EN 1993-1-8 Explained

Section 1

What EN 1993-1-8 Covers and When It Applies

EN 1993-1-8 is Part 1-8 of the Eurocode 3 series (Design of steel structures), first published in 2005 with Amendment AC:2009. It is the definitive normative standard for the design of joints in steel buildings and civil engineering works across EU member states and in countries that have adopted the Eurocode suite. It covers three broad families of joints:

Bolted connections in shear (§3.6) — bearing-type connections where bolts resist transverse loads through shear and bearing. Applies to fin plates, angle cleats, shear tabs, and end plates under shear-only loading.
Bolted connections in tension (§3.4, §6.2.4) — connections where bolts resist axial tension, including the T-stub model for prying effects in end plates and column flanges.
Welded connections (§4) — fillet welds and butt welds, designed using the directional method or the simplified method.
Base plates (§6.2.8) — bearing pressure on concrete, anchor bolt tension, and weld to column.

The standard works in conjunction with EN 1993-1-1 (general rules for structural analysis and member design) and EN 1993-1-5 (plated structural elements). EN 1993-1-1 provides the member checks and section classification; EN 1993-1-8 provides the joint-level checks. The boundary between the two is deliberate: member capacity is checked under EN 1993-1-1, and the joint's moment capacity M_j,Rd is computed using the component method in EN 1993-1-8 §6 before being passed back to the structural model as a boundary condition.

The central intellectual contribution of EN 1993-1-8 is the component method (§6), which replaced the prescriptive tables of the 1992 ENV version with a physically consistent mechanical model. Any geometry — including non-standard connections that previously fell outside the scope of tables — can now be analysed with confidence.

Section 2

The Component Method — EN 1993-1-8 §6

The component method (§6.1) is the analytical framework used to calculate the moment resistance M_j,Rd and initial rotational stiffness S_j,ini of a beam-to-column joint. It works by:

Decomposing the joint into discrete mechanical components — each bolt, each plate in bending, each weld, the column web panel in shear.
Assigning each component a stiffness coefficient k_i (for S_j,ini) and a strength F_Rd (for M_j,Rd).
Assembling all components in series and parallel to give an equivalent spring network representing the joint's moment-rotation response.

The key insight is that the joint behaves like a system of springs: the stiffest component controls the deformation, and the weakest component controls the strength. By modelling each part separately, the engineer can identify exactly which component governs and redesign accordingly — without changing the global structural model.

The component method replaced the prescriptive table approach of ENV 1993-1-1:1992, which listed a fixed set of standard joint configurations and gave a moment capacity for each. That approach could not handle extended end plates, partial-depth plates, haunches, or any other non-standard geometry. The component method has no such limitation.

The output of the component method is two numbers: M_j,Rd (design moment resistance) and S_j,ini (initial rotational stiffness). These are used to classify the joint as rigid, semi-rigid, or nominally pinned using the boundaries in §5.2.2, which depend on frame type (sway vs. non-sway, braced vs. unbraced).

Key components in a typical end-plate moment connection

Column web panel in shear — §6.2.7.1, k_w = 0.9 for web in shear.
Column flange in bending — §6.2.6, T-stub effective length from Table 6.3.
End plate in bending — §6.2.4, T-stub with Mode 1/2/3 failure modes.
Bolt in tension — §3.4, F_t,Rd = k₂·f_ub·A_s/γ_M2.
Bolt in shear — §3.6, F_v,Rd = α_v·f_ub·A_s/γ_M2 (threaded, shear plane through thread).
Fillet weld — §4.5.3.3, F_w,Rd = a·f_u/(√3·β_w·γ_M2).

Section 3

Bolted Connections in Shear — EN 1993-1-8 §3.6

Bolted connections in shear resist transverse forces through two mechanisms: bolt shear (the bolt shank resists transverse load in single or double shear) and bolt bearing (the bolt bears against the plate, deforming it locally). EN 1993-1-8 Table 3.4 gives the design resistance for both, expressed as forces per bolt.

Shear resistance — F_v,Rd (§3.6.1, Table 3.4)

For a bolt in single shear through the threaded portion (conservative, used unless the shear plane is specifically known to pass through the shank):

F_v,Rd — Design shear resistance per bolt EN 1993-1-8 Table 3.4

F_v,Rd = α_v · f_ub · A_s / γ_M2
// α_v = 0.6 for grade 8.8 bolts (shear through thread); f_ub in N/mm²; A_s in mm²

Bearing resistance — F_b,Rd (§3.6.1, Table 3.4)

Bearing is direction-dependent: k₁ is the coefficient perpendicular to the load direction (depends on edge distance e₂), and α_b is the coefficient in the load direction (depends on end distance e₁ and the ratio f_ub/f_u):

α_b — bearing coefficient in load direction EN 1993-1-8 Table 3.4

α_d = e₁ / (3 · d₀) (end distance e₁, hole dia. d₀)
α_b = min(α_d, f_ub/f_u, 1.0)

k₁ — bearing coefficient perpendicular to load EN 1993-1-8 Table 3.4

// Edge bolt row: e₂ is measured to the plate edge
k₁ = min(2.8 · e₂/d₀ − 1.7, 2.5) // for edge bolt row

F_b,Rd — Design bearing resistance per bolt EN 1993-1-8 Table 3.4

F_b,Rd = k₁ · α_b · f_u · d · t / γ_M2
// f_u = ultimate strength of the plate material; d = bolt diameter; t = plate thickness

End and edge distances — Table 3.3

Minimum end and edge distances (§3.6.1, Table 3.3) are set to prevent premature splitting and to ensure adequate bearing. For S355 steel, the minimum end distance e₁ ≥ 1.5·d₀ and edge distance e₂ ≥ 1.5·d₀. Recommended values for normal structural use are e₁ = 2.0·d₀ and e₂ = 1.5·d₀. The actual distances used in the worked example below — e₁ = 55 mm and e₂ = 50 mm for M24 bolts in a 12 mm S355 plate — exceed these minima comfortably.

Section 4

Bolt Rows in Tension and the T-stub Model — EN 1993-1-8 §3.4, §6.2.4

Under moment loading at a beam-to-column joint, the top bolt row is in tension. The end plate bends away from the column flange under this tension, generating prying forces that add to the direct bolt tension. EN 1993-1-8 §6.2.4 models this as a T-stub: the flange of the T is the plate between the bolt line and the weld toe; the stem of the T is the bolt. Three failure modes are evaluated (Table 6.2):

Mode 1 — Complete yielding of the plate flanges. The bolts snap when the plastic hinge mechanism forms. Maximum prying. Governing equation: F_T,1,Rd = 4·M_pl,Rd / m.
Mode 2 — Bolt failure concurrent with partial yielding of the plate. Prying develops partially. Governing equation: F_T,2,Rd = (2·M_pl,Rd + n·ΣF_t,Rd) / (m + n).
Mode 3 — Bolt failure with no plate yielding. No prying. Governing equation: F_T,3,Rd = ΣF_t,Rd.

The T-stub effective length l_eff (§6.2.4, Table 6.6) is the width of plate that participates in the yield line mechanism. For a bolt row in an end plate, it depends on the bolt layout and whether the pattern is circular or non-circular. For the bolt pitch p = 90 mm and a non-circular pattern, the effective length equals the bolt pitch — this is the value used in the worked example below.

Key geometric parameters

m — Distance from bolt centreline to weld toe on the beam flange (mm). Controls the lever arm of the plastic hinge.
n — Effective edge distance, n = min(e₂, 1.25·m). The 1.25 multiplier accounts for the actual stress distribution at a plate edge.
M_pl,Rd — Plastic moment capacity of the plate per unit width: M_pl,Rd = 0.25·l_eff·t_ep²·f_y/γ_M0.

Section 5

Welds — EN 1993-1-8 §4

EN 1993-1-8 §4 covers the design of fillet and butt welds. For structural steel fabrication, fillet welds are almost always used for shop and site connections. The design resistance of a fillet weld is given by §4.5.3.3 (Table 4.1):

F_w,Rd — Design weld resistance per unit length EN 1993-1-8 §4.5.3.3, Table 4.1

F_w,Rd = a · f_u / (√3 · β_w · γ_M2)
// a = throat thickness (mm); f_u = ultimate strength of parent metal (N/mm²);
// β_w = correlation factor (0.90 for S355, Table 4.1); γ_M2 = 1.25

The directional method (§4.5.3.3) resolves the applied force vector into weld-normal (σ⊥) and weld-parallel (τ) components. Each component is checked independently against the weld resistance, and the vector sum is also checked:

Directional method — combined normal and shear EN 1993-1-8 §4.5.3.3

// σ⊥ = normal stress on throat plane; τ = shear stress parallel to weld axis
σ_vm = √(σ⊥² + τ²) ≤ f_u / (√3 · β_w · γ_M2) // equivalent stress check

For a weld group under combined loading (normal force plus shear), each individual weld leg in the group is checked separately. The most stressed weld leg governs the group capacity.

Weld leg vs. throat — a common source of error

EN 1993-1-8 uses throat thickness a, not leg length. For a fillet weld with leg size s, the throat a = s / √2. A 6 mm fillet weld (leg = 6 mm) has throat a = 6/√2 = 4.24 mm. Confusion between leg and throat leads to systematic overestimation of weld capacity by a factor of √2 ≈ 1.41 — and an unsafe design.

Section 6 — Worked Example

HEA300 + IPE400 Flush End-Plate — Full Numerical Workings

Connection geometry: IPE400 beam frames into the flange of an HEA300 column. A 12 mm S355 end plate is shop-welded to the beam with 6 mm fillet welds and site-bolted to the column flange with four M24 grade 8.8 bolts in two rows of two. Design loads: V_Ed = 120 kN, M_Ed = 45 kN·m.

HEA300 Column (S355)

h290 mm

b300 mm

t_f (flange)14.0 mm

t_w (web)8.5 mm

r (root)27 mm

f_y355 N/mm²

f_u510 N/mm²

IPE400 Beam (S355)

h400 mm

b180 mm

t_f (flange)13.5 mm

t_w (web)8.6 mm

r (root)21 mm

f_y355 N/mm²

f_u510 N/mm²

Connection Detail

BoltM24 gr 8.8

Bolt layout2 rows × 2

End plate t12 mm S355

e₁ (end dist.)55 mm

e₂ (edge dist.)50 mm

p (bolt pitch)90 mm

Hole d₀26 mm

Weld a6 mm fillet

Bolt & Material Parameters

A_s (M24 bolt tensile stress area)353 mm²

f_ub (bolt ultimate strength, gr 8.8)800 N/mm²

k₂ (for tension, Table 3.4)0.9

α_v (for shear, Table 3.4)0.6

γ_M01.0

γ_M21.25

β_w (S355, Table 4.1)0.90

Step 1 — Bolt Shear F_v,Rd EN 1993-1-8 Table 3.4

F_v,Rd = α_v · f_ub · A_s / γ_M2
// α_v = 0.6 (grade 8.8, shear through thread); f_ub = 800 N/mm²; A_s = 353 mm²; γ_M2 = 1.25
F_v,Rd = 0.6 × 800 × 353 / (1.25 × 1000) = 135.6 kN per bolt

Applied: V_Ed = 120 kN / (4 bolts) / 135.6 kN

0.42 — PASS

Step 2a — Bearing coefficients α_b and k₁ EN 1993-1-8 Table 3.4

// End distance e₁ = 55 mm, hole d₀ = 26 mm
α_d = e₁ / (3 · d₀) = 55 / (3 × 26) = 0.705
α_b = min(α_d, f_ub/f_u, 1.0) = min(0.705, 800/510, 1.0) = 0.705
// Edge distance e₂ = 50 mm
k₁ = min(2.8·e₂/d₀ − 1.7, 2.5) = min(2.8×50/26 − 1.7, 2.5) = min(3.68, 2.5) = 2.5

Step 2b — Bolt Bearing F_b,Rd EN 1993-1-8 Table 3.4

F_b,Rd = k₁ · α_b · f_u · d · t / γ_M2
// f_u = 510 N/mm² (S355 plate); d = 24 mm (bolt dia.); t = 12 mm (end plate)
F_b,Rd = 2.5 × 0.705 × 510 × 24 × 12 / (1.25 × 1000) = 208.0 kN per bolt

Applied: V_Ed / (4 bolts) / F_b,Rd

0.29 — PASS

Step 3a — T-stub geometric parameters m, n, l_eff EN 1993-1-8 §6.2.4

// m = distance from bolt centreline to weld toe on beam flange
m = e₁_plate − t_fb/2 − a_w·√2
m = 55 − 13.5/2 − 6×1.414 = 55 − 6.75 − 8.49 = 39.8 mm
e = e₂ = 50 mm (edge distance to plate edge, column flange side)
n = min(e, 1.25·m) = min(50, 1.25×39.8) = min(50, 49.7) = 49.7 mm
// Effective length for bolt row group — non-circular pattern, Table 6.6
l_eff = p = 90 mm (bolt pitch governs interior row)

Step 3b — T-stub plastic moment M_pl,Rd EN 1993-1-8 §6.2.4, Table 6.2

M_pl,Rd = 0.25 · l_eff · t_ep² · f_yep / γ_M0
// l_eff = 90 mm; t_ep = 12 mm; f_yep = 355 N/mm² (S355); γ_M0 = 1.0
M_pl,Rd = 0.25 × 90 × 144 × 355 / 1.0 = 1,154,700 N·mm

Step 3c — T-stub Mode 1/2/3 resistance EN 1993-1-8 §6.2.4, Table 6.2

// ΣF_t,Rd for 2 bolts per row = 2 × 203,700 = 407,400 N
F_T,1,Rd = 4·M_pl,Rd / m = 4 × 1,154,700 / 39.8 / 1000 = 116.1 kN Mode 1 — plate yielding
F_T,2,Rd = (2·M_pl,Rd + n·ΣF_t,Rd) / (m+n) / 1000
F_T,2,Rd = (2,309,400 + 49.7×407,400) / 89.5 / 1000 = 251.3 kN Mode 2 — partial prying
F_T,3,Rd = ΣF_t,Rd / 1000 = 407.4 kN Mode 3 — bolt failure
F_T,Rd = min(116.1, 251.3, 407.4) = 116.1 kN → Mode 1 governs

Applied: F_t,Ed (row) / F_T,Rd = 90 kN / 116.1 kN

0.78 — REVIEW

Mode 1 governs — the 12 mm end plate yields before the bolts fail. Increasing plate thickness to 15 mm raises F_T,Rd to 181 kN and shifts the governing mode to Mode 3 (bolt failure).

Step 4 — Fillet Weld Resistance F_w,Rd EN 1993-1-8 §4.5.3.3, Table 4.1

F_w,Rd = a · f_u / (√3 · β_w · γ_M2)
// a = 6 mm (throat, same as leg for this weld size); f_u = 510 N/mm² (S355);
// β_w = 0.90 (S355, Table 4.1); γ_M2 = 1.25
F_w,Rd = 6 × 510 / (1.732 × 0.90 × 1.25) = 3060 / 1.948 = 1.571 kN/mm

Weld utilisation (top flange): N_fl = 189 kN / (2 × 168 mm × 1.571)

0.36 — PASS

Results Summary

Check	Resistance (kN)	Demand	Utilisation	Status	Clause
Bolt shear F_v,Rd	4 × 135.6 = 542.4	V_Ed = 120	0.22	PASS	Table 3.4
Bolt bearing F_b,Rd	4 × 208.0 = 832.0	V_Ed = 120	0.14	PASS	Table 3.4
T-stub Mode 1 (end plate, prying)	F_T,Rd = 116.1	F_t,Ed = 90	0.78	REVIEW	§6.2.4 Table 6.2
Fillet weld (flange group)	528.4	N_fl = 189	0.36	PASS	§4.5.3.3

The T-stub check at 0.78 utilisation is the governing constraint. FrameAI flags this amber and recommends increasing the end-plate thickness from 12 mm to 15 mm, which reduces utilisation to 0.49 and shifts the failure mode from Mode 1 (plate yielding) to Mode 3 (bolt failure) — the preferred sequence for a ductile connection.

Section 7

Common Detailing Pitfalls

The following errors appear regularly in practice and in subcontractor delivery checks. Each is a direct violation of EN 1993-1-8 or the associated execution standard EN 1090-2.

Wrong bolt grade callout on the drawing — EN 1993-1-8 Table 3.1 defines f_ub for each grade. A drawing that says "M24 8.8" is insufficient without specifying the standard (EN ISO 4017) and the proof load. Using the wrong f_ub in the F_v,Rd formula makes the calculated resistance unreliable.
Missing washers under slotted holes — EN 1090-2 §8.5.3 requires flat washers (minimum 3 mm thick) under the head of a bolt in a slotted hole to distribute the bearing load. Omitting the washer increases local plate stresses and violates the bearing resistance calculation.
Confusing weld leg size with throat thickness — EN 1993-1-8 §4 uses throat a. A 6 mm fillet weld has a = 4.24 mm (not 6 mm). Designing for a = 6 mm overstates weld resistance by √2 ≈ 1.41 — unsafe and non-compliant.
Missing block tearing check for bolt groups — §3.10.2 checks the net section and the shear failure path along the bolt group perimeter. It is frequently omitted from connection calculations. For a 4-bolt group under shear, V_eff,1,Rd = 542 kN (using the net area Anet = t·(p + 2·e₂ − 3·d₀) = 12×(90+100−78) = 1344 mm²).
Using γ_M2 = 1.1 instead of 1.25 for bolt shear — §3.6.1 requires γ_M2 = 1.25 for design resistance of bolts and welds. Some older practice notes show γ_M2 = 1.1 for pre-2005 work; this no longer applies to EN 1993-1-8:2005 designs.

Section 8

How FrameAI Automates EN 1993-1-8 Connection Design

FrameAI reads a structural PDF through GPT-4o vision, extracts every beam, column, bolt, and weld mark, and runs the full EN 1993-1-8 check suite automatically. The engineer reviews the utilisation table — no manual calculation required.

GPT-4o Geometry Extraction

InputStructural PDF

ExtractsProfiles, bolts, welds, plates

Confidence flaggingLow-confidence fields highlighted

LanguagesEN, NL, DE, FR, SV

EN 1993-1-8 Check Engine

Bolt shearTable 3.4 — α_v method

Bolt bearingTable 3.4 — k₁, α_b

Prying§6.2.4 T-stub Mode 1/2/3

Weld resistance§4.5.3.3 directional method

Block tearing§3.10.2 V_eff,Rd

References

Normative References

EN 1993-1-8:2005 — Design of steel structures: Design of joints. Eurocode 3 Part 1-8. CEN, Brussels.
EN 1993-1-8:2005/AC:2009 — Corrigendum 1 to EN 1993-1-8:2005.
EN 1993-1-1:2005 — General rules for structural analysis and member design (§5.8 member buckling).
EN 1090-2:2008 — Execution of steel structures: Technical requirements for the execution of steel structures.
SCI P398 — Joints in Steel Construction: Moment-Resisting Joints to Eurocode 3. SCI, London.
EN ISO 4017 — Fasteners — Hexagon head screws — Product grades A and B (bolt proof loads).
EN 10365:2017 — Hot-rolled steel channels, I and H sections — Tolerances on shape and dimensions.

FAQ

Frequently Asked Questions

What does EN 1993-1-8 cover? ▾

EN 1993-1-8 is Part 1-8 of Eurocode 3 (Design of steel structures). It covers the design of joints and冰冷的连接 in steel structures — bolted connections in shear and tension, fillet and butt welds, and base plates. Its central methodology is the component method (§6), which decomposes a joint into individual springs (end plate, column flange, bolts, welds) and combines them to give moment resistance M_j,Rd and rotational stiffness S_j,ini. It applies to buildings and civil engineering works throughout the EU and in countries that have adopted the Eurocodes.

How does the component method for joints work? ▾

The component method (§6) models each part of a joint — the end plate, the column flange, each bolt row, each weld — as an independent spring with a defined stiffness and strength. The bolts in shear are represented by their shear resistance (F_v,Rd per bolt); the end plate in bending is represented by a plastic moment capacity (M_pl,Rd) and a yield-line pattern; the column flange in bending follows the same T-stub logic. All springs are assembled in series and parallel to give the joint's moment-curvature response. This replaced the prescriptive tables of ENV 1993-1-1 (1992) with a physically based, consistent framework that covers any geometry.

What is the T-stub model for bolted connections in tension? ▾

The T-stub model (§6.2.4) represents the part of the end plate or column flange that is in bending as a T-shaped yield-line mechanism. The stem of the T is the plate between the bolt and the weld; the flanges of the T are the tension bolts. Three failure modes are evaluated: Mode 1 — the plate yields completely and the bolts snap (full prying, lowest resistance); Mode 2 — the plate yields partially and bolts fail (partial prying); Mode 3 — bolts fail without any plate yielding (no prying). The governing resistance is the minimum of the three modes. Prying forces develop when the plate deflects and bears against the column flange or beam flange, adding to the bolt tension — this is implicit in Modes 1 and 2.

How do you calculate fillet weld resistance per EN 1993-1-8 §4? ▾

For a fillet weld of throat thickness a and length l, the design resistance per unit length is F_w,Rd = a · f_u / (√3 · β_w · γ_M2) (§4.5.3.3, Table 4.1). f_u is the nominal ultimate strength of the parent metal (510 N/mm² for S355); β_w is the correlation factor, 0.90 for S355; γ_M2 = 1.25. The directional method resolves the applied force into weld-normal (σ⊥) and weld-parallel (τ) components, checks both against f_u/(√3 · β_w · γ_M2), and also verifies the vector sum σ_vm = √(σ⊥² + τ²) ≤ f_u/(√3 · β_w · γ_M2). For a group of welds under combined loading, each individual weld leg is checked separately.

How do you design a bolted end-plate moment connection? ▾

A bolted end-plate moment connection is designed in six steps: (1) Establish geometry — section sizes, plate thickness, bolt grade and layout; (2) Bolt shear check — F_v,Rd per bolt from Table 3.4, shear per bolt from V_Ed / n_bolts; (3) Bolt bearing check — F_b,Rd per bolt, using k₁ and α_b from Table 3.4 for the end/edge distances; (4) Bolt tension + T-stub prying — F_t,Rd per bolt, then M_pl,Rd of the end plate, then Mode 1/2/3 resistances from Table 6.2; (5) Weld resistance — F_w,Rd per mm from §4.5.3.3, applied to the flange and web weld groups; (6) Summation — convert bolt-row tensions to a moment M_j,Rd using lever arms h_r from the neutral axis to each row. All checks must satisfy utilisation < 1.0 (or < 0.9 for γ_M0 = 1.0 with S355 when overridden by NA).