Evaluation of Compressive Strength of Ultra-High-Performance Fiber-Reinforced Concrete Using Non-Destructive Tests

Several non-destructive test (NDT) methods have been used to assess the quality of built concrete. The most popular NDT methods used for in situ testing of concrete are the rebound hammer and ultrasonic pulse velocity techniques. Several factors related to the composition of concrete mixtures affect the estimation of strength and other properties of concrete by using the values of rebound number (RN) and ultrasonic pulse velocity (UPV). However, the calibration charts/empirical models developed for converting RN and UPV into compressive strength of the conventional concrete cannot be utilized for ultra-high-performance fiber-reinforced concrete (UHPFRC) because the composition of UHPFRC is very different from the conventional concrete. Therefore, the research toward developing calibration charts/models suitable for evaluating the in situ compressive strength of UHPFRC would be of great importance. This paper presents an experimental investigation to explore the possibility of developing empirical models that can be used to calculate the in situ compressive strength of UHPFRC by substituting the measured values of RN or/and UPV with a fair degree of accuracy. For this purpose, different sets of UHPFRC specimens were prepared and tested destructively and non-destructively, considering the water/binder ratio, micro-silica content, and curing period as variable factors. Experimental data were analyzed statistically using the analysis of variance (ANOVA) method to examine the significance of the variables to the compressive strength and NDT indicators of UHPFRC. Empirical equations in different forms were obtained to evaluate the compressive strength of UHPFRC using the NDT results, and the relative accuracies and suitability of these equations were discussed. Incorporation of the curing period in the proposed models improved their accuracies in predicting UHPFRC compressive strength as indicated by the increase in R² values, e.g., 0.79–0.91, 0.59–0.98, and 0.96–0.99 for the models in terms of RN alone, UPV alone, and RN and UPV together, respectively.