Author(s): Tommaso Attili; Richard J. S. Whitehouse; Nick Tavouktsoglou; Jingjing Yan

Linked Author(s):

Keywords: NN; Burial; Modelling; Scour; UXO

Abstract: UneXploded Ordnances (UXOs) and Discarded Military Munitions (DMMs) frequently appear in coastal and offshore regions, representing a threat for maritime engineering works and for the public. Hydrodynamic and morphodynamic forcings can cause these objects to bury and/or mobilise, making their detection challenging. Hence there is a need for reliable approaches predicting burial and mobilisation of UXOs to support the risk assessment of contaminated sites. The present work proposes two models to predict the time evolution scour burial of isolated objects on granular soils. The first model is referred to as DRAMBUIE 3.0. DRAMBUIE 3.0 couples recent empirical equations for the scour burial equilibrium depth with a time stepping approach. A new Artificial Neural Network (ANN) predictor for equilibrium burial depth was also developed and coupled with a time evolution model, making DRAMBUIE-ANN. Both models have been compared with small-scale and large-scale experiments, and with field measurements. Despite some discrepancies between the observations and predictions, both models showed the capability to predict UXO burial for a range of objects and hydrodynamic conditions. Future work should focus on further validating the models and extending their range of applicability. DRAMBUIE 3.0 and DRAMBUIE-ANN can be used to support site manager decisions for the remediation of UXO and DMM-contaminated sites.

DOI: https://doi.org/10.64697/978-90-835589-7-4_41WC-P1699-cd

Year: 2025