Static properties and impact resistance performance of U-shaped PU-modified concrete under repeated drop-weight impact load

This study investigated the impact strength of U-shaped polyurethane (PU)-modified concrete using repeated drop-weight impact tests. A U-shaped specimen was introduced in this study to modify the testing procedure of the ACI 544-2R committee for the drop-weight impact test aiming to minimize the scatter result of this testing method. The static characteristics and impact resistance of U-shaped PU-modified concrete were subjected to repeated weight drops of 0.877 kg. The effects of PU binder on the mechanical and impact strength performance were evaluated in four groups of PU-modified concrete containing 0%, 10%, 20%, and 30% prepared to produce normal concrete-polyurethane (NC-PU) specimens. Machine learning (ML) techniques, which include artificial neural network (ANN) and support vector regression (SVR), were employed to train and test the experimental dataset. The result showed that PU binders significantly decrease mechanical behavior and improve concrete impact strength. At the optimal content, flexural strength was increased by 3.6%, and the compressive strength of the control specimen (NC-PU0) was 49.17% higher than the concrete containing 20% PU (NC-PU2) specimen. The U-shaped specimens reduce the scatter result of repeated drop-weight impact test results suggested by the ACI 544-2R testing method. Adding 10% PU and 20% PU binder improved the first crack strength of NC-PU1, NC-PU2, and NC-PU3 by 685.71%, 1528.6%, and 157% compared to the NC-PU0 specimen. The machine learning techniques accurately predict the N2. ANN and SVR demonstrate a strong relationship between the experimental N2 and predicted N2 with determination correlation (R2) values of 0.9944 and 0.9981 at the training and testing phase, respectively.