No abstract

This article presents a numerical method for determining tortuosity in porous beds consisting of randomly packed spherical particles. The calculation of tortuosity is carried out in two steps. In the first step, the spacial arrangement of particles in the porous bed is determined by using the discrete element method (DEM). Specifically, a commercially available discrete element package (PFC 3D ) was used to simulate the spacial structure of the porous bed. In the second step, a numerical algorithm was developed to construct the microscopic (pore scale) flow paths within the simulated spacial structure of the porous bed to calculate the lowest geometric tortuosity (LGT), which was defined as the ratio of the shortest flow path to the total bed depth. The numerical algorithm treats a porous bed as a series of four-particle tetrahedron units. When air enters a tetrahedron unit through one face (the base triangle), it is assumed to leave from another face triangle whose centroid is the highest of the four face triangles associated with the tetrahedron, and this face triangle will then be used as the base triangle for the next tetrahedron. This process is repeated to establish a series of tetrahedrons from the bottom to the top surface of the porous bed. The shortest flow path is then constructed geometrically by connecting the centroids of base triangles of consecutive tetrahedrons. The tortuosity values calculated by the proposed numerical method compared favourably with the values obtained from a CT image published in the literature for a bed of grain (peas). The proposed model predicted a tortuosity of 1.15, while the tortuosity estimated from the CT image was 1.14.

Real-time smoke detection is of great significance for early warning of fire, which can avoid the serious loss caused by fire. Detecting smoke in actual scenes is still a challenging task due to large variance of smoke color, texture, and shapes. Moreover, the smoke detection in the actual scene is faced with the difficulties in data collection and insufficient smoke datasets, and the smoke morphology is susceptible to environmental influences. To improve the performance of smoke detection and solve the problem of too few datasets in real scenes, this paper proposes a model that combines a deep convolutional generative adversarial network and a convolutional neural network (DCG-CNN) to extract smoke features and detection. The vibe algorithm was used to collect smoke and nonsmoke images in the dynamic scene and deep convolutional generative adversarial network (DCGAN) used these images to generate images that are as realistic as possible. Besides, we designed an improved convolutional neural network (CNN) model for extracting smoke features and smoke detection. The experimental results show that the method has a good detection performance on the smoke generated in the actual scenes and effectively reduces the false alarm rate.

No abstract

No abstract