1.2+07 N
62/67
What is important to underline is not only the
possibility to catch the collapse due to the failure of
the connection, but moreover to classify the cause
of the collapse which, in this case, happened
because of the achievement for one of the
connection elements of the maximum capacity in
compression.
CONCLUSIVE CONSIDERATIONS
CONCLUSION
Connection
member
Load
ratio
Tension or
compression
1 0.28 Compression
2 0.56 Tension
3 1.00 Compression
4 0.02 Tension
5 0.41 Tension

I-35W Bridge was subjected constantly to inspection to assess its safety but even with that people
in charge did not notice that the bridge was about to fail. A future work could be to develop
parametric study on some particular shapes of gusset plates in order to identify some “critical”
points where the monitoring of the out-plane displacements, could give to the owners of the
bridges a warning of what it is happening in the connection. An idea of monitoring could have been
done with a technique of monitoring developed by NIST who focuses its research on two areas of
structural health monitoring:
•development of non-destructive techniques; and
•analysis for determining the degraded condition of infrastructural components and their
subcomponents.
FURTHER DEVELOPMENTS
•FEA results•Results from monitoring **
chiara.crosti@uniroma1.itchiara.crosti@uniroma1.it
CONCLUSION
•FHWA test

CONCLUSION

I-35W SAINT ANTHONY FALLS BRIDGE (September 2008)
CONCLUSION

•100-year life span
•Design-build project complete in 339 days.
•Designed to be aesthetically pleasing and fit in with its environment
•The bridge was constructed with high-strength concrete
•189 feet wide—the previous bridge was 113 feet wide
•10 lanes of traffic, five in each direction—two lanes wider than the former bridge
•13 foot wide right shoulders and 14 foot wide left shoulders, the previous bridge had no
shoulders
•Light Rail Transport-ready which may help accommodate future transportation needs
•There are 323 sensors that regularly measure bridge conditions such as deck
movement, stress, and temperature
I-35W SAINT ANTHONY FALLS BRIDGE (September 2008)
CONCLUSION

Opening day was six years ago, and the I-35W bridge is needing repairs — some that come from
our harsh winters, but some from improper installations.
CONCLUSION

http://www.startribune.com/new-35w-bridge-already-is-
aging/268746561/
CONCLUSION

ACKNOWLEDGMENT
The author would like to acknowledge:
•Professor Franco Bontempi and his team, www.francobontempi.org, for the support
and the help,
•the Metallurgy division of the National Institute of Standard and Technology (NIST) in
particular Dr. Dat Duthinh for the support and the help,
•Eng. Piergiorgio Perin for providing the use of the finite element code Straus, and
•NTSB and FHWA for allowing the access to the detailed FE model used in the
investigation of the collapse of the I-35 W Bridge.
CONCLUSION

Ing. Chiara Crosti: "La lezione di Minneapolis", Ispezione dei ponti IABMAS ITALIA 2015

Recommandé

Recommandé

Contenu connexe

Similaire à Ing. Chiara Crosti: "La lezione di Minneapolis", Ispezione dei ponti IABMAS ITALIA 2015

Similaire à Ing. Chiara Crosti: "La lezione di Minneapolis", Ispezione dei ponti IABMAS ITALIA 2015 (20)

Dernier

Dernier (20)

Ing. Chiara Crosti: "La lezione di Minneapolis", Ispezione dei ponti IABMAS ITALIA 2015