Flawed Building Codes Are Cited in Collapse
Quake: Defects found at Northridge Meadows, where 16 died,
but defendant’s experts say it met ’70 standards.
Los Angeles Times - June 11, 1994
For almost a month, they chain-sawed and crow-barred their way into the flattened walls of the Northridge Meadows Apartments, where 16 people died in the Jan. 17 Northridge earthquake.
In a pathos-filled exercise in urban archeology conducted amid the litter of rotting food, crushed automobiles and scattered memorabilia, dozens of lawyers, engineers and city building specialists picked apart the rubble, seeking to learn why the three-story complex collapsed.
Attorneys representing former tenants and families of the deceased found what they allege to be design flaws and construction defects from the foundation to the roof of the 163-unit complex. Poorly secured stucco, missing shear walls, skimpy nailing, undersized connectors and materials substitutions showed a pattern of shoddy workmanship, they contend.
Although conceding the existence of some such defects, experts for the defendants downplayed their significance. They said the 25-year-old building met the standards of its time, and was largely a victim of geography, standing nearly above the epicenter amid a cluster of heavily damaged or destroyed apartment buildings.
"It's a simple case, infuriatingly simple," said Seb Ficcadenti, engineer for Northridge Meadows' builder Heller Construction. "A couple hundred yards around the site, five buildings with partial collapse."
Beyond the pretrial posturing, however, the board-by-board dismantling of Northridge Meadows--conducted under court order and costing an estimated $500,000--produced some lessons that are widely accepted.
Foremost is the inadequacy of building codes for two- and three-story wooden structures. The inspection showed that even today's standards, though tougher than those of 25 years ago, fail to provide the level of safety they should, Los Angeles building officials have concluded.
Secondly, the building that killed 16 people was not a lot different from other buildings that survived the Jan. 17 quake, suggesting that there could be a loss of life in future earthquakes.
"There was nothing different," said Nicolino Delli Quadri, chief of the Training & Emergency Management division of the Los Angeles Department of Building and Safety. "Typical construction for that time. There's nothing at this time that we found atypical."
The common ground is that the causes of the Northridge Meadows failure include inadequate seismic resistance built into the walls and over-reliance on structural elements that have been shown to provide less strength than was thought in the 1970. Not only are such flaws endemic to buildings of that era, but, in some cases, today's code still allows them.
Delli Quadri, spokesman for the Department of Building and Safety's post-earthquake analysis team, is also a member of the committee examining wood frame construction. He said the group is preparing recommendations for further tightening of the code for new buildings and a program to strengthen older buildings.
If adopted, a retrofit program would require the owners of hundreds--and possibly thousands--of buildings of the early 1970s and before to remove gypsum board or similar outdated materials from critical walls and resurface them with stronger materials such as plywood.
"A lot of us in the building department had a high confidence in wood frame buildings," Delli Quadri said. Now, they are "looking back, asking questions about the buildings we approved in the past. How safe are they? They could be safer."
Although the final analysis of what made most of the first floor of Northridge Meadows collapse must await extensive calculations based on data collected over the past month, several of the experts now agree on a rough explanation.
They say its design relied upon a medley of structural elements to work together in the building's shear-wall system of seismic resistance: stucco exterior walls, diagonal wood bracing, gypsum board interior walls, plywood walls and a partial steel frame.
But most of those building materials were inherently weaker than the strength rating they were then given in the building code, and, reacting differently to the stresses, the weaker elements failed first, increasing pressure on the stronger one, several of the experts said.
"The code allowed the architect to put in stucco, drywall, plywood," Delli Quadri said. "The stucco and drywall gave up. The rest of the load went into plywood panels. Overwhelmed, they failed."
As litigation progresses, the questions attorneys raise about Northridge Meadows will most likely focus narrowly on the building standards of 1970: Did the design satisfy the requirements of the code? Did the builder follow the design?
Joel Castro, the lead attorney in a tangle of lawsuits filed by several tenants and the families of dead tenants, asserted that the evidence resoundingly answers both questions no.
As his experts wrapped up their inspection late last month, Castro presented their findings as a catalogue of flaws.
"You're looking at what we consider some very substantial construction problems," Castro said. "And then the design--it's just inadequate shear walls to support these loads."
Among the problems Castro alleged:
- Design of tuck-under parking. The architect designed the first floor to be divided between living area and parking, leaving the second and third floors suspended over long open spaces instead of rigid walls. In two sections, cars could be parked two deep, leaving more than half the first floor open, without the honeycomb support of the interior walls. Special plywood shear walls were supposed to add strength. They were inadequate, however, and failed.
- Use of steel frame is design. Plans called for a large steel I-beam to run lengthwise across the parking areas. One end rested on a heavy steel column, but the other end of the beam was spliced into a large wood beam. The other end of the wood beam rested on another steel column. In the earthquake, the wood beam broke away from the supporting steel column. Smaller pipe columns that supported the beam in the middle either bent or disconnected. Castro contends that the steel frame could have provided horizontal resistance to an earthquake, but wasn't designed properly.
- Installation of stucco. A thick layer of plaster laid over wire mesh, stucco provides a rigid exterior skin that resists distortion of a building's shape. At Northridge Meadows, large sections of stucco separated from the wooden frame when the staples holding the wire mesh to the framing pulled loose. Investigators found that in many cases, there were no staples along the top and bottom, the critical points.
- Installation of drywall. Gypsum drywall on interior walls helps to hold the building together. Because the chalk-like filling crumbles easily, nails must be driven at close intervals to provide cumulative strength. In hallways, required by code to have double thickness drywall, the lower layer was only tacked in place, rather than nailed thoroughly. No nails were placed along the top and bottom edges. Investigators found that many sheets of drywall were torn off the walls, leaving nails behind.
- Foundation attachments. The building code required bolts a minimum of every 6 feet to secure the framing to the foundation. In spots, investigators found bolts spaced almost 10 feet apart. In addition, brackets designed to secure walls to the foundation at corners, which were not necessarily required by code, were not used or specified in the design, investigators found.
- Framing. In places, only minimal nailing was used to attach floors to wall, allowing the joint to give. The joints could have been strengthened with short 2-by-4s, called horizontal shear blocking, nailed into both parts. In addition, Castro's team also identified several deviations from the architect's plan, such as wood planking substituted for plywood, the use of smaller metal hangers, the absence of plywood in some walls, and structural breaks caused by an elevation change in the foundation.
In interviews, engineers representing defendants in the lawsuit rebutted Castro's asserting one by one, either minimizing the effect of the alleged deficiency or justifying it as an accepted practice of the time.
For example, they pointed out, the code did not require drywall to be nailed on top and bottom. The building code also allowed the wall-to-floor joint to be secured by nailing only, without the short 2-by-4 block Castro said should have been used.
Code revisions since 1970 have increased the maximum seismic forces that engineers must take into account and reduced the seismic capacities of stucco, drywall and plywood walls. Diagonal bracing, one of the main lateral stiffening methods in the complex, was judged after the 1971 Sylmar earthquake to be of little seismic value and is no longer allowed in calculations of seismic resistance.
The current code also prohibits the practice of mixing different types of rigid material in a shear wall. Thus, walls at Northridge Meadows that relied on stucco, diagonal bracing, plywood and steel together to meet seismic requirements would be considered to have significantly less seismic resistance under current code.
Further, experts for the defendants said, the code in force in 1970 made no distinction between plywood or diagonal planks for floor sheathing and did not specify which joist hangers should be used.
Bill Grinell, a contractor and architect hired by the attorneys for building owner Shashikant J. Jogani, contended that the floor joist--as well as the I-beam splice--separated under the weight of the falling building rather than from the shaking of the earthquake.
Some questions about the construction of Northridge Meadows remain because neither its architect, Morris Brown, nor its engineer, David Weiss, have spoken publicly about it.
Brown's attorney, Jay Sherman, said he had only been retained last week (5/and could not make detailed comments. He said the design complied entirely with the code then in force. Sherman said the structural engineer, not Brown, was responsible for the building's ability to withstand the earthquake.
Weiss has not produced his calculations, and attorneys for both sides said they don't expect to obtain them. Neither Weiss nor his attorney returned calls from The Times.
Without the original calculations, observers can only speculate how the engineer intended the building would resist seismic forces.
But the tuck-under parking, which leaves one side of the building with no lateral strength, has emerged as a concept that all agree posed special problems. Ficcadenti said he believes the engineer came up with the steel beam system, embedded in a reinforced concrete footing to make up for the lack of interior and exterior walls. Thus, one side of the building relied on steel for its rigidity, and the other on conventional wood and plaster.
In the earthquake, the different materials reacted differently under stress, causing the weaker elements to fail first. Ficcadenti said the understanding of the dangers of this phenomenon called "vertical stiffness differential," were derived from the experience of the 1971 earthquake. Since then, the code has been changed to require shear walls to be made of only one material.
Independent experts have also focused on tuck-under parking--prevalent in apartment designs of the 1970s--as a significant problem.
"I guess we all have come to the same conclusion," said Helmut Krawinkler, co-director of the John A. Blume Earthquake Engineering Center at Stanford University. "In all those practices where the first floor is wide open because of tuck-under parking…it is really no surprise that those soft stories collapse.
"In other buildings I looked at, the entire first floor was carport. People didn't get crushed. In Northridge Meadows, part [was] parking and part apartments. It was that mixture that caused the casualties."
Although Northridge Meadows may not have been unique in its failings, the tragedy of its 16 deaths may help the city take steps more swiftly and decisively that after the 1971 quake, when the major alarm was raised over unreinforced masonry buildings.
After examining dozens of apartment buildings damaged or destroyed Jan. 17, Los Angeles city building officials are preparing recommendations for a major revision of the code covering two- and three-story wood construction.
Among the revisions Delli Quadri said the department is considering:
- Downgrading the strength rating given to stucco, drywall and plywood walls in seismic calculations.
- Prohibiting the use of staples to secure stucco.
- Requiring the manufacturers of connecting and fastening devices to subject them to cyclical testing, which measures performance under repeated stress, rather than the static testing currently done.
- Requiring all framing on first floors of three-story wood buildings to be made of boards three inches thick instead of the 2-by-4s or 2-by-6s that are currently standard. Often, the nails holding plywood sheathing to the framing merely pulled out when the studs split, Delli Quadri said.
But all the proposals, he anticipates, will run into opposition from some builders and apartment owners who don't want the added expense and complication.
"We're going to run into all kinds of hell to change it," he said. "They don't like to do it."