欢迎光临《计算力学学报》官方网站！

Uncertainty analysis of creep deformation of compacted loess considering parameters randomness

DOI：

 作者 单位 邮编 胡长明 西安建筑科技大学土木工程学院 710055 汪芳芳 西安建筑科技大学土木工程学院 710055 朱武卫 陕西省建筑科学研究院有限公司 袁一力 西安建筑科技大学土木工程学院 杨晓 陕西省建筑科学研究院有限公司 柳明亮 陕西省建筑科学研究院有限公司

现有蠕变模型的计算值为确定值，未能考虑岩土材料和参数评估过程中的随机性。为反映岩土材料的变形随机特性，本文基于多次重复的一维固结蠕变试验，研究了分级和分别加载条件下压实黄土的蠕变特性，建立了考虑参数随机性的改进Merchant模型，提出了一种蠕变变形的不确定性预测方法并通过延安高填方工程进行了验证。结果表明：压实黄土具有明显的衰减蠕变特性，引入非线性元件的Merchant模型能更好描述土体的应变时间关系，考虑参数随机性的模型计算值近似服从正态分布；通过不确定性预测方法可以获得蠕变变形值的概率云图，且工程实测蠕变变形值落在预测范围的95%的置信区间内，说明了考虑参数随机性的蠕变变形不确定性预测方法的有效性。

The calculated value of existing creep model is determined value, which fails to consider the randomness in the evaluation process of geotechnical materials and parameters. To reflect the deformation random characteristics of geotechnical materials, this paper studied the creep characteristics of compacted loess under graded and separate loading conditions based on repeated 1D compression creep tests. An improved Merchant model considering parameters randomness were established. An uncertainty prediction method for creep deformation was proposed and verified by a high fill project in Yan'an. The results show that the compacted loess has obvious decaying creep characteristics. The Merchant model with nonlinear elements can better describe the strain-time relationship of soil. The calculated value of the model considering parameters randomness approximately obeys the normal distribution. The uncertainty prediction method can obtain the probability cloud map of the creep deformation value, and the measured creep deformation value of the project falls within the 95% confidence interval of the predicted range. It shows that the creep deformation uncertainty prediction method considering parameters randomness is effective.
查看/发表评论  下载PDF阅读器