Review Reports - CFD Study on the Influence of Oblique Underflow Baffles on Bedload Transport in Rectangular Channels

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The initial investigations on underflow baffles presented in this paper were carried out only as an initial CFD investigation with the aim of assessing the feasibility and potential of this research, as CFD provides a cost-effective approach for exploring these processes prior to investing them using the more expensive physical model studies.

(1) The abstract fails to compare the technical advantages of oblique underflow baffles with those of traditional bedload control structures, making it impossible to highlight the innovative entry point and engineering application value of this study.

(2) The introduction only generally mentions the hazards of bedload deposition at water intake structures, without quantifying the defects of existing technologies or clarifying the core technical pain points to be addressed in this study, resulting in insufficient pertinence of the research objectives.

(3) In the model setup section, only the baffle submergence depth is fixed at 20% of the water depth, and the influences of different submergence depths (10%, 30%), baffle thicknesses, and baffle material roughness on water flow vortices and bedload transport are not investigated, leading to insufficient integrity of the research content.

(4) In the turbulence model section, only a single turbulence model is adopted, without comparing the simulation accuracy of models such as large eddy simulation for vortex structures. Some latest research work related with this topic can be referred. Diffusion evolution rules of grouting slurry in mining-induced cracks in overlying strata. Study on the energy evolution mechanism and fractal characteristics of coal failure under dynamic loading.

(5) In the discussion section, only CFD numerical simulations are carried out, and the numerical results are not verified by physical flume experiments, so the numerical conclusions lack the support of measured data.

Author Response

Comment 1: The abstract fails to compare the technical advantages of oblique underflow baffles with those of traditional bedload control structures, making it impossible to highlight the innovative entry point and engineering application value of this study.

Response 1: The authors are thankful for the remarks. To better highlight the innovative entry point and engineering application of this study, some minor improvements were made in the abstract, and the following sentence was added during the revision “As discharge increases, the bedload free zone expands, resulting in greater effectiveness at higher flows; an effect not observed with conventional near-bed bedload control structures.”

Comment 2: The introduction only generally mentions the hazards of bedload deposition at water intake structures, without quantifying the defects of existing technologies or clarifying the core technical pain points to be addressed in this study, resulting in insufficient pertinence of the research objectives.

Response 2: Thank you for the comment. The second last paragraph in the introduction section is revised accordingly, while highlighting one major issue with the conventional bedload deposition management techniques, the local scouring, since they are embedded. Additionally, the potential and advantages of the underflow baffle are highlighted.

Comment 3: In the model setup section, only the baffle submergence depth is fixed at 20% of the water depth, and the influences of different submergence depths (10%, 30%), baffle thicknesses, and baffle material roughness on water flow vortices and bedload transport are not investigated, leading to insufficient integrity of the research content.

Response 3: Thank you so much for highlighting this important point. We plan to investigate the above-listed and some more parameters in future research. In order to elaborate this, we added the following sentence in the model set up part: “As this paper presents the first investigation on the influence of oblique underflow baffles on bedload transport, the primary objective was to explain the underlying mechanism responsible for the observed effect. Other geometric and material parame-ters, such as submergence ratio, baffle angle, baffle thickness, and sediment properties, were not examined here but will be addressed in subsequent research.”

Comment 4: In the turbulence model section, only a single turbulence model is adopted, without comparing the simulation accuracy of models such as large eddy simulation for vortex structures. Some latest research work related with this topic can be referred. Diffusion evolution rules of grouting slurry in mining-induced cracks in overlying strata. Study on the energy evolution mechanism and fractal characteristics of coal failure under dynamic loading.

Response 4: We have cited some well-verified work where RNG k-ε turbulence model was used to simulate bedload transport and scour around hydraulic structures and therefore presume that our results won’t differ much from laboratory results. Furthermore, we added the following text in chapter 2. Study Methodology during the revision: “As reported in these referenced studies, the choice of turbulence model had only a minor influence on bedload results, primarily affecting the magnitude of transport and deposition rather than the overall bedload deposition pattern. Since this study represents the first investigation into the influence of oblique underflow baffles on bedload transport, the focus was placed on identifying general bedload deposition patterns; therefore, alternative turbulence models were not examined.”

We would also like to thank you for the suggested papers (Diffusion evolution rules of grouting slurry in mining-induced cracks in overlying strata. Study on the energy evolution mechanism and fractal characteristics of coal failure under dynamic loading.). After reviewing these studies, we felt that although they are very interesting, they are not related to our study.

Comment 5: In the discussion section, only CFD numerical simulations are carried out, and the numerical results are not verified by physical flume experiments, so the numerical conclusions lack the support of measured data.

Response 5: We appreciate your comment. We added the following sentence in the conclusion part: “The research presented in this paper constitutes an initial investigation into the influence of oblique underflow baffles on bedload transport and deposition patterns, conducted exclusively through CFD simulations. These simulations were performed to evaluate the feasibility of using oblique baffles to control bedload transport and to identify the underlying hydrodynamic mechanisms before undertaking more costly physical model experiments.” Also, this limitation is discussed in section 5.3 Summary of findings and limitations.

Reviewer 2 Report

Comments and Suggestions for Authors

The study uses 3D CFD simulations to analyze how oblique vertical underflow baffles influence bedload transport in a rectangular open channel. By varying flow discharge, baffle angle, and channel width coverage, the authors show that oblique baffles generate a downstream vortex that creates a bedload-free zone on one side of the channel an effect not seen with orthogonal baffles. Before publication, following comments should be addressed:

The study relies entirely on CFD simulations without any physical validation. Given that bedload transport and vortex-driven sediment patterns are highly sensitive to turbulence modeling and mesh quality, the absence of even preliminary experimental comparison significantly limits the credibility of the conclusions. How do you convince readers for validity of current results? Sentence “Physical model experiments are planned to be underway soon..” is not acceptable.
Why did you use RNG k–ε RANS model compared to other models. Talk more about expected RANS limitations and other models like LES, DNS etc. Following article can be helpful: https://doi.org/10.1080/10407782.2024.2368277
Although limitations are briefly acknowledged, important issues are not addressed such as the accuracy of the Nielsen bedload model in non-coastal, vortex-dominated flows, or the possible numerical diffusion in FLOW-3D’s VOF method affecting free-surface dynamics behind the baffle.
The manuscript lacks an evaluation of y⁺ values along the channel bed and sidewalls. Please add it. The suggested article can be helpful.
Only one baffle submergence level (20% of flow depth) and a handful of angles and discharges are tested. The claim that the method is broadly effective is not supported by the narrow parameter range, especially given that submergence is a key parameter controlling both underflow and vortex formation. If possible please improve it.
Sediment is introduced as a point-source at 0.05 m above the bed. This non-physical method may distort the natural hydrodynamics of sediment entrainment and deposition, especially in a uniform channel. No rationale is provided for using a point source instead of prescribing an equilibrium or uniform upstream bedload transport rate. Please elaborate about it.

Author Response

Comment 1: The study relies entirely on CFD simulations without any physical validation. Given that bedload transport and vortex-driven sediment patterns are highly sensitive to turbulence modeling and mesh quality, the absence of even preliminary experimental comparison significantly limits the credibility of the conclusions. How do you convince readers for validity of current results? Sentence “Physical model experiments are planned to be underway soon..” is not acceptable.

Response 1: Thank you for your comment. Since we don’t have the validation of the results with the physical model yet, we added chapter 5.2 Numerical Stability Control and Flow Development to ensure the reader that the vortex investigated in this research is stable throughout the whole simulation and appears for a steady state and for fully developed flow. We also added/changed the conclusion part as follows: “The research presented in this paper constitutes an initial investigation into the influence of oblique underflow baffles on bedload transport and deposition patterns, conducted exclusively through CFD simulations. These simulations were performed to evaluate the feasibility of using oblique baffles to control bedload transport and to identify the underlying hydrodynamic mechanisms before undertaking more costly physical model experiments. Future work will focus on validating the CFD results with physical model experiments before expanding the current numerical study for numerous channel conditions and applications.”

Comment 2: Why did you use RNG k–ε RANS model compared to other models. Talk more about expected RANS limitations and other models like LES, DNS etc. Following article can be helpful: https://doi.org/10.1080/10407782.2024.2368277

Response 2: Thank you for highlighting this point, which is also related to the comment 4 of reviewer 1. We have referenced some well-verified work where RNG was used to simulate bedload transport and scour around hydraulic structures and therefore presume that our results won’t differ much from laboratory results. We added the following text in chapter 2. Study Methodology: “As reported in these referenced studies, the choice of turbulence model had only a minor influence on bedload results, primarily affecting the magnitude of transport and deposition rather than the overall bedload deposition pattern. Since this research rep-resents the first investigation into the influence of oblique underflow baffles on bedload transport, the focus was placed on identifying general bedload deposition patterns; therefore, alternative turbulence models were not examined.”

However, we acknowledged the limitations of RANS models and wrote in the limitation part: “While RANS models are widely used and computationally efficient, they can have limitations in accurately resolving complex flow structures and transient bedload dynamics, particularly in the vortex-dominated flows. Consequently, the precision of the predicted bedload transport patterns should be interpreted with caution.”. Additionally, LES is going to be computationally very demanding and will not serve the purpose of the current initial numerical study on the obliqued underflow baffle. We believe DNS is still beyond our reach considering our hydrodynamic conditions.

We would also like to thank you for the suggested paper (https://doi.org/10.1080/10407782.2024.2368277). After reviewing the article, we felt that although it is a very interesting study, it is neither related to our study nor covers any discussion on the limitations of RANS or the application of LES or DES.

Comment 3: Although limitations are briefly acknowledged, important issues are not addressed such as the accuracy of the Nielsen bedload model in non-coastal, vortex-dominated flows, or the possible numerical diffusion in FLOW-3D’s VOF method affecting free-surface dynamics behind the baffle.

Response 3: Thank you for the remarks. Previously we wrote the following text in chapter 2. Study Methodology regarding the usability of Nielsen’s bedload model “Nielsen [32] studied bedload movement over a rough, moving bed with riffles by generating uniform waves in a physical model experiment. However, this approach has been found effective in modeling river and channel hydrodynamics [29,34,35].” We also added the following text in the same chapter: “Although Nielsen’s transport equation was originally developed for bedload motion in coastal regions, it has been successfully utilized to simulate bedload transport over non-erodible beds in channels characterized by vortex-dominated flows, as demonstrated by Kostić et al. [33], who studied bedload transport in channel bifurcations influenced by helical flow patterns similar to those observed in the present research.”

Regarding the limitation of the used VOF on the free-surface dynamics downstream of the baffle, we added the following text in the chapter 5.3 Summary of findings and limitations: “Lastly, one-fluid VOF was used in FLOW-3D and the air-entrainment mode was not enabled considering that any considerable air-water mixing and any splashing are unexpected in this study since the flow is in subcritical condition, and that this VOF approach can reduce the computational time. However, two-fluid VOF can be investigated in the future to inspect any possible air-entrainment and its influence on vortex formation, if any. ”. Furthermore, FLOW-3D provides a sharp interface, thus, avoiding numerical diffusion in the downstream of the baffle.

Comment 4: The manuscript lacks an evaluation of y⁺ values along the channel bed and sidewalls. Please add it. The suggested article can be helpful.

Response 4: Thank you for the comment. The y⁺ values are added during the revision and explained in the last paragraph of chapter 2. Study Methodology.

Comment 5: Only one baffle submergence level (20% of flow depth) and a handful of angles and discharges are tested. The claim that the method is broadly effective is not supported by the narrow parameter range, especially given that submergence is a key parameter controlling both underflow and vortex formation. If possible please improve it.

Response 5: Thank you for highlighting this excellent point, which is also related to the comment 3 of reviewer 1. We plan to investigate the above-listed and some more in future research. In order to elaborate this, we added the following sentence in the model set up part: “As this paper presents the first investigation on the influence of oblique underflow baffles on bedload transport, the primary objective was to explain the underlying mechanism responsible for the observed effect. Other geometric and material parameters, such as submergence ratio, baffle angle, baffle thickness, and sediment properties, were not examined here but will be addressed in subsequent research.”

Comment 6: Sediment is introduced as a point-source at 0.05 m above the bed. This non-physical method may distort the natural hydrodynamics of sediment entrainment and deposition, especially in a uniform channel. No rationale is provided for using a point source instead of prescribing an equilibrium or uniform upstream bedload transport rate. Please elaborate about it.

Response 6: The sediment was introduced through a sediment source which was spanning across the full width of the channel, not just as a point source. To highlight this, we slightly changed the last paragraph of the chapter 3. Model Set Up. In Figure 6, it is visible that the sediment is deposited across the whole channel width upstream of the underflow baffle. The sediment was also added uniformly throughout the simulation with a prescribed sediment concentration. Another way to add sediment in FLOW-3D would be to add it at the inlet defining a prescribed sediment concentration. However, since the first 9 meters of the model only had the purpose to achieve developed flow conditions and since it would be too costly to include this first 9 meters in the sediment transport simulations, we needed to add the sediment source near the inlet boundary of the restart model.