A New Three-dimensional Fractal Dimension Model to Describe the Complexity of Concrete Pores

The pore structure of concrete is highly complex and random. Fractal theory is used to describe the characteristics of its pore structure. Ds (surface fractal dimension) is commonly used to characterize the surface roughness of the pores. However, the existing Ds model is limited to two dimensions and there are unreasonable assumptions about the shape of the pores and the method of measuring pores is not applicable. Therefore, this paper proposes a three-dimensional (3-D) model for calculating the Ds. The new parameter Dss represents the chaotic degree of element morphology in the 3-D pore system. When Dss=0, the pore morphology tends to be the same. The larger Dss value is, the more divergent and diversified pore morphology is. The specimens in two states of standard curing and high temperature drying were prepared and the pore structure parameters of the specimens were calculated by combining CT scanning, Deep Learning and 3-D reconstruction. By comparing Dss and Dv (volume fractal dimension) and other parameters, it is found the 3-D pore system of concrete is more chaotic and becomes more complex under the effect of high temperature drying. It also confirms the reliability of the 3-D fractal dimension model proposed in this paper.