Reproducing Fig. 3(a) (mesh independence) from Lin et al. (2022): systematic conversion from self-propelled eel2d to tethered tandem hydrofoils
Vinod Thale
Dear all,
I hope this message finds you well. I am new to IBAMR and am using it to reproduce published results as part of a validation and learning study.
I am attempting to reproduce Fig. 3(a) (mesh independence study) from the following paper:
Lin, Z., Bhalla, A. P. S., Griffith, B. E., Sheng, Z., Li, H., Liang, D., Zhang, Y.
How swimming style and schooling affect the hydrodynamics of two accelerating wavy hydrofoils,
Ocean Engineering (2022), https://doi.org/10.1016/j.oceaneng.2022.113314
The target nondimensional parameters are:
-
Re = 5000, St = 0.4
-
Aₘₐₓ = 0.1, λ = 2
-
ϕ/π = 1, G = 0.3, D = 0.5
Key methodological point
My starting point was the self-propelled eel2d example in IBAMR.
To match the configuration in Lin et al. (2022), I performed a systematic conversion from a self-propelled swimmer to a tethered tandem-hydrofoil setup, including:
-
disabling rigid-body translation and rotation,
-
imposing a uniform free-stream velocity,
-
keeping only prescribed deformation kinematics,
-
fixing the foil locations in space,
-
modifying the control-volume and COM-update logic accordingly,
-
extending the setup to two interacting foils.
While the simulation runs successfully, this conversion is nontrivial, and I would like to confirm that my interpretation and implementation are consistent with best practices in IBAMR.
Current setup
-
IBAMR version: 0.18.0
-
Based on the eel2d example, modified to simulate two tethered wavy hydrofoils
-
Lagrangian meshes generated using a MATLAB-based script
-
3224 Lagrangian points per hydrofoil
-
.vertex files written in full double precision
-
Adaptive mesh refinement enabled
I have attached:
-
input2d
-
my modified source file (example.cpp)
Questions
1. Tethered formulation
For a tethered hydrofoil undergoing prescribed undulatory motion, is it correct in IBAMR to:
-
prescribe only the deformation velocity, and
-
disable all rigid-body translation and rotation (i.e., no penalty stiffness or imposed forces)?
2. Mesh independence study
To reproduce Fig. 3(a), is the standard IBAMR approach to:
-
vary the Eulerian grid resolution (Δx), while
-
maintaining Δs/Δx ≲ 1 for the Lagrangian mesh?
3. Reference configurations
Besides eel2d, is there an IBAMR example closer to a tethered oscillating foil or hydrofoil that you would recommend as a reference?
4. Time-step control and CFL
How do we select and decide DT_MAX the recommended approach for this class of problems?
My goal is to ensure that this self-propelled → tethered conversion is both numerically correct within IBAMR and faithful to the methodology of the paper.
Thank you very much for your time and for developing and maintaining IBAMR.
Best regards,
Vinod