Preheat temperature to EN 1011-2 Method B
Preheat is applied for one reason: to slow the cooling rate of the joint so that the heat-affected zone does not transform to untempered martensite and so that diffusible hydrogen has time to escape before the weld reaches ambient temperature. It is not about drying the steel, although raising the surface above the dew point is a useful side effect. Preheat works by extending the time the joint spends between 800 and 500 °C and by keeping the joint above roughly 100 °C long enough for hydrogen to diffuse out.
Four variables set the required temperature. Composition, expressed as CET, fixes how hardenable the steel is. Combined thickness — the sum of the plate thicknesses conducting heat away from the joint, measured within about 75 mm of the weld — fixes how fast the joint sinks heat; a fillet on a T joint has three limbs and cools far faster than a butt in the same plate. Diffusible hydrogen from the consumable is the crack agent itself, and ranges from over 15 ml/100 g for a damp rutile electrode to under 3 ml/100 g for a properly baked or vacuum-packed low-hydrogen consumable. Heat input already supplies part of the thermal energy, so a high-energy run needs less preheat than a stringer bead.
EN 1011-2 Method B combines all four into a single expression: Tp = 697·CET + 160·tanh(d/35) + 62·HD^0.35 + (53·CET − 32)·Q − 328. The hyperbolic tangent term saturates around 90 mm, reflecting that beyond a certain thickness the joint is already an effectively infinite heat sink and further thickness adds nothing. The hydrogen term is a fractional power, so halving hydrogen from 10 to 5 ml/100 g buys about 30 °C — usually the cheapest single intervention available.
The equation is valid for CET 0.20 to 0.50 %, combined thickness 10 to 90 mm, hydrogen 1 to 20 ml/100 g and heat input 0.5 to 4.0 kJ/mm. A negative or near-zero result means no preheat is calculated as necessary, but codes still impose a minimum — typically the steel must be above 5 °C and free of moisture before welding. Preheat must be measured at the required distance from the joint (usually 4t or 75 mm, on the face opposite the heating source where possible) and maintained through the whole joint, including any interruption.
Tp = 697·CET + 160·tanh(d/35) + 62·HD^0.35 + (53·CET − 32)·Q − 328 [°C] CET = C + (Mn+Mo)/10 + (Cr+Cu)/20 + Ni/40 [wt %] d = combined thickness [mm] · HD = diffusible hydrogen [ml/100 g] · Q = heat input [kJ/mm] Validity: CET 0.20–0.50 · d 10–90 mm · HD 1–20 ml/100 g · Q 0.5–4.0 kJ/mm
- Combined thickness is the sum of all limbs at the joint, not the plate thickness. A T joint in 20 mm plate has a combined thickness of 60 mm.
- Measure preheat at 4t or 75 mm from the joint, on the face opposite the heat source wherever access allows.
- Maintain preheat and interpass temperature through the whole joint, including breaks. Losing it mid-joint is a common cause of cracking.
- Halving the diffusible hydrogen is usually cheaper and more effective than raising preheat. Check consumable storage and baking.
- For prequalified minimum preheat under AWS D1.1, ASME B31.3 or ASME VIII, use the governing code table - this calculation does not replace it.
Reference: EN 1011-2:2001+A1:2003 Annex C.2 (Method B); hydrogen scales A-E per EN 1011-2 Clause 5; HD determined to ISO 3690


