Main participants: MA Huaifa, CHEN Houqun, XU Wenxiang, WU Jianping, ZHANG Lihong, LV Wei, LIU Guoqing, ZENG Xinxiang, ZHU Hongdong, HE Jiantao, CHANG Tinggai, XU Lianghua
1.1 Background
· Dynamic damage mechanism and mesoscopic numerical analysis of dam concrete materials;
· The solving method of contact nonlinearity and material nonlinearity of concrete dam:
· The artificial boundary of semi-infinite foundation and ground motion input method;
· Algorithm research and program development of a static and dynamic parallel computing system for high concrete dams;
· Seismic safety evaluation indicators and approaches for concrete dams.
1.3 Achievements
· A method for generating a random convex polyhedral (convex polygon) gravel aggregate model for fully-graded concrete and a topological method for generating meso-structure model of star-shaped solid reinforced particulate composite material are proposed. Based on the proposed meso-mechanics theory of concrete, the mechanism of enhancement effect of pre-static loading on dynamic tensile strength of concrete is revealed.
· A surface-to-surface contact algorithm with pre-partition for parallel computation and a step-by-step contact algorithm with small storage capacity and easy parallelization are presented. A fully implicit iterative method for the elasticoplastic damage of concrete-like materials are proposed, which can describe the nonlinear deformation characteristics of concrete materials more comprehensively.
· The principle of artificial viscoelastic virtual displacement of semi-infinite rock foundation and the principle of dynamic consolidation of semi-infinite saturated soil foundation are established respectively, and the symmetrized splitting operator method for dynamic consolidation problem of saturated porous semi-infinite foundation is proposed.
· A parallel preprocessing technique of factor combination for general sparse linear equations is developed, and a coarse grid correction algorithm for domain decomposition based on parallel preconditioners is proposed. On this basis, a general parallel preconditioned subspace iterations (GPPS) for solving general sparse linear equations is proposed, which provides an efficient technical means for large-scale numerical simulation of high concrete dams.
Related research results have been published in Computers and Structures, Composites science and technology, Powder technology, Composite Structures, Soil Dynamics and Earthquake Engineering and some other renowned international academic journals. As of now, the developed GPPS solver has been applied to geophysical, space technology, weather forecasting and other related research fields. The parallel calculation and analysis system for seismic response of concrete dam system (PCDSRA) established has been applied to seismic safety evaluation of major hydropower projects such as Xiaowan, Xiluodu, Baihetan, etc., and has built a high-performance computing platform for the seismic safety evaluation of high concrete dams.
Figure Mesoscopic numerical simulation of concrete
Figure: Parallel computing of seismic response of high arch dams