Search

The bond strength between ultra-high-performance concrete (UHPC) and normal-strength concrete (NC) plays an important role in governing the composite specimens' overall behaviors. Unfortunately, there are still no widely accepted formulas targeting UHPC–NC interfacial strength, either in their specifications or in research papers. To this end, this study constructs an experimental database, consisting of 563 and 338 specimens for splitting and slant shear tests, respectively. Moreover, an additional 35 specimens for "improved" slant shear tests were performed, which could circumvent concrete crushing and trigger interfacial debonding. Additionally, for the first time in our tests, the effect of casting sequence on UHPC–NC bond strength was identified. Based on the database, an artificial neural network (ANN) model is proposed with the following inputs: namely, the normal stress perpendicular to the interface, the interface roughness, and the compressive strengths of the UHPC and NC materials. Based on the ANN analyses, the explicit expression of UHPC–NC bond strength is proposed, which significantly lowers the prediction error. To be fully compatible with the specifications, the conventional shear-friction formula is modified. By splitting the total force into adhesion and friction forces, the modified formula additionally takes the casting sequence into account. Although sacrificing accuracy to some extent compared to the ANN model, the modified formula relies on a solid physical basis and its accuracy is enhanced significantly compared to the existing formulas in specifications or research papers. [ABSTRACT FROM AUTHOR]

Linear feature extraction is crucial for special objects in semantic segmentation networks, such as slot marking and lanes. The objects with linear characteristics have global contextual information dependency. It is very difficult to capture the complete information of these objects in semantic segmentation tasks. To improve the linear feature extraction ability of the semantic segmentation network, we propose introducing the dilated convolution with vertical and horizontal kernels (DVH) into the task of feature extraction in semantic segmentation networks. Meanwhile, we figure out the outcome if we put the different vertical and horizontal kernels on different places in the semantic segmentation networks. Our networks are trained on the basis of the SS dataset, the TuSimple lane dataset and the Massachusetts Roads dataset. These datasets consist of slot marking, lanes, and road images. The research results show that our method improves the accuracy of the slot marking segmentation of the SS dataset by 2%. Compared with other state-of-the-art methods, our UnetDVH-Linear (v1) obtains better accuracy on the TuSimple Benchmark Lane Detection Challenge with a value of 97.53%. To prove the generalization of our models, road segmentation experiments were performed on aerial images. Without data argumentation, the segmentation accuracy of our model on the Massachusetts roads dataset is 95.3%. Moreover, our models perform better than other models when training with the same loss function and experimental settings. The experiment result shows that the dilated convolution with vertical and horizontal kernels will enhance the neural network on linear feature extraction. [ABSTRACT FROM AUTHOR]