eva‎luation of liquefaction potential based on Spatial 3d spatial variability

11/18/2024 12:00:00 AM

In engineering, the proposed method based on the results of CPT cone penetration test is widely used to eva‎luate soil characteristics and soil liquefaction potential. But since the cone penetration test is done in a scattered and limited way in a specific site, therefore, considering the limited number of cone penetration tests performed in one site and the recorded results of the cone penetration test are only dependent on the depth, and also with Paying attention to the inherent variability of the soil, we face uncertainty in determining the characteristics of the soil. Considering the advantages of the cone penetration test, which provides the desired values in the form of a continuous signal and helps in correctly diagnosing the variability of soils, as well as the possibility of estimating and eva‎luating the liquefaction potential of soils with data There are results from this experiment, it is possible to do three-dimensional modeling of the desired soil environment and of course to eva‎luate the potential of soil liquefaction in each position of this three-dimensional environment and from the simplifications related to two-dimensional analysis including the generalization of the behavior A section prevented the entire soil. In a general view, random field is a method for generating continuous or discrete data in space, which is widely used for data analysis, especially complex data. In this research, using several borehole logs from the cone penetration test based on the assumption of resolution in the autocorrelation function and based on the assumption of a dense network, the assessment of liquefaction potential in a three-dimensional random field is investigated, and the considered parameters are better investigated in the soil. Numerical errors associated with large matrices are also reduced, and a new efficient method for simulating conditional random fields in three dimensions with Bayesian statistics approach or in other words sparse Bayesian learning SBL is used, which can be calculated in practice. CPT checks only in no. A limited number of locations are conducted at a given site, and soil properties and conditions at locations without CPT data must be interpreted from sporadic CPT surveys.

روانگرايي خاك , يادگيري بيزي پراكنده

Soil liquefaction , Sparse Bayesian learning

Alireza Abarghoie

Dr. Mehran Karimpour fard