Home Research Facilities People Courses Vita


Links Below Access Subdivisions

Former Res.
Ongoing Res.
Future Res.

Introduction. My research interests currently focus on understanding the mechanics of earthquakes. My research involves a combination of laboratory experiments on the frictional behavior of rock and computer modeling of earthquakes that employs the constitutive relations that arise from the laboratory experiments. A focus of the computer modeling is to determine if there are any signals that might be used for sort- or intermediate-term earthquake prediction. The laboratory experiments are focused on trying to understand the processes responsible for the observed frictional behavior and what constitutive relations can be used to describe this behavior. This is of intrinsic fundamental interest, and it is also necessary to have this understanding of the operative processes if the laboratory experiments are to be applied to natural faults with any confidence.

Much of my research is also summarized in the Department of Geological Sciences description of our program in Structural Geology.
The laboratory experiments are conducted in a unique rotary shear rock deformation apparatus that I have designed and built at Brown. It allows unlimited shear displacement at elevated confining pressures, a capability that produces deformation structures that resemble those found in natural fault zones.

Rate and State Friction. The reasons for studying rock friction are both to understand it for its own sake and to use that knowledge to help understand earthquakes. In order to see why it is so interesting and important and to provide background for our friction research, a brief description of the frictional behavior of rock is presented in the next paragraph.

Most rocks have similar coefficients of friction (~0.6), but the coefficient itself in not what is important for whether sliding will be unstable (in the Earth this means earthquakes) or stable (fault creep in the Earth). There are second order effects of time, slip velocity, and slip displacement on the friction coefficient that turn out to have first order influence on sliding stability. Most people are familiar with the idea that sliding friction is lower than sliding friction. It turns out that static friction increases with the time of being static, and that sliding friction frequently decreases with increasing steady-state sliding velocity. The effect of velocity is complicated, in that there are two competing effects with opposite signs, and the net steady state effect depends on their relative size. Rocks for which the friction increases with sliding velocity (velocity strengthening) cannot initiate earthquakes, whereas if the resistance decreases with velocity (velocity weakening), earthquakes are the expected behavior. There is also a dependence of friction on the slip displacement, following changes in velocity. The displacement over which the friction evolves with slip is called Dc. The process that causes static friction to increase with static time is the same one that causes velocity weakening. This effect is sometimes called the evolution effect since the friction evolves with time and displacement. Some aspect of the frictional surface called its "state" is said to evolve. Thus, it is important to understand the meaning of this state, or equivalently the cause of the evolution effect, and how to extrapolate behavior in the laboratory to natural faults This extrapolation depends on knowing what the process is, so we can be sure it is the same in both settings, and it depends on knowing how Dc changes in going from the lab to the field, since Dc determines important details of fault stability and earthquake nucleation. Consequently, many of our lab experiments are focused on trying to understand the origin of the evolution effect, and both those and our field studies have focused on trying to understand how to scale Dc. A more detailed summary of rate and state friction can be found in my review papers (Tullis, 1988; 1996), listed under Publications.

Computer Modeling of Earthquakes. The computer modeling is done on a variety of computer facilities: two IBM RS/6000 workstations in my laboratory,  a SP supercomputer at Brown, and supercomputers at remote national and international facilities.

Natural Faults. In addition to the laboratory-experimental and computer-modeling studies, I am interested in studies of natural faults to help us understand the faulting and earthquake process. Study of faults not only tells us about the Earth, it aids in guiding what is done in the lab and modeling studies so they are relevant to understanding natural processes.

Other interests. I have a variety of other research interests that do not currently involve active projects, but that I have either worked on in the past and/or may work on in the future. Follow the links to former, ongoing and future research interests/projects to see the range of my interests and that student's and visitors might pursue. Publications resulting from former and ongoing research can be seen by clicking here. Another way to learn more about various research projects is to click on the links in the list below:

Topics Discussed in More Detail in Subdivisions Listed to Left 

Friction at high slip velocity
Numerical earthquake simulation problems
Friction in the presence of fluids
Friction at elevated temperatures
Implications of fault-normal displacements for origin of evolution effect
Fault roughness
Fault roughness and Dc
Large displacement behavior of gouge and bare surfaces
Earthquake prediction, former
Earthquake prediction, future
Field studies of pseudotachylites
Fault permeability

Study of samples from a drill hole into the San Andres Fault
Measurement of stress in the Earth
The driving forces of plate tectonics
The origin of slaty cleavage and schistosity
Thermodynamics of nonhydrostatic stress