Numerical Simulation of Dissimilar Metal Surfacing Between Low Alloy Steel and Stainless Steel

Literature Overview

The paper by Jiang Xiaohua, Dai Deping, Cai Jianpeng, and Deng De'an from Chongqing University (2016), funded by the National Natural Science Foundation of China (Grant No. 51275544), presents a finite element analysis of temperature field and residual stress distribution in dissimilar metal surfacing welds between low alloy steel and stainless steel. The study specifically addresses nuclear power equipment applications where dissimilar metal welds are ubiquitous due to the need for different material properties at different locations within a single component.

Core Technical Framework

The authors developed a thermal-elastoplastic finite element model using ABAQUS software with a Gaussian surface heat source to simulate the moving heat source during surfacing. The model captures:

Heat Source Model

The Gaussian surface heat source represents the arc energy input distributed over the weld pool surface. The power density distribution follows:

q(r) = (3Q / πa²) × exp(-3r²/a²)

where Q is the effective arc power and a is the characteristic radius. The model was validated by comparing simulated weld pool geometry with actual production weld bead profiles, demonstrating good agreement.

Key Findings on Residual Stress Distribution

The simulation results reveal several critical findings:

Interface Stress Discontinuity

Due to the significant differences in material properties between low alloy steel (e.g., 18MnMoNb) and stainless steel (e.g., 304/316), a pronounced discontinuity in residual stress exists at the metallurgical interface. This occurs because:

1. The coefficient of thermal expansion (CTE) mismatch creates differential thermal strains during cooling
2. Different yield stress temperatures lead to asymmetric plastic deformation
3. The thermal conductivity difference causes non-uniform heat flow patterns

Multi-Pass Annealing Effect

One of the most practically significant findings is that subsequent weld passes exert a tempering effect on previously deposited passes. This self-annealing mechanism:

• Reduces peak residual stresses in earlier passes
• Can modify the microstructure of the HAZ in prior passes
• Creates a non-uniform stress profile through the weld thickness
• Must be considered when predicting fatigue life and creep behavior

Standards and Engineering Context

Dissimilar metal welds in nuclear and process piping are governed by several critical standards:

The residual stress state predicted by this simulation directly impacts:

• Stress corrosion cracking (SCC) susceptibility in the weld region
• Creep rupture life under sustained high-temperature loading
• Fatigue crack initiation and propagation behavior
• The effectiveness of post-weld heat treatment (PWHT)

Engineering Practice Integration

For pipe fitting engineers, dissimilar metal surfacing is encountered in several scenarios:

• CRA overlay on carbon steel pipe: Overlaying corrosion-resistant alloy (e.g., 310SS, 6Mo) on carbon steel or low alloy steel substrates for sour service or high-temperature oxidation protection
• Stainless steel weld overlay on low alloy pipe: Common in high-temperature service where the base material provides strength while the overlay provides oxidation resistance
• Transition fitting fabrication: Where different material grades must be joined in a single component

The simulation results emphasize that residual stress at the interface is the primary driver of interface cracking and delamination. Engineers should consider:

1. PWHT optimization: The temperature and duration of PWHT must be sufficient to relieve the stress discontinuity without sensitizing the stainless steel overlay
2. Interlayer design: Buffer layers or graded interlayers can reduce the property mismatch
3. Weld sequence planning: The multi-pass annealing effect should be leveraged to reduce overall residual stress
4. NDE requirements: Interface inspection (e.g., TOFD or PAUT) is essential to detect any pre-existing or developing interface cracks

Key Questions and Reflections

The paper raises important questions about the applicability of continuum-based FEA models to dissimilar metal welds where the microstructure varies significantly over distances of micrometers. The macro-element approach used here averages material properties over the element volume, which may mask critical local effects at the interface.

Furthermore, the model does not explicitly account for:

• Solidification cracking at the interface due to composition segregation
• Phase transformation effects in the HAZ of the low alloy steel substrate
• The influence of weld sequence on solidification mode (columnar vs. equiaxed)

These limitations suggest that while the macro-FEA provides valuable engineering-level predictions, complementary micro-FEA or phase-field simulations are needed for detailed microstructural predictions.

Study Insights and Implications

This work demonstrates the growing importance of numerical simulation in welding process development and qualification. For engineers involved in dissimilar metal weld design, the ability to predict residual stress distributions without extensive experimental testing significantly reduces development time and cost.

The finding regarding the multi-pass annealing effect has direct implications for welding procedure specification (WPS) development. When qualifying multi-pass surfacing procedures on dissimilar substrates, the interpass temperature and pass sequence should be explicitly controlled and documented, as they significantly influence the final residual stress state.

The research also highlights a gap in current standards: most welding codes specify acceptance criteria based on mechanical properties and NDE but do not address residual stress requirements. For critical applications such as nuclear piping, incorporating residual stress predictions into fitness-for-service assessments represents a valuable extension of current practice.