- Volumes 84-95 (2024)
-
Volumes 72-83 (2023)
-
Volume 83
Pages 1-258 (December 2023)
-
Volume 82
Pages 1-204 (November 2023)
-
Volume 81
Pages 1-188 (October 2023)
-
Volume 80
Pages 1-202 (September 2023)
-
Volume 79
Pages 1-172 (August 2023)
-
Volume 78
Pages 1-146 (July 2023)
-
Volume 77
Pages 1-152 (June 2023)
-
Volume 76
Pages 1-176 (May 2023)
-
Volume 75
Pages 1-228 (April 2023)
-
Volume 74
Pages 1-200 (March 2023)
-
Volume 73
Pages 1-138 (February 2023)
-
Volume 72
Pages 1-144 (January 2023)
-
Volume 83
-
Volumes 60-71 (2022)
-
Volume 71
Pages 1-108 (December 2022)
-
Volume 70
Pages 1-106 (November 2022)
-
Volume 69
Pages 1-122 (October 2022)
-
Volume 68
Pages 1-124 (September 2022)
-
Volume 67
Pages 1-102 (August 2022)
-
Volume 66
Pages 1-112 (July 2022)
-
Volume 65
Pages 1-138 (June 2022)
-
Volume 64
Pages 1-186 (May 2022)
-
Volume 63
Pages 1-124 (April 2022)
-
Volume 62
Pages 1-104 (March 2022)
-
Volume 61
Pages 1-120 (February 2022)
-
Volume 60
Pages 1-124 (January 2022)
-
Volume 71
- Volumes 54-59 (2021)
- Volumes 48-53 (2020)
- Volumes 42-47 (2019)
- Volumes 36-41 (2018)
- Volumes 30-35 (2017)
- Volumes 24-29 (2016)
- Volumes 18-23 (2015)
- Volumes 12-17 (2014)
- Volume 11 (2013)
- Volume 10 (2012)
- Volume 9 (2011)
- Volume 8 (2010)
- Volume 7 (2009)
- Volume 6 (2008)
- Volume 5 (2007)
- Volume 4 (2006)
- Volume 3 (2005)
- Volume 2 (2004)
- Volume 1 (2003)
In the present study, five different invert trap configurations (rectangular with and without lids on both sides; trapezoidal, trapezoidal with rectangular base and rectangular with trapezoidal base with lids on both sides) were simulated for both two-dimensional (2D) and three-dimensional (3D) flow conditions for three sediment types (sand, styrocell and plastic beads) at six flow rates (0.35, 0.70, 1.05, 1.35, 4.55 and 9.95 L/s) for each trap. Computational fluid dynamics (CFD)-based modeling using FLUENT software with Renormalization Group (RNG) k–ɛ model along with discrete phase model (DPM) were used in the simulations. A hexagonal/tetrahedral and map-type non-uniform grid was chosen to discretize the entire computational domain and a control volume finite difference method was used to solve the governing equations. The flow rates selected in the present study cover the entire range of flow rate expected for dry weather and monsoon. The simulation is capable of differentiating between 2D and 3D modeling of particle trajectories, the effects of flow rate and trap geometry on flow patterns developed in the trap. The sediment retention ratio for 2D is higher than that for 3D modeling for all flow conditions, particle types and model geometry due to inclusion of lateral effects in 3D modeling. The invert trap having rectangular shape with trapezoidal base is found to be the most efficient configuration in both 2D and 3D modeling.