Experimental study of creep Poisson ratio of hydraulic concrete under multi-age loading-unloading conditions

Elastic creep theory, which is widely used in concrete dam engineering, assumes that the elastic Poisson ratio and the creep Poisson ratio equal the effective creep Poisson ratio. However, differences in creep test conditions and stress state make this assumption controversial. To accurately clarify the Poisson effect of hydraulic concrete, uniaxial compression creep tests with four loading ages (3d, 14d, 28d, and 60d) and conventional triaxial compression creep tests with three loading ages (14d, 28d, and 60d) are conducted under unified test conditions. And high-precision creep data of transverse and longitudinal for multi-age loading and unloading are obtained. We then denoise the creep data by using a Savitzky-Golay filter. Furthermore, based on the denoised creep data, we calculate the Poisson ratio at each stage for uniaxial compression and conventional triaxial compression. Finally, we compare and analyze the Poisson ratio of each stage in different stress states. The results show that the creep tests have the same noise when conducted under the same test conditions, and the optimal denoising by the Savitzky-Golay filter is obtained with parameters W (window width) = 9 and N (polynomial order) = 6. Under different stress states, the elastic Poisson ratio, creep Poisson ratio and effective creep Poisson ratio are not equal to each other, and the elastic Poisson ratio is greater than the creep Poisson ratio. In addition, under the conventional triaxial stress state, the effective creep (creep recovery) Poisson ratio in each direction is different, which indicates that the creep (creep recovery) in each direction is dependent under multi-axial stress state.