Abstract: Experimental database of 318 recycled aggregate concrete (RAC) columns confined by fiber reinforced polymer (FRP) is established in this paper, the applicability of three representative calculation models (Lim’s model, Xiao’s model and Wei’s model) in predicting compressive strength and ultimate strain by the database is investigated. The results show that the prediction accuracy of the three calculation models for compressive strength is higher than that for ultimate strain. Among the three models for predicting compressive strength, Lim’s model shows best accuracy, the average absolute error AAE is only 9.4%, and the average value of the ratio between the predicted value and the experimental value is 100.8%. The predicted value of the compressive strength by Wei’s model is generally lower than the experimental value (Mean=78.9%), and the error is larger (AAE=21.2%) compared with the experimental value. Prediction accuracy of Xiao’s model (AAE=12.8%, Mean=109.8%) is between that of Lim’s model and Wei’s model. When three calculation models predict compressive strength, Lim’s model has the highest accuracy (AAE=21.1%, Mean=95.5%), followed by Wei’s model which underestimates the ultimate strain of RAC confined by FRP (AAE=28.5%, Mean=78.1%). Compared with the experimental value, the predicted value of Xiao’s model for ultimate strain is overall higher (Mean=125.9%), and the error is relatively large (AAE=40.1%). In general, Lim’s model has the best accuracy among the three calculation models in predicting both compressive strength and ultimate strain, and it is recommended to predict the limit state of calculation models in predicting both compressive strength and ultimate strain, and it is recommended to predict the limit state of recycled aggregate concrete confined by FRP.

Key words: FRP confinement, recycled aggregate concrete, compressive strength, ultimate strain, prediction accuracy, composites