基于随机骨料数学模型的混凝土弹性模量预测

按照混凝土试件的实际配比计算出各粒径区间的骨料体积含量,对骨料颗粒进行随机投放,生成随机骨料的数值模型。对不同形状骨料的随机骨料模型建立的体积表征单元(RVE)施加均匀位移边界条件,通过有限元方法计算该RVE中的应力和应变场,然后由细观力学数值均匀化方法预测RVE的有效弹性模量。数值计算结果和试验值的对比表明,基于随机骨料模型的混凝土弹性模量的预测是有效的,骨料含量为40%和60%时预测值与试验值相近;由多边形骨料模型得到的弹性模量略高于圆形骨料,骨料的空间分布对于预测结果的影响甚微。假定材料的细观单元力学性质服从Weibull分布时,预测的有效弹性模量结果偏低,且随均质度的减小而减小。

Prediction of concrete elastic modulus based on numerical ranndom aggregate model

The numerical random aggregates in the model are generated by means of determining the locations of aggregates randomly.The volumes of aggregates for different grain size intervals are calculated according to the real mixture ratios of the specimens for experimental study.The homogeneous strain boundary condition is applied to the representative volume elements(RVE) established by the aggregates model with various shapes,and stress-strain field of these RVE are calculated by FEM.On this basis,the effective elastic modulus can be predicted by using numerical homogenization method at mesoscopic level.The comparison of calculation result with experimental data indicates that the difference is acceptable.The approximation is satisfactory when volume faction of aggregate is 40% or 60%. The predicted results obtained by the model with polygonal aggregate are slightly higher than those with circular aggregate.The effect of aggregate distribution is not remarkable.The predicted result will be slightly lower than the reality if the mechanical properties of mesoscopic element are assumed to obey the Weibull distribution and will be lower and lower if homogeneity degree is decreasing.