SCALED EXPERIMENTAL MODELS OF EXTENSION: DRY SAND VS. WET CLAY

For more than a century, geologists have used scaled experimental models to simulate extensional deformation. Today, dry sand and wet clay are the most common modeling materials. Three common basal boundary conditions are pre-cut blocks that simulate a dipping normal fault at depth, overlapping plates that simulate a detached normal fault at depth, and a rubber strip that simulates distributed extension at depth. We have conducted identical experiments with these three boundary conditions using sand and clay, the results of which show broad similarities and significant differences. In both sand and clay models, normal faults generally strike perpendicular to the extension direction. However, the faults are steeper and more planar in sand models than in clay models. Fault spacing and fault-zone width are greater in sand models compared to clay models. Faults rapidly propagate upward and along strike in sand models, whereas fault growth is gradual and involves more linkage in clay models. Folds (e.g., fault-propagation folds and relay ramps) are poorly developed in sand models compared to clay models. Deformation is more localized in sand models and more distributed in clay models. For example, in models with overlapping plates, visible faults accommodate 85% of the deformation in sand models, compared to 44% in clay models. In the sand models, a few major antithetic normal faults accommodate most hanging-wall deformation. Most layers, although faulted, remain flat. The effective shear angle (60 - 65°) is the same as the dip of the antithetic faults. In the identical clay models, numerous minor normal faults (antithetic and synthetic) and cataclastic flow accommodate most hanging-wall deformation. The deformed layers are faulted and folded. The effective shear angle (35 - 50°) is considerably less than the dip of the antithetic normal faults.