Asymmetric Friction Connections (AFCs) are bolted connections used to dissipate seismic energy via friction. AFCs have been experimentally validated as seismic dissipaters for different steel structural systems. However, the AFC experimental validation has not considered the effects of: fire, corrosion, sliding surface treatments, bolt length variation, the use of non-metallic sliding surfaces such as brake pads, or bolt inclination during sliding. This thesis evaluates the above effects on the hysteretic behaviour of AFCs subject to reversed cyclic experimental testing under increasing displacements, and proposes simple models to quantify the average AFC strength considering these effects.

Experimental results showed that the average AFC strength reduced with increasing heating temperature and bolt length. Depending on the surface treatment, the average AFC strength was greater or less than the average AFC strength for standard steel surfaces. For AFCs subjected to a cyclic corrosion testing, the AFC strength increased in the initial sliding cycles while the corrosion product was removed and became almost constant approaching that of the non-corroded condition. As a result of bolt inclination, the lowest value of AFC strength was observed in the initial sliding cycles while the bolts were near vertical. However, the strength increased until the bolt reached its maximum inclination, after which the AFC strength became almost constant. For brake pads, the AFC strength decreased as the cycles increased until becoming constant when the brake pads reached their steady wear state.

Models proposed using the dry friction theory of Coulomb, and experimentally determined friction coefficients, represented the average experimental AFC strengths well considering the above effects.

Finally, AFC design recommendations are made based on the experimental results and models presented in this thesis.