Geng Integrity Review: "Undulatory vs. gliding locomotion: Effects on underwater object detection" (Gong et al., Journal of Ocean Engineering and Science, 2026)

Academic fraud report · Geng Detector

Summary

This report translates a Chinese-language academic-integrity review (Geng) of the paper "Undulatory vs. gliding locomotion: Effects on underwater object detection" by Shixian Gong, Dixia Fan, Linlin Kang, and Weicheng Cui, published in the Journal of Ocean Engineering and Science (DOI: 10.1016/j.joes.2025.11.012; available online 29 November 2025, listed as a 2026 issue). The reviewer concludes the paper is CLEAN at the textual level. The submission timeline (submitted August 2025, accepted November 2025, published 2026) is consistent with the check date of 28 June 2026. Methodological choices — NACA0012 airfoil, Reynolds number range 230–4360, and an immersed-boundary lattice-Boltzmann method (IBM-LBM) — are standard in bionic fluid-dynamics research. Reported numerical claims (e.g., ~2.3× pressure-signal amplification by undulation, improved SNR with tail-beat frequency) are described as plausible and free of suspiciously round or fabricated patterns. Confidence is moderate: pixel-level inspection of Figures 1–12, particularly pressure-signal traces (Figure 3) and vorticity/pressure fields (Figures 10–12), was not possible from text-only input, so image-based manipulation cannot be ruled out.

Verdict

Text-level verdict: Clean (no textual anomalies detected). Image-level verdict: Inconclusive — no figure data available for review.

Key findings

Submission timeline is coherent. Submitted August 2025, accepted November 2025, formal publication in the 2026 volume of *Journal of Ocean Engineering and Science*. No temporal inconsistencies relative to the review date (2026-06-28).
Methodology is internally consistent and field-standard. NACA0012 airfoil, Reynolds number range 230–4360, and an immersed-boundary lattice-Boltzmann method (IBM-LBM) are conventional in bionic underwater-vehicle fluid-dynamics studies.
Numerical claims are plausible. Reported ~2.3× amplification of effective pressure signal by undulatory motion and SNR increasing with tail-beat frequency are physically reasonable; the text retains "imperfect" features (e.g., weak but non-zero fin-phase influence, exponential decay of the signal), which argues against fabricated over-smoothing.
No suspicious data patterns detected in text. No identical anomalous data columns or implausible statistical distributions were identified.
Image forensics not performed. Only textual/parial-extract content was provided; pixel-level checks (background noise, splicing seams, local rotation/reuse) on Figures 1–12, particularly Figure 3 (pressure signals) and Figures 10–12 (vorticity and pressure fields), were impossible.

Evidence highlights

DOI: 10.1016/j.joes.2025.11.012
Submission→acceptance→publication chain: 2025-08 → 2025-11 → 2026, consistent with review date 2026-06-28.
Quantitative claim verified in text: "average 2.3 factor" amplification of effective pressure signal for undulatory motion.
Reynolds-number envelope stated in manuscript: 230–4360.
Source PDF supplied to reviewer: 1-s2.0-S2468013325000981-main.pdf.
No instances of citations to non-existent or future-dated literature detected.

Notes

This is a translation/summary of an AI-assisted Geng report; it is not an official institutional finding.
The reviewer explicitly recommends, if higher-resolution originals and complete datasets become available, follow-up image-level inspection using tools such as Forensically to test for local stretching, splicing, or copy-rotation artefacts.
Limits: text-only input precludes any conclusion on image duplication or manipulation; false negatives remain possible for both text-level and image-level fraud.
Any formal determination of misconduct must come from an authorized institutional investigation.