Weld Residual Stress: Sources, Measurement, and Structural Integrity Assessment

Every fusion weld contains a self-equilibrating field of residual stress — stresses present in the structure without any applied external load. These stresses arise inevitably from the severe, localised thermal cycle of welding, can reach the material yield strength in tension near the weld centreline, and have profound consequences for fatigue life, stress corrosion cracking susceptibility, brittle fracture resistance, and the outcome of fitness-for-service flaw assessments. This article covers the physical origins of weld residual stress, the distributions that result in common joint geometries, all major measurement techniques from surface XRD to bulk neutron diffraction, post-weld heat treatment for stress relief, and the methodology for incorporating residual stress into structural integrity assessments under BS 7910, API 579, and the R6 procedure.

Physical Origins of Weld Residual Stress

Residual stresses in welded structures are self-equilibrating — they must satisfy equilibrium with no external load applied. This means that every tensile zone must be balanced by a corresponding compressive zone. Understanding why the weld centreline is always in tension requires tracing the thermal and metallurgical history of the weld from pool to ambient temperature.

Thermal Mismatch Mechanism

During welding, the material immediately beneath the arc or beam is heated above the liquidus and forms a weld pool; the surrounding material remains at progressively lower temperatures with increasing distance from the heat source. On heating, the hot material near the weld expands but is constrained by the cooler, stiffer surrounding material. Since the hot material cannot expand freely, it yields in compression during heating — a phenomenon called thermal upsetting. The compressively yielded material has effectively been plastically compressed.

On subsequent cooling, this plastically compressed material attempts to contract more than it was allowed to expand. Again constrained by the surrounding material, it cannot contract freely, so it is placed in tension. By the time the component returns to ambient temperature, the weld metal and near-HAZ contain tensile residual stresses in the longitudinal direction that reach or approach the ambient-temperature yield strength.

Simplified one-dimensional analysis (Masubuchi model):

During heating (at peak temperature T_peak):
  Free thermal strain: ε_free = α × (T_peak − T_ambient)
  Constraint prevents expansion → plastic compression strain: ε_p ≈ ε_free − σ_y(T_peak)/E(T_peak)

During cooling to ambient:
  Additional elastic contraction required: Δε = ε_free (not available → tensile stress)
  Residual stress: σ_res ≈ E × ε_p  ≈  σ_y(ambient)   [upper bound]

where:
  α = coefficient of thermal expansion (steel: ~12 × 10⁻⁶ /°C)
  E = Young's modulus  (steel: ~210 GPa at ambient)
  σ_y(T) = temperature-dependent yield strength (drops sharply above ~400°C)

This is why the peak residual stress in most steel welds approximates the room-temperature yield strength regardless of the exact heat input — the plastic strain accumulated during the thermal cycle is sufficient to saturate the residual stress at this level over a wide range of welding parameters.

Phase Transformation Contribution

For steels that undergo solid-state phase transformations on cooling through the iron-carbon phase diagram, the transformation strains add to or subtract from the thermally generated residual stress field. The two most important transformations in the context of weld residual stress are:

Martensite transformation: The martensite transformation involves a diffusionless shear that produces a volume expansion of approximately 2–4% relative to the austenite parent phase. When this expansion occurs at low temperature (below the martensite start temperature M_s, typically 200–450°C depending on composition), the surrounding material is already rigid and resists expansion. The result is a compressive residual stress contribution that partially offsets the tensile thermal residual stress. In high-hardenability steels with low M_s temperatures, this transformation is the basis for low-transformation-temperature (LTT) consumables specifically designed to maximise the compressive contribution.

Bainite transformation: The bainite transformation also involves a volume change but occurs at higher temperatures (typically 250–550°C) where the material is less rigid, so the compressive contribution from bainite is smaller than from martensite. Understanding this is important for bainitic HAZ microstructures in modern pipeline steels where both thermal and transformation residual stresses must be considered.

Type I, II, and III Residual Stresses

Residual stresses in materials are classified by their length scale of self-equilibration:

Structural integrity assessments (BS 7910, R6, API 579) deal exclusively with Type I residual stresses. The macro-stress field is what contributes to the crack driving force for embedded flaws in welds and is the appropriate input to fracture mechanics calculations.

Residual Stress Distributions in Common Weld Joint Geometries

Plate Butt Welds

The characteristic distribution for a plate butt weld (shown in Fig. 1) shows longitudinal residual stress peaking in tension at the weld centreline, reaching approximately σ_y (room temperature yield strength) or above in multi-pass welds. Moving laterally away from the weld centreline, the stress decays and transitions to compression at approximately 1–2 times the weld width, before returning to near-zero in the far-field base plate. The requirement for global equilibrium ensures the compressive area under the curve exactly balances the tensile area.

Through the thickness of a thick plate butt weld, the distribution is not uniform. The last pass deposited on each face thermally cycles the previously deposited material; in multi-pass welds, this creates a complex layered residual stress history. The through-thickness profile for a plate butt weld per BS 7910 Annex Q is represented as:

BS 7910 Annex Q — Longitudinal residual stress profile (plate butt weld, ferritic steel):

σ_res(z) = σ_y × [ a₀ + a₁(z/t) + a₂(z/t)² + a₃(z/t)³ ]

where z = through-thickness coordinate (z=0 at last-welded surface)
      t = plate thickness

Coefficients (upper-bound, as-welded):
  a₀ = +1.0   (tensile at surface welded last)
  a₁ = -3.4
  a₂ = +4.8
  a₃ = -2.4

Note: Profile self-balances over thickness t (membrane + bending equilibrium)
Upper-bound assumption: peak = σ_y (room temperature yield strength)

Pipe Girth Welds

Pipe girth welds (circumferential butt welds) are among the most structurally critical weld joints in oil and gas pipelines, pressure vessels, and nuclear plant. The residual stress distribution is more complex than a flat plate because the pipe hoop and axial stiffnesses are coupled, and the root pass deposits in a constrained internal geometry often inaccessible for repair. Key features of girth weld residual stress are:

The axial stress (transverse to the weld run, acting in the pipe axis direction) at the weld root is often tensile on the inner bore surface, frequently reaching yield strength. This is the crack-opening stress for an axially aligned flaw at the root — the most common fatigue and stress corrosion cracking initiation site in girth welds. The hoop stress (longitudinal, parallel to weld run) is typically tensile on the outer surface and compressive or lower-tensile on the inner bore, creating a bending distribution through the wall thickness.

T-Butt and Fillet Welds

In T-butt joints (attachment welds, nozzle connections), the transverse residual stress perpendicular to the weld run is of primary interest because it acts to open any fatigue crack initiating at the weld toe. The stress concentration at the weld toe combines geometric stress concentration (K_t typically 2–4) with the tensile residual stress field to create conditions highly susceptible to fatigue crack initiation. The residual stress at an as-welded toe is typically at or near the yield strength in tension, which is why as-welded fatigue performance is significantly inferior to the base material — and why post-weld improvement is often needed in cyclically loaded structures.

Residual Stress Measurement Techniques

No single technique is optimal for all applications. The choice of measurement method depends on the required depth of penetration, spatial resolution, specimen accessibility, whether the component can be destructively sampled, and the available measurement infrastructure.

X-Ray Diffraction (XRD)

XRD is the most widely used technique for surface residual stress measurement. It exploits Bragg’s Law to detect the shift in lattice plane spacing caused by elastic strain:

Bragg's Law:     n·λ = 2·d·sin(θ)

Strain from d-spacing shift:
  ε = (d − d₀) / d₀   =  −cot(θ₀) × Δθ   [from differentiating Bragg's Law]

Biaxial stress from strain (sin²ψ method):
  ε(ψ) = [(1+ν)/E] × σ × sin²ψ − (2ν/E) × (σ₁ + σ₂)

  Plot ε vs sin²ψ → slope gives σ directly

where:
  λ    = X-ray wavelength (CuKα: 0.154 nm; CrKα: 0.229 nm)
  d    = measured lattice spacing (strained)
  d₀   = stress-free reference lattice spacing
  θ    = Bragg angle
  ψ    = tilt angle of specimen relative to beam
  E, ν = elastic constants for the diffraction plane used
  σ₁, σ₂ = principal stresses in measurement plane

CuKα radiation penetrates approximately 5–25 μm into steel; CrKα is preferred for steel because it diffracts from the {211} planes at a high Bragg angle (approximately 156° 2θ), giving better peak resolution and lower peak shift uncertainty. The sin²ψ method requires measurements at a minimum of 5–7 tilt angles; modern portable XRD systems (e.g. PULSTEC μ-X360, Stresstech XSTRESS 3000) can complete a measurement in under 5 minutes on flat surfaces.

Key limitations: XRD is surface-only — it cannot resolve through-thickness gradients without electrochemical layer removal, which is destructive and introduces stress relaxation corrections. The coarse-grained microstructure of the weld metal (columnar austenite grains in stainless steel, large prior-austenite grains in ferritic steel HAZ) causes peak broadening and reduced measurement accuracy compared with fine-grained base material.

Synchrotron X-Ray Diffraction

High-energy synchrotron X-rays at energies of 50–150 keV achieve penetration depths of 5–50 mm in steel, allowing non-destructive through-thickness residual stress mapping of medium-thickness weldments. Synchrotron facilities (Diamond Light Source, ESRF, SPring-8, APS) provide beams that are ~10⁶–10¹² times more intense than laboratory sources, enabling gauge volumes of 0.05–1 mm³ and measurement times of seconds per point. Synchrotron XRD gives spatial resolution superior to neutron diffraction but requires access to a large-scale facility and cannot match the full-thickness penetration of neutrons in sections above ~20 mm thick.

Neutron Diffraction

Neutron diffraction uses the same Bragg’s Law principle as XRD but with thermal neutrons (wavelength ~0.1–0.3 nm) instead of X-rays. The key advantage is penetration depth: thermal neutrons have a mean free path of 20–100 mm in steel, allowing full through-thickness stress measurement of pressure vessel walls, pipeline girth welds, and structural beams without sectioning.

Neutron diffraction gauge volume (defined by slits and collimators):
  Typical: 1–8 mm³ (rectangular box in specimen)
  Spatial resolution: ~1–3 mm

Stress tensor determination:
  Measure ε in three orthogonal directions (longitudinal, transverse, normal):
  ε_L = (1/E)[σ_L − ν(σ_T + σ_N)]
  ε_T = (1/E)[σ_T − ν(σ_L + σ_N)]
  ε_N = (1/E)[σ_N − ν(σ_L + σ_T)]

  Solve simultaneously for σ_L, σ_T, σ_N (full 3D stress tensor)

Key facilities: ISIS (UK), ILL (France), NIST Center for Neutron Research (USA),
                ANSTO OPAL (Australia), MLZ FRM II (Germany)

The most critical practical issue in neutron diffraction of weld specimens is determination of the stress-free reference lattice spacing d₀. Because the weld metal composition and microstructure differ from the base material, d₀ varies across the weld cross-section even in the absence of stress. Neglecting this compositional variation produces systematic errors in the calculated stress. Best practice is to use comb-shaped or matchstick specimens cut from the weld region to measure d₀ as a function of position, or to use peak-fitting corrections based on known composition-d₀ relationships.

Hole-Drilling Method (ASTM E837)

The hole-drilling method is a semi-destructive technique standardised in ASTM E837 and EN 13521. A small hole (typically 1.8–3.2 mm diameter) is drilled incrementally through the surface of the component, and the resulting strain relief around the hole is measured by a strain gauge rosette bonded to the surface. The residual stress is back-calculated from the measured strain relief using calibration coefficients derived from FE analysis or tabulated in the standard.

Uniform stress case (ASTM E837 Eq. 1):
  σ_max, σ_min = − E / (4A) × (ε_1 + ε_3) ± E / (4B) × √[(ε_1 − ε_3)² + (ε_1 + ε_3 − 2ε_2)²]
  α = 0.5 × arctan[(ε_1 + ε_3 − 2ε_2) / (ε_1 − ε_3)]

where:
  ε_1, ε_2, ε_3 = measured strains at 0°, 45°, 90° rosette gauges
  A, B   = calibration constants (material-independent, depend on hole geometry)
  α      = angle of maximum principal stress to ε_1 gauge direction

Measurement depth capability: ~0.5–1.0 × hole diameter (typically 1–5 mm)
Limitation: invalid if residual stress > 0.6 × material yield strength at surface
Standard: ASTM E837-20; EN 13521:2022

Incremental hole-drilling (IHD) extends the method to non-uniform stress fields by drilling in small depth increments (0.025–0.1 mm) and using integral method regularisation (Tikhonov regularisation) to recover the depth-resolved stress profile. IHD is widely used for weld residual stress characterisation in production components where neutron diffraction access is impractical, and its modest equipment requirements (portable motorised drill, strain gauge data logger) make it the most accessible quantitative technique for field applications.

The Contour Method

The contour method (Prime, 2001) is a destructive technique that provides the full two-dimensional map of one residual stress component across an entire cross-section — a capability unique among residual stress measurement techniques. The procedure is:

Contour Method Protocol:
  1. Carefully wire-EDM cut along measurement plane (cutting wire ≤ 0.2 mm, slow feed)
     → Residual stresses relax elastically; cut surfaces deform
  2. Measure deformed surface contours (CMM or laser profilometry, ±1–5 µm accuracy)
  3. Average the two opposite cut-surface contours (removes cutting artefacts)
  4. Apply averaged contour as NEGATIVE displacement boundary condition on
     FE model of half-specimen
  5. Resulting FE stress field = original residual stress normal to cut plane

Spatial resolution: ~1–2 mm over full cross-section
Accuracy: ±20–50 MPa typical for steel
Stress component: ONLY the stress normal to the cut plane
Limitation: Fully destructive; requires precision EDM cutting and high-accuracy profilometry
Application: Thick weld sections, pipe girth welds, pressure vessel nozzles

The contour method is particularly valuable for thick weld sections (50–200 mm) where neutron diffraction beam time is limited or where the specimen cannot be transported to a neutron facility. It has been widely applied to validate finite element welding simulations, providing a gold-standard cross-sectional stress map against which the FE model is calibrated.

Other Methods: Magnetic Barkhausen Noise, Ultrasonic, and Deep Hole Drilling

Several additional techniques are in industrial use. Magnetic Barkhausen noise (MBN) is sensitive to magnetic domain wall motion which is influenced by the local stress state in ferromagnetic steels; it is non-destructive and fast, but requires extensive empirical calibration and is sensitive to microstructure, hardness, and surface condition. It is used for qualitative stress mapping (identifying high-stress zones) rather than accurate quantitative measurement. Ultrasonic methods exploit the acoustoelastic effect — the dependence of ultrasonic wave velocity on stress state — to measure bulk-averaged residual stress through the material thickness; the method requires knowledge of the acoustoelastic constants, which vary significantly with microstructure. Deep-hole drilling (DHD) extends the hole-drilling principle to depths of 10–100 mm by measuring the diameter change of a reference hole after core trepanning; it is used for thick pressure vessel sections and produces through-thickness stress profiles.

Post-Weld Heat Treatment for Stress Relief

Post-weld heat treatment (PWHT) is the most effective and widely mandated method for reducing weld residual stresses in structural and pressure-containing components. Its primary function is stress relief through localised creep at elevated temperature, but it also delivers secondary metallurgical benefits including HAZ hardness reduction, hydrogen effusion, and improved impact toughness in the HAZ.

Mechanism of Stress Relief

PWHT stress relief operates primarily through creep relaxation, not simple yielding. When the component is heated to the stress relief temperature, the yield strength drops markedly (to 30–60% of the ambient value for carbon steel at 600°C), but more importantly, the material begins to creep at the residual stress level present. Research has shown that the dominant stress relief occurs during the heating ramp — even before the soak temperature is reached — as the decreasing yield strength intersects the existing residual stress magnitude, causing localised plastic flow that redistributes the stress to a lower, self-equilibrating level. Extended soak time provides further (diminishing-return) relaxation through continued creep.

Larson-Miller creep relaxation parameter:
  P = T × (C + log t_h) × 10⁻³   [°C·h units]

where:
  T   = soak temperature (°C)
  t_h = soak time (hours)
  C   = Larson-Miller constant (C ≈ 20 for carbon steel)

Stress relief efficiency (empirical, carbon steel):
  At P ≈ 16–18: ~60–70% residual stress reduction
  At P ≈ 18–20: ~75–85% residual stress reduction
  At P ≈ 20+:   ~85–95% residual stress reduction

ASME VIII Div.1 P-No.1 minimum: T = 595°C, t_h = 1 h/25 mm thickness
→ For 50 mm plate: P = 595×(20+log 2)×10⁻³ ≈ 17.8 → ~75% relief expected

PWHT Requirements by Material Group

Local PWHT for Welds on Large Structures

When furnace PWHT is impractical — on large vessels, in-situ piping systems, offshore structures, or nuclear plant — local PWHT is performed using flexible electric resistance heating pads or induction heating coils. Local PWHT must heat a band extending at least 75 mm (or 2.5√(Rt) for pipes, where R = bore radius and t = wall thickness) either side of the weld centreline to the specified soak temperature, while the thermal gradient to the un-heated structure is managed to avoid introducing new stresses from differential expansion. ASME VIII Non-Mandatory Appendix RW-350 and BS PD 5500 Annex D provide guidance on soak band width and temperature monitoring requirements for local PWHT.

Residual Stress in Structural Integrity Assessment

The most important engineering application of weld residual stress knowledge is in fracture mechanics-based fitness-for-service (FFS) assessment of flaws found in or near welds during inspection. The three principal procedures are BS 7910:2013, the R6 procedure (EDF/Rolls-Royce/Jacobs, used widely in nuclear), and API 579-1/ASME FFS-1.

Residual Stress as a Secondary Load

In fracture mechanics, residual stress is classified as a secondary stress — a self-equilibrating stress system that does not contribute to global plastic collapse but does contribute to the crack driving force (stress intensity factor K) and therefore to fracture initiation. This distinction is fundamental to the failure assessment diagram (FAD) methodology used in all three procedures:

Failure Assessment Diagram (FAD) coordinates:
  K_r = K_total / K_mat   [fracture ratio — proximity to fracture]
  L_r = P_ref / P_L       [load ratio — proximity to plastic collapse]

where:
  K_total = K_I^primary + K_I^secondary (residual)   [total stress intensity factor]
  K_mat   = material fracture toughness (K_Ic, K_Jc, or J converted)
  P_ref   = reference stress from applied loading only (residual stress excluded)
  P_L     = limit load (plastic collapse of net section)

Residual stress contribution to K (secondary SIF):
  K_I^sec = ∫ σ_res(z) × h(z, a) dz   [weight function integral]

where h(z, a) = weight function for specific crack/geometry configuration

The key point is that residual stress shifts the K_r coordinate (horizontally in the FAD) but does not shift L_r — residual stress does not cause plastic collapse. This contrasts with primary membrane stresses from pressure or weight, which affect both K_r and L_r. A structure with a high tensile residual stress field can be assessed for plastic collapse separately from the fracture assessment, which is physically correct.

Residual Stress Profiles in BS 7910 Annex Q

BS 7910:2013 Annex Q provides upper-bound through-thickness residual stress profiles for the following weld joint geometries, expressed as polynomials in the normalised through-thickness coordinate z/t:

The engineer performing an FFS assessment may substitute measured residual stress data (from neutron diffraction, hole drilling, or the contour method) for the Annex Q upper-bound profiles. This residual stress reduction approach can significantly reduce conservatism and avoid unnecessary weld repair or component decommissioning. The measured distribution must be decomposed into membrane (σ_m), bending (σ_b), and self-balancing (σ_sb) components using the weight function approach, from which the secondary stress intensity factor K_I^sec is calculated.

Plasticity Correction Factor Q (Residual Stress)

When both primary and secondary stresses are present, the interaction between them introduces a plasticity correction factor V (in R6) or ρ (in BS 7910), which accounts for the non-linear interaction between the primary plastic zone and the secondary stress crack-opening contribution. At high primary loads (L_r > 0.8), the growing plastic zone under primary load partially relieves the residual stress, reducing the effective secondary contribution. BS 7910 Clause 7.3.2 and R6 Section IV.4 both provide ρ correction methods that depend on L_r and the ratio of secondary to primary K. Applying the ρ factor can substantially reduce the effective K_r contribution from residual stress at high primary load ratios — a significant deconservatism that must be carefully justified.

Residual Stress Engineering: Managing Stress for Performance

Reducing Tensile Residual Stress: Weld Procedure Controls

Several weld procedure parameters influence the magnitude and distribution of residual stress, giving the engineer tools to reduce peak tensile stress without resorting to PWHT:

Low heat input reduces the volume of thermally affected material and the resulting shrinkage forces, generally reducing residual stress magnitude. However, very low heat input increases cooling rate and may produce harder, more crack-susceptible HAZ microstructures; the optimum is a compromise.

Backstep and intermittent welding sequences allow partial elastic recovery between weld passes, reducing the cumulative restraint. Preheating reduces the thermal gradient between the weld and base metal, lowering the differential contraction that drives residual stress.

Temper bead technique uses successive weld passes to thermally temper the previous layer, reducing HAZ hardness and partially stress-relieving previous passes without PWHT — applicable where conventional PWHT is impractical (in-service repair welds on large structures).

Introducing Compressive Residual Stress: Post-Weld Improvement

Post-weld improvement techniques deliberately introduce compressive residual stresses at the weld toe — the highest fatigue stress concentration site — to improve fatigue performance. The most industrially proven methods are:

TIG dressing: A TIG arc without filler metal is run along the weld toe, remelting the toe region, smoothing the notch geometry, and creating a thin heat-affected zone with beneficial compressive transformation stresses on cooling. Fatigue life improvements of 2–3× are typical.

Ultrasonic Impact Treatment (UIT) / High-Frequency Mechanical Impact (HFMI): A pin tool vibrating at 20–60 kHz impacts the weld toe surface, inducing compressive residual stresses to depths of 1–2 mm and reducing the toe notch angle. IIW document XIII-2832-18 provides design guidance for HFMI-treated joints, with fatigue class improvements of 1–3 FAT classes depending on steel yield strength. HFMI is now widely used in offshore wind monopile welded joints where fatigue life is the critical design driver.

Shot peening and needle peening: Impact of high-velocity steel shot or needle tips introduces compressive residual stresses to depths of 0.5–1 mm. Shot peening is extensively used in aerospace for structural weldments; needle peening is simpler to apply in restricted-access positions.

Finite Element Simulation of Weld Residual Stress

Computational weld modelling (CWM) using finite element analysis has become an important tool for predicting residual stress distributions in geometries where measurement is impractical, for optimising welding procedures before physical trials, and for providing input to fracture mechanics assessments of thick-section joints. Validated FE models can replace or supplement neutron diffraction measurements in fitness-for-service assessments, provided the model is calibrated against experimental data from representative specimens.

A complete thermo-mechanical-metallurgical weld FE simulation involves three coupled analyses: a transient thermal analysis using a moving heat source model (double ellipsoid Goldak model for arc welding); a mechanical analysis using the temperature history from the thermal model and temperature-dependent elastic-plastic material properties; and optionally, a metallurgical analysis tracking phase fraction evolution (austenite, ferrite, martensite, bainite) using kinetic models such as the Leblond-Devaux or Johnson-Mehl-Avrami approach, with coupling back to the mechanical analysis through transformation plasticity and transformation strain terms.

Goldak double-ellipsoid heat source (Goldak 1984):
  q_f(x,y,z,t) = 6√3 · f_f · Q / (π√π · a · b · c_f)
                 × exp(−3x²/a² − 3y²/b² − 3z²/c_f²)   [front quadrant]

  q_r(x,y,z,t) = 6√3 · f_r · Q / (π√π · a · b · c_r)
                 × exp(−3x²/a² − 3y²/b² − 3z²/c_r²)   [rear quadrant]

where:
  Q    = η × V × I  = net heat input (W)
  η    = thermal efficiency (~0.80 for GTAW; ~0.75 for GMAW; ~0.60 for SMAW)
  a, b = semi-axes in x and y (cross-weld dimensions)
  c_f, c_r = semi-axes in weld direction (front and rear)
  f_f, f_r = fraction of heat in front/rear (f_f + f_r = 2; typically f_f=0.6, f_r=1.4)

Calibration: adjust a, b, c_f, c_r to match thermocouple data and bead geometry
Validation: compare predicted HAZ hardness, bead width/depth, and residual stress
            distribution against experimental measurements

Industrial Case Studies

Pipeline Girth Weld: Reeling Installation Residual Stress Redistribution

Deepwater pipeline installation by the reel-lay method subjects girth welds to plastic strains of 1–3% as the pipe is reeled onto and off the lay vessel. This plastic deformation redistributes the weld residual stress field: tensile residual stresses are partially relieved as the material yields, but upon elastic springback after each reel cycle a complex redistribution occurs that can leave the inner bore weld root in compression in the short-transverse direction — potentially beneficial for hydrogen-induced cracking resistance in sour service. BS 7910 and API 579 do not fully address post-reel residual stress redistribution; industry practice is to use coupled FE simulation of the reel-lay cycle applied to a pre-computed weld residual stress field, with experimental validation by neutron diffraction or contour method on specimens representative of the as-installed condition.

Pressure Vessel Nozzle: PWHT Bypass Assessment under API 579

A refinery pressure vessel of 50 mm C-Mn steel (SA-516 Gr 70) had a nozzle replacement weld identified during turnaround inspection as having been made without the required PWHT, due to a documentation error. The engineering team performed an API 579-1/ASME FFS-1 Level 2 assessment to determine whether the vessel was fit for continued service without performing PWHT. The assessment included: (a) neutron diffraction measurement of the as-welded residual stress field at the nozzle-to-shell junction (peak axial stress = 0.92 × σ_y on inner bore); (b) UT inspection to quantify flaw population (two embedded weld flaws, maximum height 4 mm, length 25 mm); (c) fracture mechanics assessment per API 579 Annex C including the as-measured residual stress as secondary stress input; and (d) fatigue assessment for the remaining design life at the known pressure cycling history. The assessment demonstrated acceptable flaw tolerance with the as-welded residual stress for 5 years of additional service, after which planned shutdown for PWHT and full re-inspection was scheduled. This avoided an immediate unplanned outage at an estimated cost saving of $4.2 million.

Recommended Reference Books

Disclosure: MetallurgyZone participates in the Amazon Associates programme. 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.