Research On Shear Capacity Of Seawater Sea-sand And Self-Compacting Concrete Beams With GFRP Bars And Stirrups Based On Neural Network

The reduction of the structural bearing capacity caused by the corrosion of steel bars and the degradation of the overall structural performance has become a major disease of construction structures.Under this background,Fiber Reinforced Polymer(FRP)bar is encouraged to replace steel bar because of the light weight,high strength,fatigue resistance and extreme corrosion resistance of the FRP material.Because of the dilemma of extreme shortage of freshwater river sand resources required for engineering materials.The use of seawater and sea sand to fabricate concrete has become an option.Self Compacting Concrete is prepared by introducing cementitious materials such as fly ash and limestone powder,which can effectively reduce the amount of cement and improve the fluidity of concrete.At present,the combination of FRP composite materials,sea water and sea sand has become a research hotspot.The research on flexural behavior of FRP reinforced concrete structures has made significant progress,but the research on shear behavior of concrete structures reinforced with FRP stirrups is relatively lack.On the other hand,various countries proposed the shear design formulas for FRP reinforced concrete structures,but the results obtained are mostly conservative and have certain limitations.Due to the contribution of each core parameter to the mechanism of shear resistance is not clear,and also the brittle failure nature of shear failure.A more efficient and accurate method to predicte the ultimate shear capacity of FRP reinforced concrete structures is also waited.The Artificial Neural Network(ANN)algorithm has been proved to be an effective technique for predicting the shear strength of FRP reinforced concrete members.The type of concrete matrix has no significant effect on the ultimate shear capacity,so relevant research can be carried out by using the existing FRP reinforced beam shear database.This paper intends to conduct a neural network-based experimental study on the shear performance of Seawater Seasand and Self-Compacting Concrete(SS-SCC)beams with FRP rebars and stirrups.The main contents include:(1)Four-point loading tests were carried out on nine SS-SCC beams with GFRP rebars and stirrups,around the multiple core issues such as shear strength,failure mode,failure mechanism and calculation model,the focus was on shear span ratio,beam height,and GFRP longitudinal reinforcement ratio,the presence or absence of GFRP stirrups and the type of concrete matrix on the shear resistance of SS-SCC beams with GFRP rebars and stirrups and the law of influence,analysis of the cracking load,crack development and extension,beam deformation law,GFRP longitudinal reinforcement,stirrup strain and beam failure mode.And comparative and analysis of the existing theoretical calculation models for the shear capacity of FRP reinforced concrete beams,and evaluation based on the existing shear test database and the test data in this paper.(2)Relying on a large sample test database,a prediction model for the shear capacity of concrete beams with FRP rebars is constructed through machine learning methods such as artificial neural networks,and the feasibility of using the neural network model for prediction of the shear capacity is evaluated.This model is used for parameter analysis to study the influence of different parameters on the shear capacity of a SS-SCC beam with GFRP rebars and stirrups.(3)Based on the sample database of reinforced concrete beams with FRP bars,a neural network model that can be established to predict the shear capacity of beams with and without FRP stirrups.Compare it with the prediction results of national standards and scholars’ calculation models to verify the feasibility of the network model.Based on the neural network model,select the required parameters to study the size effect of the shear strength of concrete beams without FRP stirrups,and analyze the influence law of parameters such as concrete strength,shear span ratio,FRP bar longitudinal reinforcement ratio,and FRP bar elastic modulus and component width on the size effect of concrete beams with FRP bars without web reinforcement,and a prediction formula for shear bearing capacity considering the size effect is proposed.(4)Using the constructed neural network model to peel off the contribution of FRP stirrups to the shear capacity.Then,carry out parameter analysis of shear capacity of stirrups.Discuss the influence of various parameters on the shear contribution of FRP stirrups,and reveal the coupling effect between FRP stirrups and various parameters.Finally,a prediction formula of the shear capacity of FRP stirrups considering the coupling between various parameters and FRP stirrups is proposed.Which is compared,analyzed,and also evaluate the prediction accuracy of shear capacity.(5)Through the shear test database to evaluates prediction expression of the shear capacity of concrete beam with FRP rebars and stirrups based on parameter study and regression analysis and considering the size effect and the coupling effect.