Predictive methods for the evolution of oil well cement strength based on porosity

Abstract

The oil well cement undergoes various physical and chemical changes during the hydration process, leading to the formation of pores of different sizes within the cement stone. These pores can affect the mechanical properties of the cement stone. In the civil engineering field, extensive attempts have been made to predict the mechanical properties of concrete based on pore parameters, yielding good results. This paper explores in detail the methods for predicting the strength of oil well cement based on porosity and pore size distribution. Through referencing the strength prediction methods for concrete in civil engineering, porosity and pore size distribution are used as prediction parameters. The accuracy of predictions made by empirical models and deep learning models is compared, and it is concluded that neither empirical formulas nor ordinary deep learning models can provide accurate fitting results. However, due to the optimization of its algorithm and structure, the KAN model can give more accurate predictions of the pore-size-strength relationship of cement stone. Additionally, the quantitative relationship between pore size and strength of cement stone is explored. The application of the KAN model in strength prediction provides strong guidance for monitoring and optimizing cementing quality during the construction process.