Hydrogen-Assisted Cracking in a Thick EH36 Deck Butt Weld
A 40 mm thick EH36 deck butt weld is being made in a shipyard in winter. The joint uses a SMAW root and FCAW fill/cap, with 280 A, 26 V and 300 mm/min travel. The ambient temperature is 4°C and relative humidity 88%. The electrodes were taken from an open quay-side store and used without baking; the preheat applied is only 50°C, measured 75 mm from the groove. The steel composition is C 0.14, Mn 1.35, Si 0.30, P 0.012, S 0.005, Cu 0.25, Ni 0.30, Cr 0.05, Mo 0.02 and V 0.02 wt%. Twenty-four hours after welding, magnetic-particle inspection finds a toe crack in the root region. The yard proposes to increase heat input to 2.5 kJ/mm and continue without changing storage or preheat practice, arguing that the extra heat will drive out hydrogen. The surveyor asks for a carbon-equivalent check and a revised welding procedure.
1. Calculate the carbon equivalent and the PCM of the steel, and assess the cracking risk for a 40 mm section.
Read the worked answer
CE = C + Mn/6 + (Cr + Mo + V)/5 + (Cu + Ni)/15 = 0.14 + 1.35/6 + (0.05 + 0.02 + 0.02)/5 + (0.25 + 0.30)/15 = 0.14 + 0.225 + 0.018 + 0.0367 ≈ 0.42.
PCM = C + Si/30 + (Mn + Cu + Cr)/20 + Ni/60 + Mo/15 + V/10 = 0.14 + 0.30/30 + (1.35 + 0.25 + 0.05)/20 + 0.30/60 + 0.02/15 + 0.02/10 = 0.14 + 0.010 + 0.0825 + 0.005 + 0.0013 + 0.002 ≈ 0.24.
PCM is relatively low, but CE ≈ 0.42 is significant for a 40 mm section, especially with high hydrogen and a cold, damp environment. Thick sections increase restraint and allow hydrogen to concentrate. The combination of CE 0.42, 40 mm thickness, 4°C ambient, 88% RH, unbaked electrodes and only 50°C preheat is a classic cold-cracking risk. A crack appearing after 24 hours is consistent with delayed hydrogen-assisted cracking, not with a hot tear.
2. Calculate the nominal heat input. Is the yard's proposal to raise it to 2.5 kJ/mm a sound remedy?
Read the worked answer
Heat input = 60VI/(1000v) = 60 × 26 × 280 / (1000 × 300) = 436,800 / 300,000 ≈ 1.46 kJ/mm.
This is a moderate heat input, not unusually low. Raising it to 2.5 kJ/mm would slow cooling and may reduce the risk of a hard, crack-sensitive heat-affected zone, but it is not a cure for hydrogen. The main sources of hydrogen here are unbaked electrodes and moisture in the joint. Increasing heat input also coarsens the HAZ, can reduce toughness in EH36, and may increase distortion and residual stress. It can make the weld look better while leaving a hydrogen problem in place.
The proposal is therefore not sound as written. Heat input should be controlled within a qualified range, but the essential changes are low-hydrogen consumables, baking, clean dry groove and adequate preheat. Raising heat input alone would not justify continuing with the same storage and preheat practice.
3. Outline the revised welding procedure and inspection requirements you would require before welding restarts.
Read the worked answer
The procedure should be re-qualified with hydrogen control as the primary measure. Use basic-coated low-hydrogen electrodes or a hydrogen-controlled FCAW consumable, and bake electrodes according to the manufacturer's instructions, typically around 300–350°C for basic electrodes, then hold them in a heated quiver at the workplace. Discard electrodes that have been exposed to the open store. Dry the groove and remove rust, scale, paint and moisture.
Apply and measure preheat on both sides of the joint, at a distance from the groove that reflects the procedure, and maintain it through the root and until the joint has cooled under controlled conditions. For 40 mm EH36 with CE ≈ 0.42 and a high-hydrogen risk, a preheat of at least 100°C, and probably 100–150°C, should be established by procedure qualification rather than guessed. Control interpass temperature and heat input within the qualified range.
Carry out NDT after a suitable delay, at least 48 hours for this thick section, and inspect the root region specifically. If cracks are found, remove them completely, re-qualify the repair procedure and re-inspect. Record electrode baking, preheat and interpass temperatures, since the surveyor will need evidence that the hydrogen source and cooling rate have both been controlled.