Aluminium Catamaran · CE Category A Structural Verification with FreeCAD FEM

1 · What CE Category A actually demands

Under the EU Recreational Craft Directive 2013/53/EU (RCD), a Category A craft is designed for extended voyages in conditions up to roughly Beaufort 10 and significant wave heights around 7 m. The design categories are:

Category	Description	Wind (Beaufort)	Significant wave height
A — Ocean	Extended voyages, self-sufficient	up to 10	up to ≈7 m
B — Offshore	Offshore voyages	up to 8	up to 4 m
C — Inshore	Inshore / coastal	up to 6	up to 2 m
D — Sheltered	Sheltered waters	up to 4	up to 0.5 m

Structurally, you demonstrate compliance through the harmonised EN ISO 12215 series (hull construction & scantlings). The standards give you the design loads; FreeCAD FEM verifies that your actual structure — plate thicknesses, frame spacing, bridgedeck beams, weld details — carries those loads with adequate margins:

EN ISO 12215-5 — design pressures (bottom, sides, deck) for monohull panels
EN ISO 12215-7 — multihull-specific loads: bridgedeck slamming, diagonal/torsional case
EN ISO 12215-6 — structural arrangements, stiffener design, buckling
EN ISO 12215-9 — sailing-craft appendages (rudders, boards)
EN ISO 12215-10 — rig loads & rig attachments (sailing cats)
ISO 12215-2 — material design strengths (welded vs unwelded aluminium)
ISO 12216 — windows & portlights (pressure-rated glazing)
ISO 12217-1/-2 — stability & buoyancy (not an FEM task, but part of the Cat A file)

Conformity assessment: craft with hull length under 12 m may use module A (self-declaration), but 12–24 m craft must involve a Notified Body. Even below 12 m, an independently reviewed Cat A structural file is strongly recommended. FEM reports fit neatly into the Technical Construction File as “state of the art” engineering justification.

2 · Verification strategy

Don’t try to derive slam pressures from first principles in FEM — the ISO pressures are semi-empirical and already include safety factors. Your workflow is:

Compute the ISO design loads (pressures, rig forces, torsion case) for your boat → spreadsheet.
Build two model tiers in FreeCAD:
- Global model — whole boat, midsurface shells, coarse-ish mesh → global cases, hot-spot hunting.
- Local models — single panel bays, chainplates, mast step, rudder casing, refined mesh → detail stresses.
Run each load case, compare von Mises stress, deflection and buckling factors against the allowables in Section 6.

ISO 12215 loads→ Global shell model→ Hot spots found→ Local refined models→ vs allowables (weld / HAZ)

Example boat used for the illustrative numbers below (yours will differ — recompute per the standards): LOA 10.5 m · hull centreline spacing 4.6 m · sailing cruising cat · fully loaded mass 6.5 t · bottom 5 mm 5083-H321 · frames at ~350–400 mm · bridgedeck 5 mm.

Run matrix — the tests that matter for Cat A

#	Load case	Standard	Illustrative load	Priority
1	Hull-bottom slamming (panel + full bottom)	ISO 12215-5 / -7	35–65 kN/m²	Core
2	Bridgedeck / cross-deck slamming	ISO 12215-7	25–55 kN/m² (upward)	Critical
3	Diagonal torsion (“catamaran twist”)	ISO 12215-7	T ≈ 145 kN·m	Critical
4	Rig & sailing loads (mast compression, chainplates)	ISO 12215-10 / -9	P ≈ 150 kN; shrouds ≈ 30 kN	Core (sailing)
5	Rudder & daggerboard cases	ISO 12215-9	F ≈ 20–45 kN side force	Core
6	Hoisting, slings & shore blocking	Good practice	1–2 g	Optional
7	Local & detail cases (tanks, windows, deck gear)	ISO 12215-2/-5, ISO 12216	5–30 kN/m² patches	Core
8	Buckling follow-on (from cases 2, 3, 4)	ISO 12215-6	λ from preload	Critical
9	Modal analysis — model validation	—	—	Optional

3 · Units & material cards (get these right first)

FreeCAD FEM is unit-agnostic internally. Use the standard mm–N–MPa–tonne system:

Quantity	Value to use	Note
Geometry length unit	mm	Model in mm
Young’s modulus	70 500 MPa	Enter as 70.5 GPa in the material editor
Poisson’s ratio	0.33	Aluminium
Density	2700 kg/m³	Editor converts to tonne/mm³ for the solver
Gravity constraint	9810 mm/s²	FEM → Constraint gravity
Pressure example	50 kN/m² = 0.05 MPa = 0.05 N/mm²	Enter `0.05` in the pressure constraint

Marine aluminium candidates

Alloy	Rp_0.2	R_m	Welded (HAZ) behaviour	Comment
5083-H111 / H321	≈145 / ≈215–230 MPa	≈275–315 MPa	Moderate softening; retains usable strength	The marine choice for welded plate
6082-T6	≈250 MPa	≈300 MPa	Severe HAZ softening near welds	OK for extrusions/bolted parts; weld with care
6061-T6	≈240 MPa	≈290 MPa	As 6082	Alternative where available

Confirm all values against your mill certificates and the design-strength tables in ISO 12215-2 — these are typical handbook figures.

FreeCAD setup: create a FEM Material (mechanical) card with E, ν and density, and assign it to geometry. For shell models, assign one material card per plate-thickness group (5 mm bottom, 5 mm deck, 6 mm keel strake…) and set the shell thickness on each card. Then verify the exported *SHELL SECTION lines in the CalculiX .inp — if the thickness is missing, your model is wrong.

4 · Building the FEM model in FreeCAD

4.1 Geometry (the hard part — do it well)

Model plates as surfaces (midsurfaces), shipyard-style: hull skin, frames, stringers, web frames, bridgedeck, beams, bulkheads. Stiffeners = web surface + flange surface. Skip fillets, hardware, cosmetic details.
Join everything with Part → Boolean Fragments (then Refine) so adjacent plates share edges. Without this, Gmsh meshes each face independently and panels end up not connected in the analysis — a classic silent failure.
Split faces where loads/supports land (Boolean Fragments with cutting sketches/boxes): sling pads, support patches, mast step, chainplates. Constraints map to geometry, so patches must exist as faces.
Group faces by thickness for the material cards.

4.2 Mesh

FEM → Mesh from shape (Gmsh), set Element dimension = 2D for shell models.
Second-order elements ON (default) — first-order elements badly underestimate plate bending.
Element size ≈2–3× plate thickness at hot spots (e.g. 10–15 mm for 5 mm plate); 30–80 mm on the global model away from details.
Full-boat global shell models typically land around 0.3–1.5 M DOF — CalculiX handles this in minutes to ~1 hour on a desktop.
Convergence check: halve the mesh size at the critical bay; stress should move <5 %.

4.3 Constraints & loads

Fixed constraint and Displacement for supports and symmetry.
Constraint pressure for distributed loads; Constraint force for point/patch forces; Constraint gravity for weight (9810 mm/s²). Apply equipment masses (engine, tanks, rig, batteries) as forces at their real CG positions.
Never apply point loads directly to shell nodes — always load a small face patch.
Check load direction with the displacement plot before reading any stress.
Keep one Analysis per load case in the same document — results stay side by side, and because the solver is linear-static you may superimpose results between cases.

4.4 Solver

CalculiX (ccx): static linear for the main programme; enable geometric nonlinearity for the bridgedeck cases (thin plates stiffen by membrane action, which matters); frequency analysis for validation; buckling via a manual .inp edit (Test 8).

5 · The test programme — setups in detail

All load magnitudes below are illustrative for the example boat. Compute the real values from EN ISO 12215-5/-7/-9/-10 for your dimensions, displacement and speed — the standards are copyrighted and must be purchased.

Test 1 — Hull-bottom slamming Core

Why	Category A wave impact on the bottom sets plating thickness and frame spacing. This is the daily bread-and-butter case.
Standard	EN ISO 12215-5 (pressures), -7 (multihull factors). Pressure depends on the design area (panel size), speed regime and position along the hull.
Load	Illustrative distribution: 0.06 MPa (60 kN/m²) forward third, 0.045 MPa amidships, 0.035 MPa aft. Apply as pressure on the outer bottom faces, pushing inward. Also run a single-panel check with the local design-area pressure.
Model A (fast)	One framing bay (e.g. 350×500 mm) plus half-height frames; clamp the frame lines. Mesh 8–15 mm. Use for rapid thickness/spacing iteration.
Model B (final)	Whole bottom with frames, stringers, floors, bulkheads; constrain at bulkhead cuts; pressure map applied. Mesh 15–40 mm.
Watch	Mid-panel von Mises (unwelded allowable), stress at frame weld toes (HAZ allowable), panel deflection, load direction sign.
Pass	σ_vm ≤ ≈150 MPa mid-panel (5083-H321), ≤ ≈110–130 MPa at welds, deflection ≤ ~1.5 % of short span (see Section 6).

Test 2 — Bridgedeck (cross-deck) slamming Critical for cats

Why	Wave slap on the underside of the structure between the hulls is the sizing case for catamarans and a fatigue driver at every hull–deck weld. Worst where bridgedeck clearance is smallest and forward.
Standard	EN ISO 12215-7 bridgedeck slamming pressure (function of slamming speed, position, design area).
Load	Upward pressure on the bridgedeck underside: illustrative 0.05 MPa forward third tapering to 0.02 MPa aft; plus an envelope run at uniform 0.055 MPa. Pressure pushes the deck up (sign check!).
Model	Cross-deck skin, deck stringers, cross-beams (webs + flanges as surfaces), hull bulkheads and the hull–bridge junction structure. Either the full global model or a bridge section with two frame stations of each hull included.
BCs	Clamp the hull cut sections (far enough away that junction stress is unaffected), or run inside the global model. Enable geometric nonlinearity (membrane action).
Watch	Deck skin mid-panel stress; stringer end welds; beam web shear; deflection of the slam panel; combined case with rig compression (Test 4) for the top skin.
Pass	As Test 1 allowables; webs checked with von Mises including shear; buckling of the compression skin → Test 8.

Test 3 — Diagonal torsion (“catamaran twist”) Critical

Why	Balance the boat on a diagonal wave — one bow and the opposite stern supported, the rest of the mass hanging — and the cross-structure carries a large torque. Classic multihull global case.
Standard	EN ISO 12215-7 diagonal/torsional case. First-order design torque: T ≈ Δ·g·s/2 → example: 6.5 t × 9.81 × 2.3 m ≈ 145 kN·m.
Setup A (recommended)	Global model. Fix two support patches (~300×600 mm) on the bottom of the port bow and the starboard stern. Apply gravity plus point forces for engines, tanks and rig at their real CGs so the hanging mass — and thus the torque — is realistic. Check that the support reactions reproduce the expected torque.
Setup B (self-equilibrated)	Apply ±force couples at the four hull ends to generate pure torque, with a minimal “3–2–1” restraint (one node fixed in xyz, one in yz, one in z). Fewer support artifacts; more fiddly.
Run twice	Mirror the diagonal (port/stbd swapped) — any structural asymmetry (mast step, daggerboard case) shows up only in one direction.
Watch	Bridgedeck shear, the hull–bridge junction welds, bulkhead ends, twist angle (report it as a stiffness metric), compression flange of the cross-beams.
Pass	σ_vm ≤ welded allowable at junctions; buckling factor of compressed skins λ ≥ 2 (Test 8).

Test 4 — Rig & sailing loads (sailing cats) Core if sloop/cat rig

Why	Mast compression, shroud and sheet loads all land on the cross-beams, deck and hull — on a cat the rig loads and the torsion case interact.
Standard	EN ISO 12215-10 (rig loads; input righting moment from ISO 12217-1), -9 (board loads).
Loads (illustrative)	Mast compression P ≈ 150 kN downward spread over the step patch; shroud tension ≈ 30 kN per side along the stay axis at the chainplate faces; mainsheet ≈ 25 kN at the traveller; daggerboard case loads from the righting moment.
Cases	(a) Upright, max righting moment. (b) Rig loads + diagonal supports (Test 3 geometry) — usually the worst case for the cross-beam. (c) Rig-load-only local models for chainplate panel and mast step.
Watch	Cross-beam under the mast (bending + compression, buckling of the compression flange), chainplate local panel and its deck reinforcement, hull shell around fittings, crushing/bearing under the step pad.
Pass	σ ≤ welded allowable at step and chainplate boundaries; buckling λ ≥ 2 for beam flanges and deck panels under compression.

Test 5 — Rudder & daggerboard (appendages) Core

Why	Appendage failures are common and dangerous offshore; ISO 12215-9 sizes them explicitly.
Standard	EN ISO 12215-9 (design forces for rudders, boards; includes speed, area and category factors).
Loads (illustrative)	Rudder side force F ≈ 20–45 kN at the blade centroid. Cases: forward-speed maximum lift; hard-over; reverse/astern. Boards: RM-derived side force; plus a grounding/beaching reaction if applicable.
Model	Mixed: solid second-order tets for the stock and metal fittings, shells for blade and surrounding hull; bearings as contact or bearing-arc patches; include casing floors and local frames.
Watch	Stock von Mises and tip deflection (binding), bearing pressure into the casing, pintle/gudgeon brackets, hull shell stress at the casing welds.
Pass	Stock stress within its material allowable (stainless/duplex per certificate); hull stresses within welded aluminium allowables; deflection small enough that clearances are maintained under load.

Test 6 — Hoisting, slings & shore blocking Optional but cheap

Why	Travelift lifts at launch and winter storage on blocks load the hull girder differently than the sea — and the owner’s manual (RCD requirement) must state safe lifting/blocking points.
Setup	Vertical supports (`Displacement`, u_z = 0) on sling-pad patches at ≈0.2·L and 0.8·L, minimal lateral restraints elsewhere. Gravity at 1× for storage and 2× for hoisting (dynamic factor).
Watch	Bottom plating and floors at the pads, global hogging, deck-edge tension/compression.
Pass	No local yielding at pads at 2 g; sensible reactions distribution between slings.

Test 7 — Local & detail cases Core

Tanks	Fuel/water bulkheads: hydrostatic head to the vent plus a slosh/slam allowance (illustrative +25 kPa in slam-exposed zones). Check stiffeners and their welds.
Windows & portlights	Frame and surrounding shell loaded with the ISO 12216 pressure for Cat A (illustrative 15–30 kN/m²); the acrylic itself is assessed per ISO 12216 test methods.
Deck	Crew/gear pressure per ISO 12215-5 (≈5 kN/m² walkable; anchor well higher). Apply fittings’ rated loads (cleats, winches, sheet tracks) as patch loads with realistic lever arms.
Watertight bulkheads	Head from the swamped/flotation condition per ISO 12217-2 requirements.
Watch	Every one of these has a weld nearby — evaluate against the welded allowable, not base metal.

Test 8 — Buckling follow-on Critical

Why	Aluminium catamaran skins and cross-beam flanges are thin and highly stressed in compression (torsion case, rig case). Stress can be “fine” while the panel is one wave away from instability.
How in FreeCAD	FreeCAD’s GUI does not expose CalculiX’s linear buckling step (as of v1.0), so:

Run the static preload case normally in FreeCAD.
In the CalculiX solver task panel, click “Write .inp” and note the file path.
Edit the file and append a buckling step:

# --- appended after FreeCAD's static step (*END STEP) ---
*STEP
*BUCKLE
6
*END STEP

4. Run it from a terminal: ccx <jobname> 5. Open the resulting .frd in FreeCAD (File → Import) or convert with ccx2paraview for 3D-mode animations.

Run for	Top/bottom bridgedeck skins under Test 3; cross-beam flanges under Test 4; bottom shell under Test 6 hogging.
Pass	Lowest eigenvalue (load multiplier) λ ≥ 2 on the ISO factored loads (ISO 12215-6 / high-speed-craft practice typically demands 1.5–3 with imperfection knock-downs).

Test 9 — Modal analysis (model validation) Valuable

Why	A frequency run costs nothing and validates the whole model: global torsion and bending modes reflect the true stiffness/mass distribution.
How	FEM → Frequency analysis in the GUI (CalculiX *FREQUENCY). Compare the first torsion and bending modes with a hammer-tap test on the real boat (or a sister design): agreement within ~10–15 % confirms the model. Bonus: screen against engine and prop excitation.

6 · Acceptance criteria — what “strong enough” means

Stress allowables (verify against ISO 12215-2 and your certs)

Location	5083-H321	6082-T6	Basis
Mid-panel, unwelded	σ ≤ ≈130–150 MPa	σ ≤ ≈150 MPa	≈0.6 × Rp_0.2
Weld toes / HAZ (within ~25 mm of a weld)	σ ≤ ≈100–130 MPa	σ ≤ ≈80–110 MPa	HAZ design strength
Support/point-load artifacts	Ignore stress within ~2 elements of fully-fixed supports — use far-field values.

Deflection, stability, fatigue

Check	Target	Comment
Plate panel deflection (slam)	w ≤ ~1.5 % of short span	Serviceability; plating is usually stress-governed
Stiffener / frame tip deflection	δ ≤ L/100	Under slam pressure
Linear buckling factor	λ ≥ 2	On ISO-factored loads (Tests 3, 4, 8)
Fatigue at welds (qualitative)	Δσ_nominal vs IIW FAT class (alu fillet details typically FAT 22–32)	Slam zones see 10⁶⁺ cycles; detail design + weld quality matter more than the number

Key concept — the HAZ: welding locally softens heat-treatable and strain-hardened aluminium. Your FEM mesh cannot see the heat-affected zone, so you handle it with the reduced allowable at weld locations — this is usually the governing acceptance number on a welded alloy boat, not the shiny base-metal mid-panel stress.

7 · FreeCAD-specific tips & pitfalls

Non-conforming shell meshes: if you skip BooleanFragments + Refine, adjacent plates get duplicate nodes and the boat falls apart in the solver with no error message.
Shell thickness: one material card per thickness; confirm *SHELL SECTION in the .inp.
Units: pressures in N/mm² (0.05 = 50 kN/m²), forces in N, gravity 9810 mm/s².
Second-order elements always; size ≈2–3× thickness at hot spots.
Over-constraint: don’t clamp entire cut planes in local models — model one extra frame bay beyond the region of interest.
No symmetry on torsion or diagonal cases.
Superposition: results are linear — run cases separately and combine (e.g. bridgedeck slam + rig compression) rather than building monstrous combined models.
Post-processing: plot von Mises with a clip plane to see inside; use the Displacement plot to verify load signs; archive .inp/.frd for the technical file.
Advanced: plastic collapse margin via a manually added *PLASTIC card in the .inp (research-grade; beyond CE needs).

Scripting a load sweep (Python console)

# FreeCAD Python console — quick post-processing after a solve
res = [o for o in App.ActiveDocument.Objects
       if o.isDerivedFrom("Fem::FemResultObject")][-1]
print("max displacement:", max(res.DisplacementLengths), "mm")
print([p for p in res.PropertiesList if "Stress" in p])  # explore available fields

Property and object names vary slightly between FreeCAD versions — the shipped FemExamples workbench is the best reference for scripted analyses.

8 · Anchor the FEM with hand calculations

Before trusting any big model, verify one panel bay by hand. Clamped rectangular plate under uniform pressure:

sigma_max = beta * p * (s / t)^2        # s = short span, t = thickness
beta: 0.31 (square, clamped) ... 0.50 (long, clamped) ... 0.75 (long, simply supported)

Worked example (Test 1 bay): p = 0.05 MPa, s = 350 mm, t = 5 mm → s/t = 70 → σ ≈ 110–180 MPa depending on edge fixity; w ≈ 2–6 mm. Your full-model FEM must land between these bounds — if it doesn’t, fix the mesh or constraints before believing anything else.

Stiffener: q = p × stiffener spacing; M ≈ q·L²/10; σ = M/Z with the effective flange width.
Torque check: support reactions from Test 3 must reproduce Δ·g and the design torque.
Convergence: refine the critical bay 2× — accept Δσ ≤ 5 %.
Modal correlation (Test 9) as the final model-validation gate.

9 · Limits of FEM, and the technical file

What FEM cannot certify for you

Stability, freeboard, buoyancy & swamped-recovery → ISO 12217 (physical inclining test).
Glazing performance → ISO 12216 test pressures.
Watertightness → hose tests per the RCD essential requirements.
Weld quality itself → welding procedure qualifications (e.g. ISO 15614-2), NDT plan, filler certs.
Hydrodynamic slam physics — FEM verifies against ISO-derived loads; it does not predict them.

Technical Construction File checklist

Load-calculation spreadsheet per EN ISO 12215-5/-7/-9/-10 (the source of every number you entered).
FEM reports per case: model description, mesh + convergence study, BCs, load maps, von Mises and displacement plots, buckling eigenvalues, and the raw .inp/.frd archive.
Scantling drawings; material certificates (EN 485 plates, EN 755 extrusions); weld procedures & NDT records.
Declaration of Conformity, HIN, owner’s manual (including safe lifting/blocking points per Test 6).
Independent review by a naval architect — mandatory with a Notified Body for hulls of 12–24 m, and simply good engineering for any Category A ocean-going catamaran.

Buy the standards. The pressure and force formulas behind every illustrative number here live in EN ISO 12215-5, -6, -7, -9, -10 and ISO 12216 (available from ISO or your national body). They are the contractual source of truth for Category A.