This naval architecture CFD tutorial chains the hull into the wave study, total resistance, propeller selection and its CFD verification.

Start with tutorial no. 1

Sizing a propulsion system is not about calculating a propeller : it is about establishing what the ship demands, then checking that the propeller delivers it. Naval architecture CFD therefore starts upstream of the propeller — with the hull.

The chain : the hull and its waves, total resistance, the operating point, the propeller, CFD verification. Heliciel covers all five links, and this tutorial walks through them in order — starting from the hull geometry, not from coefficients copied out of a table.

• Prerequisites : the hull geometry — an STL or IGES file of the bare hull, or one of the hulls supplied with the software — and the target speed ;
• Duration : a few seconds per wave calculation, the Drag tab is instant, CFD verification takes the time of a Phase 1 ;
• What you get out of it : a thrust requirement justified by the hull, and a propeller verified for that thrust.

• Tutorial outline:

1. The hull in the wave study
2. Total resistance in the Drag tab
3. From thrust requirement to operating point
4. Choosing and setting the propeller
5. Verifying by CFD calculation
6. Inserting the hull or an appendage into the domain
7. What this chain does not tell you

1: The hull in the wave study

Open Edit > Hydrodynamics > Ship waves (Michell / Kelvin). The window is self-contained : it neither reads nor modifies the propeller project open in the designer. On first opening, it displays a sample project supplied with the software — a hull already computed, which you can explore before inserting your own.

The window has two tabs : Model and wave, with the 3D view and the parameter panels, and Drag, the subject of the next step.

Loading the hull. Three routes in the Hull group of the Model panel :

• Insert a hull (STL, IGES)… : your own geometry. An IGES file is converted and mirrored automatically, and the conversion report is displayed. Insert the bare hull : keel and rudder would distort the hydrostatics, they are handled through their wetted surface in step 2 ;
• the list of analytical hulls and the Generate hull button : the Wigley hull, the Michell strut, an ellipsoid, a prismatic planing surface, or a tabulated Series 60 hull. These are exact shapes, useful for learning the tool and for checking that it compares well against itself ;
• the Delft Systematic Yacht Hull Series (DSYHS) : published sailing yacht models, loaded in one click. They measure about two metres : it is the File scale field that brings them up to full size.

Positioning the hull. File scale (units → m) for an STL in millimetres, Draft (m) for the depth of the keel below the waterline, then Roll x, Pitch y and Yaw z. The draft follows the scale on its own : a model scaled up by 5 does not float like a cork. A draft larger than the hull is clamped to its height, and the log states it. The displayed hull follows every setting ; the wave field, on the other hand, waits until you release the wheel.

Setting the flow. In the Simulation parameters panel, Flow group : Speed (m/s), Density (1 000 in fresh water, 1 025 at sea) and Water depth, 0 for deep water. The Froude number is displayed immediately : it governs the wave pattern and the share of wave resistance in the total.

Calculate and display. The button in the toolbar slices the hull into stations, forms the free-wave spectrum, and draws the Kelvin wake on the water : the transverse and divergent waves, in the 19.47° V that every surface-piercing source produces. The speed slider above the view lets you watch the pattern tighten or widen ; the Sweep Froude checkbox steps through it.

The hull's Kelvin wake, seen from above : transverse and divergent waves, V apex at the bow

Reading the information panel. Below the panels, the calculation returns its numbers, and each of them will be used further on :

• the Froude number and the transverse wavelength λ0, which give the size of the waves ;
• the Michell wave resistance (N) and the resulting Cw coefficient : the share of resistance the hull spends making waves, according to linear thin-ship theory. An order of magnitude and a cross-check, not a measurement : section 7 returns to this ;
• the hydrostatics : wetted surface, volume, displacement, waterline Lwl × Bwl × T, coefficients CB / CM / CP / CWP and LCB. These are exactly the quantities the resistance methods of the next step call for — and they are measured on your hull, not estimated.

A point of method : these coefficients hold for the bare hull. A concave section gets filled in by the slicing, an appendage widens the hull. That is the reason to load the hull alone, and to reserve appendages for their own surface.

The information panel : Michell wave resistance, Cw, then the hydrostatics measured on the hull

2: Total resistance in the Drag tab

The Study this hull's drag button lights up as soon as a calculation has succeeded. It refreshes the Hull readout, preselects a ship type based on where the hull came from, and switches to the Drag tab. Nothing is calculated yet : the calculation remains your own action.

Choosing the ship type. This is the choice that matters : the type decides the method, and the Method used is displayed with the sentence stating how far it holds.

Separating the readout from the input. The tab separates two blocks, and that is its whole value. The Hull readout, read-only, carries S, ∇, displacement, Lwl, Bwl, T, CB, CP, CM, CWP, LCB, density and speed : all of this comes from the hull, you do not type it in. The To be entered block asks only for what no immersed geometry can give :

• Appendage wetted surface (m²) : keel and rudder, added to the bare hull's surface. A DSYHS hull proposes the surface of its standard appendages on its own, at scale ;
• Maximum speed (knots) for the curve, prefilled with the study speed ;
• Form factor (1+k) for displacement methods, Viscosity ν, optional Roughness ;
• Cw ×10⁻³ for the generic ITTC method, optional : the Michell Cw read in step 1 is a candidate, at the study speed ;
• Deadrise β and LCG for Savitsky.

Calculating the drag. The curve plots total drag R_T with its two components, viscous R_V and wave R_W, against speed in knots. The result text gives R_T and the effective power P_E = R_T × V at the maximum speed — and, alongside it, the Michell Rw for the same hull at the study speed. Two theories on a single geometry : it is a cross-check, not a substitution. If they diverge strongly, one of the two is outside its domain, and the table above says which.

For Savitsky, also read the alerts : speeds excluded for lack of a planing regime do not appear in the curve, and the text says so.

The Drag tab : the readout comes from the hull, the input covers only what it does not give, and the curve separates friction from wave

The window's File menu saves the whole thing into a wave project (.hlvag), settings and hull included : it reopens as is on another machine. It also exports the view image and the ζ(x, y) sheet as CSV.

3: From thrust requirement to operating point

At the target speed, the required thrust equals the total resistance R_T read in step 2, increased by the usual margins : sea state, fouling, ageing. That is the first figure the propeller will have to deliver. The operating point is then closed with three further quantities : the advance speed in the propeller plane, the available rotation speed, and the admissible diameter.

None of the three is free. The diameter is bounded by the draft — which you now know — and the skeg clearance ; the rotation speed by the engine and its gearbox ; the advance speed differs from the ship's speed, because the propeller works in the hull's wake.

This last point is the subtlest in the chain. The water reaching the propeller has been slowed by the hull's passage : the propeller therefore advances, relative to the fluid, more slowly than the ship. Conversely, the propeller's suction increases the hull's resistance. These two interaction effects are traditionally handled through coefficients established by experience, and they are neither in the Drag tab nor in an open-water propeller calculation.

The wave study shows a trace of this : the Thruster (thrust disc) group of the Model panel places a disc at the thrust you have just established, at the propeller's position — underwater, negative x towards the stern. The calculation then returns the wave resistance of the hull alone, of the thruster alone, and of the combination : the difference is the interference between the wake and the suction hollow, visible on the water. The group's title displays the useful power thrust × speed, in kW and horsepower : this is not the shaft power, the propeller efficiency is not included.

For the rotation speed, Heliciel offers sweep search tools in the designer : find the rotation speed that gives a target thrust, or the one that maximises efficiency. They work in BEM, hence instantly, and that is where exploring pays off.

4: Choosing and setting the propeller

Two routes are open to you in Heliciel's designer, and they are not mutually exclusive.

Starting from a reference propeller. The Wageningen B-series, built in since version 12.1.2, provides a complete, documented geometry : number of blades, area ratio, pitch ratio. This is a sound starting point, and a benchmark for comparison — tutorial no. 18 makes use of this second quality.

Designing from scratch. Heliciel's BEM optimises twist and chord distribution for your precise operating point, which a systematic series cannot do by construction. The boat propeller design tutorial details the steps ; the thrust to enter is the one from step 3.

In both cases, two checks are called for before moving to CFD : the cavitation risk, which depends on the shaft's immersion depth — enter it, it will be reused by the CFD calculation — and the blade's mechanical strength, which Heliciel's structural module evaluates in bending and in centrifugal tension.

The propeller chosen for this hull : the operating point comes from the ship, not from an assumption

5: Verifying by CFD calculation

CFD only comes in here, on the selected configuration or configurations. File > New CFD case (from current model), preset Free Level 5 — Study at minimum, Phase 1. Tutorial no. 13 follows a complete marine case from BEM to wake.

Three checks are worth the calculation time :

• thrust and torque, checked against the BEM — taking care not to compare a total that includes the hub against a blades-only value ;
• the pressure map on the suction side, where cavitation shows up first ;
• the flow at the blade tip and root, two zones where a marine propeller loses efficiency without the BEM showing it.

If your propulsion is ducted — nozzle, azimuth thruster, bow thruster — Heliciel will ask you to specify the configuration : duct in free-stream flow, or closed duct. A tug's Kort nozzle falls under the first case ; a bow thruster in a through-tunnel falls under the second, covered in tutorial no. 15.

CFD verification : the field the BEM does not produce, on the chosen configuration

6: Inserting the hull or an appendage into the domain

A ship propeller never works alone : there is a shaft, a strut, a rudder, sometimes a tunnel. The Insert tab of the CFD test bench lets you add an object to the computation domain.

Two routes : importing a geometry with Insert .STL model, or building a primitive with Insert Cube, Insert Sphere or Insert Cylinder. Primitives are created at unit size and centred ; the desired dimension is then set with the Scale X/Y/Z fields, the Uniform scaling checkbox linking the three axes. The cylinder is created with a vertical axis : to turn it into a shaft section aligned with the flow, lay it down with the Pitch slider.

The wave study supplies the hull geometry ready for this purpose : Export hull to STL…, in the Hull group, writes the hull as positioned — scale, attitude and draft applied, waterline at z = 0. The original file itself stays untouched. One point of rigour before inserting it : Heliciel's CFD calculation has no free surface (section 7). A full hull would be wetted up to the deck, and its wave wake would not exist. Insertion therefore makes sense for appendages and the shaft line, which condition the flow arriving at the propeller ; for the full hull, it does not replace the interaction coefficients of step 3.

Positioning is done as a percentage of the domain, on the three axes, with three rotations. After placing it, remesh : the object must appear as a boundary in its own right in the meshing report, and its forces will appear separately in the results table. If the object does not appear in the boundary list, it is not in the domain, and the calculation will run without it — with no error, and no effect.

The Insert tab : an object placed in the domain, with its position as a percentage and its rotations

7: What this chain does not tell you

Four limits to state before delivering a report :

• Michell's wave resistance is a linear model. Thin-ship theory, free waves downstream, deep water by default : the hull shape only enters through the distribution of its widths, and the field is not validated against a measurement. The information panel displays a domain indicator — slenderness, validity angle — to read before citing the figure. It is a cross-check against the statistical method, not a towing-tank value ;
• Each drag method has its own domain. Holtrop & Mennen on a planing hull, or Savitsky on a cargo ship, produces a meaningless number. The ship type is there to prevent that : choose it for what your ship is, not for the method you prefer ;
• The CFD calculation is open-water. The hull is not in the domain, and there is no free surface : the calculation is single-phase and without gravity, without waves, without an interface, without ventilation. Hull-propeller interactions — wake, suction — remain handled through empirical coefficients. A deeply immersed propeller copes with this well ; a surface-piercing thruster or a surface propeller does not ;
• Cavitation is assessed, not simulated. The calculation tells you where the pressure drops below the vapour threshold ; it models neither the formation nor the collapse of bubbles, and therefore predicts neither the thrust loss nor the erosion.

These limits do not disqualify the chain : they delimit its use. A report that states them is worth more than one that stays silent about them — that is the subject of tutorial no. 20.

The full set of twenty CFD tutorials

« Previous: Calibrating on a reference  |  Next: The deliverable calculation file »

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
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 (you are here)
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.