Performance of lift joints subjected to thermal fatigue cycles

Northern Manitoba is host to many pieces of infrastructure that have important roles in the province’s hydroelectric network. These generating stations and spillways are subjected to harsh fluctuations in environmental conditions each year, which lead to the development of damaging stresses due to thermal gradients in the mass concrete structures. An experimental program was conducted at the University of Manitoba to better understand the change in performance of the infrastructure over time, particularly at lift joint interfaces. This thesis presents the extent and results of said experimental program, which involved an investigation into the performance of concrete-to-concrete cold joints in cyclic thermal fatigue conditions. Small cylinders with joints cast horizontally for direct tensile specimens or at a 35 ̊, 40 ̊, or 45 ̊ angle for slant shear specimens were installed in transformation- restricting frames, stored in an environmental chamber, and subjected to temperature cycles alternating between -25 ̊C and 40 ̊C, thus simulating the combined stress and temperature conditions experienced by Manitoba’s northern infrastructure. The direct tensile and slant shear bond strength of these specimens was then evaluated after experiencing 50, 150, and 250 thermal fatigue cycles. The results indicated that a significant increase in tensile bond strength occurred after 50 cycles, followed by a subsequent decrease after 150 cycles and no further changes after 250 cycles. This trend was not shared by the slant shear specimens; slight increases in i bond strength was still observed at later cycling stages in specimens with higher normal- to-shear stress ratios. Overall, the performance of cold jointed specimens in both direct tension and slant shear after experiencing 250 thermal fatigue cycles was relatively consistent with the performance of control samples. Bond cohesion was also discovered to demonstrate a similar behaviour to the direct tensile bond strength results; however, neither dataset could be effectively represented by second-order polynomial models. Interface friction, despite appearing to follow a parabolic trend, maintained values that corresponded approximately with the same roughness level as per Canadian design codes and was determined to be negligibly affected by up to 250 cycles.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.