More Like this

1. Three-phase flow simulation of local scour around a submerged horizontal cylinder

   Huang, Shijie; Huang, Zhenhua (October 2020, Proceedings of the Thirtieth (2020) International Ocean and Polar Engineering Conference)

   null (Ed.)

   Wave-induced scour plays a key role in the stability analysis of coastal structures, submarine pipelines or cables. There is a rich literature in current-induced scour, but more research is needed to understand the characteristics of wave-induced scour and the mechanisms that are important to the scour process. Sediment transport and flow-induced scour are three-phase (air-water-sediment) flow problems in nature and multi-phase flow simulation is a useful tools that can provide information difficult to obtain from physical tests. Most existing numerical models developed for simulating local scours are based on one-way coupling, which neglects effects of sediment phase on hydrodynamics of the flow. The present study uses a three-phase (air, water and sediment) flow model, which allows for a two-way coupling, to simulate wave-induced local scour problems. The three-phase flow model captures the air-water interface using a modified VOF method, and uses an improved rheology for the sediment phase for better results. The model is validated and verified using one set of existing experiment results for local scour around a submerged horizontal pipe. The detailed flow fields of both the sediment phase and the water phase around the scour are analyzed to understand the scour process. All three-phase flow simulations flow simulations on XSEDE’s Stampede2 supercomputers. The applicability of the model to other local scour problems is also discussed. 

   more » « less
2. An Eulerian two-phase model investigation on wave-induced scour around a vertical circular cylinder

   Benjamin Tsai, Antoine Mathieu (September 2023, International Conference on Scour and Erosion)

   A three-dimensional Eulerian two-phase flow solver, SedFoam, has been developed for various sediment transport applications. The solver has demonstrated success in modeling sheet flow and bedforms driven by oscillatory flows using a Reynolds-averaged Navier–Stokes (RANS) formulation. However, the accuracy of the RANS formulation for more complex flows, such as scour around structures, requires further evaluation. SedFoam has recently been enhanced to incorporate two-phase large-eddy simulation (LES) capability. In this study, RANS and LES approaches are tested via a three-dimensional case of wave-induced local scour around a single vertical circular pile. Two laboratory experiments, one with an erodible bed and the other with a rigid bed, were chosen for simulation, with both experiments having a Keulegan-Carpenter (KC) number of 10. The k-ω turbulence closure was selected for the RANS simulation, and the dynamic Lagrangian subgrid closure was chosen for the LES simulation. Numerical results reveal that both RANS and LES simulations can resolve lee-wake vortices, although the vortices are significantly weaker in the RANS simulation. In comparison with the LES results, the RANS approach fails to predict horseshoe vortex with sufficient intensity, leading to an underestimation of scour hole depth development. Although the scour depths develop at a very similar rate in the early stage, the scour depth predicted by the RANS simulation quickly reaches equilibrium, while the LES simulation follows the measured trend. These findings indicate that a turbulence-resolving methodology, i.e. LES, is necessary for accurate scour simulations. 

   more » « less
3. On the need for physical constraints in deep learning rainfall–runoff projections under climate change: a sensitivity analysis to warming and shifts in potential evapotranspiration

   https://doi.org/10.5194/hess-28-479-2024  

   Wi, Sungwook; Steinschneider, Scott (January 2024, Hydrology and Earth System Sciences)

   Abstract. Deep learning (DL) rainfall–runoff models outperform conceptual, process-based models in a range of applications. However, it remains unclear whether DL models can produce physically plausible projections of streamflow under climate change. We investigate this question through a sensitivity analysis of modeled responses to increases in temperature and potential evapotranspiration (PET), with other meteorological variables left unchanged. Previous research has shown that temperature-based PET methods overestimate evaporative water loss under warming compared with energy budget-based PET methods. We therefore assume that reliable streamflow responses to warming should exhibit less evaporative water loss when forced with smaller, energy-budget-based PET compared with temperature-based PET. We conduct this assessment using three conceptual, process-based rainfall–runoff models and three DL models, trained and tested across 212 watersheds in the Great Lakes basin. The DL models include a Long Short-Term Memory network (LSTM), a mass-conserving LSTM (MC-LSTM), and a novel variant of the MC-LSTM that also respects the relationship between PET and evaporative water loss (MC-LSTM-PET). After validating models against historical streamflow and actual evapotranspiration, we force all models with scenarios of warming, historical precipitation, and both temperature-based (Hamon) and energy-budget-based (Priestley–Taylor) PET, and compare their responses in long-term mean daily flow, low flows, high flows, and seasonal streamflow timing. We also explore similar responses using a national LSTM fit to 531 watersheds across the United States to assess how the inclusion of a larger and more diverse set of basins influences signals of hydrological response under warming. The main results of this study are as follows: The three Great Lakes DL models substantially outperform all process-based models in streamflow estimation. The MC-LSTM-PET also matches the best process-based models and outperforms the MC-LSTM in estimating actual evapotranspiration. All process-based models show a downward shift in long-term mean daily flows under warming, but median shifts are considerably larger under temperature-based PET (−17 % to −25 %) than energy-budget-based PET (−6 % to −9 %). The MC-LSTM-PET model exhibits similar differences in water loss across the different PET forcings. Conversely, the LSTM exhibits unrealistically large water losses under warming using Priestley–Taylor PET (−20 %), while the MC-LSTM is relatively insensitive to the PET method. DL models exhibit smaller changes in high flows and seasonal timing of flows as compared with the process-based models, while DL estimates of low flows are within the range estimated by the process-based models. Like the Great Lakes LSTM, the national LSTM also shows unrealistically large water losses under warming (−25 %), but it is more stable when many inputs are changed under warming and better aligns with process-based model responses for seasonal timing of flows. Ultimately, the results of this sensitivity analysis suggest that physical considerations regarding model architecture and input variables may be necessary to promote the physical realism of deep-learning-based hydrological projections under climate change. 

   more » « less
4. Experimental study of bed evolution around a non-slender square structure under combined solitary wave and steady current actions

   https://doi.org/10.1016/j.oceaneng.2022.112792  

   Sogut, E. (December 2022, Ocean engineering)

   This paper presents the findings from the laboratory wave flume experiments designed to investigate the formation and evolution of scour around a non-slender, square vertical structure, under three flow conditions, solitary wave, combined solitary wave and steady following current, and combined solitary wave and steady opposing current. The structure was placed on a sandy berm, either fastened to the flume wall or positioned at the centerline of the flume. For the wave only case, the scour on the seaside edge turned out to be deeper than the one on the leeside regardless of the structure’s position. The analyses showed that the depth, width, volume, and location of the scour were all significantly influenced by the introduction of steady currents. The following current, for example, deepened the seaside scour, while leading to shallower leeside scour holes as a result of the backfilling process. Contrary to the opposing current, which shifted the scour area in the upwave direction, the scour was transported downwave under the effect of the following current. The scour depth was determined to be a function of the structure position and the Keulegan-Carpenter number, whereas the scour width mostly depended on the structure’s position. In this regard, the structure fastened to the wall experienced the widest scour area and the largest volume regardless of the flow condition. 

   more » « less
5. The Performance of LSTM and BiLSTM in Forecasting Time Series

   https://doi.org/10.1109/BigData47090.2019.9005997  

   Siami-Namini, Sima; Tavakoli, Neda; Namin, Akbar Siami (December 2019, 2019 IEEE International Conference on Big Data (Big Data))

   Machine and deep learning-based algorithms are the emerging approaches in addressing prediction problems in time series. These techniques have been shown to produce more accurate results than conventional regression-based modeling. It has been reported that artificial Recurrent Neural Networks (RNN) with memory, such as Long Short-Term Memory (LSTM), are superior compared to Autoregressive Integrated Moving Average (ARIMA) with a large margin. The LSTM-based models incorporate additional “gates” for the purpose of memorizing longer sequences of input data. The major question is that whether the gates incorporated in the LSTM architecture already offers a good prediction and whether additional training of data would be necessary to further improve the prediction. Bidirectional LSTMs (BiLSTMs) enable additional training by traversing the input data twice (i.e., 1) left-to-right, and 2) right-to-left). The research question of interest is then whether BiLSTM, with additional training capability, outperforms regular unidirectional LSTM. This paper reports a behavioral analysis and comparison of BiLSTM and LSTM models. The objective is to explore to what extend additional layers of training of data would be beneficial to tune the involved parameters. The results show that additional training of data and thus BiLSTM-based modeling offers better predictions than regular LSTM-based models. More specifically, it was observed that BiLSTM models provide better predictions compared to ARIMA and LSTM models. It was also observed that BiLSTM models reach the equilibrium much slower than LSTM-based models. 

   more » « less