Fracture Resistance of Wire-Wrapped Cylinders

Abstract

Steel cylinders wrapped with steel wire have the capability of offering not only an economical high-strength structure, but also a structure with improved fracture resistance compared with that of an unwrapped cylinder of equivalent strength. Accordingly, 2000-psi-pressure hydraulic burst tests were therefore conducted to determine the fracture resistance of 36-in-dia, 60-ksi yield-strength, 1000-psi-pressure wire-wrapped cylinders at different levels of shell notch ductility, which was varied by testing at different temperatures. The cylinders were prestressed with 1/4-in-dia cold-drawn wire, and the shells contained part-through-wall flaws. A similarly flawed unwrapped cylinder was tested for comparison. The working-stress level was 72 percent of the specified minimum yield strength in the shell and 60 percent of the minimum tensile strength in the wire. The results showed that at a pressure double that of the unwrapped shell, no crack extension occurred at a temperature at which the steel exhibited fully ductile shell behavior (+110 deg F). A 2-ft crack extension occurred at a temperature (+10 deg F) at which the steel was still in the transition temperature range from ductile-to-brittle behavior (about 20 percent shear fracture), but a brittle crack (−70 deg F) propagated to the end of the wire-wrapped shell. Except for the brittle propagating crack, wire wrapping appears to provide sufficient constraint of a shell defect or propagating crack to limit bulging and crack-opening displacement. A model based on the compatibility in displacements between the crack opening and the local wire strain is presented for calculating the arrest conditions of the propagating crack in the test at 10 deg F. The same flaw size was critical at the constant failure pressure for all test temperatures, and showed that, as predicted, ductile initiation occurs even at the −70 deg F temperature in both the wrapped and unwrapped-cylinder tests. A circumferential flaw was shown to be less critical than a longitudinal flaw of the same size.