Queen Mary University of London University of Cambridge

Prediction of the Microstructure of Submerged-Arc Linepipe Weld

Advances in Welding Technology and Science, ASM, Metals Park, Ohio, U. S. A., 1987, pp. 225-229. H.K.D.H. Bhadeshia, L.-E. Svensson and B. Gretoft

A model, based on phase transformation theory, has been developed to allow the prediction of steel weld microstructures. It requires the input of just chemical composition and welding conditions; this facilitates the calculation of the appropriate part of the phase diagram (for steels containing C, Mn, Si, Ni, Mo, Cr, V in any reasonable combination), needed to define the relevant paraequilibrium tie-lines for detailed kinetic analysis. TTT and CCT curves are also computed, together with various transformation start and finish temperatures. These data in turn lead to the estimation of the volume fractions of allotriomorphic ferrite, Widmanstatten ferrite and acicular ferrite. The theory is found to be in good agreement with experimental data.

The provided documents detail a mathematical model designed to forecast the primary microstructure of low-alloy steel fusion zones during the welding process. By integrating phase transformation theory with variables such as chemical composition and heat input, the researchers can estimate the volume fractions of different ferrite phases.

This specific study highlights how changes in silicon and manganese concentrations influence the resulting metallic structure, often yielding results that contradict simple isothermal experiments. The technical analysis confirms that the theoretical predictions align closely with experimental data from manual metal-arc welds.

Ultimately, the work offers a rigorous framework for understanding the complex metallurgical transitions that occur as weld deposits cool and solidify.

Download publication

Audio podcastAudio podcast

Study Guide: Prediction of Microstructure in Multicomponent Steel Weld Deposits

Review of research by H.K.D.H. Bhadeshia, Lars-Erik Svensson, and B. Gretoft regarding low-alloy steel weld metals.

Short-Answer Quiz

Instructions: Answer the following questions using 2–3 sentences based on the provided research context.

  1. What are the primary inputs required by the model to predict steel weld metal microstructures?
  2. How does the solidification process of low-alloy steel welds begin at the fusion boundary?
  3. In the context of the model, how is the morphology of the austenite (γ) grains represented?
  4. Describe the mechanism by which allotriomorphic ferrite (α) grows from the austenite boundaries.
  5. What differentiates the nucleation of acicular ferrite (αa) from Widmanstätten ferrite (αw)?
  6. What happens to the small amount of remaining austenite as the temperature reaches the martensite start (Ms) temperature?
  7. How is the cooling rate of the weld deposit typically approximated in manual-metal-arc (MMA) welds?
  8. Why can simple isothermal experiments be misleading when predicting the effect of alloying elements like Silicon on weld microstructures?
  9. How does the model address the segregation of alloying elements during the solidification process?
  10. What was the conclusion regarding the correlation between the model's calculated volume fractions and experimental data?

Answer Key

Topic Detailed Answer
Inputs The model requires the input of only the chemical composition and the welding conditions. These inputs allow for the calculation of relevant phase diagram data, paraequilibrium tie-lines, and Time-Temperature-Transformation (TTT) and Continuous Cooling Transformation (CCT) curves.
Solidification Solidification begins with the epitaxial nucleation of δ-ferrite at the fusion boundary. The melt then solidifies through the cellular growth of δ, which eventually transforms into a columnar austenite (γ) grain structure.
Morphology The morphology of the γ grains is represented as a honeycomb of hexagonal prisms. The c-axes of these prisms are oriented approximately parallel to the direction of maximum heat flow during the solidification process.
Diffusional Growth Allotriomorphic ferrite grows via a diffusional transformation mechanism at the γ/γ boundaries. As the temperature falls, this growth becomes more difficult, eventually giving way to the nucleation of Widmanstätten ferrite.
Nucleation Comparison Widmanstätten ferrite nucleates at the α/γ boundaries and grows into the austenite via a displactive transformation mechanism. In contrast, acicular ferrite nucleates on inclusions within the austenite grains and grows in the form of thin plates.
Microphases The remaining austenite decomposes into "microphases," which consist of mixtures of martensite, retained austenite, or degenerate pearlite. Because the volume fraction of these phases is relatively small, they are often not measured or calculated separately in the model.
Cooling Rate The cooling rate (dT/dt) is approximated using an equation that considers the arc transfer efficiency, the interpass temperature, and the heat input (Q). For MMA welds, the cooling rate between 800–500°C is a critical range for these approximations.
Alloying Effects Isothermal experiments do not account for the continuous cooling conditions of actual welding, where changes in transformation start temperatures (Th) can alter the time available for different phases to grow. For example, while Silicon may theoretically accelerate ferrite nucleation, the model shows its effect is tempered by the reduced time available for growth during continuous cooling.
Segregation The model uses a partition coefficient (ki) to identify the ratio of the mole fraction of an alloying element in δ to that in the liquid at a given temperature. This allows the researchers to calculate the compositions of solute-depleted regions, which are the sites where ferrite nucleation is enhanced.
Validation The model demonstrated good agreement with published experimental results, showing a correlation coefficient of 0.96 when plotting calculated versus experimental data. The mean differences between calculated and experimental volume fractions for various ferrite types were consistently low, ranging from 0.05 to 0.09.

Essay Format Questions

Instructions: Use the provided source context to develop detailed responses to the following prompts.

Glossary of Key Terms

Term Definition
Acicular Ferrite (αa) A microstructure consisting of thin plates that nucleate on inclusions within austenite grains; it is known for its contribution to toughness.
Allotriomorphic Ferrite (α) Ferrite that forms at the prior austenite grain boundaries via a diffusional mechanism; its shape does not reflect its internal crystal structure.
Columnar Austenite (γ) The grain structure that forms when δ-ferrite transforms during cooling; modelled as hexagonal prisms in this research.
Displactive Transformation A phase transformation mechanism involving the coordinated movement of atoms, characterised by the growth of Widmanstätten and acicular ferrite.
Epitaxial Nucleation The process where new crystals grow with a specific crystallographic orientation relative to the substrate (in this case, the fusion boundary).
Inclusions Small particles within the weld metal (often oxides or silicates) that serve as nucleation sites for acicular ferrite.
Microphases Minor constituents in the microstructure, including martensite, retained austenite, and degenerate pearlite, that form from remaining austenite.
Paraequilibrium A state where the ratio of iron to substitutional alloying elements remains constant across the transformation interface, while carbon reaches local equilibrium.
Partition Coefficient (ki) The ratio of the mole fraction of an element in the solid phase (δ) to its mole fraction in the liquid phase during solidification.
Widmanstätten Ferrite (αw) Ferrite that nucleates at α/γ boundaries and grows into austenite grains as plates at a rate controlled by carbon diffusion.