1. Why Weld Sizing Matters in Offshore Structures

Welds are the most inspected, most scrutinised connections in any offshore steel structure. A fillet weld that is one millimetre too thin can fail a classification review; a weld that is two millimetres too thick wastes material, increases distortion, and adds unnecessary residual stress to the joint.

The challenge is that two major codes govern weld sizing depending on your project jurisdiction. European and NORSOK-governed projects follow EN 1993-1-8 (Eurocode 3, Part 1-8: Design of joints). US-flagged and many FPSO topsides projects follow AWS D1.1 Structural Welding Code. The two codes use fundamentally different approaches to resolve weld stress, and getting them confused is one of the fastest ways to trigger a design query.

A proper weld sizing calculator needs to handle both codes, compare results, and flag electrode compatibility issues before they reach the fabrication yard. That is what this guide — and the Leide Weld Calculator — are designed to do.

2. Fillet, Butt, and Partial Penetration Welds

Fillet welds

The workhorse of structural steelwork. Fillet welds join two surfaces at roughly right angles without edge preparation. They are classified by their leg length (the visible dimension on each plate face) and their throat thickness (the shortest distance from the root to the hypotenuse face).

For an equal-leg fillet weld with leg length a, the effective throat is a / sqrt(2), approximately 0.707 * a. This relationship is fundamental to every weld sizing calculator.

Butt welds (full penetration)

Full penetration butt welds achieve complete fusion through the joint thickness. The effective throat equals the thinner plate thickness. These welds require edge preparation (V, K, X, or J grooves) and are typically more expensive to fabricate and inspect. Their strength equals the parent material, so they do not usually need a separate weld check — the connected plate governs.

Partial penetration butt welds

Where full penetration is unnecessary or impractical, partial penetration welds are used. EN 1993-1-8 treats them similarly to fillet welds for design purposes, with the effective throat measured from the actual penetration depth. These are often misunderstood — the throat depends on the welding process, position, and groove angle, not just the specified preparation depth.

Practical note Partial penetration welds in primary structural members carrying fatigue loads require specific approval from the classification society. DNV, for example, restricts their use in joints with Design Fatigue Factor (DFF) above 3. Always check the project specification before defaulting to partial penetration.

3. Effective Throat Thickness

The effective throat a is the single most important parameter in weld design. Every code uses it as the basis for stress calculation. Getting it wrong invalidates every subsequent check.

Weld type Effective throat (a) Notes
Equal-leg fillet leg / sqrt(2) Most common case; leg = z dimension on drawing
Unequal-leg fillet min(z1, z2) / sqrt(2) Conservative; some codes allow geometric throat
Deep penetration fillet a + penetration Only if verified by WPS/PQR qualification
Full penetration butt min(t1, t2) Throat = thinner plate thickness
Partial penetration butt actual penetration depth Depends on groove angle and process
EN 1993-1-8 clause 4.5.2 For submerged-arc welding (SAW) with groove angle >= 60 deg and leg length >= 10 mm, EN 1993-1-8 allows the throat to be increased by the depth of penetration beyond the root. This "deep penetration" credit must be verified by macro-section testing during WPS qualification.

4. EN 1993-1-8 Directional Method

The directional method (EN 1993-1-8 clause 4.5.3.2) resolves applied forces into stress components on the throat plane of the fillet weld. It is the more precise of the two Eurocode methods and generally yields smaller required weld sizes.

The throat plane is defined perpendicular to the weld axis, passing through the root. The applied stress is decomposed into:

  • sigma_perp — normal stress perpendicular to the throat
  • tau_perp — shear stress perpendicular to the weld axis (in the throat plane)
  • tau_par — shear stress parallel to the weld axis
$$\sqrt{\sigma_\perp^2 + 3\left(\tau_\perp^2 + \tau_\parallel^2\right)} \leq \frac{f_u}{\beta_w \cdot \gamma_{M2}}$$ $$\sigma_\perp \leq \frac{0.9 \, f_u}{\gamma_{M2}}$$
EN 1993-1-8 Eq. 4.1 — both criteria (combined stress and normal stress) must be satisfied simultaneously.

Both criteria must be satisfied simultaneously. The correlation factor beta_w depends on the steel grade of the parent material:

Steel grade f_u (MPa) beta_w f_vw,d (MPa) at gamma_M2 = 1.25
S235 360 0.80 207.8
S275 430 0.85 233.1
S355 510 0.90 261.2
S420 520 1.00 239.6
S460 540 1.00 249.4
Why S355 is the sweet spot S355 with beta_w = 0.90 provides the highest design weld strength per unit throat area of any common structural steel. At S420 and above, beta_w jumps to 1.00, which actually reduces the weld design strength relative to the parent material. This is why most offshore jacket and topside designs default to S355 — the weld efficiency is optimal.

Worked example: directional method

Consider a 200 mm long fillet weld with 8 mm leg, connecting an S355 bracket to a column. The weld carries a transverse force of 120 kN and a longitudinal shear of 40 kN.

$$a = \frac{8}{\sqrt{2}} = 5.66 \text{ mm}, \quad A_w = 5.66 \times 200 = 1131 \text{ mm}^2$$ $$\sigma_\perp = \frac{120 \times 10^3}{1131} \cdot \cos 45° = 75.0 \text{ MPa}$$ $$\tau_\perp = \frac{120 \times 10^3}{1131} \cdot \sin 45° = 75.0 \text{ MPa}, \quad \tau_\parallel = \frac{40 \times 10^3}{1131} = 35.4 \text{ MPa}$$ $$\sqrt{75.0^2 + 3\left(75.0^2 + 35.4^2\right)} = 163.9 \text{ MPa}$$ $$\frac{f_u}{\beta_w \cdot \gamma_{M2}} = \frac{510}{0.90 \times 1.25} = 453.3 \text{ MPa} \quad \Rightarrow \quad \frac{163.9}{453.3} = \textcolor{#2EC4B6}{\textbf{0.36 (36\% utilisation)}}$$ $$\sigma_\perp = 75.0 \leq 0.9 \times \frac{510}{1.25} = 367.2 \text{ MPa} \quad \checkmark$$
Worked example: 200 mm fillet weld, 8 mm leg, S355 bracket-to-column, transverse 120 kN + longitudinal 40 kN.

5. EN 1993-1-8 Simplified Method

The simplified method (clause 4.5.3.3) avoids decomposing forces into individual throat-plane stress components. Instead, it checks the resultant of all forces per unit length against a single design shear strength.

$$F_{w,\text{Rd}} = f_{vw,d} \cdot a$$ $$f_{vw,d} = \frac{f_u}{\sqrt{3} \cdot \beta_w \cdot \gamma_{M2}}$$ $$F_{w,\text{Ed}} \leq F_{w,\text{Rd}} \quad \text{(force per unit length)}$$
where: $F_{w,\text{Rd}}$ = design weld resistance per unit length, $f_{vw,d}$ = design shear strength of weld, $a$ = effective throat thickness.

The simplified method is always conservative compared to the directional method. It treats all loading as pure shear on the throat, regardless of the actual load direction. For transverse-loaded welds, the directional method can give up to 22% more capacity, because transverse loads are partially carried in normal stress.

When to use which method Use the simplified method for quick preliminary sizing or when load paths are complex and decomposition is uncertain. Use the directional method for final design, especially on weight-sensitive offshore structures where every millimetre of throat saves cost. The Leide Weld Calculator computes both simultaneously, so you can see the difference in utilisation immediately.

6. AWS D1.1 Weld Strength

AWS D1.1 (Structural Welding Code — Steel) takes a different approach from the Eurocode. Weld strength is based on the electrode classification number (FEXX) rather than the parent material ultimate strength.

The nominal strength of a fillet weld loaded in shear is:

$$\phi \, R_n = \phi \cdot 0.60 \cdot F_{\text{EXX}} \cdot A_w$$ $$F_w = 0.60 \cdot F_{\text{EXX}} \cdot \left(1.0 + 0.50 \cdot \sin^{1.5}\!\theta\right)$$
where: $\phi = 0.75$ (LRFD), $\Omega = 2.00$ (ASD), $F_{\text{EXX}}$ = electrode classification strength, $\theta$ = angle of loading relative to weld axis.

The critical difference: AWS D1.1 includes a directional strength increase factor of (1.0 + 0.50 * sin^1.5(theta)). A transverse-loaded fillet weld (theta = 90 deg) is 50% stronger than a longitudinally loaded one. This factor is built into the Leide calculator when AWS D1.1 is selected as the governing code.

Electrode FEXX (ksi) FEXX (MPa) Typical parent steel
E60XX 60 414 A36, A53
E70XX 70 483 A572 Gr.50, A992
E80XX 80 552 A514 (quenched & tempered)
E90XX 90 621 High-strength Q&T steels

7. EN 1993-1-8 vs AWS D1.1 Comparison

Engineers working on international offshore projects frequently need to compare results between the two codes. The table below shows design shear strength of a 1 mm throat fillet weld for common steel/electrode combinations:

Steel / electrode EN 1993-1-8 f_vw,d (MPa) AWS D1.1 phi*0.6*FEXX (MPa) Ratio EN/AWS
S355 / E70XX 261.2 217.4 1.20
S275 / E70XX 233.1 217.4 1.07
S460 / E80XX 249.4 248.4 1.00
Key takeaway For longitudinal shear, EN 1993-1-8 simplified method gives slightly higher capacity than AWS D1.1 LRFD for the same weld size. However, when the AWS directional strength increase is applied for transverse loads, AWS can give up to 25% more capacity. Always specify which code applies to avoid mismatched calculations during design verification.

8. Weld Group Analysis: Elastic Vector Method

Real connections rarely have a single straight weld. Brackets, gussets, and pad-eyes use weld groups — multiple weld lines forming a pattern that resists combined direct force, in-plane moment, and torsion. The elastic vector method (also called the instantaneous centre of rotation method for the elastic case) distributes forces to each weld element based on the group geometry.

The four standard weld patterns

The Leide calculator supports four standard weld group configurations with pre-computed section properties:

Pattern Description Typical use
Two parallel lines Two straight welds on opposite sides Beam-to-column web, bracket sides
C-shape Three-sided weld (two flanges + web) Channel or angle brackets
Rectangular (box) Four-sided continuous weld Tube-to-plate, pad-eye base
Circular Continuous ring weld Pipe-to-plate, trunnion collars

How the elastic vector method works

For each weld group, the calculator determines the centroid and computes the polar moment of inertia (J_w) of the weld pattern treating each weld line as having unit throat area. Forces and moments are resolved at the centroid, and the maximum stress at the critical weld element is found by vector addition:

$$f_{\text{direct}} = \frac{F}{\sum L_i \cdot a}$$ $$f_{\text{moment}} = \frac{M \cdot r}{J_w}$$ $$f_{\text{resultant}} = \sqrt{\left(f_{d,x} + f_{m,x}\right)^2 + \left(f_{d,y} + f_{m,y}\right)^2}$$
where: $f_{\text{direct}}$ = direct shear (uniform across group), $f_{\text{moment}}$ = moment-induced stress at distance $r$ from centroid, $J_w$ = polar moment of inertia of the weld group. Compare $f_{\text{resultant}}$ against code capacity.
Practical tip The critical weld element is not always obvious. For eccentric loads on a C-shaped weld group, the worst-case point is where the direct shear and moment-induced shear are additive — typically the far end of the longer weld line. The Leide calculator identifies this point automatically and reports the utilisation ratio at every weld segment.

9. Electrode Matching and Consumable Selection

Selecting the right welding consumable is not a fabrication decision — it is a design decision. The weld metal must have strength that matches or exceeds the parent material, but overmatching by too much introduces brittleness risk.

Parent steel Min. electrode UTS (MPa) Recommended AWS class EN ISO 2560 class
S235 / A36 360 E60XX / E70XX E 35 / E 42
S275 430 E70XX E 42
S355 / A572-50 510 E70XX / E80XX E 50
S420 520 E80XX E 55
S460 540 E80XX / E90XX E 55 / E 69
Overmatching warning For steels S420 and above, electrode overmatching of more than one strength class can cause hydrogen-assisted cracking in the heat-affected zone (HAZ). NORSOK M-101 specifically limits the weld metal overmatch to 100 MPa above the parent material specified minimum UTS. The Leide consumable database flags this automatically.

10. Lamellar Tearing Susceptibility

Lamellar tearing is a through-thickness fracture mechanism that occurs when welding pulls on a plate in the short-transverse (Z) direction. It is caused by planar inclusions (manganese sulphides) aligned with the rolling direction, and it occurs in the parent material, not in the weld itself.

EN 10164 defines through-thickness quality classes based on reduction of area (Z-value) in through-thickness tensile tests:

Quality class Min. Z-value (%) Application
Z15 15 Low restraint, plate thickness < 25 mm
Z25 25 Moderate restraint, cruciform joints
Z35 35 High restraint, thick plate T-joints, offshore nodes

The susceptibility depends on three factors: joint restraint, weld size relative to plate thickness, and sulphur content of the steel. As a rule of thumb:

  • Fillet welds with leg > 0.7 * plate thickness on restrained T-joints require Z25 minimum
  • Full penetration cruciform joints in plates > 40 mm require Z35
  • Sulphur content below 0.005% significantly reduces risk regardless of Z-class

The Leide calculator assesses lamellar tearing susceptibility based on EN 10164 and flags joints that require Z-quality steel, saving the designer from overlooking a material procurement requirement that can delay a project by weeks.

11. Common Weld Sizing Mistakes

Mistake 1: Using leg length instead of throat

The most frequent error in manual calculations. An 8 mm fillet has a throat of 5.66 mm, not 8 mm. Using the leg length directly overestimates capacity by 41%. Every weld sizing calculator must clearly distinguish between leg and throat.

Mistake 2: Ignoring beta_w for high-strength steels

For S420 and S460, beta_w = 1.00, which means the weld design strength does not proportionally increase with the parent material UTS. Engineers who interpolate beta_w linearly from S355 underestimate the penalty and produce unconservative weld sizes.

Mistake 3: Mixing code approaches

Using the AWS directional strength increase factor with EN 1993-1-8 design resistance, or applying Eurocode partial safety factors to AWS nominal strength. These are different calibrated systems — mixing them invalidates the safety margin.

Mistake 4: Neglecting minimum weld size rules

Both codes specify minimum fillet weld sizes based on the thicker plate joined. EN 1993-1-8 requires a >= 3 mm. AWS D1.1 Table 3.4 requires minimum leg sizes from 3 mm (for plates up to 6 mm) to 8 mm (for plates over 19 mm). A weld that passes the strength check can still fail the minimum size check.

Mistake 5: Forgetting effective length deductions

EN 1993-1-8 clause 4.5.1 requires that the effective length of a fillet weld is reduced by twice the throat at each end (to account for start/stop craters), unless the weld is returned around a corner. For short welds, this deduction is significant — a 50 mm weld with 6 mm throat loses 24 mm of effective length (48% reduction in capacity).

Minimum effective length EN 1993-1-8 also requires that the effective length of a fillet weld shall not be less than 30 mm or 6 times the throat thickness, whichever is larger. Welds shorter than this should be ignored in the strength calculation entirely. The Leide calculator enforces both deductions and minimum length rules automatically.

Size your welds in seconds, not spreadsheets

The Leide Weld Calculator handles EN 1993-1-8 and AWS D1.1 simultaneously — directional method, simplified method, weld group analysis, electrode matching, and lamellar tearing checks. Input your loads and geometry once; get utilisation ratios for both codes instantly.

Open Weld Calculator → More articles