Effect of a Shallow Weak Zone on Fault Rupture: Numerical Simulation of Scale-Model Experiments
Effect of a Shallow Weak Zone on Fault Rupture: Numerical Simulation of Scale-Model Experiments
复制标题
浅层弱带对断层破裂的影响：比例模型实验的数值模拟
DOI：
--
复制
发表时间：
2002
期刊：
影响因子：
--
通讯作者：
G. Ely
中科院分区：
文献类型：
--
作者：
S. Day;G. Ely
比例模型地震实验提供了自然地震无法获得的破裂传播和断层运动的详细地下记录，从而为测试数值地震模拟方法提供了机会。利用实验数据实现的优势包括最佳传感器位置、精确了解介质的体积和表面性质、详细了解初始应力状态以及实验可重复性。我们进行了数值模拟，精确地再现了布鲁恩和阿努谢普（1998）在泡沫橡胶地震实验中记录的靠近断层表面的加速脉冲的形状和持续时间。在事件成核区之外，实验和模拟的破裂速度几乎无法区分。在实验测量的约束下调整静摩擦系数，可以使加速度的绝对振幅接近一致，通常在几十个百分点之内。当断层上部的摩擦强度发生变化时，模拟结果与实验结果吻合较好。数值与实验结果的吻合验证了离散数值模型准确地反映了自爆问题的连续动力学。数值模拟也有助于进一步解释比例模型实验。他们支持一种解释，即泡沫实验中的断层位移主要是通过传统的摩擦滑动机制发生的，而不是在断层张开期发生的，这导致了裂缝状而不是脉冲状的滑动模式。模拟进一步提出了对断层弱带的表观破裂速度测量的重新解释，即破裂减慢而不是加速。他们还预测，薄弱带在一定距离内会减弱地表加速度和速度（相对于统一的断层模型），该距离与薄弱带的深度成比例，并可能在中距离处略微增强振幅。
Scale-model earthquake experiments provide detailed, subsurface recordings of rupture propagation and fault motion that are unavailable for natural earthquakes and thereby offer an opportunity to test numerical earthquake simulation methods. Among the advantages realized from the use of experimental data are optimal sensor locations, precise knowledge of bulk and surface properties of the medium, detailed knowledge of the initial stress state, and experimental repeatability. We perform numerical simulations that closely reproduce the shape and duration of the acceleration pulses recorded adjacent to the fault surface in the foam rubber earthquake experiments of Brune and Anooshehpoor (1998). Outside of the event nucleation zone, experimental and simulated rupture velocities are nearly indistinguishable. With adjustment of the static friction coefficient within the constraints imposed by experimental measurements, the absolute amplitudes of the accelerations can be brought into close agreement, typically within a few tens of percent. Close agreement between simulation and experiment is also maintained when the frictional strength of the upper part of the fault plane is varied. The agreement of the numerical and experimental results verifies that the discrete numerical model accurately represents the continuum dynamics of the spontaneous rupture problem. The numerical simulations also facilitate further interpretation of the scale-model experiments. They support an interpretation in which fault displacement in the foam experiments occurs predominantly through a conventional frictional sliding mechanism rather than during fault opening episodes, resulting in a cracklike rather than pulselike mode of slip. The simulations further suggest a reinterpretation of the apparent rupture velocity measurements in the fault weak zone, with rupture slowing rather than accelerating. They also predict that the weak zone diminishes surface accelerations and velocities (relative to a uniform fault model) out to a distance that scales with weak-zone depth and may enhance amplitudes slightly at intermediate distances.
Manuscript received 23 October 2001.