Abstract:
The major objective of this research is to advance the state of art interpretation and application of results obtained from in-situ testing of soil, in particular Cone Penetration Test and Standard Penetration Test. CPT-SPT based soil characterization requires reliable empirical correlations among in-situ and laboratory obtained soil parameters. A large part of the research involves development of various CPT-SPT correlations for studied soil type. Correlations have been developed for the use of CPT and SPT data to evaluate shear strength parameters (c-) as well as settlement parameter (cc). qc/N ratio has been determined for different soil types and also as a function of index properties like Soil Behavior Index (Ic), Mean Grain Size (d50) and Fines Content (fc). New developments have been compared with previously established correlations. CPT based soil classification/stratigraphy has been prepared for riverine and coastal regions of Bangladesh. A new approach has been taken to correlate plasticity characteristics of soil with in-situ test parameters.
For this study, a total of 38 pairs of CPT and SPT have been performed in different locations from northern-most riverine embankment to the southernmost coastal embankment. Each pair of CPT and SPT has been carried out as closely as possible, maximum horizontal distance not being greater than 1m. SPT N value varies from 1-57, CPT cone tip resistance has a variation of 0.4 to 25 MPa, Soil behavior Index (Ic) varies from 1.5 to 3.5. This research demonstrates that continuous and reliable information on soil strata can be achieved through CPT. Analyses show good agreement between previously established and newly developed correlations for studied soil type. This study also illustrates that in-situ soil characterization aids easy and less conservative structural design for Bangladesh.
Statistical analysis of the data results in the formulation of empirical correlations between the various soil parameters, with often reasonably high correlation coefficients considering the wide variation of soil types tested. However, the confidence limits for these correlations range from 30-60% after performing a data filtering for the regression analysis. Due to the moderate size of the database, the proposed correlations for CPT and SPT can be considered moderately reliable for Bangladeshi embankment soils; the consistent uniformity of testing procedures is also ensured as testing has been carried out using standard techniques.
