Intelligent meshing for two dimensional overland flow modelling; a comparison of two methodologies

Mulet-Marti, J. and Underhill, C. and Gutierrez Andres, J. and Woolhouse, G. (2009) Intelligent meshing for two dimensional overland flow modelling; a comparison of two methodologies. In: International Conference on Urban Flood Management, 26-27 November 2009, Paris.

PDF HRPP427_Intelligent_meshing_for_two_dimensional_overland_flow_modelling.pdf Available under License Creative Commons Attribution. Download (847kB)

Abstract

Two dimensional flow modelling is becoming widely used for urban pluvial flood modelling and is a very effective tool for assessing overland flood flow paths. Given the high resolution LiDAR ground elevation data now widely available, the accuracy of the hydraulic model is dependant largely upon the refinement of the model mesh or grid in representing the ground model topography. This paper considers methods for the accurate representation of overland flow paths in unstructured triangular meshed hydraulic models without the requirement to model the entire system at a high level of detail. The methods involve refinement of the 2D mesh in areas of complex topography and allow less refinement in less complex areas. This approach attempts to maximise the efficiency of the model in terms of representing flow paths while minimising computer run times. To perform the study presented in this paper two simple intelligent mesh approaches have been selected to be applied in a case study test in conjunction with the 2D hydraulic modelling platform InfoWorks Collection System (IWCS). The numerical results provided by the hydraulic modelling engine have been compared using different mesh resolutions in terms of flood extent and run time. The hydraulic results given by the intelligent mesh methods show an improvement in some cases when compared with their equivalent baseline scenarios. It must be noted that these results presented here are the outcome of limited sensitivity testing and represent an initial phase of research. Further work is ongoing to understand fully the variables which influence the performance of the methods.

Actions (for site administrators only - login required)

View Item