Compressive strength and intelligent prediction modeling of early-strength cement-based grouting materials considering age, shape, and size effects

As an essential construction material, early-strength cement-based grouting materials are extensively applied in structural reinforcement and related engineering fields. To investigate the effects of specimen age and size on the compressive strength of early-strength cement-based grouting materials and to develop an intelligent prediction model, five specimen sizes were prepared. Compressive strength tests were performed on specimens of different sizes and ages. A backpropagation (BP) neural network model optimized using a particle swarm optimization (PSO) algorithm was employed to train and validate machine learning models based on the experimental data, followed by a comparative analysis of prediction accuracy.The results indicate that the compressive strength of early-strength cement-based grouting materials increases with age, exhibiting a growth trend similar to ordinary concrete. The strength rises rapidly at early stages and slows progressively over time, demonstrating pronounced early-strength characteristics and a faster strength development rate compared with ordinary concrete. The specimen shape significantly influences the compressive strength of early-strength cement-based grouting materials. Moreover, a pronounced size effect is observed, with the size effect coefficient generally decreasing as the overall strength increases. Analysis of experimental data and model accuracy comparisons demonstrate that the PSO-BP neural network predicts the compressive strength of early-strength cement-based grouting materials more effectively than other machine learning models. The proposed approach achieves high prediction accuracy for the compressive strength of early-strength cement-based grouting materials and offers valuable insights for future studies in this field.