Load Path

A load path is the route a force takes through a structure from the point of application to the foundation. Every pound of weight, every gust of wind, every footstep, every snowflake on the roof must travel through the structural members — slab to beam to column to footing to soil — in an unbroken chain, and the chain must be designed, each link specified, each connection detailed, and if any link is missing or inadequate, the load has nowhere to go, and a load with nowhere to go is the definition of a structural failure.

Ada read the drawings at her kitchen table.

She had printed them at the FedEx Office on Illinois Street because she did not trust a screen for this work, she needed the paper, needed to spread the sheets across a surface and see them in relation to each other, the way you see a building in relation to the ground, and her kitchen table was four feet by three feet, which was not large enough, so the sheets overlapped and she moved them like a dealer shuffling cards, pulling one out, laying another down, cross-referencing the plan views against the sections against the details against the general notes, building the structure in her mind the way the contractor had built it in the world, one element at a time.

The structural drawings comprised twenty-six sheets. Sheet S-1 was the general notes — the design loads, the material specifications, the code references, the special inspections required. Sheet S-2 was the foundation plan. Sheets S-3 through S-8 were the framing plans for each level. Sheets S-9 through S-14 were the framing elevations. Sheets S-15 through S-26 were the structural details — the connection details, the reinforcing details, the post-tensioning details, the miscellaneous details that specify how the abstract intention of the framing plans becomes the physical reality of steel and concrete.

Ada started with the general notes.

The design dead load for the parking levels was 105 pounds per square foot, which included the self-weight of the double-tee floor members, the topping slab, and an allowance for the waterproofing membrane and the MEP systems. The design live load was 40 pounds per square foot, which was the code-mandated minimum for a parking garage, representing the weight of the vehicles and the people inside them. The wind load was calculated per ASCE 7-02, which was the version of the load standard in effect when the building was designed, and the basic wind speed was 90 miles per hour, and the exposure category was B — urban and suburban terrain — which was different from the Exposure C that Ada would have used, and she noted this, noted it as a possible underestimation, because the building was on Wabash Avenue with a clear exposure to the east where the old Illinois Central rail yards were still open land in 2002 when the design was done, and Exposure B assumed more sheltering from adjacent buildings than the site actually had, and this meant the wind loads were lower than they should have been, and lower wind loads meant smaller members and smaller connections, and smaller connections meant less capacity, and less capacity meant less margin.

She noted this on a yellow legal pad beside the drawings. She wrote: Wind exposure — B used, C arguable. Verify.

She moved to the framing plans.

The structure was regular. Six levels of parking, the first level at grade, the sixth level the roof. The structural grid was thirty feet in the east-west direction and sixty feet in the north-south direction, and the double-tee floor members spanned the sixty-foot direction, bearing on inverted-tee beams that spanned the thirty-foot direction between columns. The columns were precast concrete, sixteen inches square, and the beams were precast inverted-tees, thirty-six inches deep with a twelve-inch-wide ledge on each side for the double-tees to bear on.

This was standard. Ada had seen this system in a hundred parking structures. The system worked. It had been working for fifty years. The components were well-understood, the load paths were clear, and the behavior under load was predictable. The system did not fail unless something in the system was wrong.

She moved to the framing elevations.

She studied the east elevation — the side that had collapsed — and she traced the load path from the roof down. The roof double-tees bore on the sixth-floor beams. The sixth-floor beams bore on the sixth-floor columns. The sixth-floor columns bore on the fifth-floor beams by way of a pin connection at the column splice. And so on, down through the building, each element delivering its load to the element below, the chain unbroken from the roof to the footings.

Except.

Ada looked at the third-floor framing plan, sheet S-5, and she looked at the east wall of the structure, the wall that ran along grid line F at the eastern edge of the building, and she saw that the wall was identified in the general notes as a shear wall — a wall designed to resist lateral loads, wind and seismic, a wall that was part of the building's lateral force-resisting system — and she saw that the wall was also carrying gravity loads, because the double-tees on the east side of the building bore directly on the wall at grid line F rather than on an inverted-tee beam, which meant the wall was performing two functions: lateral resistance and gravity support.

A wall that carries gravity loads is a bearing wall.

A wall that resists lateral loads is a shear wall.

A wall that does both is both, and it must be designed for both, and the design must account for the interaction between the two load types, because a wall that is simultaneously carrying the weight of the floors above it and resisting the sideways push of the wind is a wall under combined stress, and the combined stress is greater than either stress alone, and the capacity of the wall must be checked for the combination.

Ada looked at the structural calculations. She did not have them yet — Brin was still working on getting the full calculation package from the building department — but she could see from the drawings what the wall was supposed to be. The east wall at grid line F was drawn as a twelve-inch-thick precast concrete wall panel, reinforced with number-five bars at twelve inches on center in each face, and it was connected to the foundation with dowels and to the adjacent panels with welded connections at each floor level.

She looked at the connection detail for the double-tee to wall connection on sheet S-22.

The detail showed the double-tee bearing on a steel angle embedded in the wall panel, with a bearing pad beneath the tee stem and a field weld connecting the tee's embed plate to the wall's embed plate. The bearing length was four inches. The weld was specified as a three-sixteenths-inch fillet weld, four inches long.

Ada measured the detail with the scale ruler she kept in her desk drawer. She checked the dimensions. She compared the detail on sheet S-22 to the framing plan on sheet S-5.

Something was wrong.

She looked at the framing plan again. The east wall at grid line F was identified by a note: "12" PRECAST SHEAR WALL — SEE DETAIL 4/S-22." She looked at detail 4 on sheet S-22. The detail was titled "DOUBLE-TEE TO SHEAR WALL CONNECTION." The detail showed the connection as a lateral connection — a connection designed to transfer shear forces from the floor diaphragm to the wall — but it did not show the connection as a gravity connection. The bearing angle was there, but the weld was a shear transfer weld, sized for the lateral load only, not for the gravity load.

The wall was carrying gravity load, but the connection was designed as if it were not.

Ada set down her pencil. She looked at the detail. She looked at the framing plan. She looked at the general notes. She looked at the legend, where the symbols for bearing walls and shear walls were defined, and she saw that the legend had one symbol for shear walls — a solid triangle — and one symbol for bearing walls — a hatched triangle — and she looked at the east wall on the framing plan and she saw the solid triangle.

The wall at grid line F was identified as a shear wall only.

It was not identified as a bearing wall.

But it was carrying gravity loads. The double-tees bore on it. The loads were there. The loads were real. The loads did not care what the legend said.

This was the error.

Someone — and the someone was the engineer of record, the person whose stamp was on the drawings, Martin Hardin — had classified the east wall as a shear wall only, had designed the connections for lateral loads only, had sized the welds for shear transfer only, and had not accounted for the gravity loads that the wall was carrying because the double-tees were bearing on it. The wall was a bearing wall that had been identified as a non-bearing wall, and the connections had been designed accordingly, and the connections were therefore undersized for the actual loads, and the loads had been there for twenty years, permanent, patient, waiting.

The dead load does not hurry.

It does not surge or spike. It is there on the first day and it is there on the last day and every day between, pressing down with the same force, and if the connection that receives the load is adequate, the load is invisible, a silent transaction between elements, but if the connection is inadequate, the load is a slow violence, a constant over-stress that manifests first as creep — the gradual deformation of the material under sustained load — and then as fatigue — the weakening of the material under repeated loading cycles — and then as fracture — the sudden failure that appears sudden only if you were not watching.

Ada had been trained to watch.

She stood up from the table and walked to the window. Her apartment faced north, toward the Chicago River, and she could see the Wrigley Building and the Tribune Tower and the Equitable Building and behind them the darker mass of the Merchandise Mart, and each building was a collection of load paths, each load path a chain of elements, each element a promise, and she stood at the window and thought about promises.

Hardin had been her teacher. Hardin had written the textbook she had studied. Hardin had sat in his office in the Wishnick building and told her she was his best student and she had carried that weight for fifteen years, the dead load of his approval, and now she was looking at his drawings and seeing an error that was not a typo, not a misplaced dimension, not a minor inconsistency, but a fundamental misclassification of a structural element, a bearing wall called a shear wall, a gravity load path ignored, and the error had been there for twenty years, embedded in the concrete, and three people were dead.

She went back to the table.

She spent the next three hours going through every sheet, checking every connection, verifying every load path, building the case the way you build a structure — from the foundation up, one element at a time, each element verified before the next is placed. She found four additional connections where the gravity load path was inadequately detailed. She found two instances where the reinforcing in the wall panels was insufficient for the combined loading. She found a note on sheet S-1 that specified a concrete strength of 5,000 psi for the wall panels but a note on sheet S-15 that specified 4,000 psi for the same panels, a discrepancy that could mean a twenty-percent reduction in capacity.

Each finding was a link in the chain. Each link pointed in the same direction.

At 11:00 p.m. she closed the drawings and stacked them in a neat pile on the table and sat back in her chair and looked at the pile and thought about what she knew and what she did not yet know.

What she knew: the east wall at grid line F was a bearing wall that had been classified as a shear wall. The connections were undersized. The welds were inadequate. The collapse had initiated at the third level on the east side, precisely where the load path was most compromised.

What she did not yet know: whether the error was in the design only or in the construction also. Whether the connections had been built as drawn — undersized — or built differently. Whether there had been an inspection during construction that should have caught the discrepancy. Whether Hardin had known about the error. Whether anyone had known.

She needed the construction inspection records. She needed the shop drawings — the detailed fabrication drawings prepared by the precast manufacturer, which would show exactly what was built as opposed to what was designed. She needed the post-tensioning records, the concrete test reports, the weld inspection reports. She needed the full documentary history of the building, every piece of paper that existed between the first design sketch and the certificate of occupancy.

She needed to talk to Hardin.

She did not want to talk to Hardin.

She opened her laptop and composed an email to Richard Alford, the principal of the firm, her boss, a man she respected for his precision and feared for his impatience. She wrote:

Rich — Preliminary findings from Lake-Wabash parking structure. Have identified what appears to be a design error in the original structural drawings. The east wall at grid line F is functioning as a bearing wall but was classified and detailed as a shear wall only. Connections appear to be undersized for the actual gravity loads. Need full calculation package, shop drawings, and inspection records to confirm. Note that the engineer of record is Hardin & Keane. Martin Hardin stamped the drawings. I want to flag this early given the implications. Will have preliminary report by Wednesday. — Ada

She sent the email. She closed the laptop.

She sat at the table in her apartment on the seventh floor and thought about the apartment on Pulaski, the six-flat with the unreinforced masonry bearing walls, the building where she had spent the first eleven years of her life, and she thought about the walls in that building and how she had not known, as a child, that the walls were doing anything, had not known that the walls were working, that they were holding the building up, that the weight of the apartments above hers — the weight of the furniture and the people and the books and the dishes and the radiators and the plumbing and the bathtubs full of water — was flowing down through the walls of her apartment, through the bricks laid in 1923, through the mortar that was now probably crumbling, and the walls had been carrying this weight her entire life and she had not known, because that is what a bearing wall does, it carries the load invisibly, and you only notice it when it fails.

The phone rang at 11:30. She looked at the screen and did not recognize the number, but it was a 312 area code, Chicago, and she answered.

"Ms. Nowak? This is Martin Hardin."

The voice was older than she remembered. Of course it was. Fifteen years had passed since her thesis defense, since the last time she had sat in his office, since the last time she had heard this voice speaking about load paths and failure modes and the responsibility of the structural engineer, and the voice had aged the way a structure ages — the same fundamental character, the same resonant frequency, but with a roughness at the edges, a deterioration in the fine detail.

"Professor Hardin," she said.

"I heard about the collapse. I heard Alford & Associates is doing the assessment."

"We are."

"I want you to know that I am available to assist. I have my files on that project. I want to cooperate fully."

Ada looked at the stack of drawings on her table. She looked at the yellow legal pad with her notes. She looked at the detail on sheet S-22, the shear wall connection that was carrying a gravity load it was not designed for.

"I appreciate that," she said.

"It's a terrible thing. Three people."

"Yes."

"I designed that structure twenty years ago. My associate did most of the calculations — Frank Keane, you may remember him — but I reviewed them. I stamped the drawings. The stamp is mine."

"I know," Ada said.

"I want to understand what happened. I want to help you understand what happened."

There was something in his voice that Ada could not quite identify, a quality that was either sincerity or its careful imitation, and she thought about the difference between a bearing wall and a shear wall, how on the drawings they look similar — both are walls, both are concrete, both have reinforcing — but their function is different, their purpose is different, and the difference is invisible from the outside, the difference is in the load path, in what the wall is carrying, and you cannot determine the function of a wall by looking at it, you can only determine it by looking at the drawings and the calculations, by understanding what loads are applied and how they flow, and people are like this too, opaque, their structural systems hidden, their load paths internal.

"Professor Hardin, I should tell you that my preliminary review of the drawings has identified some concerns regarding the classification of the east wall. I'll need to verify these concerns with the full calculation package before I can discuss them in detail."

A silence. Then: "What kind of concerns?"

"I'd rather not say until I've completed the review. But I wanted to be transparent that concerns exist."

Another silence. Longer.

"I understand," he said. "I'll have my files couriered to your office tomorrow. Everything I have. The calculations, the correspondence, the shop drawing reviews, the inspection reports, everything."

"Thank you."

"Ada." He said her name the way he had said it in his office fifteen years ago, with a particular emphasis on the first syllable, a stress that gave the name a weight it did not normally carry, and she remembered how it had felt to be addressed by this voice, to be the student addressed by the teacher, and the feeling was a dead load, permanent, structural.

"Yes?"

"I need you to know that I designed that building in good faith. I designed it to the code. I designed it to be safe."

"I understand," she said, and she did understand, she understood the statement and she understood the need behind the statement, the need of a structural engineer to assert that the structure he built was sound, because the stamp is not just a signature, it is a declaration — I, the undersigned, have reviewed these documents and believe them to be in conformance with applicable codes and standards — and to have that declaration proven wrong is not a professional embarrassment, it is an existential one, it is a failure of the self, of the load-bearing capacity of the self.

"Good night, Professor Hardin."

"Good night, Ada."

She set the phone down. She looked at the drawings. She looked at the detail on sheet S-22. She looked at the note on her legal pad: Wind exposure — B used, C arguable. Verify.

She added a second note: Hardin called. Cooperative. Files coming tomorrow. Aware of assessment. Tone: concerned, maybe something else.

She put the cap on the pen. She turned off the light over the table. She walked to the window and looked out at the city and the city looked back with its ten thousand lit windows, each one a room, each room a space inside a structure, each structure a collection of load paths, and the load paths were continuous or they were not, and the buildings were standing or they were not, and the people inside them were alive or they were not, and it was 11:45 on a Sunday night in March in Chicago, and Ada Nowak was standing at her window thinking about bearing walls, and the wind off the lake was pushing against the glass with a force she could calculate but chose not to, and the glass held, and the frame held, and the wall held, and the building stood, and she stood inside it, carrying what she carried, which was the dead load of everything she had learned and everything she had seen and everything she had not yet determined, and the weight was familiar, and the weight was permanent, and she carried it to bed.