三维凸型混凝土骨料随机投放算法

Random 3-D aggregate structure for concrete

为了提供三维混凝土细观强度研究的数值模型,进行了三维混凝土随机骨料的投放算法研究。以体积为标度建立了一般多面体和凸型多面体两种形状骨料的侵入判别准则。并以凸多面体为研究对象建立了凸型骨料的生长模式和骨料"凸性"条件,给出相应的随机投放算法。该算法中,投放骨料初始形状为任意空间八面体,并始终在最长边上不断生长。在整个生长和投放过程中,骨料始终保持"凸"型。投放实例表明,与球形骨料假定建立的数值混凝土模型相比,凸多面体随机骨料模型能够更广泛地模拟混凝土材料结构。简单的算例说明了该模型可用于混凝土不均质性研究。

A 3D numerical model was developed for microscale concrete strength analysis using an algorithm for random 3D aggregate packing. Aggregate location conflicts and overlaps were eliminated with conflicting and overlapping criteria in the algorithm. Growth pattern and convexity conditions were analyzed for a convex polyhedron aggregate. Each aggregate grew from a random spatial octahedron and always maintained a convex shape during the growth and packing processes. The convex polyhedron aggregate model more accurately simulates the concrete material structure than the spherical aggregate model. Numerical tests show that the 3D random aggregate model provides sufficiently accurate results for concrete inhomogeneity studies.