FhSim  3.1.0
Marine systems simulation
Loading...
Searching...
No Matches
0159 — Cylinder law has no KC dependence: C_m fixed at 2 and C_d from Re only (Keulegan and Carpenter 1958, A5)
ID 0159
Class KNOWN-LIMITATION
Severity 2
Status blocked

Models: — (annotated on the fhsim_environment library group)

Found: 2026-11-18, validation case A5 (fhsim_validation, cases/hydrodynamics/cylinder_load/, plan HP D8)

Decision needed: the owner chooses whether the cylinder law gets a KC-dependent C_d and C_m (for example a table from Keulegan and Carpenter 1958 or Sarpkaya 1976), or whether the law documents a model range (steady flow, large KC) and A5 records the gap.

Evidence

include/fhsim_environment/hydrodynamics/CylinderHydro.h:307: kCylinderAddedMassCoefficient = 1.0, a constant, so C_m = 1 + C_a = 2 at every KC. NormalDragCoefficient (CylinderHydro.h:96-105) depends on Re only. KC (Keulegan-Carpenter number, U_m T / D) appears nowhere.

Case A5 (tracked, 15 %, the scatter at equal KC) runs the law cylinder_oscillating of fhsim_validation/laws/LawTable.cpp. It takes the Morison force from DragForce and MorisonInertiaForce over one cycle and applies the Fourier averages of Keulegan and Carpenter (Eqs. 30, 31). Table 3 of Keulegan and Carpenter 1958 (57 cylinder runs) gives measured C_m down to about 0.7 near KC 15 and up to 2.5 at KC 120, and C_d up to about 2.4 near KC 11 to 15. The law gives C_m = 2 and C_d about 1.0 to 1.2. First run: C_d 15 of 57 rows pass (median |error| 0.26), C_m 18 of 57 (0.23).

Effect

Loads on a cylinder or cable in oscillating flow (waves, a moving fairlead) use the steady-flow C_d and a constant C_m. Near KC 10 to 20 the measured drag is about twice and the measured inertia about a third of the law's values.

Possible fix

  • A KC-dependent C_d and C_m in CylinderHydro.h, with the KC of the local flow. Needs an owner physics ruling and a way to know the flow period.
  • Or document the model range (large KC, steady flow) in the user doc and keep A5 as the record.

Test that would prove it

The A5 rows of fhsim_validation pass their 15 % band (ctest -R fhsim_validation_laws_tracked, case A5).

Risk

A KC-dependent law moves every cable and cylinder result in waves (base, ME, fishery, maritime). Report the moves in the moves.md of those repos.