FLOOD10391 (2) - Laserfiche WebLink

Home Browse Search

I I I I I I I I I I I I I I I I I I I AYRES ASSOCGAYES Finite Element vs. Finite Difference Models Ayres Associates has experience in several 2-dimensional models, including the Corps of Engineers RMA-2 model, the Federal Highway Administration FESWMS model and the proprietary FLO-2D model. Each of these models has specific strengths and weaknesses. Therefore, we select the most appropriate model based on the physical conditions of the project and the needs of the client. The first two models are finite element (FE) models and the third is a finite difference (FD) model. The two FE models are well suited for detailed hydraulic and hydrodynamic modeling but can be very challenging to apply to some conditions. RMA-2 is not recommended for this project because it cannot accurately simulate areas with near-critical and supercritical flow. Given the steep grade of the main channel, side channels and floodplain areas, near-critical flow is expected and RMA-2 would probably not maintain numerical stability. FESWMS can accurately simulate sub- and supercritical flow conditions and would be our recommendation if a steady state model would be required by the City of Boulder. We do not recommended FESWMS for this project because we expect significant numerical instability problems for dynamic flow simulation as areas of the model wet during the rising limb of the hydrograph and dry as the hydrograph recedes. Large areas of wetting and drying expected for South Boulder Creek would be a significant technical obstacle for either FESWMS or RMA-2. The two FE models are also less suitable when the main channel is very small, as is the case at South Boulder Creek. FLO-2D is the most appropriate model for this project. Although the finite difference (FD) approach has some limitations related to the level of detail that can be achieved, this model should be numerically stable and accurate for South Boulder Creek conditions. In FLO-2D, a uniform grid is required for the entire model, so some.small topographic variation could be obscured depending on the grid size selected. The FD approach in FLO-2D uses slightly simplified equations of motion, but is more numerically stable and, therefore, much better for this project. FLO-2D includes overland flow, channel flow, structures, levees, and embankments, and can include rainfall, infiltration, and sediment transport if necessary. FLO-2D includes good flow visualization, and Ayres Associates has experience with importing FLO-2D output into the Microstation CADD program for additional floodplain delineation and enhanced graphical output. Ayres Associates has extensive experience in post-processing 2- dimensional model output for additional analysis and flow visualization purposes. This experience will be applied to FLO-2D results by importing the results into the graphical environment used for finite element modeling. Therefore, this project will have the numerical advantages of finite difference modeling combined with the flow visualization advantages associated with finite element modeling. Topographic Mapping for Hydraulic Analysis We recommend that new topographic mapping be developed for hydraulic modeling and floodplain delineation. The new topographic mapping should be at least to a 2-ft contour interval and preferably to a 1-ft contour interval to very accurately identify split flow paths and other topographic features. Direct survey of the South Boulder Creek channel, ditches, other channels, roadways, and levees should also be conducted for model development. The first step in developing the FLO-2D model will be developing a HEC-RAS low-flow model of the South Boulder Creek channel. A FLO-2D model 22