【作者】 刘振平；

【导师】 迟世春；

【作者基本信息】 大连理工大学 ， 岩土工程， 2013， 博士

【摘要】 土石坝是当今坝工建设中使用最多的一种坝型。我国已建和在建的高土石坝大多位于地震高烈度区。随着西部大开发战略的实施和西电东送工程的推进,在我国西南和西北还将兴建一批高土石坝。但这些高土石坝所在的西南和西北地区的坝址区域地震强度高、活动频繁,地震设防动参数高,一旦因地震而失事,将造成灾难性的后果。目前土石坝的抗震研究还落后于工程实践。其抗震分析基本上还停留在20世纪70年代提出的等价线性化方法上,这种方法在非线性程度不高的情况下,对地震加速度和剪应力的预测能取得较好的效果。目前,由于实际遭受过地震作用的现代碾压土石坝很少,缺乏强震区高土石坝的地震反应记录和实际震害资料；再加上坝料大应变非线性动力本构模型及其求解方法的理论研究尚未成熟,使得目前通过理论分析和数值计算得到高土石坝地震动力破坏响应与破坏模式的难度很大,难以理清高土石坝的地震破坏过程与机理。从而无法对地震作用下高土石坝的抗震安全性进行准确评价,严重影响了高土石坝抗震设计的可靠性、经济性和科学性。针对以上土石坝抗震研究中亟待解决的关键问题,本文开展了心墙堆石坝振动台试验、紫坪铺面板堆石坝在汶川地震中的残余应变分析和土石坝坝料动力参数反演等研究工作。其主要研究内容和结论如下：(1)鉴于土石料等散粒体材料应变测量的困难性,尝试发展了数字图像应变测量技术。把有限元数据平滑方法引入到数字图像应变测量中,开发了一套适合于土工试验模型应变测量计算的程序,并通过数字散斑图和三点弯曲梁实验进行了验证,显示该方法显著提高了应变的计算精度。将其应用到心墙堆石坝振动台模型试验的应变测量中,获得了模型坝的位移场和应变场分布规律,探讨了模型坝的地震破坏机理,初步研究表明破坏过程大致分三个阶段,即整体变形阶段、坝坡滑移变形阶段和破坏阶段,其中第二个阶段的标志是坝体由整体运动转为沿上下游坝坡分别向两侧滑动。同时表明空库状态下的破坏模式是上、下游坝坡的浅层滑塌破坏,满库状态下是坝体上部上游侧的变形破坏。(2)应用有限元数据平滑方法,对紫坪铺面板堆石坝在汶川地震中的监测位移资料进行分析,得到了紫坪铺大坝在汶川地震中的残余应变分布规律,并据此解释了大坝的一些震害现象,为进一步分析土石坝的地震破坏机理奠定基础。分析结果显示,坝体最大震陷率发生在2/3坝高附近；大坝发生剪胀的范围为坝顶区域和从坝顶往下约30m的上下游坝坡处；堆石料地震残余剪应变过大并形成下滑趋势是造成面板水平施工缝错台的直接原因；沿坝轴向在两岸坝肩附近区域受拉,河谷中央部位受挤压,整个坝体向河谷收缩；坡面残余压应变集中是造成面板垂直挤压破坏的主要原因；左岸坝肩周边缝张开和坝面张裂缝是由于坡面残余拉应变过大造成的。(3)实际土石坝坝料的动力性质与室内试验得到的动力参数有一些差异。本文提出了两种根据实测土石坝地震加速度记录信息反演筑坝土石料的动力参数反演方法,以探讨目前室内外坝料动力参数确定方法的合理性。①联合加速度峰值和反应谱为目标函数,研究了土石坝动力参数反演方法。结合鲤鱼潭心墙堆石坝在集集地震中的加速度响应,反演了该坝土石料的动力参数,并与室内外试验成果进行比较,表明筑坝土石料的室内动三轴试验得到的动剪模量系数K偏小。②以小震时土石坝动力特性为目标函数,研究了土石坝动力参数反演方法。利用紫坪铺面板堆石坝在汶川地震余震中的加速度记录,反演了该坝堆石料的动力参数,并与室内试验成果进行比较,也表明坝料的室内动三轴试验得到的动剪模量系数K偏小。更多还原

【Abstract】 Earth-rockfill dams are the most commonly constructed type of dams. Most of the high earth-rockfill dams existing or under construction in China are located in areas of high seismic intensity. With the implementation of West China Development strategy and the West-to-East Electricity Transmission project, a new patch of high earth-rockfill dams will be built in the high-intensity areas in Southwest and Northwest China. However, if these dams fail upon the occurrence of earthquakes, disastrous consequences will emerge.Currently, the research on seismic resistance of earth-rockfill dams is far lagged behind engineering practice. Anti-seismic analysis basically depends on the equivalent linearization method proposed in1970s, which is accurate in evaluating the distribution of seismic acceleration and shear stress only under low nonlinearity. The seismic deformation should still be estimated by using semiempirical methods. However, the seismic injury and damages of earth-rockfill dams can still not be reliably predicted. There is no reference for designing of earth-rockfill dams with a height over200m at home and abroad. Currently, because few modern compacted earth-rock dams have undergone earthquakes, there are few records of real seismic reactions or few seismic damage data from high earth-rock dams in high intensity regions. Moreover, because the nonlinear dynamic constitutive models for large-strain dam materials, and the theoretical research on solving methods are immature, it is difficult to obtain the dynamic damage response of high earth-rock dams by using theoretical analysis and numerical calculation, or to clarify their seismic damage mechanism. Therefore, it is hard to accurately evaluate their anti-seismic security under earthquakes. This severely affects the reliability, economy and science in anti-seismic design of high earth-rock dams, and restricts the development of their construction. Therefore, to meet the rapid development of high earth-rock dams in high-intensity regions in China, these key anti-seismic problems should be solved immediately.On account of some key problems in studies about earthquake resistance of earth-rockfill dams, earth core rockfill dam shaking table test, residual strain of Zipingpu Concrete Faced Rockfill Dam (CFRD) after Wenchuan earthquake, and back-analysis of earth-rockfill dam dynamic parameters were carried out. The contents and conclusions are as follows:(1) Based on the data smoothing method which combines the finite element method and general interaction validation, a set of programs suitable for strain measuring and calculation in geotechnical test models were developed. The three point bending beam experiment verifies the correctness and reliability of the programs. This method is applied to the strain measuring and calculation of the earth core rockfill dam shaking table model test, so the displacement field and strain field of model dam were obtained. Meanwhile, the earthquake damage mechanism of core rock-fill dam was discussed. The results show that the failure process can be roughly divided into three stages:overall deformation stage, dam-slope slipping deformation stage, and failure stage. The second stage starts when the overall movement turns to sliding along the upstream and downstream slopes. The failure pattern of the model dam is shallow damage along upstream and downstream slopes under empty reservoir, and is deformation failure along upper downstream slope under full reservoir.(2) Finite element smoothing was used to process the seismic residual deformation data from the Zipingpu CFRD and thereby to obtain permanent deformation displacement fields. Then, the seismic residual strain field was calculated, including vertical seismic residual strain, volumetric strain, shear strain of the dam body, and slope residual strain, which can be used to explain various seismic damages observed on the dam body. Therefore, the method in this study is reliable. The precision of the acquired residual strain satisfies engineering requirements.The results show that the maximum settlement rate on the dam body occurs at about2/3of the dam height, and dilatancy occurs approximately from the dam crest to30m in the upstream and downstream slopes, so this region should be reinforced by anti-seismic measures. The immediate cause of the dislocation of horizontal construction joint in face slabs is the excessive residual shear strain. The two bank abutments are under axial tension; the valley is extruded axially and the entire dam body shrinks to the valley.(3) Real earth-rockfill dam materials dynamic parameters different slightly from that obtain from indoor laboratory test.In order to probe the rationality of current detemination method of earth-rockfill dynamic parameters, two back-analysis methods of earth-rockfill dynamic parameters were presented according to observed seismic acceleration information.①The back-analysis approach of earth-rockfill dynamic parameters based on response spectra and acceleration peak is presented. According to the acceleration response information of Liyutan Dam in the Chi-chi earthquake, the dynamic parameters were obtained by applying the back calculation approach. The results demonstrate that the dynamic shear modulus coefficient K of the dam materials obtained from laboratory dynamic triaxial test is smaller than the true value, suggesting that it should be adjusted.②The back-analysis approach of earth-rockfill dynamic parameters based on earth-rockfill dam dynamic characteristics is presented. According to the acceleration response information of Zipingpu CFRD in the aftershock of Wenchuan earthquake, the dynamic parameters of dam materials were back-analyzed. The results indicate that K of the dam materials obtained from laboratory dynamic triaxial test is smaller than the true value, suggesting that it should be adjusted.更多还原