Search for a command to run...
Monitoring While Drilling (MWD) is a relatively new technology where several sensors are installed on drill rigs and used to continuously monitor and record drilling parameters. Common drilling parameters measured in geotechnical engineering projects include penetration rate, rotation speed, thrust (down pressure), and torque applied to the bit. Different physics-based or empirical equations have been suggested to combine these individual drilling parameters into compound parameters that can be potentially used to predict different mechanical properties of the subsurface material. The correlations between these compound parameters and the soil/rock mechanical properties are not universal and depend on many factors including the geology of the project site (soft or hard material), drilling type, energies used by different tools, and operational procedures. Therefore, until a universal correlation normalized based on the energies used by different tools is developed for different geomaterials, individual correlations are required for different drilling operations in every region. As more data becomes available over time, the regional data and correlations can be used to develop a more universal correlation. In this study, MWD data from seven different boreholes (drilled using hollow stem augers in the state of Nebraska, USA) have been analyzed to develop correlations between various compound parameters and the undrained shear strength of fine-grained soils encountered in these boreholes. The compound parameters examined in this study included specific energy (Es), drillability (Ds), Somerton index (SD), soil-rock resistance (RSR), and alteration index (AI). The results indicate reliable correlations can be found between MWD compound parameters and the undrained shear strength of fine-grained soils. Specific energy, drillability, and Somerton index parameters showed the most reliable correlations with the highest coefficients of determination. The main challenges encountered in developing these correlations, including the required preprocessing of the data, and the solutions to overcome these challenges are discussed in detail.