Experimental Quantification of High‐Strength Bolt Fracture Uncertainty under Varying Loading Rates

Abstract Bolt tensile rupture constitutes an ultimate failure mode in bolted steel connections undergoing extreme loading events such as strong earthquakes and progressive collapse scenarios. As such, bolt failure controls the connection's ductility and that of the structure. To be able to study the upper and lower bounds of structural ductility as part of performance‐based and probabilistic simulations, it is important to quantify the uncertainty associated with bolt failure. Towards that end, a large experimental study is conducted comprising a total of 200 zinc‐plated and galvanized high‐strength bolt assemblies. The bolts covered diameters ranging from M12 to M24, grades ranging from 8.8 to 10.9, thread lengths from 5mm to 130mm. The bolts were subjected to uniaxial tension applied at varying loading rates ranging from 0.05 mm/sec to 80mm/sec. The experimental data was used to quantify the uncertainty in bolt elongation at the onset of necking and at complete rupture. Observed correlations between the bolt's critical response parameters and its geometric and material parameters are then used to establish an empirical trilinear numerical model that can be used to represent bolt components in mechanics‐based and continuum finite element simulations. The model is used to demonstrate the implications of bolt fracture uncertainty on steel connection ductility.