南海珊瑚泥一维压缩分数阶元件模型研究

    Fractional order element model for one-dimensional compression of coral mud in the South China Sea

    • 摘要: 南海珊瑚泥广泛分布于中国南海海域,研究珊瑚泥的沉降变形规律对南海的开发有着重要意义。南海珊瑚泥的一维压缩具有典型的软土变形特征,同时由于其特殊的性质,其变形速度与常见软土相比较快,适合进行变形模型的研究。针对软土的一维压缩变形常使用元件模型进行模拟,但由于元件模型中反映长期变形的元件参数为定值,并不能准确反映延时塑性变形。应用分数阶理论的Abel黏壶比牛顿黏壶能更反应土体延时塑性变形特征,故引入Burgers模型进行修正,得到分数阶修正Burgers模型。对近似天然条件下的珊瑚泥进行每一级历时8~14 d的一维固结压缩试验,使用前11 h的读数分别对Merchant模型,Burgers模型和分数阶修正Burgers模型进行自拟合,并使用拟合得到的模型参数对11 h后的变形进行预测。对比3种模型11 h内的自拟合精度和11 h后的预测精度,并对各个模型的参数及各部分变形情况进行分析。研究发现,分数阶修正Burgers模型的自拟合精度为Merchant模型的23~65倍,为Burgers模型的6~18倍;预测精度为Merchant模型的5~35倍,为Burgers模型的262~6213倍。

       

      Abstract: Coral mud is widely distributed in the South China Sea. Studying its deformation law is of great significance for the development of the South China Sea. One-dimensional compression of coral mud in the South China Sea has the typical deformation characteristics of soft soils. At the same time, due to its special properties, the deformation speed of coral mud is faster than that of common soft soil, therefore coral mud is suitable for the study of deformation model. One-dimensional compression deformation of soft soils is often simulated by element models. However, element parameters reflecting the long-term deformation in the element model are fixed, which cannot accurately reflect the delayed plastic deformation. Abel dashpot appling the fractional order theory can better reflect the delayed plastic deformation characteristics of soils than Newton dashpot. Therefore, Burgers model is introduced to correct the fractional order-modified Burgers model. In this study, one-dimensional consolidation compression tests on the coral mud under approximate natural conditions lasting for 8~14 days per stage are carried out. The first 11 hours of readings are used to self-fit Merchant model, Burgers model and fractional order-modified Burgers model, respectively. The model parameters obtained by fitting are used to predict the deformation after 11 hours. The self-fitting accuracy of the three models within 11 hours and the prediction accuracy after 11 hours are compared, and the parameters of each model and the deformation of each part are analyzed. It is found that the self-fitting accuracy of the fractional order-modified Burgers model is 23~65 times that of Merchant model, and 6~18 times that of Burgers model. The prediction accuracy is 5~35 times of Merchant model and 262~6213 times of Burgers model.

       

    /

    返回文章
    返回