Optimized Venturi-Ejector Adsorption Mechanism for Underwater Inspection Robots: Design, Simulation, and Field Testing
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe authors present a contribution of interest in the topic.
The dynamical model of the system is not approached. Moreover the imperfections are not discussed. The system is nonlinear and the imperfections of course could improve the behavior of it.
Make some consideration abut looking at the following contribution:
Bucolo, M., Buscarino, A., Famoso, C., Fortuna, L., & Gagliano, S. (2021). Imperfections in integrated devices allow the emergence of unexpected strange attractors in electronic circuits. IEEE Access, 9, 29573-29583.
Author Response
Response to Reviewer Comments
Manuscript ID: [JMSE-3881095]
Title: Optimized Venturi-Ejector Adsorption Mechanism for Underwater Inspection Robots: Design, Simulation, and Field Testing
Dear Reviewer,
We sincerely appreciate your positive assessment of our work and your valuable comments regarding the dynamical modeling and the role of imperfections in our nonlinear adsorption system. Your insightful suggestions have opened up important directions for deepening our study. We have carefully considered your points and provide a detailed response below.
Comment 1: The dynamical model of the system is not approached. Moreover the imperfections are not discussed. The system is nonlinear and the imperfections of course could improve the behavior of it.
Response:
We thank the reviewer for these critical observations, which highlight significant avenues for future research beyond the current scope of our manuscript, which primarily focused on the static adsorption force and parametric optimization.
1.Regarding the System Dynamical Model: We agree that a complete dynamical model describing the force-motion coupling during the adsorption process was not established. Our current model (Section 2.2) focuses on the steady-state relationship between adsorption force and key geometric parameters. In direct response to your comment, we have added a paragraph in the revised conclusion section outlining our plan for future work. This includes the development of a coupled "fluid-structure-motion" dynamical model. This model will incorporate dynamic parameters (e.g., rate of adsorption force change, contact stiffness) to analyze transient responses, such as the robot's stability under sudden flow disturbances.
- Regarding the Discussion of Imperfections: We fully agree with the reviewer's profound insight that imperfections can positively influence nonlinear systems. While we observed related phenomena (e.g., micro-pores in the sealing layer affecting leakage on rough surfaces), a mechanistic discussion was lacking. In the revised conclusion, we now explicitly state our intention to systematically investigate imperfections (e.g., manufacturing tolerances, material inhomogeneity, wear). We plan to employ numerical simulation and experiment to analyze how specific imperfection parameters affect performance, exploring potential positive effects, such as enhanced negative pressure stability via a controlled "throttling effect" from micro-pores.
Comment 2: Make some consideration about looking at the following contribution: Bucolo et al. (2021).
Response:
We are grateful for the reviewer's recommendation of the highly relevant work by Bucolo et al. (2021). We have studied this paper carefully.
We acknowledge that their demonstration of how imperfections can lead to unexpected complex dynamics (strange attractors) in electronic circuits provides a groundbreaking perspective for our research on fluidic adsorption systems. Inspired by this, we have added a sentence in the revised conclusion stating that our future research will draw on this "imperfection-induced nonlinear dynamical behavior" approach. We aim to explore whether introducing imperfection parameters (e.g., sealing layer porosity) into our dynamical model could reveal similar nonlinear phenomena, potentially offering a new theoretical basis for designing robust systems by strategically managing imperfections.
Once again, we extend our deepest appreciation for your guidance. Your comments have been instrumental in helping us frame the future trajectory of this research. We have revised the conclusion to reflect these plans and hope you find our response satisfactory.
Sincerely,
The Authors
Reviewer 2 Report
Comments and Suggestions for Authors## Comments on the suction cup structure and material configuration:
The dual-layer sponge suction cup appears to be arranged concentrically; however, the manuscript does not provide sufficient details regarding the dimensions and material layout. Could the authors clarify the following points:
- What are the thicknesses and heights of the inner and outer sponge layers, respectively?
- If both layers have equal heights, the inner sponge may interfere with the deformation of the outer one, especially if it is stiffer. Was there any intentional height difference to mitigate this?
- How were the material hardnesses and thicknesses of the two sponge layers determined? Were they experimentally optimized, or based on prior knowledge?
## Comments on vacuum pressure and suction flow rate:
The paper states that the vacuum pressure generated by the venturi-type ejector was −75 kPa, and the suction flow rate was 50 L/min. However, it is unclear whether this flow rate was measured under actual experimental conditions or is simply the rated value of the ejector. I recommend that the authors clarify:
- Whether the 50 L/min value was actually verified during operation.
- Whether any fluctuations in flow rate occurred during experiments and whether they impacted suction performance.
## Comments on durability of the sponge material:
Since the system is intended for use on rough surfaces, the durability of the sponge—particularly the outer layer—is a critical concern. Could the authors provide:
- Any experimental data or observations on the wear resistance of the sponge material over repeated use?
- Information on whether the suction performance degraded after multiple trials?
Author Response
Response to Reviewer Comments
Manuscript ID: [JMSE-3881095]
Title: Optimized Venturi-Ejector Adsorption Mechanism for Underwater Inspection Robots: Design, Simulation, and Field Testing
Dear Reviewer,
We sincerely appreciate your time and valuable comments on our manuscript. Your insightful questions have helped us identify areas where our descriptions could be clearer. We have carefully considered each point you raised and provide the following clarifications based on the original design and experimental data.
Comment 1: The dual-layer sponge suction cup appears to be arranged concentrically; however, the manuscript does not provide sufficient details regarding the dimensions and material layout. Could the authors clarify the thicknesses and heights of the inner and outer sponge layers, respectively? If both layers have equal heights, the inner sponge may interfere with the deformation of the outer one, especially if it is stiffer. Was there any intentional height difference to mitigate this? How were the material hardnesses and thicknesses of the two sponge layers determined? Were they experimentally optimized, or based on prior knowledge?
Response:
We thank the reviewer for these important questions regarding the suction cup design, which highlight a need for clearer description. We apologize for any confusion caused by the lack of detailed dimensions in the manuscript. We clarify the actual design as follows:
- The dual-layer EPDM sponge sealing ring adopts a superimposed structure, where a closed-cell sponge layer is axially stacked on an open-cell sponge layer. Both layers are concentric rings.
- The closed-cell sponge has a thickness (axial height) of 20 mm, and the open-cell sponge has a thickness of 10 mm. The radial width of both layers is 20 mm, and the overall outer diameter is 120 mm.
- The intentional height difference (20 mm vs. 10 mm) was designed to allow the more compressible open-cell layer to adapt to surface roughness without being fully constrained by the stiffer closed-cell layer, which primarily ensures sealing integrity and recovery.
- The selection of material types (closed-cell vs. open-cell) and their thickness ratio was based on prior knowledge and theoretical analysis of pressure transmission efficiency, aiming to minimize negative pressure attenuation. Experimental optimization of these parameters represents valuable future work.
We recognize that including these dimensional details in the manuscript would improve clarity. We will incorporate this information into the final version upon acceptance.
Comment 2: The paper states that the vacuum pressure generated by the venturi-type ejector was −75 kPa, and the suction flow rate was 50 L/min. However, it is unclear whether this flow rate was measured under actual experimental conditions or is simply the rated value of the ejector. I recommend that the authors clarify whether the 50 L/min value was actually verified during operation, and whether any fluctuations in flow rate occurred during experiments and impacted suction performance.
Response:
We thank the reviewer for highlighting this point and apologize for the confusion. We wish to clarify that the value "50 L/min" was incorrectly mentioned in our previous response letter and does not appear in the manuscript. The manuscript focuses on and reports only the vacuum pressure data (e.g., achieving up to -92.17 kPa under optimal conditions). The suction flow rate was not measured in this study. We acknowledge that investigating flow rate characteristics and their impact on stability is an important aspect for future research.
Comment 3: Since the system is intended for use on rough surfaces, the durability of the sponge—particularly the outer layer—is a critical concern. Could the authors provide any experimental data or observations on the wear resistance of the sponge material over repeated use? Information on whether the suction performance degraded after multiple trials?
Response:
We appreciate this critical point regarding practical application. Currently, our study has not included specific durability tests (e.g., wear resistance or cyclic adsorption-desorption performance). Therefore, we cannot provide experimental data on performance degradation over repeated use. This is a recognized limitation. In future work, we plan to systematically evaluate sponge durability through reciprocating friction tests on rough surfaces and multi-cycle adsorption tests, and we agree that such data are essential for assessing long-term viability.
Once again, we extend our deepest appreciation for your rigorous and constructive feedback. Your comments have been invaluable in identifying areas for clearer communication and future investigation.
Sincerely,
The Authors
Reviewer 3 Report
Comments and Suggestions for AuthorsThe article titled "Optimized Venturi Ejector Adsorption Mechanism for Underwater Inspection Robots: Design, Simulation, and Field Testing" submitted to the Journal of Marine Science and Engineering was reviewed. The submitted article was determined to be suitable for the journal's subject and objectives. The study was produced from a project funded by the Changzhou Science and Technology Program.
Six keywords were used. Technical terms that highlight the importance of the topic can be replaced with keywords instead of quoting from the title.
A brief explanation of why the Venturi injector is recommended in the abstract would be more appropriate.
26 references were used. These references are up-to-date, but citations may be needed in other sections, such as the simulation section. The authors are advised to increase the number of references.
The visuals used in the article are clear and understandable.
The formulas are highly consistent with the scientific data.
The numerical data obtained appear to be reproducible and reliable.
A few statements encouraging readers and new research could be added to the conclusion.
Author Response
Response to Reviewer Comments
Manuscript ID: [JMSE-3881095]
Title: Optimized Venturi-Ejector Adsorption Mechanism for Underwater Inspection Robots: Design, Simulation, and Field Testing
Dear Reviewer,
We sincerely appreciate your time and valuable comments on our manuscript. Your insightful suggestions have been instrumental in helping us improve the quality and clarity of our work. We have carefully considered each point you raised and have revised the manuscript accordingly. Below, we provide a point-by-point response to your comments, detailing the changes made.
Comment 1: Six keywords were used. Technical terms that highlight the importance of the topic can be replaced with keywords instead of quoting from the title.
Response:
We thank the reviewer for this suggestion. In response, we have revised the keywords to focus on technical terms that emphasize the innovation and core advantages of the study. The original keywords, which overlapped highly with the title (e.g., "Underwater inspection robot" and "Venturi-ejector"), have been replaced with more specific terms: "Venturi-ejector adsorption", "Dual-layer EPDM seal", "CFD parametric optimization", "Underwater non-magnetic adhesion", "Low-power negative pressure", and "Rough surface adaptability". These new keywords better highlight the technical contributions and meet the journal's academic standards.
Comment 2: A brief explanation of why the Venturi injector is recommended in the abstract would be more appropriate.
Response:
We agree with the reviewer and have added a concise explanation in the abstract. The supplementary text now reads: "Owing to its 'zero-mechanical-vacuum-pump' fluid entrainment characteristics, the Venturi ejector possesses core advantages including an extremely simple structure, rapid response (forming a stable negative pressure zone within milliseconds), and low sealing requirements. It can efficiently generate negative pressure through hydrodynamic entrainment. Compared with traditional negative pressure adsorption schemes relying on mechanical vacuum pumps, it is more suitable for long-term stable operation in complex underwater environments, thus being selected as the core component for negative pressure generation in the adsorption system." This addition clarifies the rationale for choosing the Venturi ejector and enhances the technical logic of the abstract.
Comment 3: 26 references were used. These references are up-to-date, but citations may be needed in other sections, such as the simulation section. The authors are advised to increase the number of references.
Response:
We thank the reviewer for this important suggestion. We have significantly expanded the references and added citations in key sections:
- Added 7 new references (References 27–33), covering theoretical support (e.g., Schar, 1993 for Bernoulli's theorem) and CFD methodology (e.g., Kim et al., 2021; Cloete et al., 2015).
- Citations have been incorporated in Section 2.2 (Theoretical Analysis) and Chapter 3 (CFD Simulation) to strengthen the theoretical and methodological foundation.
- The total number of references has increased from 26 to 33, ensuring timeliness (1993–2024) and broader coverage of theoretical, simulation, and experimental aspects.
Comment 4: A few statements encouraging readers and new research could be added to the conclusion.
Response:
We have added encouraging and prospective statements at the end of the conclusion: "Moreover, the proposed Venturi-ejector adsorption mechanism not only provides a reliable solution for current underwater inspection tasks but also opens up new possibilities for the application of fluid-driven adhesion in other challenging environments, such as marine infrastructure monitoring, underwater welding, and aquaculture facility maintenance. We encourage researchers and engineers to further explore the potential of this mechanism in different operational scenarios and to continue optimizing its performance through advanced materials, adaptive control strategies, and multi-physics modeling. Such efforts will contribute to the broader adoption of intelligent adhesion technologies in unstructured aquatic environments." This addition enhances the guidance for future research and the manuscript's academic impact.
Once again, we extend our deepest appreciation for your time and expertise. Your comments have greatly improved the manuscript. We hope you find our responses satisfactory.
Sincerely,
The Authors
Reviewer 4 Report
Comments and Suggestions for AuthorsI recommend publication of this article after major changes.
The main question addressed by this research is the development of EPDM suction cup used for underwater adhesion of robots to dams. Numerical CFD study was used for geometry optimization, and verification was performed with experimental testing on various surfaces, two different pumps, three EPDM materials (open-cell, closed-cell, and their composite), and five (or is it six) ejector prototypes. I will return to this later.
This topic is original since, unlike conventional vortex suction cups [https://doi.org/10.3390/jmse12040662], this paper showcases the usage of a Venturi effect-composite suction cup. This paper is relevant in the field of remotely operated vehicles (ROVs) and autonomous underwater vehicles (AUVs) design, because it enables the combination of the propulsion and suction assemblies, thus simplifying and streamlining the design. The researchers filled the gap in the field by exploring the potential of different EPDM materials for suction cup design, not just cup diameter, as was done in [https://doi.org/10.3390/jmse12040662].
Compared with other published material, this paper offers a comprehensive study that includes everything from theoretical analysis, grid sensitivity, nozzle and throat diameter optimization, lab experiments, and field testing.
However, there are some very specific improvements that the authors must consider regarding the methodology before this paper is published. Figure 14 shows great discrepancies between simulated and experimental adsorption forces for prototypes H1–H5, which are not properly addressed.
“This discrepancy is attributed to two primary factors: firstly, the actual operating power of the selected submersible system (88.9 W) was lower than its rated capacity (100 W), reducing kinetic energy driving the jet flow; secondly, additional turbulent dissipation occurred due to surface roughness in the metal fluid channels, further diminishing negative pressure generation eficiency.”
Simulations should be re-run with the actual pump properties (88.9 W). “Surface roughness” is a questionable justification, as the authors used a no-slip wall boundary condition in their simulation.
Also, the authors run numerous simulations and determined that the optimum combination is a = 4.4 mm, b = 5.8 mm. For experimental testing, they used five prototypes (H1-H5), but they did not use the sixth prototype: H3.5, with the optimum combination of a = 4.4 mm, b = 5.8 mm. According to the simulations shown in Figure 14, H4 combination is better than H3, but the experiments showed H3 to be superior in terms of negative pressure and adsorption force. The optimum combination H3.5 is not shown here. Maybe the robot would be better with H3 (a = 4.0 mm, b = 5.8 mm)
Numerical simulations imply approximations; the more complex the simulated model is, the less accurate the results are; however, the trends at least should be the same.
There is a discrepancy in Figure 5 as well. Why is negative pressure so random, while the adsorption force shows almost a linear relation to pad diameter?
In Figure 1 top section is not appropriate. Either the whole robot should be shown, or better yet, the unmarked section above (a) and (b) should be removed. I would also like to see a short description of robot movement. I see a small wheel in Figure 1 (a), so I assume the robot traverses the dam wall on these wheels, and that the suction cups are only activated when the robot needs to hold a specified position.
Figures 6,7,9 10 are too small, with pressure and velocity contours looking the same for all 8 cases, with the only difference being the range in the legend. The authors should choose only 2 images (one for pressure and one for velocity) and plot 8 results on two diagrams(one diagram for pressure and one for velocity). Maybe the overlapping would be too much; in that case, they could plot the 4 most distinct cases.
For Figure 13, the section description should be done in order of appearance. Section (a) should be described before sections (b) and (c), or Figure 13 should be rearranged.
Figure 16 should be replaced with a table, as all of the elements in Figure 16 are already shown in Figure 13.
Table 1 should be formatted differently to keep the Values tested in one line.
Author Response
Response to Reviewer Comments
Manuscript ID: [JMSE-3881095]
Title: Optimized Venturi-Ejector Adsorption Mechanism for Underwater Inspection Robots: Design, Simulation, and Field Testing
Dear Reviewer,
We sincerely appreciate your time and valuable comments on our manuscript. Your insightful suggestions have been instrumental in helping us improve the quality and clarity of our work. We have carefully considered each point you raised and have revised the manuscript accordingly. Below, we provide a point-by-point response to your comments, detailing the changes made.
Comment 1: Figure 14 shows great discrepancies between simulated and experimental adsorption forces for prototypes H1–H5, which are not properly addressed. Simulations should be re-run with the actual pump properties (88.9 W). “Surface roughness” is a questionable justification.
Response:
We thank the reviewer for this critical observation. We agree that a more detailed explanation was necessary and have taken the following actions to address this discrepancy thoroughly:
Supplementary Experiment: We have conducted additional experiments to include the optimal parameter combination (a=4.4 mm, b=5.8 mm, designated as H4) that was identified by the CFD simulations but was missing from the initial experimental matrix.
Enhanced Explanation: We have significantly expanded the discussion in the revised manuscript (Section 4.2, Page 16). The explanation now more clearly states that the simulation uses the pump's theoretical characteristic curve under ideal, rated power (100 W), while the experimental system operates at a lower actual power (88.9 W) due to inevitable system losses. This is the primary reason for the systematic quantitative difference.
Validation of Trend Prediction: We emphasize that despite the quantitative difference, the CFD model successfully predicted the relative performance trend across all parameter configurations. Crucially, the experimental results for the newly tested H4 configuration confirmed it as the best-performing prototype, achieving a peak adsorption force of 417.9 N, which aligns perfectly with the simulation's prediction. This consistency validates the model's utility for parametric optimization.
Clarification on Surface Roughness: We have refined the wording to clarify that the "hydraulically smooth" wall assumption in CFD is an idealization. The microscopic roughness of real machined surfaces contributes to additional energy dissipation not captured by the model, which is a secondary factor compounding the discrepancy.
These changes have strengthened the discussion and validity of our conclusions regarding the CFD model's optimization capability.
Comment 2: The authors did not use the sixth prototype: H3.5, with the optimum combination of a = 4.4 mm, b = 5.8 mm. According to the simulations... H4 combination is better than H3, but the experiments showed H3 to be superior.
Response:
We thank the reviewer for identifying this important omission in our initial submission. This has been rectified.
As mentioned in response to Comment 1, we have fabricated and tested the H4 prototype (a=4.4 mm, b=5.8 mm). The experimental results now clearly show that the H4 configuration yields the highest adsorption force (417.9 N), outperforming the H3 configuration (403.1 N). This result is consistent with the CFD predictions and resolves the previous apparent contradiction. The relevant figures, tables, and text in Sections 4.1 and 4.2 have been updated accordingly. We sincerely appreciate this suggestion, which has significantly improved the experimental rigor and conclusiveness of our study.
Comment 3: There is a discrepancy in Figure 5 as well. Why is negative pressure so random, while the adsorption force shows almost a linear relation to pad diameter?
Response:
We sincerely thank the reviewer for highlighting this issue. The inconsistency observed in the initial version of the figure was valid and primarily due to the simulation setup not achieving ideal convergence in the first attempt.
In direct response to your comment, we have performed a rigorous re-simulation of the parameter study presented in Figure 5. The recalculated results now clearly demonstrate that the negative pressure within the cavity remains essentially constant as the suction cup diameter increases, while the adsorption force shows a clear linear growth due to the increase in the effective area. This aligns perfectly with the theoretical expectation that Adsorption Force = Negative Pressure × Area.
The revised Figure 5 accurately reflects this physical relationship. We apologize for the oversight and appreciate your review, which has led to an improvement in the quality and accuracy of our data presentation.
Comment 4: Figure 1 top section is not appropriate. Either the whole robot should be shown, or better yet, the unmarked section above (a) and (b) should be removed. I would also like to see a short description of robot movement.
Response:
We agree with the reviewer's suggestion regarding the figure composition.
We have removed the extraneous blank space above subfigures (a) and (b) in Figure 1, resulting in a cleaner and more professional presentation.
Comment 5: Figures 6,7,9,10 are too small, with pressure and velocity contours looking the same for all 8 cases... The authors should choose only 2 images... and plot 8 results on two diagrams.
Response:
We appreciate this excellent suggestion for improving the clarity and informativeness of the figures.
Following your advice, we have completely redesigned these figures:
For each parameter study, we now present only one representative pressure contour plot and one velocity contour plot for reference.
The core information is now conveyed through newly created line plots that extract and overlay the axial pressure and velocity profiles for all 8 parameter cases.
Furthermore, we have added detailed zoomed-in views of the critical nozzle/throat regions to highlight the subtle differences between configurations.
This significant revision has resulted in a much more effective and reader-friendly presentation of the simulation results.
Comment 6: For Figure 13, the section description should be done in order of appearance... Figure 16 should be replaced with a table.
Response:
We thank the reviewer for these practical suggestions.
Figure 13: We have reordered the subfigures and their labels in Figure 13 to ensure they are described in the logical order of appearance ((a) Surface specimens, (b) Test platform, (c) Ejector prototypes, (d) Sealing layers, (e) Power profiles).
Figure 16: We have replaced the original Figure 16 with a new Table 2. This table succinctly presents the negative pressure retention performance data, making it easier to compare and interpret.
Conclusion
Once again, we extend our deepest appreciation for your time and expertise. Your comments were exceptionally insightful and have guided us in making substantial improvements to the manuscript. We believe the revised version is significantly stronger due to your input. We hope that you find our responses and the accompanying modifications satisfactory.
Sincerely,
The Authors
Round 2
Reviewer 4 Report
Comments and Suggestions for AuthorsThe newly added section seems incomplete: "Future work will focus on Moreover," Please finish it before publication.
