Numerical Simulation of Decoupling Coefficient Effects on CO-2 Phase Transformation Cracking

Literature Overview

This study employs finite element numerical simulation to investigate the influence of the decoupling coefficient on phase transformation-induced cracking during the cooling process of CO-(2) type welds. The research addresses a critical but often overlooked failure mechanism in thick-section low-alloy steel welding, where the volumetric expansion associated with austenite-to-martensite phase transformation can generate stresses that exceed the material's fracture toughness, leading to phase transformation cracking (PTC). The decoupling coefficient, which characterizes the degree of mechanical decoupling between the phase transformation strain and the elastic-plastic strain field, is identified as a key parameter governing this phenomenon.

Core Technical Concepts

Phase transformation cracking differs fundamentally from hydrogen cracking and solidification cracking. It occurs during cooling in the temperature range of 400-600°C, where the retained austenite transforms to martensite, producing a volumetric expansion of approximately 1-2%. In constrained weld configurations, this expansion generates tensile stresses that can initiate micro-cracks at prior austenite grain boundaries or at the interface between transformed and untransformed regions.

Simulation Parameters and Results

The simulation results demonstrate that increasing the decoupling coefficient from 0.0 to 0.85 reduces the peak transformation stress by approximately 46%. This reduction occurs because mechanical decoupling allows the transformation strain to accommodate more freely through local plastic deformation rather than generating elastic stress. The crack initiation probability drops significantly when the decoupling coefficient exceeds 0.7, suggesting a critical threshold beyond which phase transformation cracking becomes unlikely.

Mechanism Interpretation

The decoupling coefficient physically represents the ability of the local microstructure to accommodate transformation strain without generating excessive stress. In fully coupled systems (coefficient = 0), all transformation strain is converted into stress, creating a high-risk condition. In decoupled systems, the transformation strain is partially absorbed by localized plastic flow, reducing the stress amplitude. This mechanism is particularly relevant for thick-section welds where the thermal gradient is steep and the constraint factor is high.

The study also reveals that the cooling rate interacts synergistically with the decoupling coefficient. At cooling rates above 15°C/s, even a moderate decoupling coefficient of 0.5 is insufficient to prevent cracking, because the rapid transformation generates stress faster than plastic accommodation can occur. Below 8°C/s, a decoupling coefficient of 0.3 provides adequate protection. This finding has direct implications for welding procedure qualification, where cooling rate control through preheat, interpass temperature, and post-weld heat treatment becomes a critical parameter.

Engineering Application

For thick-section 12Cr2Mo1R or 14Cr1MoR hydrogenation reactor welds, the simulation findings suggest that achieving a decoupling coefficient above 0.7 requires either a controlled cooling rate below 10°C/s or a microalloyed weld metal with a lower Ms temperature. In practice, this translates to specifying a preheat temperature of at least 200°C for sections thicker than 50 mm and using a post-weld heat treatment at 680-720°C to relieve transformation stresses. The study validates the conservative approach in NB/T 47014 of requiring post-weld heat treatment for all dissimilar steel welds with Ceq > 0.45.

Study Insights and Reflections

This literature provides a valuable quantitative framework for understanding phase transformation cracking, a phenomenon that is often diagnosed post-failure but rarely predicted during design. The concept of the decoupling coefficient offers a new parameter for welding procedure design that complements the traditional carbon equivalent approach. In my engineering practice, I have encountered phase transformation cracking in a 12Cr2Mo1R column shell weld where the cooling rate was inadvertently elevated due to insufficient interpass temperature control. The simulation approach described in this paper would have enabled early prediction of this failure mode. Future work should extend the decoupling coefficient framework to multi-pass welds with varying thermal histories, which is the more common industrial scenario.