Stanford University

News Service



CONTACT: Stanford University News Service (650) 723-2558

Shake test of low-cost adobe reinforcement system takes place at Blume Earthquake Center


At the count of zero, the scale model adobe building begins shaking violently. After a few seconds, one of the building's walls falls with a loud crash. As it goes, it pushes over the wooden latticework that serves as a roof, which in turn sends the gable at the other end of the building plummeting to the ground.

The model, one-fifth the size of a typical one-story adobe building, failed disastrously when subjected to scale forces equivalent to a magnitude 6.5 earthquake. That response is typical of actual unreinforced adobe structures, and is the reason why adobe has a reputation for being an extremely hazardous building material in earthquake zones.

The unreinforced model, which was tested to the point of destruction Wednesday, May 10, at the Blume Earthquake Engineering Center, was a control against which to compare a series of tests performed on an identical but reinforced structure that did not collapse, despite experiencing accelerations comparable to those of a magnitude 8 great earthquake.

These tests strongly suggest that relatively inexpensive and unobtrusive reinforcement methods can greatly reduce the likelihood that adobe structures will collapse even in very strong quakes, said E. LeRoy Tolles, who earned his doctorate from Stanford in civil engineering and is a principal in the company Earthen Building Technologies, which conducted the tests.

If half-scale model tests that will be conducted in Macedonia next fall confirm the results of the Stanford experiments, as Tolles and his colleagues expect, the project could lead to extensive seismic retrofitting of historic adobe structures like the California missions.

In addition, the earthquake engineer maintained, this technology also could save thousands of lives in the Third World, where earthen buildings are commonplace. He said that he intends to seek funding from various sources to apply the technology for this purpose as well.

In addition to Tolles, the research team that tested the models included Frederick A. Webster, also a Stanford Ph.D. engineer, and Edna E. Kimbro, an architectural conservator, from Earthen Building Technologies; William S. Ginell, head of architecture and monuments conservation research at the Getty Conservation Institute which is funding the study; and Chris Thomas, the craftsman who built the models.

"We started this project in 1991 with the premise that what we call stability-based methods would work better with adobe than the more usual approach of trying to strengthen them, and the results of these tests basically support that," said Tolles.

The object of the stability-based approach is not to prevent the brittle adobe from cracking, but to prevent walls from overturning and to keep adjacent blocks from moving out of line from each other so the building can survive repeated episodes of ground shaking, such as occur in a major earthquake with several strong aftershocks.

"This is one of two projects in seismic retrofitting that the Getty Conservation Institute is currently supporting," said Ginell. "This project is designed to test seismic retrofitting techniques suitable for historic adobe structures." There are about 350 such structures in California, and thousands more in New Mexico and Arizona. In addition, such structures are extremely numerous in Mexico and South America.

The engineers used the one-fifth scale model to test several stability management techniques:

  • Two walls were reinforced with a series of vertical and horizontal nylon straps placed inside and outside of the adobe walls and fastened together by straps that run through holes drilled through the wall. The strap technique has an unusual origin. It was invented in 1918 in Guatemala. After a major earthquake in 1917, some residents reinforced their adobe homes by wrapping them with barbed wire. After another major earthquake in 1976, a survey of the surviving buildings found that these wire-wrapped buildings had survived remarkably well.
  • The other two walls were treated with another, slightly more expensive reinforcement method: drilling holes vertically and horizontally through the walls and inserting fiberglass rods set in epoxy grout.
  • In addition, the model was retrofitted with extra stiffening of the roof and attic rafters. The roof was tied to the adobe structure by clamping the roof edges along the top of the gable wall, while the rafters were attached to a horizontal strap that ran around the exterior of all four walls.

"We've all been surprised at how well the strap method appears to work," Tolles said. "From a life safety basis, it appears that we can use these methods to bring adobe up to the level of a wood-frame building, which is the best there is. However, the cost of repairing an adobe structure will always be much higher than for most types of buildings."

For a typical adobe structure, this kind of retrofit might run about $50,000 to $100,000. In the case of the strap method, much of the cost comes from patching and repainting the structure to hide the straps. Although center coring is more difficult and expensive, it doesn't leave any external marks that must be covered over.

Kimbro of Earthen Building Technologies plans to write a report that contains guidelines advising owners on how to seismically reinforce historic adobe buildings.

"If we can make these methods inexpensive enough, and unobtrusive enough, then the owners will have no reason not to use them," Kimbro said. "Take the state of California. It has about 70 historic adobe structures. Every day busloads of schoolchildren visit them. In their current condition, not only the buildings but also the children are at risk if an earthquake occurs. But the state is moving very slowly to reinforce these buildings because of the costs involved."

The recent tests were the last in a series of nine model tests that have been run at the Blume Center. The roughly 5-foot- square buildings were built out of special, miniature adobe blocks by Chris Thomas. Each model took about 30 hours to construct.

The miniature reinforced adobe building in the latest tests was subjected to a sequence of ground accelerations patterned on a 1952 earthquake in Bakersfield that lasted 20 seconds. In the first test, the building was subjected to a relatively low acceleration level, a scale force equivalent to a magnitude 5 earthquake. In each succeeding test, the researchers stepped up the acceleration by 30 percent. By the tenth test, the scale forces were about as strong as those in a great quake, although the duration was kept at 10 seconds, while the shaking in a magnitude 8 quake can last more than a minute.

In between each test, the engineers carefully inspected the model, marking and numbering the cracks that had developed with black felt-tipped markers. They also measured any offsets that developed between different chunks of adobe and photographed each side of the model to document the effects. By the end of the tenth test, the structure was riddled with cracks, and a few chunks of adobe had fallen out of the walls, but the building remained standing.

The final experiments in the project will take place next fall. Two half-scale models comparable to the two final fifth-scale models will be tested at a much larger shake table at the Institute of Earthquake Engineering and Engineering Seismology in Skopje, Macedonia.

"We need the larger test because gravity does not scale accurately," explained Webster of Earthen Building Technologies. For example, despite the fact that the shake table only subjected the model to horizontal accelerations, the model visibly jumped off the concrete slab it was sitting on. "That wouldn't happen with a real structure because of the weight of the adobe," Webster said.



This is an archived release.

This release is not available in any other form. Images mentioned in this release are not available online.
Stanford News Service has an extensive library of images, some of which may be available to you online. Direct your request by EMail to

© Stanford University. All Rights Reserved. Stanford, CA 94305. (650) 723-2300. Terms of Use | Copyright Complaints