Date of Award

2022

Publication Type

Thesis

Degree Name

M.A.Sc.

Department

Civil and Environmental Engineering

Keywords

Artificial neural networks, Bridge pier scour, Scour width

Supervisor

R.Barron

Supervisor

R.Balachandar

Rights

info:eu-repo/semantics/openAccess

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Abstract

Accurate equilibrium scour depth and width estimations are essential to both safe and economic designs of bridge foundations. Review of current scour estimation methods demonstrate that the empirical equations produce scour values that are overestimated, resulting in uneconomical designs. In the current investigation, artificial neural networks (ANNs) were optimized and applied to scour data under laboratory conditions, field conditions, and a combination of the two conditions. Additionally, physics-based parameters – in place of empirical parameters (e.g., shape factors) – and parameters incorporating blockage effects were introduced as input parameters to the ANNs in an attempt to improve scour predictions. Finally, ANNs were applied to scour width estimations to investigate the applicability of machine learning tools to scour width prediction. For each of the ANNs developed, a sensitivity analysis was conducted to ensure each of the input parameters selected had significant value on the prediction models. Sensitivity analyses also allow for a further understanding of each of the parameters’ influence on the models.

Share

COinS