Research on intelligent prediction method of fresh concrete slump process
The quality of concrete pouring has an important influence on the mechanical properties of complex concrete structures and composite structures.The traditional research methods usually use full-scale pouring test,which is time-consuming and labor-consuming.Based on computational fluid dynamics(CFD)method,machine learning model XGBoost and video depth generation model SimVP,an intelligent prediction method for the slump process of fresh concrete was proposed.Firstly,the slump test of concrete was simulated by CFD software,and the velocity cloud diagram in slump process was generated,which provides the data set for the prediction model.Then,intelligent prediction model of typical concrete mix ratio(including mixing time)-rheological parameters(yield stress and plastic viscosity)and rheological parameters(yield stress and plastic viscosity)-slump process were constructed respectively.The results show that the intelligent prediction model can accurately predict the slump process of concrete,and the predicted results are in good agreement with the CFD method in the two indexes of slump flow and flow velocity,and the predicted speed is 4～8 times higher than that of CFD method.
concrete pouringCFD simulationartificial intelligencevideo predictionslump of concrete
万方数据
维普
