Calibration of overtopping roughness factors for concrete armor units in non-breaking conditions using the CLASH database

Abstract

The roughness factor (γf) is a parameter used in overtopping estimators to account for the effects of armor unit geometry, the number of layers in the armor and other structural characteristics. Different values of γf for the same armors are given in the literature; however, in this study, the value of γf was calibrated for each overtopping estimator using the best available overtopping data. A methodology based on a bootstrapping technique is proposed to statistically characterize the roughness factors which best fit each formula. For each selected armor unit and overtopping formula, three percentiles (10%, 50% and 90%) of γf are given. Five sets of γf are given for five different overtopping estimators, calibrated using overtopping data from the CLASH database and additional tests with Cubipod armors in non-breaking conditions. The results indicate differences up to 20% in the optimum values of γf compared to those given in the literature. Optimum roughness factors are provided for the CLASH neural network (CLNN); the CLNN was found to be a better overtopping estimator than the other four overtopping formulas compared in this study. The γf is dependent on both the overtopping estimator and the dataset used. Armor porosity affects not only armor roughness and overtopping but also armor hydraulic stability; thus, recommended packing densities must be followed to avoid changes in porosity during lifetime. The sensitivity of the overtopping prediction to the roughness factor depends on the relative crest freeboard (Rc/Hm0); the greater the Rc/Hm0, the higher the sensitivity.