Tutorial: Calculating Minimum Preheat Temperature for Structural Steel Welding
Hydrogen-induced cold cracking (HICC) — also called hydrogen-assisted cracking (HAC) or delayed hydrogen cracking (DHC) — is one of the most serious and insidious weld defects in structural and pressure vessel fabrication. Unlike hot cracking, which occurs during solidification, cold cracking typically develops hours or even days after welding is complete, making it undetectable by in-process visual inspection. The primary engineering defence against cold cracking is the correct calculation and application of minimum preheat temperature, which slows the HAZ cooling rate and allows diffusible hydrogen to escape before the weld joint reaches the critical susceptibility temperature. This tutorial covers both analytical methods in EN 1011-2:2001 Annex C (Methods A and B) and the AWS D1.1 tabular and formula approaches, with fully worked examples for structural and pressure vessel steel applications.
Key Takeaways
- Three conditions must coexist for hydrogen cold cracking: a susceptible (hard martensitic or bainitic) microstructure, a critical concentration of diffusible hydrogen, and a tensile stress field. Eliminating any one condition prevents cracking.
- EN 1011-2 Method B uses Pcm (Ito–Bessyo parameter) and accounts for hydrogen scale HD, plate combined thickness d, and heat input Q — it is more accurate than Method A for low-carbon microalloyed steels (CE < 0.40).
- CE(IIW) = C + Mn/6 + (Cr+Mo+V)/5 + (Ni+Cu)/15 — used in Method A and AWS D1.1 Annex I. Steels with CE > 0.70 are generally considered unweldable without specialist procedures.
- The hydrogen scale classification HD is critical: switching from cellulosic SMAW (HD5, ~15 ml/100g) to basic low-hydrogen SMAW (HD3, ~5 ml/100g) typically reduces the required preheat by 50–75°C for the same steel and section.
- Combined thickness d — not just one plate thickness — determines the heat sink effect at the joint. A T-joint has much higher combined thickness than a butt weld of the same plate, requiring significantly higher preheat.
- Preheat temperature must be measured at ≥75 mm from the weld edge (or plate thickness if <75 mm) after a thermal soak of at least 2 minutes per 25 mm plate thickness to ensure equilibrium through the section.
Why Cold Cracking Occurs: The Metallurgical Mechanism
Hydrogen cold cracking (HICC, HAC, or underbead cracking) requires the simultaneous presence of three conditions: a susceptible microstructure, diffusible hydrogen above a threshold concentration, and a tensile stress field. Understanding the metallurgical basis of each condition is essential for understanding why preheat is effective and how to select the correct parameters.
Susceptible Microstructure
Martensite and lower bainite in the coarse-grained HAZ (CGHAZ) are the microstructures most susceptible to hydrogen cracking. Both are hard (high dislocation density) and have limited capacity to accommodate the stress concentrations created by hydrogen-induced lattice dilation. The susceptibility threshold is generally quoted as above approximately 350 HV10 for hydrogen concentrations >5 ml/100g, and above approximately 250 HV10 in the presence of very high hydrogen (~15 ml/100g cellulosic electrodes). Preheat slows the HAZ cooling rate, shifting the t8/5 to longer times and moving the HAZ microstructure from martensite toward softer bainite or ferrite-pearlite on the CCT diagram. For the relationship between cooling rate and microstructure, see the TTT and CCT diagram tutorial.
Diffusible Hydrogen
Hydrogen enters the weld pool from moisture in electrode coatings, flux, wire surface oxides, base metal surface contamination (oil, grease, water), and the shielding gas. Atomic hydrogen, generated by dissociation in the arc plasma, dissolves into the liquid weld pool at concentrations many times the equilibrium solid-state solubility. On solidification, the excess hydrogen is trapped in the as-deposited microstructure as diffusible hydrogen — mobile interstitial atoms that diffuse through the lattice and accumulate at stress concentrations such as inclusions, grain boundaries, and crack tips. The critical threshold for HICC in martensitic HAZ is typically 2–5 ml/100g diffusible hydrogen (measured per ISO 3690). Preheat above 100°C substantially accelerates hydrogen diffusion and its escape from the joint before the HAZ cools to the critical temperature range (<150°C) where trapping becomes irreversible.
Tensile Residual Stress
Residual stresses from weld metal contraction, joint restraint, and thermal gradient are always present in welded structures. They are the driving force for crack propagation once a hydrogen-assisted crack initiates. Preheat reduces residual stress by lowering the thermal gradient between weld and base metal, but does not eliminate it. Joint design, welding sequence, and post-weld stress relief (PWHT) are the primary tools for residual stress management. For the hydrogen-induced cracking mechanisms in detail, including the role of microstructure and stress intensity, see the linked article.
Effect of preheat on HAZ cooling time t8/5 (Rykalin 3D model): t8/5 = (6700 − 5×T₀) × Q × F(T₀) / (2π × λ) Where: T₀ = preheat temperature (°C) Q = heat input (kJ/mm) F(T₀) = [1/(500−T₀)² − 1/(800−T₀)²] λ = thermal conductivity of steel ≈ 0.04 kJ/(mm·s·K) Example: T₀ = 25°C vs T₀ = 150°C, Q = 1.0 kJ/mm: F(25°C) = [1/225625 − 1/591625] = 4.43×10⁻⁶ − 1.69×10⁻⁶ = 2.74×10⁻⁶ t8/5(25°C) = (6700−125) × 1.0 × 2.74×10⁻⁶ / 0.2513 = 71.8s F(150°C) = [1/(350)² − 1/(650)²] = 8.163×10⁻⁶ − 2.367×10⁻⁶ = 5.796×10⁻⁶ t8/5(150°C) = (6700−750) × 1.0 × 5.796×10⁻⁶ / 0.2513 = 160.8s → Preheat from 25°C to 150°C more than DOUBLES t8/5 at Q = 1.0 kJ/mm → Slower cooling shifts HAZ microstructure from martensite toward bainite
Carbon Equivalent Formulae
Carbon equivalent (CE) is an empirical single-number index that summarises the susceptibility of a steel composition to hydrogen cold cracking by weighting the contribution of each alloying element relative to carbon. Two formulae are in common use; they are calibrated for different composition ranges and the correct choice depends on the CE range of the steel being assessed.
CE(IIW) — IIW Carbon Equivalent
CE(IIW) = C + Mn/6 + (Cr+Mo+V)/5 + (Ni+Cu)/15 Where all elements are in wt% from the ladle analysis on the MTR. Interpretation: CE(IIW) ≤ 0.35: Generally weldable without preheat for t ≤ 25 mm CE(IIW) 0.35–0.45: Low preheat risk; preheat 50–100°C for t > 25 mm CE(IIW) 0.45–0.60: Preheat required; 100–200°C; use low-H process CE(IIW) 0.60–0.70: High preheat; PWHT likely needed; specialist procedure CE(IIW) > 0.70: Very difficult to weld; requires specialist assessment Standard reference: EN 1011-2 Method A; AWS D1.1 Annex I Best suited for: CE(IIW) > 0.40; medium and high-carbon alloy steels
Pcm — Carbon Equivalent (Ito–Bessyo)
Pcm = C + Si/30 + (Mn+Cu+Cr)/20 + Ni/60 + Mo/15 + V/10 + 5×B Where all elements are in wt%. Interpretation: Pcm ≤ 0.18: Very low cracking risk Pcm 0.18–0.22: Low risk; preheat may be needed for thick sections or high H Pcm 0.22–0.28: Moderate risk; preheat required in most conditions Pcm > 0.28: High risk; significant preheat mandatory Standard reference: EN 1011-2 Method B (primary formula for Method B) Best suited for: Pcm ≤ 0.40; low-carbon microalloyed and HSLA steels (e.g., API 5L X65, S460, S690Q, offshore structural grades) Note: For the same steel: If CE(IIW) < 0.40 → use Pcm (more sensitive to carbon at low CE) If CE(IIW) > 0.45 → use CE(IIW) (Pcm underestimates risk at high CE)
EN 1011-2 Method A: Chart-Based Approach
Method A (formally designated the “CEN Method A” in EN 1011-2 Annex C) is a simplified approach that produces preheat requirements directly from CE(IIW) and plate thickness using a series of tables and charts. It does not explicitly account for hydrogen content or heat input, making it inherently conservative for low-hydrogen processes and lower heat inputs. Its advantage is simplicity: given the CE and thickness, the preheat is read directly from a table.
Method A Procedure
Calculate CE(IIW) from the MTR ladle analysis
Using the formula CE = C + Mn/6 + (Cr+Mo+V)/5 + (Ni+Cu)/15. Elements not reported are taken as zero.
Determine the combined thickness d
d = sum of all plate thicknesses meeting at the joint. For a butt weld: d = 2×t. For a T-fillet: d = tweb + tflange. For a corner joint: d = t1 + t2.
Read preheat from EN 1011-2 Table C.1 (reproduced below)
The table gives minimum preheat temperature based on CE and combined thickness band. If the calculated value falls between table entries, interpolate or use the higher entry conservatively.
| CE(IIW) | d ≤ 20 mm | 20 < d ≤ 40 mm | 40 < d ≤ 60 mm | 60 < d ≤ 80 mm | d > 80 mm |
|---|---|---|---|---|---|
| ≤0.35 | 0°C | 0°C | 0°C | 50°C | 100°C |
| 0.35–0.40 | 0°C | 0°C | 50°C | 100°C | 150°C |
| 0.40–0.45 | 0°C | 50°C | 100°C | 150°C | 200°C |
| 0.45–0.50 | 50°C | 100°C | 150°C | 200°C | 250°C |
| 0.50–0.55 | 100°C | 150°C | 200°C | 250°C | 300°C |
| 0.55–0.60 | 150°C | 200°C | 250°C | 300°C | 350°C |
| >0.60 | 200°C | 250°C | 300°C | 350°C | Specialist assessment |
EN 1011-2 Method B: The Analytical Formula
Method B (CEN Method B) uses the Pcm carbon equivalent and an explicit formula that incorporates hydrogen content, combined thickness, and heat input. It is the preferred method when the steel composition is fully known from the MTR, the hydrogen classification of the consumable is documented, and the heat input is defined in the welding procedure specification (WPS).
The Method B Formula
T₀ (°C) = 697 × Pcm + 160 × tanh(d/35) + 62 × HD^0.35 − 328 × Q^0.5 − 182
Where:
T₀ = minimum preheat temperature (°C)
Pcm = C + Si/30 + (Mn+Cu+Cr)/20 + Ni/60 + Mo/15 + V/10 + 5B (wt%)
d = combined thickness (mm)
HD = diffusible hydrogen content of weld metal (ml/100g deposited)
Q = heat input (kJ/mm) [arc energy × thermal efficiency η]
Notes on the formula:
— If calculated T₀ < 0°C, minimum preheat is 0°C (no preheat needed)
but base metal must be dry and at ≥ 5°C
— The tanh(d/35) function saturates at d >> 35 mm (thick plate effect)
— The −328×Q^0.5 term: higher heat input REDUCES preheat requirement
— The +62×HD^0.35 term: higher hydrogen INCREASES preheat requirement
Valid composition range for Pcm:
%C: 0.07–0.22, %Si: 0.00–0.60, %Mn: 0.40–1.60
%Cu: 0.00–0.50, %Ni: 0.00–1.60, %Cr: 0.00–0.60
%Mo: 0.00–0.50, %V: 0.00–0.12, %B: 0.00–0.005
For steels outside this range, validate against known crack-free welds.
Hydrogen Scale Classification (HD)
| Class | HD (ml/100g) | Typical welding processes | Required electrode/wire condition |
|---|---|---|---|
| HD1 | ≤1 | Vacuum-degassed solid wire GMAW; TIG (GTAW) with clean wire | Clean, dry; no surface contamination |
| HD2 | ≤3 | Moisture-controlled solid wire GMAW; basic MIG/MAG; GTAW | Dry wire storage; no moisture contamination |
| HD3 | ≤5 | Basic low-hydrogen SMAW (E7018, E8018) — well-stored and re-dried; FCAW basic | Re-dried at 300–350°C ≥2h; held at 150°C; max 4h exposure after drying |
| HD4 | ≤10 | Basic SMAW, as-received; some FCAW rutile; SAW with dry flux | Standard storage; no active re-drying |
| HD5 | ≤15 | Cellulosic SMAW (E6010, E6011); rutile SMAW; undried flux SAW | As received; no special requirements |
| >HD5 | >15 | Cellulosic in high-humidity; wet flux; contaminated consumables | Avoid or use alternative process; high preheat mandatory |
Step-by-Step Method B Calculation
Obtain the steel composition from the MTR ladle analysis
Use the actual heat analysis values, not the maximum permitted by the standard. Using maximum values gives a conservative overestimate of preheat that may be unnecessarily costly in production. The ladle analysis is the legally certified composition per EN 10204 Type 3.1 or 3.2.
Calculate Pcm
Sum all terms in the Pcm formula. Elements not present in the steel (e.g., no copper additions) are set to zero. Check that the composition falls within the valid range stated for Method B; if not, use Method A or seek specialist advice.
Determine the combined thickness d
d is the total thickness of all steel sections meeting at the joint, measured in mm. For complex multi-plate connections (bracket, stiffener, gusset), sum all thicknesses meeting within the thermal influence zone (~75 mm from the weld). Use the actual thicknesses from the approved drawings, not nominal or minimum wall.
Classify the welding process hydrogen (HD)
Use the HD classification from the consumable manufacturer’s datasheet (diffusible hydrogen per ISO 3690 test) and verify against the storage and re-drying conditions actually achieved in production. If re-drying has not been performed, use a higher HD class (more conservative) than stated on the datasheet.
Determine the heat input Q
Use the heat input from the approved WPS: Q (kJ/mm) = (I × U × 60) / (v × 1000) × η, where η is the process thermal efficiency factor (SAW 1.0; SMAW/GMAW/FCAW 0.8; GTAW 0.6). Use the minimum heat input that will be applied in production, as this gives the most conservative (highest) preheat requirement. For the full heat input calculation, use the welding heat input calculator.
Substitute into the Method B formula and evaluate
Calculate T0 = 697 × Pcm + 160 × tanh(d/35) + 62 × HD0.35 − 328 × Q0.5 − 182. If T0 < 0°C, use 0°C minimum (but ensure metal surface is dry and ≥5°C). Round up to the nearest 25°C increment for practical application.
Apply practical minimums and code-specific rules
Check the result against any higher minimum specified by the applicable construction code: ASME B31.3 Section 330 gives minimum preheat temperatures for P-number groups; AWS D1.1 Table 4.5 gives structural steel minimums; API 1104 gives pipeline welding minimums. The governing minimum is the higher of the calculated value and the code minimum.
Worked Examples
Scenario: A T-joint fillet weld connecting a 20 mm thick web plate to a 25 mm thick flange plate, both S355J2 (EN 10025-2). Welding process: SMAW with basic low-hydrogen electrodes E7018, re-dried at 300°C. Heat input Q = 1.2 kJ/mm. Calculate minimum preheat using both Method A and Method B.
MTR ladle analysis (actual heat):
C = 0.14%, Mn = 1.42%, Si = 0.30%, P = 0.015%, S = 0.006% Cr = 0.08%, Mo = 0.02%, Ni = 0.08%, Cu = 0.12%, V = 0.004%, B = 0.000%
Step 1 — CE(IIW):
CE(IIW) = 0.14 + 1.42/6 + (0.08+0.02+0.004)/5 + (0.08+0.12)/15
= 0.14 + 0.237 + 0.021 + 0.013
= 0.411
Step 2 — Pcm:
Pcm = 0.14 + 0.30/30 + (1.42+0.12+0.08)/20 + 0.08/60 + 0.02/15 + 0.004/10 + 0
= 0.14 + 0.010 + 0.081 + 0.001 + 0.001 + 0.000
= 0.234
Step 3 — Combined thickness:
d = t_web + t_flange = 20 + 25 = 45 mm
Step 4 — Hydrogen class:
E7018 basic, re-dried at 300°C: HD = 5 ml/100g → Class HD3
Step 5 — Method A result:
CE(IIW) = 0.411 → range 0.40–0.45 d = 45 mm → column 40 < d ≤ 60 mm Method A preheat = 100°C
Step 6 — Method B result:
T₀ = 697 × 0.234 + 160 × tanh(45/35) + 62 × 5^0.35 − 328 × 1.2^0.5 − 182 = 697 × 0.234 = 163.1 + 160 × tanh(1.286) = 160 × 0.859 = 137.4 + 62 × 5^0.35 = 62 × 1.737 = 107.7 − 328 × 1.2^0.5 = 328 × 1.095 = −359.1 − 182 = −182 T₀ = 163.1 + 137.4 + 107.7 − 359.1 − 182 = −133°C Since T₀ < 0°C: minimum preheat = 0°C (no preheat required by calculation)
Conclusion: Method A gives 100°C; Method B gives 0°C. The large difference illustrates that Method A can be significantly conservative for modern low-carbon microalloyed steels welded with low-hydrogen consumables at moderate heat input. The Method B result is the more technically defensible value. In practice, a minimum temperature of 5–10°C above ambient (to ensure the surface is dry) is recommended regardless of calculation. The WPS should state 0°C minimum preheat with the Method B basis recorded.
Scenario: A 35 mm thick butt weld in ASTM A517 Gr. F quenched and tempered steel (equivalent to S690Q) for a crane boom structure. Process: cellulosic SMAW (E6010) on the root pass (field condition), Q = 0.9 kJ/mm.
Typical A517 Gr. F composition (max permitted by specification):
C = 0.20%, Mn = 0.85%, Si = 0.25%, Cr = 0.65%, Mo = 0.55%, Ni = 0.00% Cu = 0.00%, V = 0.08%, B = 0.003%
CE(IIW):
CE(IIW) = 0.20 + 0.85/6 + (0.65+0.55+0.08)/5 + (0.00+0.00)/15
= 0.20 + 0.142 + 0.256 + 0
= 0.598
Pcm:
Pcm = 0.20 + 0.25/30 + (0.85+0.00+0.65)/20 + 0.00/60 + 0.55/15 + 0.08/10 + 5×0.003
= 0.20 + 0.008 + 0.075 + 0.000 + 0.037 + 0.008 + 0.015
= 0.343
Combined thickness:
d = 2 × 35 = 70 mm (butt weld, both sides)
Hydrogen class:
Cellulosic SMAW E6010: HD ≈ 15 ml/100g → Class HD5
Method B result:
T₀ = 697 × 0.343 + 160 × tanh(70/35) + 62 × 15^0.35 − 328 × 0.9^0.5 − 182
= 697 × 0.343 = 239.1
+ 160 × tanh(2.0) = 160 × 0.964 = 154.3
+ 62 × 15^0.35 = 62 × 2.623 = 162.6
− 328 × 0.9^0.5 = 328 × 0.949 = −311.3
− 182 = −182
T₀ = 239.1 + 154.3 + 162.6 − 311.3 − 182 = +62.7°C → round up to 75°C
Comparison: switch to basic low-hydrogen SMAW (HD = 5 ml/100g, HD3):
T₀ = 239.1 + 154.3 + 107.7 − 311.3 − 182 = +7.8°C → 0°C
Conclusion: Cellulosic SMAW requires 75°C minimum preheat.
Switching to basic low-hydrogen electrode reduces preheat to 0°C.
For Q+T steel (A517), preheat > 200°C risks softening the Q+T base metal.
STRONGLY RECOMMENDED: use basic low-hydrogen (HD3) or better.
Code check (ASME B31.3 Table 330.1.1): P-No. 11B Group 1 (690 MPa SMYS), thickness 35 mm > 19 mm → minimum preheat 120°C per code. The Method B result of 75°C (cellulosic) is below the ASME code minimum of 120°C — use 120°C as the governing minimum. With basic low-hydrogen electrodes, Method B gives 0°C but the code still requires 120°C. In any case, for high-strength Q+T steels, always apply the more stringent of the code minimum and the calculated value.
AWS D1.1 Preheat Requirements
AWS D1.1 Structural Welding Code — Steel is the primary structural welding standard in North America and widely referenced internationally for steel structures, bridges, and offshore platforms. It provides preheat requirements through two routes: a simplified table (Clause 4.2.2) and a formula-based method (Annex I).
AWS D1.1 Table 4.5 (Prequalified Joints)
| Steel category | Typical ASTM grades | Minimum SMYS (ksi) | t ≤ 19 mm (3/4”) | 19 < t ≤ 38 mm | 38 < t ≤ 57 mm | t > 57 mm |
|---|---|---|---|---|---|---|
| Category I | A36, A53B, A500 Gr.B, A501, A709 Gr.36 | 36 ksi (250 MPa) | 0°C | 0°C | 66°C (150°F) | 107°C (225°F) |
| Category II | A572 Gr.42–50, A588, A709 Gr.50 | 42–50 ksi | 0°C | 66°C | 107°C | 150°C (300°F) |
| Category III | A514, A517, A709 Gr.100, A852 | 90–100 ksi | 107°C | 107°C | 107°C | 107°C |
| Category IV | A709 Gr.100W, A710 | 100 ksi (690 MPa) | 107°C | 107°C | 107°C | 107°C |
AWS D1.1 Annex I Formula Method
AWS D1.1 Annex I uses CE(IIW) and a tabular lookup, not a closed-form formula. The minimum preheat temperature is read from Annex I Table I.1 based on: — CE(IIW) in six ranges: ≤0.40, 0.40–0.45, 0.45–0.50, 0.50–0.55, 0.55–0.60, >0.60 — Combined thickness in five ranges: ≤19mm, 19–38mm, 38–64mm, 64–102mm, >102mm Approximate formula equivalent (AWS D1.1 Annex I basis): T₀(°F) = 1440 × CE(IIW) − 0.0225 × d − 25 × Q^0.5 − 130 (approximate) T₀(°C) = (T₀(°F) − 32) / 1.8 Key differences from EN 1011-2: — AWS D1.1 does NOT explicitly account for consumable hydrogen class — AWS D1.1 preheat applies to the 75mm zone each side of the joint — Interpass temperature maximum: 250°C (Category I–II), 205°C (Category III–IV) — AWS D1.1 requires measurement before EACH pass in multi-pass welds
Code Comparison: EN 1011-2 vs AWS D1.1 vs ASME B31.3
| Feature | EN 1011-2 Method B | AWS D1.1 Table 4.5 | ASME B31.3 Table 330.1.1 |
|---|---|---|---|
| Carbon equivalent used | Pcm (Method B) | CE(IIW) (Annex I) or grade category | P-Number + Group Number |
| Hydrogen content | Explicit (HD1–HD5) | Not explicit in Table 4.5 | Not explicit |
| Heat input | Explicit term in formula | Not in Table 4.5; partial in Annex I | Not explicit |
| Geometry | Combined thickness d | Single plate thickness t | Nominal wall thickness |
| Basis | Analytical formula from CCT/cracking tests | Empirical from structural steel practice | Material P-Number classification |
| Typical application | Structural, offshore, pressure equipment | Buildings, bridges, structural (North America) | ASME pressure piping |
| Max interpass T | Not fixed; typically 250°C by practice | 250°C (Cat I–II), 205°C (Cat III–IV) | Not specified (per WPS) |
| Measurement location | ≥75 mm from weld edge | 75 mm from weld edge | As defined in WPS |
| When to apply | Before each pass | Before each pass | Before first pass; maintain during welding |
Practical Preheat Application and Verification
Methods of Applying Preheat
Three main techniques are used in industrial fabrication:
- Oxy-fuel gas torch (propane or oxy-acetylene): Most common for field and shop work on smaller sections. The torch is moved in broad sweeping passes over the 75 mm zone each side of the joint until the required temperature is achieved and thermal equilibrium is confirmed. Risk: localised overheating causing HAZ damage if torch is held stationary.
- Electric resistance or induction heating: Preferred for thick sections, precision control, and documentation. Heating blankets or induction coils wrapped around the joint provide uniform heating with thermocouple feedback and data logging. Mandatory for many pressure vessel codes on thick-wall joints.
- Furnace preheat: Used for small components that can be placed in a controlled furnace. Provides the most uniform temperature distribution and is sometimes combined with stress relief in a single cycle.
Temperature Measurement and Verification
- Tempilstik (Tempil) or Crayons: Phase-change wax crayons that melt at specific calibrated temperatures. Applied to the base metal surface; melting confirms the surface has reached at least that temperature. Fast and inexpensive; sufficient for most structural work. Must be applied to the non-heated face where possible for a true through-thickness indication.
- Contact thermocouple (Type K): Calibrated thermocouple with handheld meter, used for accurate measurement. Required for critical applications. Measurement must be made after a soak period of minimum 2 minutes per 25 mm of plate thickness after the heat source is removed.
- Infrared (IR) thermometer: Non-contact, fast, suitable for heated surfaces. Less accurate than contact thermocouples due to emissivity variation; should be calibrated against a contact reference on the same surface. Not suitable for reflective or scale-free surfaces without emissivity correction.
- Thermocouple data logger: For induction or resistance heating of pressure vessel joints, data loggers provide a continuous printed record of temperature at multiple points throughout the heating and welding cycle. This record becomes part of the quality documentation.
Maintaining Preheat During Welding
The minimum preheat temperature must be maintained throughout the welding operation — not just at the start. On long multi-pass welds in cold environments, the metal behind the arc cools rapidly and may drop below the minimum preheat before the next pass begins. Monitoring with Tempilstik before each pass is required by most codes. For automatic or mechanised welding in outdoor conditions, wind shields and insulating blankets are standard practice to maintain interpass temperature. After welding is complete, a slow cool (achieved by maintaining heat and allowing gradual temperature decay, or wrapping in insulating blanket) is required for highly restrained joints or high CE steels to prevent thermal shock cracking and to ensure hydrogen has time to diffuse out before the joint cools to the critical range below 100°C.
Special Cases and Limitations
Preheat for Repair Welds
Repair welds in service-aged material require extra caution for several reasons: the base metal may have a different composition than the original specification (ageing, contamination); the existing residual stress state and weld defect may create a more severe stress concentration; and access constraints may limit heat input options. For unknown-composition service steel, PMI should be performed before calculating preheat; a conservative CE based on the worst-case composition for the grade specification should be used. Repair welds in thick sections often require elevated preheat and mandatory PWHT regardless of calculated values. See the material traceability tutorial for PMI procedures applicable to repair weld material verification.
Preheat for Dissimilar Metal Welds
When welding two different steel grades (e.g., S355 to P91, carbon steel to stainless steel, carbon steel to Inconel overlay), the preheat calculation must be performed for both base materials independently, and the higher required preheat governs. In addition, the thermal expansion difference between dissimilar materials generates additional restraint stress; this favours the use of buttering layers (applying a layer of weld metal to one material face before completing the joint) and potentially higher preheat than the calculated minimum.
High-Strength Q+T Steels: Maximum Preheat
For quenched-and-tempered high-strength steels (S690Q, A514, A517), excessive preheat (typically above 200–230°C) can soften the base metal in the HAZ by exceeding the tempering temperature of the Q+T treatment, permanently reducing the yield strength below the minimum specified. These steels therefore have both a minimum preheat (to prevent cold cracking) and a maximum preheat (to prevent Q+T softening). Maximum interpass temperature for S690Q and similar grades is typically 150°C; verify against the material manufacturer's welding guidance and the applicable construction code before exceeding this.
Frequently Asked Questions
What is preheat in welding and why is it required?
What are the two EN 1011-2 methods for preheat calculation?
What is the difference between CE(IIW) and Pcm?
What is the hydrogen scale HD in EN 1011-2?
At what distance should preheat temperature be measured?
Does higher heat input reduce the required preheat?
What is combined thickness and why does joint geometry matter?
What is the difference between minimum preheat and maximum interpass temperature?
When is PWHT required instead of or in addition to preheat?
What is the AWS D1.1 approach to preheat specification?
Recommended References
Further Reading
