On the Boundary Conditions and Internal Mechanics of Parallel Wire Strands

This dissertation analyzes the internal mechanics of parallel wire strands as found in the main cables of suspension bridges. Parallel wire strands of reduced order (7-wire, 19-wire, and 61-wire strands made of steel and aluminum) are fabricated and subjected to various boundary conditions and external loads (tension, clamping, twist, etc.). Neutron diffraction is used as an elastic strain measurement tool for its ability to penetrate bulk materials and/or layers of a multi-body system without disturbing the sample. Firstly, this thesis aims to quantify the development length – the distance over which a broken wire within a strand regains near-full service strain – as a function of various boundary conditions and failure scenarios. The feasibility of using neutron diffractometers to measure in situ elastic strains on civil-engineering-scale samples under both tensile load and radial confinement is validated using strands fabricated from steel bridge wire. Results from various 7-wire strands indicate that friction and mechanical interference on the microscopic level play a significant role in the load partitioning. Furthermore, wires that have been broken – either pre-cracked or fractured live and in situ during tensile loading – are capable of regaining significant stresses from their neighbors over a distance of tens of centimeters. The contribution of both friction force and mechanical interference on elastic strain redevelopment in broken wires should be included in analytical models designed to simulate failure processes. The second part of this thesis aims to measure the internal mechanics of larger parallel wire strands in response to various confinement (clamping) forces. 19 and 61 aluminum wire strands are fabricated and the internal strains of all constituent wires mapped in three orthogonal directions under various clamping loads. The strain distributions for both 19-wire and 61-wire strands show a surprising degree of heterogeneity. An increase in clamping force homogenizes the distribution to a degree, but only at unfeasibly high clamping forces. The results suggest that microscale variations in wire diameter dominate the internal mechanics of parallel wire strands. The stochastic distribution of wire sizes due to manufacturing tolerances throughout a strand cross-section creates a randomly ordered network of over- and under-sized wires. This imperfectly packed lattice results in large wire-to-wire variations in clamping constraint. The up-scaling in strand size from 19 to 61 wires increases the resolution of the experiment but does not reduce the heterogeneity of the strain distribution. Ergo, the assumption of perfect hexagonal packing in parallel wire strands is weak, and mean field distributions do not accurately describe the internal mechanics of such structures.