Dynamic Derivatives

This example demonstrates the computation of dynamic stability derivatives for an airfoil undergoing pitching oscillations. It employs a sliding mesh approach, where the airfoil is enclosed within a rotating cylindrical interface. The simulation strategy involves two stages: first, a steady-state simulation is performed to establish a baseline flow field, followed by an unsteady simulation where the sliding interface oscillates according to a specified sinusoidal motion to capture the dynamic effects. The script showcases the setup for both the steady initialization and the subsequent unsteady, oscillating simulation using the Flow360 Python API.

"""
Example script for calculating dynamic derivatives using sliding interfaces in Flow360
See documentation for more information:
 https://docs.flexcompute.com/projects/flow360/en/latest/tutorials/DynamicDerivatives/DynamicDerivatives.html
"""

import math

import flow360 as fl
from flow360.examples import TutorialDynamicDerivatives

TutorialDynamicDerivatives.get_files()
project = fl.Project.from_geometry(
    TutorialDynamicDerivatives.geometry,
    name="Tutorial Calculating Dynamic Derivatives using Sliding Interfaces from Python",
)
geometry = project.geometry


# show face groupings
geometry.show_available_groupings(verbose_mode=True)
geometry.group_faces_by_tag("faceName")
geometry.group_edges_by_tag("edgeName")

with fl.SI_unit_system:
    cylinder = fl.Cylinder(
        name="cylinder",
        axis=[0, 1, 0],
        center=[0, 0, 0],
        inner_radius=0,
        outer_radius=1.0,
        height=2.5,
    )
    sliding_interface = fl.RotationCylinder(
        spacing_axial=0.04,
        spacing_radial=0.04,
        spacing_circumferential=0.04,
        entities=cylinder,
        enclosed_entities=geometry["wing"],
    )
    farfield = fl.AutomatedFarfield()
    params = fl.SimulationParams(
        meshing=fl.MeshingParams(
            defaults=fl.MeshingDefaults(
                surface_max_edge_length=0.03 * fl.u.m,
                curvature_resolution_angle=8 * fl.u.deg,
                surface_edge_growth_rate=1.15,
                boundary_layer_first_layer_thickness=1e-6,
                boundary_layer_growth_rate=1.15,
            ),
            refinement_factor=1.0,
            volume_zones=[sliding_interface, farfield],
            refinements=[
                fl.SurfaceEdgeRefinement(
                    name="leadingEdge",
                    method=fl.AngleBasedRefinement(value=1 * fl.u.degree),
                    edges=geometry["leadingEdge"],
                ),
                fl.SurfaceEdgeRefinement(
                    name="trailingEdge",
                    method=fl.HeightBasedRefinement(value=0.001),
                    edges=geometry["trailingEdge"],
                ),
            ],
        ),
        reference_geometry=fl.ReferenceGeometry(
            moment_center=[0, 0, 0],
            moment_length=[1, 1, 1],
            area=2,
        ),
        operating_condition=fl.AerospaceCondition(
            velocity_magnitude=50,
        ),
        time_stepping=fl.Unsteady(
            max_pseudo_steps=80,
            steps=400,
            step_size=0.01 * 2.0 * math.pi / 20.0 * fl.u.s,
            CFL=fl.AdaptiveCFL(),
        ),
        outputs=[
            fl.VolumeOutput(
                output_fields=[
                    "Mach",
                ],
            ),
            fl.SurfaceOutput(
                surfaces=geometry["*"],
                output_fields=[
                    "Cp",
                    "CfVec",
                ],
            ),
        ],
        models=[
            fl.Rotation(
                volumes=cylinder,
                spec=fl.AngleExpression("0.0349066 * sin(0.05877271 * t)"),
            ),
            fl.Freestream(surfaces=farfield.farfield),
            fl.Wall(surfaces=geometry["wing"]),
            fl.Fluid(
                navier_stokes_solver=fl.NavierStokesSolver(
                    absolute_tolerance=1e-9,
                    linear_solver=fl.LinearSolver(max_iterations=35),
                ),
                turbulence_model_solver=fl.SpalartAllmaras(
                    absolute_tolerance=1e-8,
                    linear_solver=fl.LinearSolver(max_iterations=25),
                ),
            ),
        ],
    )
# Run unsteady case with an oscillating sliding interface to collecting the data.
project.run_case(
    params=params,
    name="Tutorial Calculating Dynamic Derivatives using Sliding Interfaces",
)

Notes

• Two-Stage Simulation: A steady-state case initializes the flow field before the unsteady simulation begins, utilizing fork_from to inherit the flow field solution.

• Sliding Interface Oscillation: The pitching motion is modeled using a RotationCylinder sliding interface. Its angular position over time is defined via AngleExpression, incorporating nondimensional time (t) and frequency.