By Mike Gaetani and M. Abraham
A large number of hospitals throughout the state were constructed prior to 1973, the watershed year in which significant new building code provisions were introduced based on lessons learned from the 1971 San Fernando earthquake. Preliminary screenings performed by structural engineers of older California hospitals exposed approximately 1,000 acute care facilities that do not currently meet life-safety protection expectations.
In an effort to develop and implement innovative engineering approaches to the seismic rehabilitation of hospitals, civil and environmental engineering professor John Wallace and his research group are working closely with KPFF Consulting Engineers, hospital owners and the California Office of Statewide Health Planning and Development to integrate their research into a viable strategy for reducing both the costs and the uncertainty associated with predicting how these aging buildings will react in an earthquake.
Assessments of the expected behavior of these buildings, as well as recommended approaches for upgrading them to meet either minimum or strict performance targets typically are conducted using guidelines established in FEMA reports. The information, however, is based mainly on research conducted prior to 1995, and proves limited in scope due to a lack of available information regarding the expected load versus deformation behavior of typical structural building components – beams, columns, walls, floors, and foundations. And full-scale application of the guidelines tends to produce costly, disruptive rehabilitations that require staged construction over a long period of time.
Wallace’s project team aims to make the process less cumbersome by coupling sophisticated computer modeling with results obtained from building-specific test programs conducted in the UCLA Structural/Earthquake Engineering Research Laboratory, located in the basement of UCLA Engineering’s Boelter Hall.
The laboratory, constructed in 2004, includes a 40-foot by 60-foot strong floor, five feet thick, and large reaction blocks with equipment for simulating earthquake loading on components. During the tests, forces, displacements, and strains are measured to capture the response of the component (e.g., column, beam-to-column connection) to a full range of simulated earthquake actions, from low-level shaking all the way up to the “Big One” and beyond.
The simulations are performed by large-capacity hydraulic actuators that push and pull the test specimen back and forth to produce forces and deformations that would be expected during an actual earthquake. The data then are translated into physical relationships that researchers use for computer modeling of the entire building. The models offer the project team yet further insight and a wide range of options to study as they prepare to develop the final design.
To date, the project team, which includes former PhD students Dr. Kutay Orakcal, Dr. Leonardo Massone, and current MS student Sarah Taylor Lange, and engineers at the Los Angeles and Irvine offices of KPFF Consulting Engineers – including John Gavan (MS 1991), Aaron Reynolds MS (1994), Ayse Kulahci (MS 1999), Peter Sarkis (MS 1997), and Mostafa Sobaih, has completed four test programs, two on reinforced concrete wall segments, one on columns, and one on beam-to-column connections for three different hospital structures in southern California. Several more testing programs are in the planning stages.
Test results so far have confirmed that intrusive, costly rehabilitation measures produced with existing FEMA guidelines are not always required, and have led to the formulation of more economical seismic rehabilitation strategies.
Costs typically can be lowered by using fewer materials at fewer locations (less disruptive) throughout the building and, more specifically, reducing foundation work in the face of high costs associated with digging under a structure to add new foundations.
In the case of one hospital building, the new rehabilitation approach eliminated the need to cut back an adjacent building by two feet over several floors; the conventional approach would have entailed making space for new perimeter concrete walls.
Though the cost of each test program usually runs between $100,000 and $200,000 and the added costs associated with the state-of-the-art computer modeling can reach $0.5 to $1.0 million per building, these costs are minimal compared with the tens of millions of dollars in estimated savings that already have been realized by one of the participating hospitals.
The group already is working on additional test programs and detailed computer modeling studies.