Seismic Impact Provides Hard Data
Civil engineer Brady Cox subscribes to a free notification system, offered by U.S. Geological Survey, that sends an e-mail or text message containing considerable seismic data within minutes of an earthquake anywhere in the world. Cox has set his own parameters – magnitude 6 and above for quakes outside Arkansas, magnitude 2 and above for activity inside the Natural State.
While in his lab on the evening of Jan. 12, about 20 minutes after the Earth ruptured in Haiti, Cox received one of these e-mails.
“I always look at the most basic information first – magnitude, depth of the hypocenter and distance from the epicenter to the nearest city,” he said. “When these basic facts sunk in – and knowing that Haiti was an impoverished country with many poorly constructed or un-engineered buildings – I knew it would be bad, very bad.”
Three weeks later, Cox, a geotechnical engineer and assistant professor of civil engineering, stood on the lawn of the Presidential Palace in Port-au-Prince. He had traveled to Haiti with nine other members of Geo-engineering Extreme Events Reconnaissance (GEER), an organization funded by the National Science Foundation to conduct reconnaissance efforts of extreme events such as earthquakes, tsunamis and hurricanes. GEER missions augment researchers’ understanding of the effect of earthquakes in general but also provide hard data that will aid the design of new earthquake-resistant structures that limit damage to buildings and could save human lives.
In Haiti, Cox and his GEER colleagues documented the geotechnical and structural impact of the earthquake by mapping and surveying damaged areas. They examined damage patterns, port facilities and coastal infrastructure. Specifically, the researchers studied examples of liquefaction – solid ground turning into liquid – lateral spreading, surface faulting, coastal uplift, road-fill performance and landslides.
As an expert in soil dynamics, earthquake loading and nondestructive material characterization using stress waves, Cox contributed to several of these areas, although he focused on damage patterns, the category with the greatest impact on human life. Cox wrote this section in the organization’s nearly 100-page mission report.
Mosaic of Destruction
For six days, Cox and other GEER researchers raked Port-au-Prince, Haiti’s capital and the focus of most of the property damage and casualties. What they found was, to use his expression, “some weird stuff,” which is typical for damage caused by major earthquakes.
Some of the worst destruction occurred in a multi-block area immediately north and west of the well-documented Presidential Palace, which itself suffered major damage. Many buildings and in some cases whole blocks of structures within this section collapsed entirely or crumbled down to their foundations. Yet some structures within this area had only minimal damage or none at all.
Another section of the city, a mostly residential area near the heavily damaged zone, experienced only minor damage to a handful of structures. While this is surprising, most of the buildings in this lightly damaged area were smaller, densely spaced shanties with tin roofs, while most of the badly damaged buildings had been large, multi-story, reinforced-concrete structures.
In other areas of the city, namely the foothills where the effects of topography came into play, damage seemed to run along distinct boundaries. Cox also observed sections in the foothills that experienced severe damage, while other, nearby sections at the same elevation suffered little or no damage.
These observations – virtually unscathed buildings in areas of heavy damage, relatively unharmed neighborhoods adjacent to destroyed areas, and damage according to seemingly arbitrary boundaries – make it difficult to determine exactly what forces caused the damage, which is a critical part of Cox’s job. Before he can identify and explain damage patterns, before any kind of rigorous analysis can occur and long before he can begin to understand the “weird stuff,” Cox must consider these critical factors:
- the distance of the damage from the fault rupture,
- the direction of the rupture,
- the building construction type and quality,
- the topography,
- and, perhaps most importantly, local soil conditions, including age, strength and stiffness of the foundation soil as well as its depth to the bedrock.
“So there are many factors that affect and really determine damage patterns during earthquakes,” he says. “And what we often find is that multiple factors combine to create a complicated mosaic of destruction.”
Shear Wave Velocity Profiles
Earthquakes produce three types of waves: compression, shear and surface. They may be thought of as cars on a train, with compression waves as the engine and surface waves as the caboose.
Energy released by the initial, violent rupture creates compression waves, which travel quickly through the earth and arrive first at the ground surface. Shear waves follow compression waves. They move slower and have larger amplification, which causes horizontal or side-to-side shaking. Surface waves arrive last and typically have smaller amplification than shear waves. Surface waves produce rolling-type motions similar to waves produced by a rock thrown into a pond. While all seismic waves can cause damage, shear waves are predominantly responsible for the violent shaking that damages man-made structures.
The reason for amplification, Cox explains, has to do with soil stiffness and layering. When fast-moving and focused shear waves traveling through rock hit soft soil layers, they immediately slow down and create a vibrating effect known as impedance contrast. In other words, the different layers of soil impede and amplify waves, very much like the effect of a hard-thrown baseball upon impact with an aluminum bat. In fact Major League Baseball players cannot use aluminum bats because the high impedance contrast between the ball and the aluminum would result in too many home runs and more injuries to infielders.
Specific characteristics of the Haiti earthquake – distance from the fault and direction of rupture propagation westward and away from Port-au-Prince – compelled Cox to focus on construction quality, topography and especially soil conditions as the best sources for explaining damage patterns. With severe time constraints and limited mobility because of widespread damage throughout the city, Cox and other researchers kept GPS track logs and identified the best possible locations to conduct spectral analysis of surface waves tests. This is a geophysical technique in which seismometers – special sensors placed on the ground surface – measure the speed of seismic waves that are manually sent through the ground by striking the surface with a sledgehammer. The technique allows researchers to obtain quantitative evidence of soil stiffness and layering.
Cox conducted several of these tests at the port, three within the heavily damaged downtown area, including two tests on the grounds of the Presidential Palace, and several more at liquefaction-failure sites along the coast west of Port-au-Prince. Although information gleaned from these tests was helpful, Cox knew they did not tell him enough about the ground under Port-au-Prince. With spectral analysis testing only, he could not thoroughly evaluate potential amplification of the seismic shaking that may have contributed to the damage patterns. Rough and suspect geologic maps of the area only complicated the picture. Cox knew that more information would be needed to properly characterize the city for rebuilding according to seismic design codes.
In late April, Cox, doctoral student Clint Wood and several other researchers returned to Haiti to collect more seismic data that may be used in rebuilding efforts. Although not an official GEER mission, the second trip was also sponsored by the National Science Foundation. This time Cox schlepped more equipment and conducted multi-channel analysis of surface waves testing, which, as the name suggests, uses multiple seismometers, or geophones, to sense surface waves propagating through the ground.
The multi-channel analysis tests allowed Cox to develop shear wave velocity profiles, a term geotechnical engineers use to define and classify soil stiffness and the expected performance of a given tract of land during shaking or vibration caused by an earthquake. The profiles tell how fast earthquake-generated shear waves travel from hard bedrock up to the ground surface. Changes in the speed of wave propagation caused by alternating layers of hard and soft soil can cause either amplification or attenuation of ground shaking. The profiles also help engineers estimate site resonances, or natural frequencies, that can cause extreme shaking at the ground surface.
Considering the basic information – poor building construction, magnitude, proximity of epicenter to population center, depth of the hypocenter – it is easy to understand why so many buildings collapsed and so many people died. Simply put, the Haiti earthquake was a powerful and shallow event less than 20 miles from a major city. But this understanding is superficial. Almost every expert will argue that the destruction did not have to be so extreme, and lives could have been spared.
How so? At the risk of directly comparing earthquakes, which rankles geologists, seismologists and geotechnical engineers almost as much as predicting earthquakes, much can be learned from the 1994 Northridge, Calif., earthquake. It too was a powerful (magnitude 6.7) and shallow event that occurred very near a large population center. Despite the similarities, the Northridge earthquake killed only 61 people.
Of course, there are many variables and subtle differences that distinguish Haiti from Northridge, but one thing is certain: Many more people would have died in Southern California in 1994 if buildings had not met the specifications of the International Building Code. While emphasizing that he is not a structural engineer, Cox said that most failed buildings he observed in Port-au-Prince did not appear to have features required in the building code.
It takes money to build to code, which is something the Haitian government will have to contend with, but Cox’s work will enable the country to make informed choices when rebuilding.
With information supplied by the multi-channel analysis tests and shear-wave velocity profiles, Cox is developing what he calls a “seismic micro-zonation” of Port-au-Prince. The system will provide specific and detailed recommendations for seismic design of structures on a site-by-site or tract-by-tract basis throughout the entire city. The recommendations will be based on site classifications – A for “hard rock” through F for “liquefiable” soils – and other profiles specified in the International Building Code.
“The classifications correspond to design features that must be included to resist forces in an earthquake,” Cox says. “Haiti can use the system to determine structural design parameters based on soil conditions.”
Cox isn’t ready to say exactly which factors – or combination of factors – caused the most damage, but it is apparent that some areas of Port-au-Prince suffered from amplification due to soil layering. However, the effect of amplification alone on damage patterns may not be as great as expected. His initial findings show that soil under the city is mostly “competent,” meaning dense and stiff.
Other parts of the city, such as those at higher elevations, clearly experienced topographical effects. Cox said it is possible that shear waves channeled themselves through valleys or ran along ridges, which might explain the arbitrary boundaries between heavily damaged areas and adjacent areas that experienced little or no effects.
And then there’s building type and quality of construction. Why, in the middle of zones that experienced massive destruction, did some buildings survive with only a scratch? And why, especially in areas that experienced the effects of amplification, were whole neighborhoods spared while adjacent areas were heavily damaged? Without committing definitively, Cox attributed these outcomes to the effects of building type and construction quality.
“Evidence of this was apparent on blocks where heavy, steel-reinforced concrete buildings collapsed while right next to them, wood-frame structures held up just fine,” he says.
Cox will continue going wherever GEER sends him. The research, he says, is meaningful on a personal level.
“I love my work, but if it saves even one life in future earthquakes, it will be even more gratifying,” he said.