Invar alloy welding heat-affected zone (HAZ) expansion analysis quantifies how welding thermal cycles alter the ultra-low coefficient of thermal expansion (CTE) of Fe-36%Ni Invar across the base metal, heat-affected zone, and weld metal. This article explains the metallurgical mechanisms behind expansion drift in welded joints and describes the sample types, test items, and standards-based methods used by accredited third-party laboratories to verify dimensional stability after welding. The analysis supports quality control for LNG containment systems, aerospace tooling, precision instruments, and other dimension-critical applications.
Technical Overview
Invar (Fe-36%Ni, UNS K93600, EN 1.3912) is a face-centered cubic iron-nickel alloy prized for its anomalously low coefficient of thermal expansion, typically 1.2 to 2.0 × 10⁻⁶/°C between room temperature and 100 °C. This "Invar effect" arises from spontaneous volume magnetostriction below the Curie temperature (approximately 230–280 °C), which counteracts normal lattice expansion. The property, however, is highly sensitive to chemical composition, atomic ordering, residual stress, and thermal history, which makes welding a critical process risk.
During welding, the fusion zone dilutes or enriches nickel relative to the base metal, while the heat-affected zone experiences peak temperatures from roughly 600 °C to the solidus, causing grain coarsening, stress relaxation and re-generation, micro-segregation, and local deviation from the optimal Fe-Ni ratio. As a result, the local CTE in the HAZ and weld metal can rise to 4–10 × 10⁻⁶/°C, two to five times the base-metal value. Because adjacent zones with mismatched expansion coefficients are mechanically constrained, cyclic service temperatures generate local stresses, distortion, and potential fatigue cracking at the fusion line.
Welding heat-affected expansion analysis therefore maps the CTE distribution across the joint, correlates it with microstructure and residual stress, and verifies that welds meet the dimensional-stability requirements of the design specification. Testing is typically performed by accredited third-party laboratories using calibrated dilatometry in accordance with ASTM E228, ASTM E831, ISO 11359, and related metallographic and mechanical standards.
Testing Samples
Butt-Welded Plate Joints
Representative plate butt welds are produced with the qualified welding procedure (process, filler metal, heat input, number of passes, and interpass temperature). Plate thicknesses commonly range from 3 mm to 20 mm. These joints provide the master samples from which all zone-specific coupons are extracted, and they preserve the as-welded thermal history for macro-examination and HAZ width measurement.
Welded Pipe and Tube Sections
Circumferential pipe and tube welds used in cryogenic or precision fluid systems are submitted as full sections or as arc segments. The curved geometry allows evaluation of welding position effects (flat, vertical, overhead) on HAZ uniformity and expansion behavior around the full circumference.
Zone-Specific Dilatometer Coupons
Small prismatic or cylindrical specimens, typically 3–6 mm in cross-section and 10–25 mm in length depending on the dilatometer, are extracted by low-stress machining or wire EDM from precisely located positions: unaffected base metal, fine-grained HAZ, coarse-grained HAZ, fusion line, and weld centerline. Specimen orientation relative to the rolling and welding directions is recorded because Invar CTE can be mildly anisotropic.
Cross-Weld and Thermal Cycling Coupons
Cross-section coupons spanning the entire joint are used for expansion profiling by incremental techniques and for metallography. Duplicate coupons are prepared for thermal cycling evaluation (for example, 20 to −196 °C to 20 °C) to measure hysteresis and permanent dimensional change over repeated cycles representative of LNG or cryogenic service.
Testing Items
- Mean coefficient of thermal expansion over defined intervals (for example, 20–100 °C, 20–200 °C, 20–300 °C) for each weld zone
- Instantaneous (differential) CTE as a continuous function of temperature, including behavior through the Curie transition
- Cryogenic expansion behavior down to −196 °C for LNG and cryogenic service conditions
- Curie temperature shift in the HAZ and weld metal as an indicator of compositional and ordering changes
- Thermal cycling hysteresis and permanent length change (dimensional stability) after repeated cycles
- HAZ width, sub-zone delineation (coarse-grained, fine-grained, partially affected), and grain size distribution per ASTM E112
- Microhardness profiles across the joint to reveal softening, hardening, or hidden secondary phases
- Residual stress distribution at the weld surface and through depth, referenced to ASTM E915 and EN 15305
- Chemical composition and micro-segregation of Ni, Fe, Mn, Si, S, P, and C across the fusion line
- Correlation of measured CTE deviation with welding heat input (kJ/mm) for procedure qualification data
Testing Methods
Push-Rod Dilatometry (ASTM E228)
Horizontal or vertical push-rod dilatometers with fused silica or alumina systems measure length change against a certified reference material such as sapphire. Specimens from each weld zone are tested under controlled heating rates (typically 3–5 °C/min) in inert atmosphere or vacuum, with calibration and uncertainty verified per ASTM E228. Measurement uncertainty of better than ±0.1 × 10⁻⁶/°C is achievable for the mean CTE in the critical 20–100 °C interval.
Thermomechanical Analysis (ASTM E831, ISO 11359)
TMA with a flat-probe expansion mode is applied to smaller coupons and thin-walled sections where full dilatometer specimens cannot be extracted. TMA also enables rapid screening of multiple positions across a polished cross-weld surface when equipped with micro-specimen capability.
Cryogenic Dilatometry
A dilatometer or TMA fitted with a liquid-nitrogen cryostat extends measurement to −196 °C, reproducing LNG service conditions. The test records contraction from room temperature to cryogenic temperature and the expansion on rewarming, quantifying hysteresis and any irreversible dimensional change at the welded joint.
Metallography and Grain Size Analysis
Cross-sections are prepared per ASTM E3 and etched per ASTM E407 to reveal the austenitic structure, fusion boundary, and HAZ sub-zones. Grain size is determined per ASTM E112, and image analysis maps grain coarsening across the HAZ, which is cross-correlated with the CTE profile.
Microhardness Profiling
Vickers or Knoop microindentation traverses per ASTM E384 (for example, HV0.5 at 0.5 mm spacing across the joint) detect hardness anomalies that indicate carbide or nitride precipitation, which can affect both expansion behavior and toughness.
Residual Stress Measurement by XRD
Composition and Micro-Segregation Analysis
Bulk composition is confirmed by optical emission or combustion methods (carbon and sulfur per ASTM E1019), while SEM-EDS line scans across the fusion line quantify nickel dilution and dendritic segregation in the weld metal, the primary driver of CTE drift in the fusion zone.
Applications
- Membrane-type LNG carrier containment systems, where weld CTE mismatch at −163 °C drives thermal fatigue assessment of corrugated Invar primary membranes
- Aerospace autoclave tooling and composite cure molds, where welded Invar frames must maintain tool geometry across repeated 180 °C cycles
- Precision metrology frames, optical benches, and interferometer structures requiring micrometer-level dimensional stability
- Cryogenic pipelines, valves, and vessel internals operating from liquid nitrogen to liquid natural gas temperatures
- Welding procedure qualification and filler-metal selection studies (GTAW, laser, electron beam) for heat-input optimization
- Failure analysis of welded Invar components exhibiting distortion, cracking, or loss of dimensional control in service
FAQ
Why does welding increase the thermal expansion of the Invar heat-affected zone?
Welding locally alters the factors that produce the Invar effect: nickel dilution or enrichment shifts the alloy away from the 36% Ni optimum, high peak temperatures coarsen grains and modify short-range ordering, and residual stresses superimpose elastic strain on thermal expansion. The combined effect reduces the magnetostrictive compensation, so the local CTE rises measurably above the base-metal value.
How much CTE deviation after welding is typically acceptable?
Acceptance limits depend on the application. For precision tooling and metrology structures, the HAZ mean CTE between 20 °C and 100 °C is commonly held below approximately 3–4 × 10⁻⁶/°C, while LNG membrane projects impose project-specific limits verified by procedure qualification testing. The governing specification for each component defines the exact criterion.
Which welding processes minimize expansion drift in the HAZ?
Low heat-input processes such as pulsed GTAW, laser beam welding, and electron beam welding produce narrow HAZs and limited dilution, and therefore the smallest CTE deviation. Matching Fe-36Ni filler metal, strictly controlled interpass temperature, and minimized pass count further preserve the low-expansion character of the joint.
Can post-weld heat treatment restore the low expansion property?
A properly specified stress-relief or re-annealing treatment can reduce residual stress and partially restore dimensional stability, but it cannot correct nickel dilution in the fusion zone and may itself cause grain coarsening if the temperature is excessive. Heat treatment parameters must therefore be validated by repeat dilatometry on treated coupons rather than assumed to be effective.
What specimen size is required for dilatometric testing of weld zones?
Conventional push-rod dilatometers typically require specimens of 3–6 mm diameter and 10–25 mm length, which is adequate for plate welds; TMA can accept considerably smaller specimens for thin sections. Because HAZ sub-zones may be narrower than the specimen length, laboratories report effective values averaged over the extracted volume and use incremental scanning methods when finer spatial resolution is required.