Artificial Neural Network-Based Relationship Modeling of Concrete Compressive Strength and Schmidt Hammer Results

The compressive strength of concrete is a crucial parameter for evaluating the safety, performance, and quality control of concrete structures. This strength develops over time, with concrete typically reaching its approximate ultimate strength after ۲۸ days under standard curing conditions. Measuring this parameter requires testing on samples. In many in-service structures, performing destructive tests to determine strength is limited, costly, and time-consuming; therefore, non-destructive methods, such as the Schmidt hammer test, have attracted attention as a rapid and cost-effective alternative. However, the accuracy of such methods is influenced by several factors, including the age of the concrete, environmental conditions, stress state, and temperature. To provide more accurate predictions and improve the modeling of concrete behavior, this study employed a feedforward artificial neural network as an intelligent tool to estimate the ۲۸-day compressive strength of concrete. The model inputs included early-age strength data (less than ۷ days) and Schmidt hammer test results (from ۷ to ۲۰ days), enabling the capture of the strength development trend from the first day up to the ۲۸th day. By training the neural network on these data, the model successfully learned the complex relationships between the input parameters and the ultimate concrete strength, providing reliable predictions. The results demonstrated that this approach can effectively predict the overall strength development of concrete and estimate its ultimate compressive strength with high accuracy, without the need for multiple destructive tests.

کلیدواژه ها:

concrete ، ultimate compressive strength ، Schmidt hammer ، artificial neural network