Racing Yacht Keel FEA — Daniel Tombleson

Overview

Structural FEA of an IMOCA-class racing keel

This study used ANSYS to simulate the structural response of a racing yacht keel — a fin-and-bulb assembly — under two distinct load cases: the sustained gravitational self-weight of the keel and bulb acting on the fin, and a dynamic frontal grounding impact. Both scenarios were assessed against IMOCA Class Rules for 2028, which mandate an overall safety factor of 5.0 and a heightened factor of 6.5 in structural attachment zones such as the root-fin and fin-bulb interfaces.

With a steel yield strength of 800 MPa, this translates to a maximum permissible von Mises stress of 123.07 MPa in the attachment regions. The simulation results were validated against independent analytical hand calculations, and a mesh convergence study was conducted across node counts from approximately 8,000 to 65,000 to confirm result reliability.

Case 1

Self-weight loading

The first load case applied the combined self-weight of the fin and bulb assembly as a gravitational body force. With the bulb mass of 2,486 kg at the fin tip, the cantilevered fin experiences significant bending stress along its length and a measurable lateral deflection at the tip.

FIG 1 — Fin lateral displacement magnitude (self-weight)

FIG 2 — Max normal stress in z-direction (fin)

Overall von Mises stress (Case 1 self-weight)

FIG 3 — Max equivalent von Mises stress (overall, self-weight)

Displacement Magnitude

42.24 mm

Lateral tip deflection of the fin under full gravitational loading

Max z-Normal Stress (Fin)

91.99 MPa

Bending-driven normal stress in the longitudinal fin direction

Max von Mises (Overall)
185.6 MPa
Peak equivalent stress — concentrated at structural attachment zones

Case 2

Frontal grounding impact

The second load case simulated a frontal grounding event — a critical scenario in offshore racing where the keel strikes a submerged obstruction. In this case, the safety philosophy shifts: rather than preventing any deformation, the criterion is to permit controlled plastic deformation that dissipates impact energy, analogous to automotive crumple zones.

With a safety coefficient of 1 relative to the 16% breaking strain, the fin attachment zone is designed to yield plastically before fracture — protecting both the vessel's hull and crew from the sudden transfer of impact force. A brittle fracture at this joint would be catastrophic.

Overall von Mises stress (Case 2 frontal impact)

FIG 4 — Overall max equivalent von Mises stress (frontal impact)

Fin equivalent von Mises stress (Case 2)

FIG 5 — Fin max equivalent von Mises stress (frontal impact)

FIG 6 — Max equivalent von Mises strain (overall)

Regulatory compliance — IMOCA 2028

Region	Safety Factor Required	Permissible Stress	Simulated Stress	Status
General structure	5.0×	160.0 MPa	185.6 MPa (Case 1)	⚠ Review
Attachment zones (root/bulb)	6.5×	123.07 MPa	125.99 MPa (fin, Case 2)	⚠ Marginal
Frontal impact — breaking strain	1.0× (16% strain)	Plastic deformation permitted	Strain = 0.001156	✓ Pass

Mesh Verification

Convergence study

To confirm the reliability of the FEA results, a mesh convergence study was performed by systematically reducing element size and recording the change in key stress outputs. As the node count increases — meaning smaller, more numerous elements — the solution converges towards the true continuum behaviour. Effective convergence was observed at approximately 40,000 nodes (mesh size 0.035 m), beyond which the stress values stabilise to a negligible difference.

Two independent datasets were compared across the same nodal range: the maximum z-direction normal stress under self-weight, and the fin equivalent von Mises stress under frontal impact. Both converge at similar nodal intervals, reinforcing the validity of the analysis as a whole.

Max z-direction normal stress vs node quantity

FIG 7 — Max z-direction stress vs node quantity

Convergence threshold

~40,000 nodes

Beyond this point, stress variation becomes negligible

Optimal mesh size

0.035 m

Balances computational cost with solution accuracy

Both datasets
Converge together
Agreement across two independent stress metrics confirms mesh reliability

Validation

Analytical verification

To verify the FEA outputs, an independent analytical calculation was performed using classical bending stress theory. The keel fin was idealised as a cantilever beam loaded by the bulb weight at its tip. The z-direction bending stress at the root section was calculated from first principles and compared against the ANSYS-predicted value of 91.99 MPa.

Analytical Method — Bending Stress

σ_z = M_x · y / I_x

M_x = moment = PL → 487.753+ N·m

y = distance from neutral axis to probe point

I_x = second moment of area

L = 4.17 m (total fin length)

c = 4.37 m (to probe point)

m_bulb = 2,486 kg · W = 24.388 kN

M_x,bulb = 24.388 × 10³ × 4.1 = 99.99 kN·m

I_x = 74.4 × 10⁻⁶ m⁴ | y = 0.075 m

σ_bulb = 99.99 × 10³ × 0.075 / 74.4 × 10⁻⁶
≈ 100.8 MPa

FEA result: 91.99 MPa | Analytical: ~100.8 MPa | Difference: ~9.6% — within expected bounds for simplified beam theory vs full 3D solid model.

Design Critique

Safety assessment & improvement recommendations

Two areas of critique were identified from the simulation results: one relating to the coefficient of safety against yield, and one addressing potential design improvements to better distribute stress concentrations and optimise material usage.

1

CoS — Yield strength at attachment zones

Under Case 1 self-weight, the overall peak von Mises stress of 185.6 MPa exceeds the general structure limit (800 MPa / 5.0 = 160 MPa). Under Case 2 frontal impact, the fin attachment zone stress of 125.99 MPa marginally exceeds the required limit of 123.07 MPa (800 MPa / 6.5). These results highlight that the current geometry produces stress concentrations at connection interfaces that require design attention before the keel can be unconditionally approved for IMOCA competition.

2

CoS — Breaking strain under frontal impact

For the frontal grounding scenario, the safety criterion was set at a coefficient of 1 relative to the material's 16% breaking strain. This philosophy — intentionally analogous to a crumple zone — allows the attachment zone to deform plastically rather than fracture, dissipating energy progressively and reducing the impulse transferred to the hull and crew. The simulated strain of 0.001156 sits well within this envelope, confirming the design's energy-absorption capability under critical grounding events.

3

Fin connection geometry — stress concentration reduction

Reducing the transition angle between keel sections and incorporating fillets at the root-fin and fin-bulb joints would smooth the stress flow-line paths, reducing the density of stress concentrations in these critical zones. A more gradual geometric transition allows load to distribute across a wider area rather than concentrating at sharp corners — directly addressing the marginal compliance in attachment zones identified above.

4

Material utilisation — weight optimisation

Given that the fin's maximum predicted stress (91.99 MPa under self-weight) is well below the 800 MPa yield strength of the steel, the material is significantly over-specified for fin loading. Two options follow from this: (a) substitution of a lower-density material — such as a high-strength aluminium alloy or titanium — where the reduced stiffness is acceptable, or (b) topological thinning of low-stress regions of the fin to remove unnecessary mass while preserving load paths. Either approach would improve the yacht's power-to-weight ratio and competitiveness.

Racing YachtKeel FEA