Licensed not to kill

September 2005 » Feature Article
The 1994 Northridge earthquake changed the rules on steel momentframe connections because the connections didn't behave the way engineers were taught they should. The Federal Emergency Management Agency (FEMA) acted swiftly, sponsoring the SAC Steel Building Project and publishing the FEMA 350 Report. As a result, the pre- 1994 connection was outlawed in all the building codes relating to new construction, but what about existing buildings?

Evaluating your moral obligation

By Gary C. Hart, Ph.D., C.E.

The 1994 Northridge earthquake changed the rules on steel momentframe connections because the connections didn't behave the way engineers were taught they should. The Federal Emergency Management Agency (FEMA) acted swiftly, sponsoring the SAC Steel Building Project and publishing the FEMA 350 Report. As a result, the pre- 1994 connection was outlawed in all the building codes relating to new construction, but what about existing buildings?

No structural engineer wants to advise clients to spend more money to make the proper repair, especially when competitors will do it the old way for less. But pleading ignorance is not acceptable.We have a moral obligation to protect lives and uphold the highest standards of our profession.

The facts Following are the facts as I see them: • Pre-1994 steel beam-column connections used for steel moment-frame buildings are now outlawed. They are not allowed in any building code because they do not provide acceptable levels of life safety.

• The public trusts structural engineers because we have a professional code of practice.

• Owners and insurance companies, as a general rule, are profit-driven businesses and usually do not want to spend one dollar more than they have to by law. The exception, however, is when they are forced to spend money for their own benefit or for protection against a lawsuit.

• The only recommendation that an engineer can justify morally and professionally when asked to “repair” existing steel moment-frame connections is one that meets accepted life safety standards made to the best of today's state of knowledge.

These facts have an impact on the real world. When a structural engineer is asked to provide forensic services for an insurance company or an attorney seeks damages from an insurance company, the structural engineer always is asked two questions. One question is, “What is the cause of the damage?”The other question is, “What is your recommended repair?” The answer to these questions usually has profound financial import, as demonstrated in the aftermath of the 1994 Northridge earthquake.

This article could present many post-Northridge cases with pictures showing the locations of weld fractures and graphs describing the damage.

However, while it would be visually interesting, it would take away from the fundamental dilemma posed by this article, which is: Should a structural engineer design a repair knowing in advance that if a code-level earthquake were to occur, people probably would die because the repair was insufficient? The findings In structural engineering, it is essential to learn as much as possible from the field investigation of damage and from the scientific study of the possible causes of that damage. Field investigations and scientific studies of damage are carried out around the world by professional society teams - such as the Earthquake Engineering Research Institute, the Los Angeles Tall Buildings Council, and the American Association of Wind Engineering - and by state and U.S. government teams. The primary objectives are to observe the damage and communicate visual images of the damage to other professionals, to determine the cause of the damage, to validate structural engineering models for estimating the damage, and to recommend changes - if necessary - to building codes.

The structural engineering profession learned many things from investigations made by professional society and government teams after the 1994 Northridge earthquake. Perhaps the most visible effect of their observations and subsequent failure analysis studies was their recommendation to revise the design criteria for steel moment-frame buildings.With unprecedented leadership and concern for public safety after the 1994 Northridge earthquake, FEMA funded an in-depth study of the steel moment-frame lateral force resisting system. This study commonly is referred to as the SAC Steel Building Project.

The SAC Steel Building Project was a combined effort of the Structural Engineers Association of California (S), Applied Technology Council (A), and the California Universities for Research in Earthquake Engineering (C). Therefore, the project involved nationally respected structural engineers, professors, and industry representatives.

The seriousness of the apparent problem with steel momentframe buildings was perhaps only exceeded by the surprise that there was even a problem. It is very important to state clearly that the typical momentresisting steel beam-column connection detail, used prior to the 1994 Northridge earthquake, did not perform as expected in the 1994 Northridge earthquake. The evidence is now sufficient to conclude that this connection has an unacceptable level of risk associated with its brittle fracture.

The FEMA 350 report says that the connection “had a number of features that rendered it inherently susceptible to brittle fracture.” The connection was shown in the SAC study to fail to provide acceptable performance for use in building design and, in the words of the FEMA 350 authors, “The connection, as it was being used in contemporary practice, was inadequate for the anticipated seismic demands.” The bottom line is that the pre-1994 steel beam-column connection is now outlawed by all building codes.

Specifically, it is not included in the set of pre-qualified steel beam-column connections that FEMA 350 recommends for use in the design of new steel moment-frame buildings.

FEMA 350 is one of those landmark reports that comes along every decade or so. It is exceptional because it identified the cause of the problem with the pre-1994 connections, it presented a group of acceptable pre-qualified connections for structural engineers to use, and it made a great step forward by presenting a scientifically sound performance evaluation methodology to evaluate any connection that is proposed by a structural engineer. For example, assume that a structural engineer is unfocused enough to use the performance evaluation methodology presented in FEMA 350 to determine if the pre-1994 steel beam-column connection and associated lateral force resisting system is acceptable. The structural engineer would find that the calculated level of confidence for protecting the building against collapse is not even close to an acceptable value.

The use of FEMA 350 to perform this calculation shows in clear and quantifiable terms that the connection fails to meet today's standards and poses an unacceptable level of risk in terms of collapse prevention and life safety.

The argument Now let's return to the beginning of this article and the real world of existing buildings, insurance companies, lawyers, private building owners, and to the fact that state and federal governments own many buildings. Assume that now we know the cause of the damage to the steel beam-column connections in a particular steel building.

The question still remains:What is the recommended repair? Prior to the publication of FEMA 350 and the associated reports, one could be justified perhaps in re-welding the damaged connections. Some call this returning the buildings to their pre-existing condition, only better.

This is what was done to ‘upgrade' many steel moment-frame buildings after the 1994 Northridge earthquake.

However, now that FEMA 350 has been published, the question must be asked again.

Consider first the case of the property owner or the insurance company.

Typically, these stakeholders only want the problem to go away at the minimum expense and inconvenience. The solution as they see it could be arrived at through the following FEMA 350 statements: • “For many years, the basic intent of the building code seismic provisions has been to provide buildings with an ability to withstand intense ground shaking without collapse.” • “Following that earthquake, a number of steel moment-frame buildings were found to have experienced brittle fracture of beam-tocolumn connections.” • “In general, steel moment-frame buildings damaged by the Northridge earthquake met the basic intent of the building code. That is, they experienced limited structural damage, but did not collapse.” Therefore, some could interpret FEMA 350 to state that a re-welding repair past or future is all that is needed.

The position could be defended further by the fact that this is an existing building and not a new building and the code addresses itself to new buildings.

No steel moment-frame building has collapsed, so the re-weld solution is acceptable. Furthermore, the money for the repair has already changed hands and all are agreed to the settlement.

This close-the-door position is appealing to the insurance company because it does not open a closed file or require any more money. It is appealing to the building owner who typically wants to sell the building and does not want to have to deal with any more steel moment-frame connections. It is appealing to the state and federal governments because they have inventories of thousands of buildings with pre-1994 steel moment-frame connections and competition for funding is limited.

The point Now consider the flip side. A structural engineer submits a repair to the building department that does not meet the well-defined minimum collapse prevention confidence level of FEMA 350. How can a structural engineer refuse to design a repair when almost certainly one of his or her fellow engineers from a “good” structural engineering firm will do the repair with no concern for collapse? On the other hand, should a structural engineer be exposed to the probable liability of a lawsuit if FEMA 350 is not considered and collapse occurs? I believe that as a structural engineer, my fundamental concern is for the life safety protection of the people in the building. It is clear that when people enter a building - especially a tall, modern, steel moment-frame building - with nice architecture and expensive furnishings, they do not believe that they are putting their lives at risk. They trust us to do what is morally right because we have a professional code of practice and we are engineers.

Therefore, the only recommendation that is morally and professionally justified is that repairs must be made to the moment-frame system that meet the accepted life safety standards to the best of today's state of knowledge. In the words of FEMA 350, this means that the repair must “be determined to provide a 95 percent level of confidence that the structure will provide Collapse Prevention or better performance for earthquake hazards with a 2 percent probability of exceedance in 50 years.” Conclusion A gray area will always exist when codes are improved only for new construction. Engineers who practice in ‘eye of the storm' states such as California and Florida should lead the charge for exceeding code requirements, if and when it makes sense to do so. Taking the line of least resistance puts the engineer in a vulnerable position.

More important, it is immoral and could lead to a tragic loss of lives.

Gary C. Hart, Ph.D., C.E., is principal and division director of Hart-Weidlinger, a Division of Weidlinger Associates, Inc. In addition, he is Professor Emeritus at the University of California, Los Angeles. He can be reached at

Upcoming Events

See All Upcoming Events