TY - JOUR
T1 - 3-D DIRECT SIMULATION OF DRIFTWOOD ACCUMULATION IN A REAL RIVER FLOOD
AU - Shen, Dawei
AU - Aoki, Takayuki
AU - Watanabe, Seiya
AU - Moriguchi, Shuji
AU - Takase, Shinsuke
AU - Sakuraba, Masaaki
N1 - Publisher Copyright:
© 2024 Japan Society of Civil Engineers. All rights reserved.
PY - 2024
Y1 - 2024
N2 - Severe damage from river flooding carrying driftwood has become a critical problem. We have developed a code for 3-D simulations of water flows carrying a lot of driftwood by using the cumulant lattice Boltzmann method (LBM) coupled with the discrete element method (DEM). Adaptive mesh refinement (AMR) makes it possible to apply the code to large areas with a high-resolution mesh, and a GPU implementation greatly accelerates the computation. A small-scale water-tank experiment with a driftwood model showed that the computational results were in good agreement with the experiment results. For an actual river-flooding disaster that occurred on the Omoto River in Iwate Prefecture, Japan, we conducted two large-scale computations of large areas with a total of 1000 pieces of driftwood and a maximum resolution of 7.8125 [cm]. The simulation results identified three driftwood accumulation processes: trapping in narrow channels, hitting building corners, and stagnation in front of buildings.
AB - Severe damage from river flooding carrying driftwood has become a critical problem. We have developed a code for 3-D simulations of water flows carrying a lot of driftwood by using the cumulant lattice Boltzmann method (LBM) coupled with the discrete element method (DEM). Adaptive mesh refinement (AMR) makes it possible to apply the code to large areas with a high-resolution mesh, and a GPU implementation greatly accelerates the computation. A small-scale water-tank experiment with a driftwood model showed that the computational results were in good agreement with the experiment results. For an actual river-flooding disaster that occurred on the Omoto River in Iwate Prefecture, Japan, we conducted two large-scale computations of large areas with a total of 1000 pieces of driftwood and a maximum resolution of 7.8125 [cm]. The simulation results identified three driftwood accumulation processes: trapping in narrow channels, hitting building corners, and stagnation in front of buildings.
KW - GPU computing
KW - adaptive mesh refinement
KW - cumulant lattice Boltzmann method
KW - discrete element method
KW - driftwood
KW - river flooding
UR - http://www.scopus.com/inward/record.url?scp=85214100093&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85214100093&partnerID=8YFLogxK
U2 - 10.2208/journalofjsce.24-00101
DO - 10.2208/journalofjsce.24-00101
M3 - Article
AN - SCOPUS:85214100093
SN - 2187-5103
VL - 12
JO - Journal of Japan Society of Civil Engineers
JF - Journal of Japan Society of Civil Engineers
IS - 1
M1 - 24-00101
ER -