Benchmark Computations of stresses in a spherical dome with shell finite elements

Antti H. Niemi

DOI: https://doi.org/10.48550/arXiv.1507.03747

2015-07-14

Abstract:We present a computational framework for analysing thin shell structures using the finite element method. The framework is based on a mesh-dependent shell model which we derive from the general laws of three-dimensional elasticity. We apply the framework for the so called Girkmann benchmark problem involving a spherical shell stiffened with a foot ring. In particular, we compare the accuracy of different reduced strain four-node elements in this context. We conclude that the performance of the bilinear shell finite elements depends on the mesh quality but reasonable accuracy of the quantities of interest of the Girkmann problem can be attained in contrast to earlier results obtained with general shell elements for the problem.

Numerical Analysis

What problem does this paper attempt to address?

The core problem that this paper attempts to solve is to evaluate and improve the accuracy of thin - shell finite - element analysis, especially when dealing with complex geometries (such as spherical domes). Specifically, the author evaluates the performance of different types of four - node shell elements in calculating stress by studying the Girkmann benchmark problem.

### Main Problem Description

1. **Accurate Simulation of Thin - Shell Structures**:
- The author aims to develop a computational framework based on the finite - element method (FEM) for analyzing thin - shell structures.
- This framework is based on a mesh - dependent shell model derived from three - dimensional elasticity theory.

2. **Girkmann Benchmark Problem**:
- The Girkmann problem involves the stress distribution of a spherical dome reinforced by a foot ring under the action of gravitational load.
- The specific task is to determine the shear force and bending moment at the connection between the dome and the foot ring, as well as the maximum bending moment in the dome.

3. **Deficiencies of Existing Methods**:
- The paper points out that the previous calculation results of the Girkmann problem using general shell elements have large deviations and cannot reach acceptable accuracy.
- The author hopes that through improved finite - element methods (such as MITC4S and MITC4C), more accurate results can be obtained.

### Methods and Technical Means

- **Finite - Element Method**:
- Multiple four - node shell elements (such as DISP4, MITC4C, MITC4S and their stabilized versions) are used.
- The locking phenomenon is avoided by reducing membrane strain and transverse shear strain.

- **Numerical Experiments**:
- Convergence studies on different mesh sequences are carried out to evaluate the performance of each method.
- The performance of different element types in the cases of membrane - dominated deformation and bending - dominated deformation is compared.

### Conclusions

- The author finds that using improved finite - element methods (such as MITC4C and MITC4S) can obtain relatively accurate results under reasonable mesh quality.
- Especially in the Girkmann problem, the performance of these methods is better than that of traditional shell elements, which can effectively avoid the locking phenomenon and provide reliable stress calculation results.

In summary, this paper is committed to improving the accuracy of stress analysis of thin - shell structures through improved finite - element methods and verifies the effectiveness of the new methods through the Girkmann benchmark problem.