Development of a new 2D structured and unstructured mesh generator for flow, sediment, temperature, groundwater, and vegetation modeling

Project ID: 1724
Principal Investigator: Jianchun Huang
Research Topic: Sediment Management and River Restoration
Funded Fiscal Years: 2017, 2018 and 2019
Keywords: None

Research Question

How to develop mixed triangular and quadrilateral meshes for two dimensional (2D) flow, sediment, temperature, groundwater, and vegetation numerical simulation?

How to dynamically link the mesh generator with a 2D model to simulate the bank erosion and channel migration (i.e., movable mesh)?

How to facilitate the model setup for 2D simulation?

Need and Benefit

Existing mesh generation software limits our capability to advance our numerical model developments, which are widely used on flow, sediment, temperature, and vegetation simulations. Current mesh generation software is costly and can't be integrated dynamically into a 2D model, which requires a moving mesh during model simulation. Army Corp of Engineer's HEC-RAS 5.0 model provides a Cartesian mesh and local triangle mesh. USGS's iRIC provides only triangular mesh. Neither can be used to provide a flexible mesh with mixed triangle and quadrilateral cells, and also can't be dynamically linked to our SRH-2D model to create a dynamical mesh that can be adjusted during numerical simulation.

The proposed research will provide software for structured and unstructured mesh generation that will be used in future 2D numerical simulations of flow, sediment, temperature, and vegetation. Numerical models require a top quality mesh generator and associated interface.

The proposed research will take into account of the channel alignment and provide a more accurate channel topography interpolation. The algorithm adopted could potentially reduce the cost of river and reservoir survey and still maintain a high quality of numerical simulation. Currently, this algorithm has been developed in S&T researches at the Sedimentation and River Hydraulics Group, but is impossible to implement in the current commercial software.

Mesh generation requires a large amount of user effort and time in 2D numerical model simulations. An in-house mesh generator will fit the need to reduce the cost of numerical model simulations, and to provide an interface to set up the boundary conditions and other input parameters. For example, a library of pre-designed large wood structures can be added into the model, and to add large wood structures into the channel for different alternative could be a few mouse button clicks. Furthermore, it will no longer be necessary to purchase mesh generation software that is currently employed for SRH-2D modeling, but lacks many functions for river applications.

The research will also provide an algorithm to advance Reclamation-owned 2D numerical models to adjust the mesh during the numerical simulation. For example, a 2D bank erosion and migration model requires fine moving meshes near the bank. During the numerical simulation, the moving mesh with finer mesh near the banks could be generated to better represent the large flow gradients near the bank. The current numerical model with pre-generated mesh poses instability problems near the bank and lacks accuracy near the bank during bank erosion and channel migration.

Contributing Partners

Contact the Principal Investigator for information about partners.

Research Products

Contact the Principal Investigator for information about these documents.


Return to Research Projects

Last Updated: 4/4/17