United States Department of Transportation
i
Development of Statistical Models to Predict the Compressibility of Florida’s Soils
-
2019-09-01
-
[PDF-2.54 MB]
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Resource Type:
-
Geographical Coverage:
-
Edition:3/30/201 – 10/31/2019
-
Contracting Officer:
-
Corporate Publisher:
-
Abstract:The magnitude of the overall settlement depends on several variables such as compression index (Cc), recompression index (Cr), and secondary compression index (Cα), which are determined by a consolidation test. However, the test is time consuming and labor intensive. Correlations have been developed to approximate these compressibility indexes to supplement but not eliminate a geotechnical sampling and testing program. In this study, a data-driven approach has been employed to estimate Cc, Cr, Cα, and coefficient of consolidation (Cv). In this project, a support vector machine (SVM) classification technique was used to determine the number of distinct models to be developed. The statistical models were then constructed through a forward-selection stepwise regression procedure. Eight variables were used, and these variables included the moisture content, automatic hammer standard penetration test (SPT) blow count, overburden stress, fines content (-200), liquid limit, percent organic content, plasticity index, and specific gravity. The best performing statistical models of soil compressibility were constructed for the classified soil types. In addition, reduced models that minimize the number of variables were constructed for more practical usage. It is recommended that only the models with R2 > 0.5 be used for engineering analysis. Lastly, the correlation between soil compressibility between CPT tip resistance (qc) was checked. There is a general trend that Cc and Cr decrease as qc increases; however, the correlation is not strong due to the limited dataset. Data collected under a controlled environment will be able to improve the correlation quality.
-
Format:
-
Funding:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type:
