Abstract:
Piezocone penetration test (CPTu), as an in situ test, is ubiquitously utilized for generating a continuous profile of geotechnical parameters and soil profile. The near-continuous measurements with depth are generated using data obtained from cone tip stress 𝑞𝑐, sleeve friction 𝑓𝑠 and pore pressure (𝑢2 in this study). These basic CPTu parameters correlate to many geotechnical parameters relating to undrained shear strength (𝑠𝑢) and consolidation properties of cohesive deposits. A review of the literature reveals an absence of an experimental research program that quanitfies the consolidation parameters from CPTu in the context of Bangladesh.
Using data collected from nineteen (19) points in Dhaka, Bangladesh, this study evaluates several CPTu interpretation methods and quantifies its estimation capability to determine consolidation parameters (constrained modulus (𝑀), stress history (OCR), compression index (𝐶𝑐), and coefficient of consolidation (𝑐𝑣)) and 𝑠𝑢 of the cohesive deposits. Empirical correlations of the consolidation parameters and 𝑠𝑢 were also formulated through statistical analysis of the data.
The newly proposed correlations indicate a reasonably high coefficient of determination (𝑅2). However, no CPTu correlations could be established based on the CPTu-derived pore pressure data. Similarly, CPTu-derived 𝑐𝑣 yielded no correlation with the laboratory-derived 𝑐𝑣. In the case of 𝑠𝑢, normalized pore pressure (in terms of effective unit weight) resulted in a slightly better correlation than just using excess pore pressure data. But a better relationship between 𝑓𝑠 and laboratory-derived 𝑠𝑢 was observed, indicating that results from unconfined compression strength test (UCT) closely matches with 𝑓𝑠.
𝑓𝑠 based correlations to estimate consolidation parameters also resulted in good 𝑅2, indicating the precision and accuracy of the subtraction-type cones. Moreover, empirical cone factor (𝑁𝑘𝑡) was found to be between 14.7 to 17.2 for this study. Also, a hyperbolic
trend was evident between 𝑞𝑐 and 𝐶𝑐. The highest correlations proposed in this study, accompanied with 𝑅2 and number of data points (𝑛) are as follows: 𝑀 = 3.61β𝜎𝑣0+607.67,
𝑅2 = 0.43 for 𝑛 = 14; 𝐶𝑐 = 0.02 β 𝑤% β 0.31, 𝑅2 = 0.78, 𝑛 = 14; 𝑂𝐶𝑅 = 0.1412 β 𝑄𝑡𝑛,
𝑅2 = 0.85, for 𝑛 = 14; 𝑠𝑢 = 34.36 β 𝐹𝑟 , 𝑅2 = 0.84, for 𝑛 = 17.
Other fair correlations obtained are as follows: 𝑂𝐶𝑅 = 0.0648β𝑄𝑡, for 𝑛 = 14, 𝑅2 = 0.69;
𝑂𝐶𝑅 = 1.1467 β 𝐹𝑅, for 𝑛 = 14, 𝑅2 = 0.73; 𝑂𝐶𝑅 = 0.0645 β (𝑞𝑡 β 𝑢2)/𝜎β² , for
𝑛 = 14, 𝑅2 = 0.67; 𝑂𝐶𝑅 = 0.0651 β (𝑞𝑡 β 𝜎𝑣0 β Ξ𝑢)/𝜎β² , for 𝑛 = 14, 𝑅2 = 0.66;
𝑂𝐶𝑅 = 0.1346 β (𝑞𝑡 + 𝑓𝑠)/𝜎𝑣0, for 𝑛 = 14, 𝑅2 = 0.74; 𝑂𝐶𝑅 = 1.1978 β ( 𝑓𝑠/𝜎β² ) + 2.526,
for 𝑛 = 14, 𝑅2 = 0.60; 𝑠𝑢 = 0.0680 β 𝑞𝑛𝑒𝑡, for 𝑛 = 17, 𝑅2 = 0.74; 𝑠𝑢 = 0.0641 β 𝑞𝑡, for 𝑛 = 17, 𝑅2 = 0.75; 𝑠𝑢 = 1.44 β 𝑓𝑠, for 𝑛 = 17, 𝑅2 = 0.82; 𝑠𝑢 = 0.0752 β (𝑞𝑡 β 𝑢2), for 𝑛 = 17, 𝑅2 = 0.71; 𝑠𝑢 = 0.0650 β (𝑞𝑡 + 𝑓𝑠 β 𝜎𝑣0), for 𝑛 = 17, 𝑅2 = 0.75; and
𝑠𝑢 = 0.0582 β (𝑞𝑡 + 𝑓𝑠 + 𝜎𝑣0), for 𝑛 = 17, 𝑅2 = 0.77.
