Concerns regarding 'Joint multi-objective optimization based on multitask and multi-fidelity Gaussian processes for flapping foil' (Ocean Engineering, 2024)

Academic fraud report · Geng Detector

Summary

This report flags a highly suspect pattern of internal inconsistencies in Zhongyuan Wang et al.'s 2024 paper published in Ocean Engineering, DOI 10.1016/j.oceaneng.2024.116862. The verdict is 'highly suspicious'. Three categories of issues are documented: (1) a mathematical/linguistic contradiction in Appendix A where the text claims the lift coefficient 'decreases from 0.0144 to 0.0242', although 0.0242 is numerically greater than 0.0144; (2) inconsistent narrative in Table A.4, where the authors claim the lift coefficient follows a 'downward trend' with increasing resolution, yet the tabulated values rise after Resolution 64 (e.g., 0.0242 → 0.0256 → 0.0273 in the 16c×16c domain); (3) catastrophic typesetting errors in the Introduction and throughout, including broken citations such as '(Wang et al., ; , )' and 'A comprehensive review by 2022)' and truncated sentences like 'Section'. No image-based misconduct was assessed. These findings strongly suggest careless or template-based writing, but cannot independently prove data fabrication. Confirmation requires the authors' raw CFD logs and code. Confidence is moderate-to-high for textual inconsistencies and low for inferring deliberate fabrication.

Verdict

🟠 Highly suspicious. Multiple textual descriptions directly contradict the paper's own reported numerical values, and the publication contains numerous broken citations and incomplete sentences. These issues are consistent with hasty compilation or template-based writing, and raise serious questions about research diligence. They do not, by themselves, constitute definitive proof of data fabrication.

Key findings

Mathematical/linguistic contradiction: Appendix A states that the lift coefficient "decreases from 0.0144 to 0.0242", which is mathematically impossible because 0.0242 > 0.0144.
Trend description contradicts data: The text claims the lift coefficient shows a "downward trend" with increasing resolution, yet Table A.4 shows values rising after Resolution 64 (16c×16c domain: 0.0242 → 0.0256 → 0.0273; 32c×32c domain: 0.0247 → 0.0303 → 0.0304).
Broken citations throughout: Multiple in-text references are malformed, including (Wang et al., ; , ), A comprehensive review by 2022), and ...propulsion optimization, 2020) and 2005) employ....
Truncated sentence: The phrase "The rest of this paper is organized as follows. Section" appears without a section number, terminating mid-sentence.
Production timeline anomaly: The paper reportedly went through Revised (Dec 25) and Accepted (Jan 22) stages yet retained these errors, which is unusual for a finalized SCI publication.

Evidence highlights

DOI: 10.1016/j.oceaneng.2024.116862
Table A.5 (Appendix A, pp. 10–11): the "decreases from 0.0144 to 0.0242" line is internally contradictory.
Table A.4 (Appendix A, p. 10):
16c×16c domain C_l sequence: 0.0393 (Res 32), 0.0283 (Res 48), 0.0242 (Res 64), 0.0256 (Res 80), 0.0273 (Res 96).
32c×32c domain C_l sequence: 0.0307, 0.0231, 0.0247, 0.0303, 0.0304.
Citation fragments observed in the Introduction and elsewhere, including missing author names and years paired with commas and parentheses.

Notes

No image-based or figure-level analyses were performed in this report; pixel-level checks are not applicable here.
The findings strongly suggest copy-paste/template reuse or extreme writing carelessness, but cannot independently confirm data fabrication. Verification requires the authors' raw CFD simulation logs, source code (LilyPad platform), and underlying datasets for Tables A.4 and A.5.
Reporters may wish to raise the contradictions on PubPeer and request clarification from the corresponding author or the journal editorial office.