The damage to the bottom flange
On 25 September 2018, a worker discovered a crack. The terminal closed that same day. Two girders over Fremont Street were affected. The exact time of fracture remained unknown. Following investigation and repairs, the Transit Center reopened on 1 July 2019. Peer review report, pp. 1 and 8
The welded I-girders supported the rooftop park and, through hangers at midspan, the bus level below. Their flanges were 4 inches, or 101.6 mm, thick. The hanger connection plate passed through a slot in the bottom flange. Beside it were additional thermally cut openings at the end of the complete-joint-penetration flange splice weld. The fractures initiated at the re-entrant corners of these openings, in the base metal near the weld. Peer review report, pp. 3–8 and 10
From small cracks to brittle fracture
LPI, Inc., the firm commissioned to investigate the failure, reconstructed the following sequence. Thermal cutting produced microcracks in a hard martensitic surface layer. During subsequent welding, larger, localised cracks known as pop-in cracks developed at two locations. Loading and welding residual stresses later led to rapid fracture. Toughness at mid-thickness was particularly low; Charpy V-notch impact values at quarter-thickness, by contrast, met the project specification. LPI, conclusions, pp. 185–186
This also explains why considering external loading alone is insufficient. LPI concluded that normal service stresses alone were insufficient to trigger the bottom-flange fracture. In the comparable First Street girders, the openings were cut only after the flange splice had been welded. The investigation attributed their different behaviour to lower residual stresses and the absence of pop-in cracks. This does not establish a general instruction to cut such openings after welding. The fabrication sequence must suit the particular detail. LPI, pp. 185–186
The peer review report explicitly describes the reconstructed sequence as a failure hypothesis. Not every fracture origin showed a pop-in crack. The entire sequence was therefore not observed continuously. Peer review report, pp. 15–16
What a toughness value tells us
The Charpy impact test measures the energy absorbed when a notched specimen breaks under an impact load. The value permits comparisons and checks against a specified requirement. It is not, however, a fracture toughness parameter that can be used directly to assess an existing crack in a component. The test temperature is also part of the result. A single value in joules without the test conditions does not answer whether a particular detail is safe. TWI: Charpy testing
A fracture mechanics assessment requires suitable material data. Appropriate tests can provide parameters such as K, J or CTOD. Their selection and applicability to the component require expert assessment. TWI: Fracture toughness testing
A shared task for design and fabrication
As an editorial lesson for practice, those involved should answer three questions together when approving the detail: Which locations are critical with respect to cracking? At what stage of fabrication can they be inspected? What decision follows from a finding?
A drawing should therefore describe more than the finished component. Where the fabrication sequence affects the result, it belongs in the agreed fabrication documents. The inspection plan must match it. If an area will become difficult to access later, inspection needs an appropriate hold point. A recorded indication is only useful once those responsible establish its significance and decide on rework, reinspection or withholding release.
This article does not assess historical contractual compliance or apply a present-day German standard retrospectively to the American structure.
Preventing the damage
For a comparable detail, the investigation and technical sources suggest the following present-day approach to inspection and decision-making. It is an editorial inference, not evidence that prevention of the historical failure could have been guaranteed.
- Critical decision point
- Approval of the opening and weld detail, and release of the fabricated component before installation.
- Suitable present-day method
- Joint review of detail design and fabrication; qualified surface crack testing of accessible cut surfaces, suitable material testing and, where needed, a fracture mechanics assessment. The investigation report documents fluorescent magnetic particle testing among other methods; its incorporation into a fabrication inspection plan requires separate qualification.
- Detectable finding
- Detectable crack indications, unfavourable detailing or material data that are insufficient for the assessment. A finding must be assessed in terms of its location and extent.
- Required action
- Suspend release when a critical finding remains unresolved. Revise the detail or fabrication appropriately, provide the necessary additional evidence and verify the effectiveness of the action.
- Limit
- Access, surface condition and the detection limit constrain crack testing. Calculations depend on assumed cracks, loads, residual stresses and material data. Residual uncertainty remains; methods and software replace neither expert assessment nor implementation of the decision.
Method references: LPI, scope of investigation, pp. 9–10, TWI: Fracture toughness testing.
Further source of illustrations
The MTC peer review report contains component sketches and photographs of the damage. No third-party images or captions are reproduced. Permission to embed these images has not been obtained.
Sources
Technical sources are linked directly beside the relevant paragraphs. The MTC announcement of 27 February 2020 also records completion of the peer review.