Prediction model for cooling time in the heat-affected zone of girth welds in fully automatic welding of X80M pipeline

Objective Due to the heat input from multi-layer and multi-pass welding of thick-wall pipelines, the microstructure in the heat-affected zone (HAZ) of girth welds experiences a cyclic heating and cooling process. To prevent cold cracks resulting from a hardened structure, it is essential to strictly control the cooling rate of the joints. Investigating the relationship between the welding procedure and the cooling time (t_8/5) required for the HAZ of girth welds to cool from 800 °C to 500 °C plays a decisive role in controlling welding heat input and enhancing joint performance.

Methods The study focused on the girth welds of a 1 219 mm×22 mm X80M pipeline. A finite element model was developed based on the fully automatic welding procedure, and a heat source program was designed to reflect single- and dual-torch welding characteristics. Heat source parameters were established using experimental data on weld layer morphology. The calculated welding temperature field data were compared with experimental thermal cycle data to validate the model’s accuracy. By analyzing the thermal cycle information of the HAZ for each weld layer and adjusting welding heat input through changes in welding current and speed, the t_8/5 variation was determined, leading to the development of a cooling time prediction formula through fitting.

Results At low heat input, increasing the welding current significantly affected t_8/5 more than decreasing the welding speed. Conversely, at high heat input, decreasing the welding speed had a greater impact on t_8/5 than increasing the welding current. However, under the same heat input change conditions, the calculation error from either adjustment did not exceed 6％.

Conclusion The developed finite element model effectively calculates thermal parameters for the welding procedure with minimal error, demonstrating both accuracy and reliability. Adjusting the welding speed results in greater fluctuations in peak temperature and t_8/5 distribution compared to adjusting the welding current; therefore, prioritizing adjustments to the welding current is recommended to maintain welding stability. A new t_8/5 prediction formula, devoid of empirical parameters, has been proposed based on simulated weld thermal cycle curves, enhancing prediction accuracy by over 10％ compared to traditional empirical formulas.