11:00 AM - 1:00 PM
[HDS10-P01] Optimization of wave-dissipating structure arrangement for land run-up tsunami using finite element analysis
Keywords:Tsunami, Wave-dissipating structure, Optimal arrangement, Froude number, Finite element analysis
Research on the evaluation of wave force acting on structures by land run-up tsunamis has been widely conducted via experiments and numerical simulations, and the relationship between water depth coefficient and Froude number has been also investigated to precisely evaluate the wave force (e.g., Asakura et al., 2000; Ikeya et al., 2013; Kihara, 2015; Sakakiyama, 2012). In addition to the studies that directly evaluate the tsunami wave forces acting on target structures, there have been active studies on the effect of the wave-dissipating structures placed in front of the target structure reducing the wave force level impacting the target structure (Ueno et al., 2020; Kanda et al., 2016). For instance, Kanda et al. confirmed the effect of wave force reduction by measuring the wave pressure on the back wall of the wave-dissipating structure consisting of four square or circular columns positioned in a single row. Although a more efficient arrangement of square (or circular) columns could be achieved, such a study has not been conducted sufficiently.
In this study, we investigate the effect of the arrangement of the square columns on the reduction of the tsunami wave force using the finite element analysis based on the shallow water equation implemented on COMSOL Multiphysics. First, we simulate the dam-breaking experiment (Gómez-Gesteira and Dalrymple, 2004) to validate the simulator by comparing the wave force obtained from the simulation and from the experiment. Next, we investigate the relationship between the water depth coefficient and the Froude number for different shapes of wave-dissipating structures (square, circular, and elliptical columns) and quantitatively compare the results with related studies (Asakura et al., 2002; Kihara et al., 2012, 2015; Sakakiyama, 2010, 2012; Ikeya et al., 2013). Through these investigations, we have confirmed that the developed numerical simulator has an accuracy reaching a certain level. Within the range of the numerical results obtained in this study, there was good agreement with the relationship between the water depth coefficient and the Froude number theoretically derived by Ikeya et al.
After confirming the simulation accuracy, three square columns are set up as wave-dissipating structures and are subjected to a water flow with various velocities (Froude numbers ranging from 2.0 to 4.5). Note that optimization analysis based on the BOBYQA algorithm (Powell, 2009) is also performed for each Froude number case using the position coordinates of the square column as design variables. As a result, the optimized shape for a Froude number of 2 or 2.5 is detected as an isosceles triangle with a spacing of about 10.0D perpendicular to the downstream direction (D is the representative length of a square) and about 5.5D in the downstream direction. We also found that an isosceles triangle of about 6.5D perpendicular to the downstream direction and 5.0D in the downstream direction is optimal for the Froude number case of 4.5. A comparison of the efficiency with that of a single row arrangement of square columns shows that the efficiency is 1.1 times higher when the Froude number is less than 2.5, while the efficiency is 1.2 times higher when the Froude number is greater than 3.0.
References
Asakura et al., 2002. Coastal Engineering 2002. WORLD SCIENTIFIC, pp. 1191–1202.
Asakura et al., 2000. Proceedings of Coastal Engineering, JSCE, , 47, pp.911-915. (in Japanese)
Gómez-Gesteira M., Dalrymple Robert A., 2004. J. Waterway Port Coast. Ocean Eng. 130, 63–69.
Ikeya et al., 2013. J.JSCE, Ser.B2, Coastal engineering, 69, 2, pp. I_816-I_820. (in Japanese with English abstract)
Kanda et al., 2016. J.JSCE, Ser.B2, Coastal engineering, 72, 2, pp. I_1069-I_1074. (in Japanese with English abstract)
Kihara et al., 2015. CRIEPI Report, O15002. (in Japanese)
Kihara et al., 2012. CRIEPI Report, N12010. (in Japanese)
Powell, M.J.D., 2009. NA Report NA2009/06, University of Cambridge, Cambridge 26–46.
Sakakiyama, 2012. J.JSCE, Ser.B2, Coastal engineering, 68, I_771–I_775. (in Japanese with English abstract)
Sakakiyama, 2010. Proceedings of civil engineering in the ocean, 26, 285–290. (in Japanese)
Ueno et al., 2020. Numerical simulation for reduction effects on tsunami inundation behind porous vertical barrier. J.JSCE, Ser.B2, Coastal engineering, 76, 2, pp. I_265-I_270. (in Japanese with English abstract)
In this study, we investigate the effect of the arrangement of the square columns on the reduction of the tsunami wave force using the finite element analysis based on the shallow water equation implemented on COMSOL Multiphysics. First, we simulate the dam-breaking experiment (Gómez-Gesteira and Dalrymple, 2004) to validate the simulator by comparing the wave force obtained from the simulation and from the experiment. Next, we investigate the relationship between the water depth coefficient and the Froude number for different shapes of wave-dissipating structures (square, circular, and elliptical columns) and quantitatively compare the results with related studies (Asakura et al., 2002; Kihara et al., 2012, 2015; Sakakiyama, 2010, 2012; Ikeya et al., 2013). Through these investigations, we have confirmed that the developed numerical simulator has an accuracy reaching a certain level. Within the range of the numerical results obtained in this study, there was good agreement with the relationship between the water depth coefficient and the Froude number theoretically derived by Ikeya et al.
After confirming the simulation accuracy, three square columns are set up as wave-dissipating structures and are subjected to a water flow with various velocities (Froude numbers ranging from 2.0 to 4.5). Note that optimization analysis based on the BOBYQA algorithm (Powell, 2009) is also performed for each Froude number case using the position coordinates of the square column as design variables. As a result, the optimized shape for a Froude number of 2 or 2.5 is detected as an isosceles triangle with a spacing of about 10.0D perpendicular to the downstream direction (D is the representative length of a square) and about 5.5D in the downstream direction. We also found that an isosceles triangle of about 6.5D perpendicular to the downstream direction and 5.0D in the downstream direction is optimal for the Froude number case of 4.5. A comparison of the efficiency with that of a single row arrangement of square columns shows that the efficiency is 1.1 times higher when the Froude number is less than 2.5, while the efficiency is 1.2 times higher when the Froude number is greater than 3.0.
References
Asakura et al., 2002. Coastal Engineering 2002. WORLD SCIENTIFIC, pp. 1191–1202.
Asakura et al., 2000. Proceedings of Coastal Engineering, JSCE, , 47, pp.911-915. (in Japanese)
Gómez-Gesteira M., Dalrymple Robert A., 2004. J. Waterway Port Coast. Ocean Eng. 130, 63–69.
Ikeya et al., 2013. J.JSCE, Ser.B2, Coastal engineering, 69, 2, pp. I_816-I_820. (in Japanese with English abstract)
Kanda et al., 2016. J.JSCE, Ser.B2, Coastal engineering, 72, 2, pp. I_1069-I_1074. (in Japanese with English abstract)
Kihara et al., 2015. CRIEPI Report, O15002. (in Japanese)
Kihara et al., 2012. CRIEPI Report, N12010. (in Japanese)
Powell, M.J.D., 2009. NA Report NA2009/06, University of Cambridge, Cambridge 26–46.
Sakakiyama, 2012. J.JSCE, Ser.B2, Coastal engineering, 68, I_771–I_775. (in Japanese with English abstract)
Sakakiyama, 2010. Proceedings of civil engineering in the ocean, 26, 285–290. (in Japanese)
Ueno et al., 2020. Numerical simulation for reduction effects on tsunami inundation behind porous vertical barrier. J.JSCE, Ser.B2, Coastal engineering, 76, 2, pp. I_265-I_270. (in Japanese with English abstract)