The construction of high-rise oblique grid buildings requires a large amount of concrete. To save resources, an oblique grid of intersecting nodes composed of steel outer tubes and steel fiber, recycled concrete inner tubes (OGSFRCIN) has been proposed. ABAQUS is used to study the mechanical properties of the nodes under axial pressure, accounting for the effects of six parameters: the oblique angle, the thickness of the stiffening ring, the thickness of the connecting plate, the concrete strength, the recycled aggregate replacement rate, and the steel fiber content. The results show that the oblique angle, connecting plate thickness, concrete strength, and steel fiber content significantly affect the ultimate bearing capacity of specimens. The reinforcing ring thickness has a significant effect on the evolution of lateral displacement. It is not advisable to use components with a replacement rate of 100% recycled aggregate in engineering practice because of insufficient lateral stiffness and ultimate strength. The specimen’s out-of-plane displacement is tightly restricted when the connecting plate’s thickness is greater than or equal to 10 mm. In practical engineering, the connecting plate and reinforcing ring thickness should not be less than 10 mm, and the recommended steel fiber content is 1.0% to 2.0%. Through the analysis of the mechanical properties of the OGSFRCIN under monotonic axial compression and reciprocating axial tension and compression loads, it can be seen that OGSFRCIN have good mechanical properties and can be applied in engineering practice. Here, the modified formulas for calculating the bearing capacity of OGSFRCIN are put forward.

/pdf/mechanical-properties-of-steel-fiber-reinforced-recycled-2elr89vp.pdf

Mechanical Properties of Steel Fiber-Reinforced, Recycled, Concrete-Filled Intersecting Nodes in an Oblique Grid

The design of intersecting nodes in high-rise oblique mesh structures is a critical issue. The existing research on the intersecting nodes of oblique meshes mainly focuses on plane intersecting nodes and monotonic axial compression loads. The plane intersecting nodes cannot consider the contribution of the node’s out-of-plane angle and floor beam to the node’s out-of-plane stiffness in actual structures. In this paper, numerical analysis using ABAQUS was conducted to investigate the mechanical performance of space intersecting nodes of oblique meshes (OMSIN) under cyclic axial tension and compression loads, to provide a reference for the engineering application of oblique mesh structures in seismic regions. Six parameters were considered: the space intersecting angle, the plane angle symmetry coefficient, the plane intersecting angle, the out-of-plane constraint restraint, the steel content of the cross-section, and the concrete strength. The study showed that changes in the thickness of the steel tube wall are unfavourable for the uniform transmission of stress. Increasing the space intersecting angle significantly weakened the seismic performance, and the space angle affects the failure mode of the node. Asymmetric arrangements of the upper and lower plane angles caused nonlinear development of out-of-plane. The ultimate load and overall compressive stiffness of the specimen were positively correlated with the plane angle, and vertical constraints should be applied to the node position of components with plane angles greater than or equal to 70°. The out-of-plane constraint was a key factor affecting the seismic performance of the node, and it was proportional to the ultimate load of the component. In structural design, if the aim is to improve the mechanical performance of the component by increasing the steel content, more enormous out-of-plane constraints should be set to control plane external displacement strictly. The concrete strength is proportional to the ultimate axial load and axial stiffness, and its influence on the mechanical performance in the axial tension direction is not significant. Finally, a dimensionless skeleton curve model of the node was established. The existing formula for the bearing capacity of CFST columns was fitted to obtain the calculation formula for the axial yield and ultimate load of the OMSIN under cyclic loads.

/pdf/seismic-behaviour-of-cfst-space-intersecting-nodes-in-an-iu0awovj.pdf

Seismic Behaviour of CFST Space Intersecting Nodes in an Oblique Mesh

Sobel is one of the most popular edge detection operators used in image processing. To date, most users utilize the two-directional 3x3 Sobel operator as detectors because of its low computational cost and reasonable performance. Simultaneously, many studies have been conducted on using large multi-directional Sobel operators to satisfy their needs considering the high stability, but at an expense of speed. This paper proposes a fast graphics processing unit (GPU) kernel for the four-directional 5x5 Sobel operator. To improve kernel performance, we implement the kernel based on warp-level primitives, which can significantly reduce the number of memory accesses. In addition, we introduce the prefetching mechanism and operator transformation into the kernel to significantly reduce the computational complexity and data transmission latency. Compared with the OpenCV-GPU library, our kernel shows high performances of 6.7x speedup on a Jetson AGX Xavier GPU and 13x on a GTX 1650Ti GPU.

Multi-directional Sobel operator kernel on GPUs

Deep convolutional neural network (DCNN for short) models are vulnerable to examples with small perturbations. Adversarial training (AT for short) is a widely used approach to enhance the robustness of DCNN models by data augmentation. In AT, the DCNN models are trained with clean examples and adversarial examples (AE for short) which are generated using a specific attack method, aiming to gain ability to defend themselves when facing the unseen AEs. However, in practice, the trained DCNN models are often fooled by the AEs generated by the novel attack methods. This naturally raises a question: can a DCNN model learn certain features which are insensitive to small perturbations, and further defend itself no matter what attack methods are presented. To answer this question, this paper makes a beginning effort by proposing a shallow binary feature module (SBFM for short), which can be integrated into any popular backbone. The SBFM includes two types of layers, i.e., Sobel layer and threshold layer. In Sobel layer, there are four parallel feature maps which represent horizontal, vertical, and diagonal edge features, respectively. And in threshold layer, it turns the edge features learnt by Sobel layer to the binary features, which then are feeded into the fully connected layers for classification with the features learnt by the backbone. We integrate SBFM into VGG16 and ResNet34, respectively, and conduct experiments on multiple datasets. Experimental results demonstrate, under FGSM attack with $\epsilon=8/255$, the SBFM integrated models can achieve averagely 35\% higher accuracy than the original ones, and in CIFAR-10 and TinyImageNet datasets, the SBFM integrated models can achieve averagely 75\% classification accuracy. The work in this paper shows it is promising to enhance the robustness of DCNN models through feature learning.

Improving the Robustness of Deep Convolutional Neural Networks Through Feature Learning

Image Sobel edge extraction algorithm accelerated by OpenCL

Mechanical Behavior of Concrete-filled Steel Tubular Space Intersecting Nodes in High-rise Oblique Diagrid Tube Structures

In this paper, the notion of vector-valued multiresolution analysis and biorthogonal vector wavelets are introduced. The existence of biorthogonal finitely supported vector-valued wavelets associated with a pair of biorthogonal finitely supported vector-valued scaling functions is discussed. A procedure for designing a class of biorthogonal vector-valued finitely supported wavelet functions is proposed by virtue of multiresolution analysis approach and filter bank theory.

Ge Ma

Papers

Image Sobel edge extraction algorithm accelerated by OpenCL

Mechanical Behavior of Concrete-filled Steel Tubular Space Intersecting Nodes in High-rise Oblique Diagrid Tube Structures

A novel algorithm for designing biorthogonal five-scale wavelets