Hydraulics & Waterways

Oral

394968 - Is a linear model suitable for predicting cohesive soil detachment?

Wednesday, June 6
4:00 PM - 2:00 AM
Location: Greenway CD
Co-Authors: Katherine Wardinski, Platteville, WI – University of Wisconsin Platteville; Garey Fox, Raleigh, NC – North Carolina State University; Celso Castro Bolinaga, Raleigh, NC – North Carolina State University

Cohesive soil detachment is complex, and simplified mathematical models are generally used to simulate erosion of cohesive material for engineering applications such as prediction of fluvial or coastal morphodynamics or erosion of dam embankments and earth spillways. The linear excess shear stress model assumes a linear relationship between the erosion rate and the applied shear stress, and the existence of a critical shear stress. The linearity of the relationship between erosion rate and shear stress has repeatedly been questioned; several non-linear models have been proposed as alternatives to the linear model. Recent studies demonstrated that over the typical range of shear stress tested in situ (from 5 to 15-20 Pa), non-linear models generally matched observations as well as the linear model. Because of the lack of data, the accuracy of soil detachment models in predicting erosion over wider ranges of shear stress have not been fully explored. The objectives of this study were to generate erosion data outside of the typical range of shear stress tested in situ and evaluate the suitability of the linear detachment model for describing erosion rates over large ranges. Laboratory mini-JETs were performed on remolded samples with sand and clay contents of tested materials ranging from 24.6 to 71.5% and 12.5 to 44%, respectively. Statistically significant breakpoints were detected for most of the trials with a segmented regression analysis showing that a unique linear model was not suitable. The large scatter in data typically observed around breakpoints in linear regressions within a dataset suggested that changes in the erosion trend were quite progressive. The shear stress values at breakpoints were of the same order as the typical range of shear stress tested in situ, and the extrapolation potential of linear models fit to erosion data obtained from standard in-situ JET procedure should therefore be questioned.

Lucie Guertault, NC

Postdoctoral Research Associate
North Carolina State University

Presentation(s):

Send Email for Lucie Guertault


Assets

394968 - Is a linear model suitable for predicting cohesive soil detachment?



Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Is a linear model suitable for predicting cohesive soil detachment?