Comparing predicted in situ 8-year concrete strength by ground penetrating radar attributes and maturity method

Concrete compressive strength is important, yet difficult to quantify without direct testing. In particular, it is difficult to obtain the mature concrete strength measurements which are necessary for safe and optimal use of existing structural capacity. Reliable measurements of mature strength using nondestructive testing methods (NDT) like ultrasonic pulse velocities depend on many factors, including the inherent material variability, sampling frequency, and quality of the NDT measurements. Methods like ground penetrating radar (GPR) and concrete maturity relationships are common for investigating the early-age properties of concrete but are rarely used for mature concrete. Using a case study of a concrete pedestrian bridge where both long term temperature data from structural health monitoring (SHM) and recent GPR surveys of the bridge are available, this work compares the predicted 8-year strength using two different indirect methods. The first uses a regression model trained on laboratory GPR attributes and material properties. The second uses the maturity method to predict strength based on 28-day cylinder tests and the temperature history recorded by the bridge’s SHM system. The maturity method predicts the correct relative trends in strength between the two phases and overpredicts the cylinder 28-day strength by 12% - 25%. The GPR predictions do not reliably capture the relative difference between the two phases, but have similar accuracy and underpredict cylinder strength by 4% - 22%. These strength comparisons from noninvasive methods motivate further improvements in GPR attribute modeling and integrating these methods with other ultrasonic models to improve spatial resolution and reliability.