论文部分内容阅读
为了测定弱运动场地响应的稳定性和试验土工技术测量预测场地响应的能力,在穿过加州帕克菲尔德附近的土耳基平地(Turkey Flat)场地效应试验区的一些场地,对经验转换函数和理论场地放大作用函数进行了比较。用傅里叶振幅频谱比方法,对给定场地测定了相对于基岩参考场地的观测场地响应。进行了两种单独的研究:剖面研究和特定场地研究。相隔2km的两个横跨谷地岩石的台站之间的表面基岩场地响应差异不超过30％。在剖面研究中,在距试验区的谷地上间隔100m的一些场地使用的探测器间隔为10和20m。跨过10m的探测器间隔,表面硬土场地响应可重复测得,差异不超过30％。在硬土场地响应中,对于20m的探测器间隔,存在偶尔出现的显著变化,在10～20Hz频带内可差2至3倍。在硬土场地响应中,对于探测器间隔为100m时存在系统的变化,在5～20Hz频带内相对于土层厚度的变化可差3至10倍。在土耳基平地发现了特定场地弱运动平均频谱比和其标准偏差,可用作硬土条件强运动转换函数的经验预测值。使用由计算机程序SHAKE得到的相对放大作用函数模拟了观测的转换函数。使用接近原地测量的阻尼值,比使用实验室实验测定的阻尼值获得了共振峰值振幅更好的拟合。土耳基平地标准土工技术模型的不精确性(Real,1988),导致了计算的弱运动转换函数与观测的弱运动转换函数相比的不精确性。观测转换函数的模拟表明,模型拟合对5％的S波速度变化敏感。土耳基平地现场测量S波速度有10％～20％的误差。这一情况表明,仅仅根据现场土工技术的测量,对弱运动的精确和一致的预测是不可能的。
In order to determine the stability of response at weak sports venues and the ability of experimental geotechnical techniques to measure predictions of site response, empirical transfer functions and theories were applied at a number of venues across the Turkey Flat site effects test area near Parkfield, California Field amplification function is compared. Using the Fourier amplitude-to-frequency ratio method, the observed site response to a bedrock reference site is determined for a given site. Two separate studies were conducted: profile studies and site-specific studies. The difference in site bedrock response between two stations across the valley between 2km apart does not exceed 30%. In the cross-section study, detector spacings of 10 and 20 m were used at some sites 100 m apart from the valley in the test area. Across the detector interval of 10 m, the response to surface hard soil can be measured repeatedly, with a difference of no more than 30%. In the hard-soil site response, there is an occasional significant change for a detector interval of 20 m, which can be 2 to 3 times worse in the 10 to 20 Hz band. In the case of a hard-soil site, there is a systematic variation for a detector spacing of 100 m and a variation of 3 to 10-fold relative to the thickness of the soil layer in the 5-20 Hz band. Weak-moving average spectral ratios for specific sites and their standard deviations found on the Turkish flatlands can be used as empirical predictions for strong motion transfer functions in hard earth conditions. The observed transfer function is modeled using the relative magnification function obtained from the computer program SHAKE. Using a damping value closer to the in situ measurement, a better fit of the peak amplitude of the formant was obtained than using the damping values determined experimentally. Inaccuracies in the Turkish Geotechnical Standard Geotechnical Model (Real, 1988) lead to inaccuracies in the calculated weak motion transfer function compared to the observed weak motion transfer function. Simulations of the observed transfer function show that the model fit is sensitive to a 5% change in S-wave velocity. Turkish flat ground S-wave velocity measurement of 10% to 20% error. This situation shows that accurate and consistent predictions of weak movements are not possible based solely on on-site geotechnical measurements.