Probabilistic Investigation of Debris Impact Forces during Extreme Hydraulic Events

Forensic engineering surveys of extreme hydraulic events, such as the 2011 Tohoku Tsunami and the 2012 Hurricane Katrina, demonstrated the importance of considering debris in the estimation of loading on structures. Building standards have historically addressed debris loading through a deterministic approach. However, due to challenges in estimating debris loading in engineering surveys (Yeh et al. 2014) and the inherent stochastic nature of debris motion (Matsutomi 2009), the probabilistic nature of debris loading is not captured within these models. This study uses experimental data to propose and validate a probabilistic model to accurately capture the motion of debris and debris impact forces. As provisions on debris loading for building design standards continue to move towards a probabilistic approach, this model provides a framework for estimating debris loading within that paradigm.

The experimental program used in the validation of the proposed model were performed in the Dam-Break Flume at the University of Ottawa (Canada). The dam-break wave was used to model a tsunami-like wave propagating over a flat, horizontal topography. A single debris, modelled as a 1:40 geometrically scaled shipping container, was placed on the bed of the flume, where it was entrained by the wave and propagated downstream. A camera-based image processing technique was used to track and analyze the motion of the debris. A square column was placed downstream of the debris site, mounted onto a load cell, which recorded the debris impact forces.

The probabilistic model is based on the approach by Lin and Vanmarcke (2010) who estimated debris hazard under extreme wind conditions. The probabilistic model shows an accurate representation of the physical processes involved in debris hazard assessment. The motion of the debris is well-captured by the Gaussian distribution and the impact forces are accurately estimated using the Beta Distribution. Further investigation will be needed to address scaling effects, particularly related to the lateral motion of the debris, as this can be influenced by local turbulence (Rueben et al. 2014).

Due to space limitations, the references were not provided here.