HR: 09:10h
AN: T21B-05 [PDF]
TI: Controls on the Growth and Maximum Size of Fault Arrays and Fault Segments--Insights from Experimental
Clay Models
AU: * Schlische, R W
EM: schlisch@rci.rutgers.edu
AF: Rutgers University, Department of Geological Sciences, 610 Taylor Road, Piscataway, NJ 08854 United States
AU: Withjack, M O
EM: drmeow3@rci.rutgers.edu
AF: Rutgers University, Department of Geological Sciences, 610 Taylor Road, Piscataway, NJ 08854 United States
AU: Ackermann, R V
EM: rolf@beicip-inc.com
AF: Beicip-Inc., 1880 S. Dairy Ashford, Suite 630, Houston, TX 77077 United States
AU: Clifton, A E
EM: amy@norvol.hi.is
AF: Nordic Volcanological Institute, Grensasvegur 50, Reykjavik, 108
Iceland
AB:
We have used a series of scaled experimental models with clay to study the nucleation, growth and linkage of fault arrays and
their constituent fault segments. We have varied the clay thickness and the magnitude, rate, width, and obliquity of
deformation. We have also varied the basal boundary conditions: 1) focussed deformation associated with the edges of moving
metal plates or metal blocks, simulating the effects of fault reactivation on a cover sequence, and 2) deformation above a
rubber sheet, simulating distributed deformation. For small amounts of strain, the lengths of fault arrays and segments
increase with increasing deformation and increase with decreasing deformation rate. Most fault arrays and segments
ultimately achieve a `maximum' length. This length depends on mechanical-layer thickness (increasing as thickness
increases), deformation-zone width (increasing as width increases), and obliquity (decreasing as obliquity increases).
Mechanical-layer thickness and deformation-zone width have an especially strong influence on the 'maximum' length of fault
segments.
The length of the fault array that develops during fault reactivation is strongly controlled by the length of the underlying
reactivated fault, whereas the length of the constituent fault segments is controlled by the factors discussed above.
Although a hard-linked fault array ultimately develops from these segments (given enough displacement on the reactivated
fault), evidence of the location of former segment boundaries is long-lived. This evidence includes breached relay ramps,
fault-displacement folds, and fault strands. The persistence of these features decreases with increasing displacement on the
reactivated fault, decreasing thickness of the cover sequence, and decreasing obliquity.
As most fault populations have a 'maximum' length, we expect that power-law size distributions will be uncommon. Our
experimental models indicate that an exponential size distribution describes most fault populations. An exception is fault
populations that form during orthogonal deformation at low strains. At higher strains, these fault populations also exhibit
an exponential size distribution, with a power-law to exponential transition occurring at higher strains in thicker
mechanical layers. As the maximum length of faults is limited in the experimental models, we expect that the scaling law
between length and displacement will change with increasing strain. This may contribute to the large scatter in experimental
and natural length-displacement data.
DE: 8010 Fractures and faults
SC: Tectonophysics [T]
MN: 2003 Fall Meeting