Document Type

Dissertation

Degree

Doctor of Philosophy

Department

Civil Engineering

First Advisor's Name

Fernando Miralles-Wilhelm

First Advisor's Title

Committee Chair

Second Advisor's Name

Hector R. Fuentes

Second Advisor's Title

Committee Member

Third Advisor's Name

Luis Prieto-Portar

Third Advisor's Title

Committee Member

Fourth Advisor's Name

Francisco R. García-Martínez

Fourth Advisor's Title

Committee Member

Fifth Advisor's Name

George Dulikravich

Fifth Advisor's Title

Committee Member

Keywords

debris, flow, two-phase, FEM, DEM, Cross, Bingham, boulders

Date of Defense

11-13-2009

Abstract

The main objective of this work is to develop a quasi three-dimensional numerical model to simulate stony debris flows, considering a continuum fluid phase, composed by water and fine sediments, and a non-continuum phase including large particles, such as pebbles and boulders. Large particles are treated in a Lagrangian frame of reference using the Discrete Element Method, the fluid phase is based on the Eulerian approach, using the Finite Element Method to solve the depth-averaged Navier–Stokes equations in two horizontal dimensions. The particle’s equations of motion are in three dimensions. The model simulates particle-particle collisions and wall-particle collisions, taking into account that particles are immersed in a fluid. Bingham and Cross rheological models are used for the continuum phase. Both formulations provide very stable results, even in the range of very low shear rates. Bingham formulation is better able to simulate the stopping stage of the fluid when applied shear stresses are low. Results of numerical simulations have been compared with data from laboratory experiments on a flume-fan prototype. Results show that the model is capable of simulating the motion of big particles moving in the fluid flow, handling dense particulate flows and avoiding overlap among particles. An application to simulate debris flow events that occurred in Northern Venezuela in 1999 shows that the model could replicate the main boulder accumulation areas that were surveyed by the USGS. Uniqueness of this research is the integration of mud flow and stony debris movement in a single modeling tool that can be used for planning and management of debris flow prone areas.

Share

COinS