This propeller sliding mesh tutorial runs the unsteady Phase 2 and compares its forces to those of Phase 1.

Start with tutorial no. 1

Phase 1 makes the propeller rotate in the equations. Propeller sliding mesh makes it rotate for real : at each time step, the mesh block containing the rotor pivots by a small angle, and slides against the fixed mesh surrounding it. This is Phase 2 of Heliciel PRO CFD/BEM.

The price is stated plainly by the interface : the checkbox that controls Phase 2 carries the note « Phase2 ≈ 10 × Phase1 », an order of magnitude (on the real cases of this series, Phase 2 lasted 2 to 27 times as long as Phase 1). This tutorial shows what you buy for that price.

• Prerequisites : a completed Phase 1, the one from tutorial no. 4 ;
• Duration : variable — that is precisely the point ;
• What you get : forces that vary over time, and a wake that is no longer frozen.

• Tutorial outline:

1. Taking back control of the sequence
2. What happens between the two phases
3. Launching Phase 2
4. Comparing the forces from both phases
5. What a revolution costs
6. When Phase 2 is worth its price

1: Taking back control of the sequence

By default, Heliciel chains everything : meshing, Phase 1, then Phase 2. For this tutorial, we want to see the seam. Parameters tab, Progress management frame : uncheck Run unsteady Phase2 after Phase1.

The calculation will therefore stop at the end of Phase 1, and you will launch Phase 2 yourself. This is also the setup to adopt when a machine is shared, or when you want to examine the steady results before committing hours of computation.

Two special cases to know : in a closed duct, Phase 2 is very costly and Heliciel automatically unchecks the chaining while warning you ; and if the mesh has the flaws that make a sliding interface diverge, the warning arrives before launch, not after three hours.

Progress management frame after the instruction: the Phase 1 box stays checked, the Phase 2 box is unchecked — so the calculation will stop at the end of Phase 1

2: What happens between the two phases

The transition is not a simple restart. Between Phase 1 and Phase 2, Heliciel runs a precise sequence :

1. the converged fields from Phase 1 — velocity, pressure, turbulence — are copied over as Phase 2's initial conditions. Phase 2 therefore does not start from fluid at rest, it starts from an already-established flow ;
2. the mesh is split along the cylinder that bounded the MRF zone. Where there was only an internal boundary, you now get two facing surfaces ;
3. these two surfaces are declared as a non-conformal coupling : they do not have matching faces, they do not need to, and the solver will interpolate fluxes from one to the other at every time step.

It is this interface that allows the rotation : the inner block rotates, the outer block stays fixed, and the coupling ensures flow continuity between the two.

The rotating-zone cylinder, meshed, around tutorial no. 13's marine propeller (pressure coloured on the blades): the cut that creates the sliding interface is made on its surface. The capture does not let you tell the two facing sides apart.

3: Launching Phase 2

With the end-of-Phase-1 marker present, the CFD calculation button now launches Phase 2. The algorithm changes : it switches from steady SIMPLEC to transient PIMPLE, with no under-relaxation, using an upwind convection scheme (second order for the velocity) for stability.

The simulated duration is counted in rotor revolutions. On a free rotor, Phase 2 simulates one revolution by default, more as soon as the rotor has four blades or fewer (the convergence monitor needs at least 4/N + 0.1 revolution, N being the number of blades). In an open duct, it is capped at 5 revolutions and stops by itself once the forces stabilise. In a closed duct, where each revolution costs much more, it is limited to 0.3 revolution for all four presets.

The time step is not fixed : a regulator adjusts it based on the observed stability, stepping it up in stages when the calculation is holding well. What you see scrolling in the Logs are these steps.

4: Comparing the forces from both phases

Here is the point of the tutorial. Results tab, switch the selector between MRF (phase 1) and Sliding mesh (phase 2), with Surface pressure checked in both cases.

Three differences jump out on an ordinary propeller :

• the wake : in Phase 1 it is smooth and steady ; in Phase 2 it carries the helical trace of each blade's passage ;
• the pressure at the blade tip, where the tip vortex takes on a structure the frozen calculation could not form ;
• the forces, which are no longer a single number but a signal. They oscillate at the blade-passing frequency, and it is the average over the last revolution that makes the result — not the instantaneous value.

Check Propeller wake in the Show results: frame to explicitly display the vortex sheet.

Oblique view of tutorial no. 13's marine propeller : the pressure on the blades and the flow lines coloured by axial velocity. This is the kind of display with which Phase 2 reveals the turbulent trace of the blades and hub, a structure a frozen-rotation calculation does not form.

5: What a revolution costs

The cost of a Phase 2 is dominated by two factors : mesh fineness and the number of revolutions simulated. The propeller preset catalogue displays a relative cost for each mesh level, from ×1 for the Express level up to ×117 for the Reference level ; it is a yardstick for comparing levels, not a duration.

Two rules of thumb : if Phase 2 diverges early, step up one notch of preset — a finer mesh gives a better sliding interface. If it is stable but too slow, step down one notch.

6: When Phase 2 is worth its price

A decision table, so as not to pay the price of Phase 2 without reason :

Tutorial no. 16 covers precisely the static-thrust case, where this decision matters most.

The « noise » block of the results file after a Phase 2 (marine propeller of tutorial no. 13, Level 1 mesh, half a simulated revolution) : the blade-passing frequency (12.25 Hz = 7 blades × 1.75 rev/s), the line amplitude at each harmonic, the total fluctuation and the tonal share relative to the broadband. On a mesh this coarse and half a revolution, these figures show the shape of the result, not a measurement.

Noise_BPF_Hz=12.25
Noise_FxTotalRms_N=84.0962
Noise_MxTotalRms_Nm=34.6244
Noise_FxBroadbandRms_N=79.0087
Noise_FxTonalToBroadband_dB=-8.76
Noise_MxTonalToBroadband_dB=-8.19
Noise_SampleCount=135
Noise_BPFx1_Hz=12.25_Fx=6.0107N_Mx=4.1303Nm
Noise_BPFx2_Hz=24.50_Fx=18.279N_Mx=7.1656Nm
Noise_BPFx3_Hz=36.75_Fx=16.284N_Mx=6.8568Nm
Noise_BPFx4_Hz=49.00_Fx=14.243N_Mx=6.3696Nm
Noise_BPFx5_Hz=61.25_Fx=15.133N_Mx=6.7829Nm
Noise_BPFx6_Hz=73.50_Fx=14.887N_Mx=6.4618Nm
Noise_BPFx7_Hz=85.75_Fx=13.828N_Mx=6.0436Nm
Noise_BPFx8_Hz=98.00_Fx=13.398N_Mx=5.9372Nm

The full set of twenty CFD tutorials

« Previous: The MRF zone and Phase 1  |  Next: Comparing BEM and CFD »

1. First CFD propeller simulation
2. Reading the OpenFOAM case folder
3. Choosing the mesh preset
4. The MRF zone and Phase 1
5. Phase 2 with sliding mesh (you are here)
6. Comparing BEM and CFD
7. Reading the convergence of a run
8. Turbulence models
9. Boundary layer and y+
10. Reading a pressure map
11. Where the forces come from
12. Checking an airfoil polar
13. Marine propeller and cavitation
14. Wind turbine and tidal turbine
15. Fan in a closed duct
16. Aircraft propeller, cruise and static thrust
17. The optimisation loop
18. Calibrating on a reference
19. From the hull to the propeller
20. The deliverable calculation file

This series accompanies the Heliciel design tutorials, which cover the BEM part: blade design, rotation speed choice, performance curves. CFD comes after them, to verify and to see.