China 
Africa 
Explore chapters and articles related to this topic
Wave and Structure Interaction Porous Coastal Structures
Published in David M Kelly, Aggelos Dimakopoulos, Pablo Higuera, Advanced Numerical Modelling of Wave Structure Interactions, 2021
The Discrete Element Method, also known as Distinct Element Method, is a numerical method that has been widely used in literature to solve discontinuous mechanics problems such as granular flow or rock mechanics simulations since Cundall and Strack (1979). More recent advances include the so-called extended discrete element method (XDEM) (Peters, 2013), which enhances DEM, allowing solving multi-physics problems such as chemical reactions (e.g., pellets acting as catalysts), heat transfer or electro-magnetic forces (e.g., magnetic particle sorting and separation). Although DEM particles are often spherical in shape, this is not a restriction, as complex shapes can be built from overlapping spheres of different sizes (Ferellec and McDowell, 2010). There are two main approaches for sphere interaction. In the inelastic hard sphere method, the spheres are rigid and particle collisions produce an exchange of momentum through impulsive forces. This method has a limitation, it does not allow sustained contact between particles, therefore, it is suitable for the so-called collisional particle regime only. The other method is the soft sphere approach, which is more common in DEM. In it, particles, although still assumed rigid, are allowed to overlap slightly. This way, the interaction between particles takes a finite time and allows simulating a sustained contact between particles (i.e., frictional forces) as well as regular collisions (Mitarai and Nakanishi, 2003). The soft sphere method will be used in the present work.
DEM Algorithm for Progressive Collapse Simulation of Single-Layer Reticulated Domes under Multi-Support Excitation
Published in Journal of Earthquake Engineering, 2019
Different from the aforementioned methods, the DEM proposed by Cundall and Strack [1979] in 1971 is derived from noncontinuum mechanics and is uniquely advantageous in the simulation of structural large deformation and fracture behaviors. The difficulty of DEM lies in establishing a reasonable contact constitutive model. Lorig and Cundall [1989] introduced DEM into the analysis of a reinforced concrete structure, and successfully simulated the fracture process of the component under static and dynamic loads. Meguro and Hakuno [1989] analyzed the progressive failure of a reinforced concrete frame structure resulting from an earthquake using DEM. Since then, DEM has been continuously extended, and many new methods, such as extended discrete element method [Meguro and Hakuno, 1989], applied element method [Tagel and Meguro, 2000], plane segment model [Utagawa et al., 1992], and layer element model [Gu et al., 2012], have been presented. Rafiee and Vinches [2013] studied the mechanical behavior of a stone masonry bridge under static load by DEM and compared it to actual failure phenomena. Olmedo et al. [2016] simulated the dynamic responses of fresh wood stems subjected to impact load. However, the above results are available for concrete and masonry structures rather than large-span spatial steel structures. Furthermore, they are mostly limited to qualitative analysis of failure phenomena, and are of low accuracy. To figure out these problems, the research group of Prof. Ye [Qi and Ye, 2013; Ye and Qi, 2017] derived a new contact constitutive model for member structures and called the member discrete element method (MDEM). Recently, it has been successfully applied to geometric nonlinearity analysis of member structures [Qi and Ye, 2013] and collapse simulation of reticulated shells [Ye and Qi, 2017]. However, the subjects of these works were small structures under uniform excitation. Spatial shell structures are large and there are various members. To effectively obtain true responses of structures, the applicability of the MDEM for large structures under multi-support excitation is a problem that must be solved.