Numerical Research of Additional Mass Based on Panel Method

Haojian Zhang, Jianing Zhang, Lei Zhang, Tao Li2016 · DOI

Abstract

This paper is based on the panel method from three-dimensional potential flow theory, where sources and sinks are distributed on surface panels to meet boundary conditions. Additional mass is obtained by solving the velocity potential. Ellipsoids of various scales are used as examples to study the influence of grid generation on additional mass accuracy across different longitudinal and vertical grid numbers. By comparing calculated values with theoretical benchmarks, the study identifies the grid generation strategies and panel counts that best satisfy accuracy requirements.

Introduction

When a submerged object moves through a fluid, it must displace the surrounding medium, an effect modeled as "additional mass" or added mass. This added inertia is a critical parameter in the design and stability analysis of marine structures, such as ships and submarines.

A common numerical approach for calculating this is the Panel Method, which discretizes an object's surface into small, flat tiles or panels. While increasing the total number of panels generally improves accuracy, it also significantly increases computational costs. This research investigates how to optimize the distribution of a limited number of panels—specifically the balance between longitudinal (K1) and vertical (K2) divisions—to maximize accuracy for different geometric shapes.

Methodology

The study employs three-dimensional potential flow theory under the assumption of an incompressible, inviscid, and irrotational fluid. The core process involves:

  1. Source-Sink Distribution: Mathematical "sources" and "sinks" are distributed across surface panels to satisfy the kinematic boundary conditions on the body's surface.
  2. Velocity Potential: The Laplace equation is solved to determine the fluid's velocity potential.
  3. Pressure and Force Calculation: Using the Bernoulli equation, the potential is used to derive fluid pressure, forces, and ultimately the additional mass coefficients.

The researchers specifically analyzed ellipsoids with varying aspect ratios ($a:b:c$): 1:1:1 (sphere), 2:1:1, 5:1:1, and 10:1:1 (slender body). These shapes provide reliable theoretical benchmarks for error analysis.

Results and Discussion

Grid Sensitivity in Spherical Geometries

For a perfect sphere (1:1:1), the study found that both longitudinal (K1) and vertical (K2) grid numbers significantly influence the calculation error for surge motion ($m_{11}$). As shown in Figure 2, increasing the number of longitudinal grids leads to a rapid decrease in error, though the rate of improvement slows as the grid becomes finer.

Calculation errors for a sphere (1:1:1) showing the relationship between longitudinal grid number (K1) and error percentage across different vertical grid counts (K2).

Grid Sensitivity in Slender Bodies

The optimal meshing strategy shifts dramatically as the object becomes more slender. For a highly elongated ellipsoid (10:1:1), the longitudinal grid number (K1) has a negligible impact on accuracy. Instead, the vertical grid number (K2) becomes the dominant factor.

Calculation errors for a slender ellipsoid (10:1:1), demonstrating that increasing the longitudinal grid number (K1) provides minimal accuracy gains compared to increasing the vertical grid number (K2).

As illustrated in Figure 5, the error curves for different K1 values are nearly horizontal, indicating that adding more longitudinal panels does not significantly improve the result. However, shifting from $K2=3$ to $K2=5$ drastically reduces the error from nearly 20% to almost zero.

Conclusion

The research demonstrates that grid generation for the panel method should not be uniform but rather adapted to the object's aspect ratio. For blunt objects like spheres, refinement in both directions is beneficial. For slender, cigar-like bodies, computational resources are best spent increasing the vertical (circumferential) grid density rather than the longitudinal density.

While these findings provide a clear path for optimizing added mass calculations, the study is limited to surge motion in an infinite fluid. Future work should incorporate free surface effects and explore other degrees of freedom, such as heave and pitch, to better model real-world maritime conditions.

Summary updated

Abstract

This paper is based on the panel method from three-dimensional potential flow theory, sources-sinks are distributed on surface panel to meet the boundary conditions. Additional mass can be obtained by solving the velocity potential. Ellipsoids under different scales are as example to study influence of grid generation on additional mass through various longitudinal and vertical grid numbers. By comparing the calculated values with the theoretical values, the way of grid generation and the number of panel that most satisfy the accuracy requirements can be obtained.
BibTeX
@misc{zhang2016numerical,
  author = {Haojian Zhang and Jianing Zhang and Lei Zhang and Tao Li},
  title = {{Numerical Research of Additional Mass Based on Panel Method}},
  year = {2016},
  doi = {10.2991/iceep-16.2016.123},
}
Figure