Wired drillpipe with along string measurement sensors is a key to drilling events prediction. An interesting objective during drilling operations is real-time assessment of the drill bit to determine an acceptable time to stop drilling and change the drill bit. In addition, dynamic modeling of the drillstring requires the calculation of forces at the bit. A novel approach to deal with this challenge involved experiments and the evaluation of the forces on the cutters based on their geometrical characteristics, rock, and drilling parameters.
In this research two empirical correlations were developed based on the experimental data from numerous works to evaluate normal and contact forces on polycrystalline diamond compact (PDC) cutters. More than 700 data points were collected and utilized to investigate the influence of parameters such as differential pressure, cutter size, cut depth, cutter wear state, rock drillability, and back rake angle on the normal and contact forces on cutters. Machine learning techniques such as deep learning, nonlinear regression approaches, and genetic algorithms were implemented to fit nonlinear equations to data points.
The outcome of this research (cutter-force equations) could be utilized to determine normal, lateral, and tangential forces on PDC cutters as inputs and constraints to model drillstring dynamics. Besides, the models could be adopted as a measure to assess bit conditions in real-time drilling and predict drilling events and issues related to the bit and mitigate drawbacks before they occur.
All available correlations and equations concerning cutter-rock interaction only consider one or two parameters or are just applicable under atmospheric conditions. The novelty of this research revolves around developing thorough cutter-force models to include all design and operational parameters under real differential fluid pressure.