【作者】 袁群；

【导师】 赵国藩；

【作者基本信息】 大连理工大学 ， 结构工程， 2000， 博士

【摘要】 本文的主要内容如下： 1）对混凝土碳化深度的的变化规律进行了研究，提出了预报碳化深度的ARIMA（1,1,0）模型。 2）揭示了混凝土裂缝分布具有的分形现象，并提出混凝土裂缝分布分维数是衡量结构老化程度的一个有效指标。 3）采用Bayes方法，对超声～回弹法测强与回弹法测强联合应用时，客观评价混凝土的强度做了一次有益的尝试分析。 4）对功率谱法分维和高差法分线作为混凝土粘结面粗糙度的定量评价方法进行了研究。 5）采用塑性极限分析中的上限定理，推出了新老混凝上粘结层剪切强度的理论解。并从破坏机理上分析了影响粘结层剪切强度的因素；及对低静水应力状态下，粘结层剪切强度与其上作用的正应力关系进行了阐述。 6）阐述了新老混凝土粘结的微观机理，建立了新老混凝土粘结劈拉强度及粘结抗折强度的BP神经网络预测模型。更多还原

【Abstract】 The main contents of the thesis are as follows: 1) The growth rule of concrete carbonization depth was studied, and the ARIMA (1,1,0) model was presented to forecast the carbonization depth. 2) It was discovered that the concrete crack distribution had the fractal character. It was presented that the fractal dimension value of crack distribution was an effective index to appraise the structural aging degree. 3) With Bayes method, a valuable analysis was attempted to evaluate objectively the concrete strength with ultrasonic - rebound method and rebound method together. 4) As quantitative evaluating methods of the concrete bonding face roughness, the power spectrum fractal method and the height difference fractal method were studied to determine the fractal dimension. 5) With the upper bound theorem of plastic limit analysis, the theoretical solution for shear strength of new-old concrete bond layer was deduced, and the factors, which affect the shear strength of bond layer, were also determined with failure mechanism. The relationship between the shear strength of bonding layer and its normal stress was expatiated under low hydrostatic stress-state. 6) The micro-mechanism of new-old concrete bond was analyzed, and the BP neural network models were established to predict the bond split- tensile strength and the bond flexure strength of new-old concrete.更多还原