Abstract
As the characteristic coefficients of carbon fiber composite pressure vessels improve, the design requires more precise finite element models. However, the varying thickness profiles of the dome caused by fiber stacking and changes in winding angle pose significant challenges. To predict the dome profile more accurately, this paper developed a three-dimensional prediction method that did not require piecewise solving. The yarn was discretized, and its centerline was computed using the geodesic equation. Discrete points in the width direction were expanded outward according to the "shortest path" principle, forming its three-dimensional winding shape. The profile of the next layer was calculated using the arcs formed by the intersection of the yarn boundary and the parallel circles on the dome. The predicted outermost profile was then compared with the CT scan results. Based on this method and the cubic spline method, finite element models were established respectively, and progressive damage analysis was conducted. Hydraulic burst tests of the pressure vessel were performed. The results indicate that the predicted shape of the yarn shows a maximum deviation of approximately 2 mm from the actual measurements at the polar hole. The maximum relative error of the thickness predicted is -14.3%. The average strength from the three hydrostatic burst tests was 40.4 ± 0.6 MPa. The relative error of the simulation result based on the proposed method was -1.9%, marking a 5.5% improvement in accuracy over the cubic spline method. The proposed method can provide valuable insight for accurate modeling of pressure vessels.
| Original language | English |
|---|---|
| Article number | 113180 |
| Journal | Thin-Walled Structures |
| Volume | 212 |
| DOIs | |
| State | Published - Jul 2025 |
Keywords
- Carbon fiber composites
- Hydraulic burst test
- Pressure vessels
- Profile prediction method
- Progressive damage method
Fingerprint
Dive into the research topics of 'A novel prediction method for pressure vessel dome profiles based on "shortest path"'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver