Rectangular Blocks vs Polygonal Walls in Archaeoseismology

Abstract

Collapsed or deformed walls in ancient structures constitute important evidence in archaeoseismology, where damage is interpreted in terms of earthquake ground motion. A large variety of wall types have been developed during the millennia in different cultural backgrounds. Often walls with polygonal-shaped building blocks are regarded as more earthquake-resistant than a wall consisting of rectangular elements and, as is sometimes speculated, that the irregular wall types were intentionally developed for that purpose. We use simply structured discrete element models of four walls with different block geometries, perfect rectangular, an Inka-type structure and two polygonal designs, to test their dynamic behavior. In addition to an analytic calculation of ground motion, we use measured strong motion signals as boundary conditions for the 3D wall models with varying height to width ratios. At peak ground accelerations between 1.0 and 9.0 m/s² and major frequencies of 0.5 to 3 Hz, numeric experiments with the horizontally applied analytic ground motions result in clear differences in the resistance of the four wall types with the rectangular block wall being most vulnerable. For more complex measured 3D motions the Inka-type wall proves more stable than the rectangular block wall; however, height to width ratio still has equally strong influence on the stability. Internal deformation of non-collapsed walls shows some correlation with the parameters of the driving motion. For simple impulsive ground motions, a peak ground displacement threshold exists between toppling and remaining upright for all four models but peak acceleration cannot be reliably back calculated.