A drill collar is a thick-walled tubular component that provides a passage to pumping drilling fluids and a mechanical protection for sensing, power supply, communication, and control devices. Multiple collars can be screwed together along with other downhole tools to make a bottomhole assembly (BHA). Radially oriented ports are often used in the wall of a collar for various reasons. These ports could be susceptible to fatigue-induced failures when a BHA has to undergo a large number of revolutions in a curved well. A cracked port could result in leakage, thereby causing flood damage to the internal devices, which are supposed to be protected from drilling fluid. Understanding the risk of fatigue cracking of a collar port is an important part of BHA design and well planning.
The total fatigue life of a port can be considered as a summation of the crack initiation life, which is consumed to nucleate a dominant crack with a minimum detectable size, and the crack growth life, which is measured as the crack grows from the minimum detectable size until it reaches the seal. Prediction of the initiation life is expected to be conservative due to the many uncertainties involved. As a result, solely relying on the predicted initiation life to retire a port and the entire collar is not cost effective. A more economical way of port fatigue management is to compute the crack growth life based on a minimum detectable crack size and use this life as the inspection interval. If a crack is detected during an inspection, a port is declared as failed because a cracked port cannot be repaired with the same strength. Otherwise, the port can last at least until the next scheduled inspection.
In this study, a fracture-mechanics-based method is developed to predict the fatigue crack growth (FCG) life of a collar port subjected to constant-amplitude cyclic bending. It is assumed that a prescribed corner crack with a minimum detectable size lies in a plane perpendicular to the collar axis. It intersects with the collar outside surface and the port wall surface. The crack front follows an elliptical function. The stress intensity factors (SIFs) along the crack front are numerically computed with finite element analysis (FEA) at the two endpoints, respectively. A response surface of the SIF is generated by assigning a set of predetermined crack fronts based on incrementally advancing positions of the two endpoints. It is then used to determine the SIFs at these points throughout all crack growth increments. The Paris law is utilized to describe the FCG rate of the collar material, whereby, along with the SIFs computed, the crack growth life and the associated crack front shape are incrementally determined. To validate the newly developed method, a test apparatus is developed to apply constant-amplitude cyclic bending to a collar specimen that contains a through-hole in the middle. The predicted growth rate for the crack on the collar outside surface agrees favorably well with the test data. The computed crack front before rupture is also in good agreement with the experimental measurement.
