The Experiment Research On Anti-corrosion By Sulfuric Acid Of Fly Ash Concrete And Silica Fume Concrete

The increasingly serious Acid rain problem, biological sulfuric acid erosion of contamination system, make the sulfuric acid erosion resistance research of concrete be of significantly importance.This paper introduces the method of artificial neural network to the sulfuric acid erosion resistance research of fly ash concrete and silica fume concrete,to explore the general rules of fly ash and silica fume isometric substitute cement in improving the concrete sulfuric acid erosion resistance performance.The following are the major work.Fitting the experiment datum according to the theory formula of common concrete and mortar acid consumption rate with software MATLAB7.1,the results showed that the fitting effect is good.that is to say, the corrosion rates of fly ash concrete and silica fume concrete varying with time share the same function family with those of common concrete and mortar.Processing the experiment datum with the software s-plus by using nonparametric regression analysis, then fitting the data before and after regression analysis with the software MATLAB7.1, and got a relational expression between acid consumption and test parameters of fly ash concrete and silica fume concrete, finally reached the action rule of sulfuric acid erosion of concrete.The basic theoretical knowledge of the BP neural network structure design based on MATLAB is introduced;The BP neural network prediction model of fly ash concrete and silica fume concrete sulfuric acid erosion resistance performance is established. Model results show that BP neural network in this study of the forecast has the very good performance. This paper introduces how to apply the established database to the actual project.