The experimental results of y-groove weld cracking tests were compared with the prediction methods of minimum preheat temperatures to avoid cold cracking in the heat affected zones, which include the BS-5135, the AWS D1.1, the JSSC procedure and the CEN chart method. The effects of steel chemical compositions, plate thicknesses, diffusible hydrogen contents and ambient temperatures on cold cracking susceptibility were investigated in this study.
The CEN chart method can predict minimum preheat temperatures most precisely as long as homogeneous preheating is conducted and the ambient temperature is 20°C. When local preheating is conducted, cooling time down to 100°C, t100, must be used. The minimum preheat temperture for local preheating is selected so that t100 of local preheating is equal to or longer than that of the minimum preheat temperature for homogeneous preheating.
Ambient temperatures greatly affect cold cracking susceptibility more than expected from its effect on t100. As ambient temperature becomes lower, minimum preheat temperature to avoid cold cracking becomes higher. However, all the previous prediction methods cannot estimate this effect. In this study, the effects of ambient temperatures on necessary preheat temperatures were converted into CEN increments, which were determined from the experimental results. Using these CEN increments, the CEN chart method can estimate the effect of ambient temperature.