Full probability conversion model for predicting concrete compressive strength using the rebound method

Abstract

The conversion model forms the basis for predicting concrete compressive strength using the rebound method and plays a crucial role in improving prediction accuracy. Traditional approaches, such as regression and calibration methods, primarily estimate the mean compressive strength while neglecting the full probabilistic relationship between the rebound number and compressive strength. To overcome this limitation, a full probability conversion model is proposed using the Copula function method, which effectively captures the joint probability distribution between the rebound number and compressive strength. In addition, a Bayesian full probability conversion model is introduced, enabling the integration of core sample data to enhance the predictive accuracy of compressive strength. To validate and compare the proposed method, 20 datasets comprising 1838 rebound number and compressive strength pairs were analysed. Results demonstrate that the proposed full probability conversion model improves the prediction accuracy, particularly when combined with the Bayesian update method. Moreover, the proposed method delivers comprehensive probabilistic information for predicting concrete compressive strength, offering a more complete and reliable understanding than traditional approaches.