Study On Strength Permeability Relationship And Formation Mechanism Of Recycled Aggregate Permeable Concrete Based On BP Neural Network

The further research of recycled aggregate permeable concrete(RAPC)help to promote construction waste comprehensive utilization industry standardization and scale,and for the sponge city construction and application has a positive role.In this paper,the broken sources of waste concrete pavement for recycled aggregate,by changing the traditional RAPC "no sand" concept,design sand ratio range is 0% ~ 28% of the patients,according to the test data to determine the optimal sand ratio of RAPC;On this basis,two kinds of fine aggregate,waste glass particles(6 gradient substitution)and waste rubber particles(7 gradient adding),were selected to analyze their influence on the strength and permeability of RAPC and their variation tendency.At the same time,by means of nuclear magnetic resonance(NMR)technology and the Image method,RAPC in sand ratio,dosage of waste glass particles and waste rubber particles under the condition of pore structure characteristic information,the relationship between the analysis and macro performance.In addition,based on the BP neural network method to establish two types of RAPC performance prediction model,analysis of the relationship between the raw material and macro performance,macro and macro performance of RAPC.The results show that:(1)With the increase of river sand content,the strength performance of RAPC shows an overall growth trend,while the permeability performance shows a continuous decline.By analyzing the test data of strength and permeability,think that when sand at a rate of 7%,top best-permeability for the looking for balance.(2)RAPC strength with the increase of waste glass particles replacement rate showed a tortuous changes,increase or decrease when the replacement rate is 60%,strength and permeability of the RAPC have got better improvement;Mixed with waste rubber particles make its strength and permeable coefficient are decreased,and the porosity is slightly increase,analysis waste rubber particles on the improvement of the RAPC effect is not obvious.(3)By means of nuclear magnetic resonance(NMR)technology test of sand at a rate of 0%,7% and 28%,waste glass replacement rate of 40%,80%,waste rubber admixture is 15%,25% of the specimen,found RAPC internal pore types of intensive to small pore far exceed big pore,and big pore ratio above 97%;With the increase of sand ratio and the dosage of waste rubber,RAPC surface pore size and porosity is on the decline;Pore area and waste glass replacement plane and plane porosity increased and performance first.(4)Based on BP neural network method to establish the two types of RAPC performance prediction model,the results show that the average relative error of different kinds of model are within 10%,and that between predicted values are very close to the measured values,the prediction accuracy is more accurate.