Why Fatigue-Critical Components Should Have Fewer Welds
The fundamental principle governing fatigue design for welded structures is deceptively simple: the best weld for a fatigue-critical component is the one you do not have to make. Every welded joint introduces a stress concentration, a zone of residual tension, a potential imperfection, and a future inspection obligation. When a component operates under cyclic loading, these combined effects reduce structural life far more severely than most designers appreciate at the concept stage — and no amount of improved welding procedure, NDT, or post-weld treatment can fully compensate for a design that unnecessarily multiplies weld count.
This principle is not a conservative opinion or a fabricator’s preference. It is a consistent, explicit requirement of every major international fatigue design code: BS 7608, Eurocode 3 (EN 1993-1-9), DNV RP C203, ASME Section VIII Division 2, ASME B31.3, API RP 2A, API RP 579, and the IIW Recommendations for Fatigue Design all state in various forms that unnecessary welds should be avoided in structures subject to cyclic loading. The codes also establish — without exception — that welded joints occupy lower fatigue classes than unwelded parent material regardless of how well the weld is made, regardless of filler metal strength, and regardless of base material grade.
This article examines the engineering and physical reasons behind this universal code requirement, the mechanism by which weld count translates directly into fatigue risk, what the codes actually say about weld minimisation, the sustainability case for fewer welds, and the practical design strategies that allow engineers to reduce weld count without compromising structural performance. The material builds on the technical background in the residual stress and fatigue life guide on WeldFabWorld and applies it to the design decision-making context.
Why Every Weld is a Fatigue Liability
To understand why weld minimisation is so central to fatigue design, it is necessary to understand what welding does to a structural member at the locations it joins. Three distinct degradation mechanisms operate simultaneously at every weld, and their combined effect is to reduce fatigue endurance by a factor that can range from two to ten compared with unwelded parent material at the same nominal stress level.
Geometric Stress Concentration at the Weld Toe
The weld toe — the junction between the weld reinforcement and the base plate surface — creates an abrupt geometric discontinuity. The local stress at this notch is higher than the nominal applied stress in the plate by a factor Kt (the stress concentration factor). For typical as-welded toe geometries with a toe angle of 40 to 60 degrees and a toe radius of 0.1 to 0.5 mm, Kt values under transverse loading range from 2 to 5. This means that while the plate might be designed for a nominal stress range of 100 MPa, the local stress range at the weld toe is 200 to 500 MPa — accelerating crack nucleation and dramatically shortening the initiation phase of the fatigue damage process. Even with a smooth, well-executed weld, the geometric notch cannot be eliminated without post-weld toe treatment such as grinding or TIG toe dressing.
Tensile Residual Stress Approaching Yield Strength
The welding thermal cycle — intense localised heating followed by constrained cooling — produces a self-equilibrating residual stress field in which the weld metal and HAZ carry tensile residual stress while the surrounding base material carries balancing compressive stress. In structural steels, the peak tensile residual stress at the weld commonly approaches the room-temperature yield strength of the material: 250 to 400 MPa in S275 to S355 steels, and proportionally higher in higher-strength grades. This residual tension acts as a permanent mean stress superimposed on the cyclic applied loading. It elevates the effective stress ratio R at the crack tip, reduces the crack closure effect, and increases the effective driving force for crack growth — even when the applied load is nominally at zero. The full technical mechanics of this process are covered in the residual stress in welded joints article.
Weld Imperfections as Pre-Existing Crack-Like Defects
No weld is geometrically or metallurgically perfect. Every production weld contains imperfections of varying severity: micro-slag inclusions, porosity, incomplete fusion at the root or sidewall, cold laps at the toe, HAZ hydrogen cracking at the fusion boundary, or surface undercut along the weld toe. In fatigue assessment, these imperfections are treated as pre-existing crack-like defects from which propagation can begin without any initiation period. The practical consequence is that fatigue life in welded joints is dominated by the crack propagation phase alone, whereas in unwelded material there is a substantial initiation phase that contributes to total endurance. This elimination of the initiation phase — because imperfections effectively provide a zero-length starting crack — is a major reason why welded joints have dramatically shorter fatigue lives than polished unwelded specimens at the same applied stress range.
The Fatigue Class System — How Welds are Penalised by Design Codes
The fatigue class (or detail category) system used by all major codes quantifies the fatigue penalty imposed by different joint types. Understanding where welds sit in this hierarchy makes the design imperative for weld minimisation concrete rather than abstract.
What a Fatigue Class Means
In BS 7608 and EN 1993-1-9, each weld detail is assigned to a fatigue class (for example, Class B through to Class W in BS 7608, or FAT 160 through FAT 36 in the IIW/Eurocode system). The fatigue class number represents the stress range in MPa at which the detail achieves a fatigue life of 2 million cycles. A higher number indicates better fatigue performance. The fatigue life at other stress ranges is calculated using the S-N curve slope, typically m = 3 for structural details.
This calculation makes the code’s message concrete: a fillet-welded attachment on a plate member reduces the achievable fatigue life at the same applied stress to less than one tenth of the unwelded plate’s performance. Every attachment weld, gusset, bracket, or stiffener welded to a primary member creates this type of fatigue class downgrade at that location. Multiplying such details across a component multiplies the potential failure sites proportionally.
The Fatigue Class Hierarchy for Common Joint Types
| Detail Type | FAT Class (IIW) | BS 7608 Class | Relative Fatigue Life at 100 MPa | Notes |
|---|---|---|---|---|
| Unwelded parent plate (rolled surface) | 160 | B | 100% (baseline) | No weld, no degradation |
| Full-penetration butt weld, flush ground | 125 | C | 61% | Residual stress remains despite grinding |
| Full-penetration butt weld, as-welded | 112 | D | 34% | Toe geometry unremediated |
| Transverse fillet weld on main member | 80 | F | 10% | Severe stress concentration at toe |
| Longitudinal attachment, fillet welded | 71 | F2 | 8.7% | Common gusset / bracket detail |
| Short stiffener, fillet welded, on flange | 63 | G | 6.3% | End of stiffener is critical initiation point |
| Cover plate end, fillet weld on flange | 50 | W | 3.1% | One of lowest classified details |
| Cruciform joint, partial penetration fillet | 36 | W / Class X | 1.2% | Root crack risk; avoid in fatigue design |
The table demonstrates that a single poorly positioned fillet weld — a cover plate end, a transverse stiffener, a welded bracket — can reduce the local fatigue life of a primary structural member to 3 to 10% of its unwelded potential. When a design contains many such details, fatigue failure becomes a question of when, not whether.
What the International Codes Say About Weld Minimisation
The following section presents the specific positions of major international fatigue design and structural codes on the question of weld minimisation. All codes are consistent: fewer welds in fatigue-critical zones is a fundamental design requirement, not a secondary consideration.
BS 7608 — Guide to Fatigue Design and Assessment of Steel Products
Welded joints always occupy lower fatigue classes than parent material. Unnecessary welded details — attachments, stiffeners, and changes of section achieved by welding — should be avoided in fatigue-critical design. Fewer welds directly means fewer fatigue initiation sites.
BS 7608 is the primary UK standard for fatigue design and assessment of steel structures. Its classification system assigns fatigue classes from B (parent plate, highest) through to W (worst welded details), with no overlap between parent material and any welded detail class. The standard’s design guidance explicitly states that avoidance of unnecessary welded details is the first and most effective measure for improving fatigue performance. It identifies welded attachment ends, cope holes, and stiffener toes as the most common fatigue initiation sites in steel bridges and offshore structures, noting that most of these details could have been avoided through more considered design at the concept stage.
Eurocode 3 — EN 1993-1-9 Fatigue
Designers shall avoid unnecessary welded joints in fatigue-loaded structures, minimise welded attachments in stress concentration zones, and ensure continuity of load-carrying members to avoid the fatigue penalty of joints and discontinuities.
Eurocode 3 Part 1-9 governs fatigue design for steel structures throughout Europe. Its detail category system (FAT classes) closely parallels BS 7608 in assigning lower fatigue resistance to every welded joint compared to parent material. The normative design guidance requires that designers avoid unnecessary welded joints and eliminate abrupt geometry changes introduced by welding — because even a small gusset or bracket welded to a high-stress region of a primary member can reduce its design fatigue life to a fraction of its potential. The Eurocode philosophy of “continuous members” — using rolled or extruded profiles to span multiple connection points rather than built-up welded sections — is directly motivated by this fatigue class penalty.
DNV RP C203 — Fatigue Design of Offshore Steel Structures
Fatigue cracks in offshore steel structures almost invariably initiate at weld toes or weld roots. The designer should reduce the number of welded details and simplify structural geometry to minimise the number of fatigue hotspots requiring assessment, monitoring, and inspection throughout the structure’s service life.
DNV RP C203 is the dominant fatigue design standard for offshore fixed and floating structures globally. It employs both the nominal stress approach (with S-N classes equivalent to those in BS 7608/Eurocode 3) and the hot-spot stress (structural stress) approach. The standard contains extensive guidance on reducing hotspot stress concentrations through geometry optimisation — smooth radius transitions, avoided attachments in high-stress regions, minimised welded reinforcing details. Its practical guidance on inspection planning makes the weld-count relationship to inspection cost explicit: every welded detail classified as a fatigue hotspot requires defined inspection intervals throughout the structure’s 20 to 30-year design life, and reducing the weld count in the design reduces this inspection burden proportionally. For submerged arc welded structures such as offshore jacket legs and risers, this is a significant operational cost driver.
DNV ST F101 — Submarine Pipeline Systems
Welds and welded attachments are the dominant fatigue-critical locations in submarine pipeline systems. Limiting the number of welds and welded fittings in dynamically loaded zones reduces fatigue sensitivity and the scope of required in-service inspection.
For submarine pipelines subject to hydrodynamic and thermal fatigue loading, the weld at each girth joint represents the lowest fatigue class in the system. Reducing the number of girth welds through the use of longer pipe sections, minimising mechanical fittings and tees in high-dynamic-loading zones, and avoiding any unnecessary couplings or weld overlays directly improves the system’s fatigue performance and reduces the inspection scope mandated by the standard.
AWS D1.1 — Structural Welding Code (Steel)
Weld toes and weld roots are inherent stress raisers in any welded joint. Weld imperfections — undercut, porosity, lack of fusion — increase fatigue sensitivity. Reducing the number of welded connections in the design improves fatigue life and simplifies inspection reliability assessment throughout the structure’s service life.
AWS D1.1 is the primary structural welding code for steel construction in North America. Its fatigue provisions and commentary reflect the same joint classification approach as BS 7608 and Eurocode 3. The commentary’s fatigue annex explains in practical terms why weld toe geometry, root condition, and residual stress govern fatigue performance independently of weld quality — motivating the design-first approach to weld minimisation rather than relying on weld quality improvement alone to achieve fatigue targets.
ASME Section VIII Division 2 and ASME B31.3
Welds and welded attachments introduce local stress concentration factors in pressure vessel and piping design. Unnecessary welds and attachments in regions subject to cyclic loading should be avoided. Every weld not made is one less location requiring fatigue assessment and future inspection under the code’s periodic examination requirements.
ASME Section VIII Division 2 uses a smooth-bar S-N curve approach combined with total stress concentration factors for fatigue assessment of pressure vessels. The code’s fatigue section explicitly penalises welded joints through elevated stress concentration factors compared to parent material, and its design by analysis methodology requires that all welds in cyclic service be individually assessed and their fatigue lives demonstrated. For pressure vessels under ASME VIII Division 2, the mechanical testing requirements and weld quality provisions exist in part to minimise the defect population that degrades fatigue life. Similarly, ASME B31.3 process piping code treatment of cyclic loading highlights weld attachments as stress concentration locations that should be avoided in piping systems with high cycle fatigue exposure.
API RP 2A and API RP 579
Welded tubular joints and welded attachments are the primary fatigue concern in offshore platform structures. Reducing the quantity and complexity of welded joints reduces the number of fatigue hotspot categories, lowers the analytical and inspection burden, and improves structural reliability over the platform’s service life.
API RP 2A governs the design of fixed offshore platforms used predominantly in the Gulf of Mexico. Its wave load fatigue analysis requires hotspot stress calculation at every welded tubular joint, with fatigue lives assessed against S-N curves analogous to those in DNV RP C203. The standard’s design guidance notes that platform reliability is strongly influenced by the number and configuration of welded connections, and that rationalising connection geometry — reducing the number of members framing into a node, avoiding eccentric overlapping joints, minimising welded stiffening details — produces disproportionate improvements in calculated fatigue life. API RP 579 fitness-for-service assessment framework notes that welds are the dominant crack initiation points in in-service pressure equipment, and that reducing weld count in the initial design reduces the number of locations requiring fitness-for-service evaluation and repair over the equipment’s life.
The Sustainability Case for Fewer Welds
Beyond the structural integrity argument, weld minimisation is increasingly framed in the language of sustainable fabrication — a lens that modern engineering practice, client specifications, and ISO 3834 quality management are applying more consistently across the industry.
Energy and Material Consumption
Every weld pass consumes electrical energy for arc heating, filler metal and flux as consumables, shielding gas, and preheat energy where required by the carbon equivalent of the base material. For a complex welded assembly with a high weld volume, the direct energy cost of welding can represent a significant fraction of the component’s total manufacturing energy. Weld volume — measured in kilograms of deposited metal — is a direct proxy for this energy consumption. A design that achieves the same structural function with half the weld volume uses half the consumables, generates half the fume requiring extraction, and produces less distortion requiring subsequent correction.
Quality Risk and Rework Multiplier
Each additional weld in a component is an additional opportunity for a weld defect — a repair, a re-test, a delay. In practice, defect occurrence probability is roughly constant per unit length of weld. A design with twice the weld length therefore produces approximately twice the expected defect count, twice the repair rate, and twice the inspection cost per production unit. Because repair welding typically requires twice the energy and twice the time of the original weld (preparation, welding, PWHT if required, re-inspection), the rework multiplier on total production cost from excessive weld count can be substantial. Design-stage weld minimisation eliminates this quality risk before it enters the production system.
Carbon Footprint and Lifecycle Assessment
Lifecycle assessment (LCA) of fabricated steel structures increasingly accounts for the energy embedded in the manufacturing process, not just the material. A component with a large weld volume has a higher fabrication carbon footprint than a functionally equivalent component using formed or rolled sections with fewer joins. For clients with net-zero or low-carbon supply chain commitments, weld volume is a design metric, and engineers who can demonstrate weld minimisation through design optimisation contribute directly to the project’s sustainability performance. The ISO 3834 welding quality management framework, while not directly addressing sustainability, provides the quality system infrastructure within which weld volume tracking and optimisation can be formalised.
Practical Design Strategies to Minimise Weld Count
Weld minimisation is not a theoretical aspiration — it is achievable through specific design choices made at the concept and detail design stages. The following strategies are applicable across structural, pressure vessel, and mechanical component design contexts.
Use Rolled and Extruded Profiles Rather than Built-Up Sections
The most effective single strategy for eliminating welds in steel structures is to use hot-rolled I-sections, hollow sections (CHS, RHS), or extruded profiles in place of built-up plate girders wherever the profile range permits. A hot-rolled UB or UC section contains no welds whatsoever; a plate girder of equivalent moment capacity requires at minimum two continuous longitudinal flange-to-web fillet welds totalling twice the member length. In fatigue design terms, the plate girder carries all the penalty of Classes F and F2 along its full length; the rolled section carries parent material Class B performance throughout. Where plates must be used for large or unusual sections, minimising stiffener frequency and concentrating stiffeners in low-stress zones reduces fatigue hotspot count substantially.
Use Cast or Forged Nodes for Complex Connections
In offshore tubular structures, crane pedestals, and complex mechanical linkages, welded tubular joints (K, T, Y, X nodes) are the dominant fatigue categories. Cast or forged nodes replace multiple high-stress weld toes with a single smooth continuous surface, eliminating the tubular joint detail entirely and replacing it with a butt weld between the cast node and the tubular member — a much higher fatigue class. While cast node costs are higher than fabricated alternatives, the inspection savings and fatigue life improvement over a 25-year offshore platform life cycle generally justify the premium for the most fatigue-critical nodes.
Design Continuous Members Through Connection Points
Where a member must pass through or alongside a connection, designing it as a continuous element rather than two butt-welded segments eliminates a full-penetration butt weld (FAT 112 as-welded) from the highest-stress point in the member. This principle applies to deck plate continuity through transverse stiffeners in ship hull girders, to chord continuity through gusset plate connections in trusses, and to beam continuity at intermediate support points in continuous bridge girders. In each case, continuity reduces weld count at the stress concentration point where fatigue life is most sensitive.
Relocate Unavoidable Welds to Low-Stress Regions
Where welds cannot be eliminated, positioning them in regions of low nominal stress dramatically increases their fatigue life. A butt weld in the tension flange at midspan of a beam is subjected to the maximum bending stress — the worst possible location for a FAT 112 detail. Moving the same butt weld to a location where the nominal stress range is reduced by 50% increases its calculated fatigue life by a factor of 8 (because life scales as stress range to the power of 3). Design rules for bridge girder splices specifically exploit this principle by requiring splices to be located away from maximum moment zones wherever structurally feasible.
Avoid Welded Attachments in High-Stress Zones
Brackets, lifting lugs, drainage holes, instrument connections, and other service attachments welded to primary structural members in high-stress regions represent some of the most common sources of in-service fatigue cracking. A welded attachment that appears minor from a fabrication standpoint creates a FAT 63 to FAT 50 detail at that location — potentially reducing the fatigue life of the primary member to less than 10% of its unwelded potential. Wherever possible, attachments should be bolted rather than welded, relocated to low-stress regions, or redesigned as integral features of the primary section. In the offshore industry, this principle motivates the practice of “clean deck” and “clean structure” detailing — minimising all attachments in fatigue-sensitive members. For detailed guidance on joint geometry and stress concentration factors, the WeldFabWorld joint types guide provides the relevant design context.
Standards Summary Table — Weld Minimisation Requirements
The following table consolidates the positions of all major international fatigue and structural codes on weld minimisation in fatigue-critical design. This table preserves and substantially expands the reference table from the original source article.
| Standard | Full Title | Weld Minimisation Position | Applicable Sector |
|---|---|---|---|
| BS 7608 | Guide to Fatigue Design and Assessment of Steel Products | States that welded joints always have lower fatigue classes than parent material; unnecessary welded details should be avoided to reduce fatigue initiation sites. Higher-strength steel provides no fatigue benefit in welded joints. | Structural / Offshore |
| EN 1993-1-9 | Eurocode 3 — Fatigue | Designers shall avoid unnecessary welded joints and minimise attachments in fatigue-loaded structures. Continuous members without welded interruptions always provide better fatigue performance. Eliminate abrupt stiffness changes introduced by welding. | Structural / Bridges |
| DNV RP C203 | Fatigue Design of Offshore Steel Structures | Fatigue cracks almost always initiate at weld toes or weld roots. Reduce the number of welded details and simplify geometry to minimise fatigue hotspot count, inspection scope, and lifecycle monitoring burden. | Offshore / Marine |
| DNV ST F101 | Submarine Pipeline Systems | Welds and welded attachments are dominant fatigue-critical locations in dynamic pipeline systems. Limiting weld count in dynamic loading zones reduces fatigue sensitivity and in-service inspection scope. | Subsea / Pipeline |
| AWS D1.1 | Structural Welding Code — Steel | Weld toes and roots act as inherent stress raisers. Reducing welded connections in design improves fatigue life and inspection reliability. Weld imperfections increase fatigue sensitivity independent of weld class. | Structural / Industrial |
| ASME BPVC VIII Div. 2 | Pressure Vessels — Design by Analysis | Welds introduce local stress concentration factors. Unnecessary welds and attachments should be avoided in regions subject to cyclic loading. Each weld requires individual fatigue assessment and periodic inspection over the vessel’s life. | Pressure Vessels |
| ASME B31.3 | Process Piping | Welds and welded attachments contribute to local stress concentration in piping under cyclic loading. Limiting welds and attachments where fatigue governs design reduces fatigue sensitivity and inspection obligations. | Process Piping |
| API RP 2A | Fixed Offshore Platforms | Welded tubular joints and attachments are the primary fatigue concerns in offshore platforms. Reducing weld quantity reduces fatigue demand, the number of hotspot categories, and the long-term inspection burden over a 20 to 30-year platform life. | Offshore / Oil & Gas |
| API RP 579 | Fitness-for-Service | Welds are common fatigue crack initiation points in pressure equipment. Reducing weld count in design reduces the number of locations requiring fitness-for-service evaluation, crack growth assessment, and repair over the equipment’s service life. | Pressure Equipment |
| IIW Recommendations | Fatigue Design of Welded Joints and Components | Fatigue design curves (FAT classes) are assigned by joint detail with no benefit for higher-strength steel. Unwelded parent material consistently outperforms all welded details. Weld minimisation is the most cost-effective design action for fatigue improvement. | All Sectors |
Recommended Reading on Fatigue Design and Structural Integrity
Disclosure: WeldFabWorld participates in the Amazon Associates programme (StoreID: neha0fe8-21). If you purchase through these links, we may earn a small commission at no extra cost to you. This helps support free technical content on this site.