On the Finite Element Analysis of Shell Structures
Summary
This paper is a Korean review of finite element analysis for shell structures. It connects three layers that must be understood together: physical shell behavior, the Basic Shell Mathematical Model, and finite element discretization. The paper focuses on thin-shell difficulty: as thickness decreases, shell problems split into bending-dominated, membrane-dominated, and mixed-dominated asymptotic behavior, and unreliable elements show Shell Locking Phenomenon in convergence curves.
The local source is a converted Markdown/image extraction: two Markdown files and 78 extracted images under .raw/쉘구조물의유한요소해석에대하여/.
Coverage Map
Section
Topic
Abstract and 1
Why shell finite element analysis needs integrated physical, mathematical, and numerical understanding
2
Basic Shell Mathematical Model from midsurface geometry, covariant bases, director kinematics, and variational equations
Uniform Optimal Convergence, ideal shell element requirements, MITC/ANS/EAS remedies, and consistency/ellipticity tradeoffs
5.3
Shell Element Benchmark Testing using basic tests, S-norm, layers, Gaussian curvature, asymptotic classes, and mesh patterns
6
Conclusion that shell mathematical models and asymptotic behavior are prerequisites for reliable shell FE interpretation
Key Takeaways
Shell FE reliability is not only an implementation issue; it depends on matching physical behavior, shell mathematical model, and discretization.
The basic shell model captures bending, membrane, transverse shear, and coupling terms and is the mathematical model beneath continuum-mechanics-based shell finite elements.
The load scaling factor rho classifies thin-shell behavior: membrane-dominated near 1, bending-dominated near 3, and mixed-dominated between them.
Locking appears as thickness-dependent loss of convergence and artificial stiffness, especially for displacement-based shell elements in bending or mixed-dominated problems.
MITC-style mixed interpolation is presented as a strong locking remedy, but the paper emphasizes the balance between locking control, consistency, and ellipticity.
Shell element benchmarking should include basic tests, global error norms, asymptotic behavior classes, Gaussian curvature, layer behavior, and mesh distortion sensitivity.
Composite source hash recorded in .raw/.manifest.json.
The converted Markdown contains OCR and encoding artifacts, but the title, authors, abstract, section structure, equations, tables, and conclusions are usable.
김경종