CN-117093955-B - Method for predicting peak intensity based on coarse-grained soil triaxial test database

CN117093955BCN 117093955 BCN117093955 BCN 117093955BCN-117093955-B

Abstract

The invention belongs to the field of geotechnical engineering, and provides a method for predicting peak intensity based on a coarse-grained soil triaxial test database. The coarse-grained soil peak strength prediction method includes the steps of S1, calculating test information data to construct a coarse-grained soil triaxial test database, S2, selecting physical characteristic parameter information of coarse-grained soil from the database as independent variables, selecting the peak strength of coarse-grained soil as dependent variables, S3, sequencing the importance of the independent variables, S4, screening key independent variables, S5, performing multiple nonlinear regression analysis to establish an empirical formula, and S6, predicting the peak strength through the physical characteristic parameters and the empirical formula of coarse-grained soil materials to be detected. Based on a large amount of test data of coarse-grained soil, the invention provides an empirical relationship between the strength parameter and the key physical characteristic parameter, establishes a prediction method of the peak strength of coarse-grained soil, and has important guiding significance and reference value for evaluating the mechanical characteristics of coarse-grained soil lacking the test data.

Inventors

ZOU DEGAO
LIU JINGMAO
LI DUO
ZHOU CHENGUANG
NING FANWEI

Assignees

大连理工大学

Dates

Publication Date: 20260512
Application Date: 20230822

Claims (2)

1. The method for predicting the peak intensity based on the coarse-grained soil triaxial test database is characterized by being capable of predicting the peak intensity through coarse-grained soil physical characteristic parameter information, and comprises the following steps of: S1, constructing a coarse-grained soil triaxial test database, namely collecting test information of coarse-grained soil under the conventional triaxial consolidation drainage shearing test working condition, wherein the test information comprises sources, physical characteristic parameters and test results of the coarse-grained soil, and forming a database containing coarse-grained soil with different physical characteristics and test results, wherein the method comprises the following steps of: The sources of the coarse-grained soil comprise engineering names and geographic positions; The physical characteristic parameters of the coarse-grained soil comprise a parent rock property, saturated uniaxial compressive strength sigma r , specific gravity G s , dry density rho d , a void ratio e 0 , a particle shape regularity index rho, test confining pressure sigma 3 and grading parameters, wherein the grading parameters comprise a non-uniform coefficient C u , a curvature coefficient C c , a maximum particle size D max , and particle size values D 80 、D 60 、D 50 、D 30 and D 10 which respectively correspond to the content percentages of soil particles smaller than a certain particle size on a grading curve of 80%, 60%, 50%, 30% and 10%; The test result comprises a stress-strain curve and a peak strength value of a triaxial consolidation drainage shear test; s2, selecting variables, namely selecting physical characteristic parameters of coarse-grained soil from the coarse-grained soil triaxial test database constructed in the step S1 as independent variables, and selecting peak strength q peak of coarse-grained soil as the dependent variables; S3, sorting the importance of the independent variables, namely evaluating and sorting the importance of the independent variables by a random forest method, wherein the importance sorting result is σ 3 、σ r 、e 0 、ρ、D 80 、C u 、D 30 、D 10 、C c 、D 60 、D 50 、D max 、G s , in sequence, and the importance is larger when the importance is higher than the former importance; s4, screening key independent variables, namely sorting according to the importance of the independent variables, and selecting the first independent variables with accumulated relative importance exceeding 95% as key independent variables, wherein the key independent variables are sigma 3 、σ r 、e 0 、ρ、D 80 、C u respectively; S5, establishing an empirical formula, namely performing multi-element nonlinear regression analysis on the key independent variables and the dependent variables, wherein the empirical formula for establishing a power function form is as follows: the coefficient R 2 is determined to be 0.96; S6, predicting peak intensity, namely substituting the physical characteristic parameter (sigma 3 、σ r 、e 0 、ρ、D 80 、C u ) of coarse-grained soil to be detected into the formula in the step S5, so that the predicted value of the peak intensity can be directly determined.
2. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the method of predicting peak intensities based on a coarse soil triaxial test database according to claim 1.

Description

Method for predicting peak intensity based on coarse-grained soil triaxial test database Technical Field The invention belongs to the field of geotechnical engineering tests, and particularly relates to a method for predicting peak intensity based on a coarse-grained soil triaxial test database. Background Coarse-grained soil is an important material widely used for engineering construction of earth-rock dams, roadbeds and the like, and accurate evaluation of mechanical properties is important for safety evaluation of engineering structures. Early studies have reserved a large amount of coarse-grained soil triaxial test data, but these data are often limited to the study of the corresponding engineering, and have not been subjected to summary analysis and expansion application, and the potential value and function of the coarse-grained soil triaxial test data have not been mined. In addition, the parameters of many engineering test data at the initial stage or without coarse-grained soil are often obtained by engineering analogy, and the reliability is difficult to ensure. Therefore, it is highly desirable to integrate a coarse-grained soil triaxial test database, and establish a prediction method of key mechanical parameters of coarse-grained soil by adopting big data analysis and artificial intelligence technology, so as to ensure that the preliminary design of an engineering structure is guided as accurately as possible on the premise of lacking test data. Disclosure of Invention The invention aims to provide a method for predicting peak strength based on a coarse-grained soil triaxial test database, which provides a relatively scientific and reasonable material mechanical parameter determination method for engineering structural design and safety evaluation related to coarse-grained soil. In order to achieve the above purpose, the invention adopts the following technical scheme: a method for predicting peak intensity based on a coarse-grained soil triaxial test database comprises the following steps of: S1, constructing a coarse-grained soil triaxial test database, namely collecting test information of coarse-grained soil under the conventional triaxial consolidation drainage shearing test working condition, wherein the test information comprises sources, physical characteristic parameters and test results of the coarse-grained soil, and forming a database containing coarse-grained soil with different physical characteristics and test results. The source of coarse-grained soil comprises an engineering name and a geographic location. The physical characteristic parameters of the coarse-grained soil comprise a parent rock property, saturated uniaxial compressive strength sigma r, specific gravity G s, dry density rho d, a porosity ratio e 0, a particle shape regularity index rho, test confining pressure sigma 3 and grading parameters, wherein the grading parameters comprise a non-uniform coefficient C u, a curvature coefficient C c, a maximum particle size D max and corresponding particle size values D 80、D60、D50、D30 and D 10 on a grading curve, wherein the content percentage of soil particles smaller than a certain particle size is 80%, 60%, 50%, 30% and 10%. The test results include stress-strain curves and peak strength values of triaxial consolidation drainage shear tests. The peak strength is the maximum of the stress in the stress-strain curve. S2, selecting variables, namely selecting physical characteristic parameters of coarse-grained soil from the coarse-grained soil triaxial test database constructed in the step S1 as independent variables, and selecting peak strength q peak of coarse-grained soil as the dependent variables. S3, sorting the importance of the independent variables, namely evaluating the importance of the independent variables (determining the relative importance) and sorting the independent variables through a random forest method, wherein the importance sorting result of the independent variables is σ3、σr、e0、ρ、D80、Cu、D30、D10、Cc、D60、D50、Dmax、Gs, in sequence through research, and the importance of the independent variables which are more forward is more important. S4, screening key independent variables, namely sorting according to the importance of the independent variables, and selecting the first independent variables with accumulated relative importance exceeding 95% as key independent variables, wherein the key independent variables are sigma 3、σr、e0、ρ、D80、Cu respectively. S5, establishing an empirical formula, namely performing multi-element nonlinear regression analysis on the key independent variable and the dependent variable, wherein the empirical formula for establishing a power function form is ：qpeak＝1.702σ30.814σr0.097e0-0.214ρ-0.106D800.043Cu0.004,, and the coefficient R 2 is determined to be 0.96. S6, predicting peak intensity, namely substituting the physical characteristic parameter (sigma 3、σr、e0、ρ、D80、Cu) of coarse-grained soil to be detected into the formula