Residual stresses evolution during strip clad welding, post welding heat treatment and repair welding for a large pressure vessel

https://doi.org/10.1016/j.ijpvp.2020.104259Get rights and content

Highlights

  • ā€¢

    An improved moving heat source model for strip cladding was proposed.

  • ā€¢

    The simulated temperature and stress field and agree with the experiment very well.

  • ā€¢

    PWHT cannot completely eliminate the cladding residual stress.

  • ā€¢

    The residual stress on bottom surface was affected by the initial stress model.

  • ā€¢

    A new PWHT method based on distinct temperature regulation was proposed.

Abstract

Strip clad welding is a key manufacturing process of large pressure vessel and repair welding was usually used to repair the defects of strip welded structure, which generates very complex residual stresses and have a great effect on structure integrity. Therefore, in this paper, the residual stress distributions and their evolutions during strip clad welding, post-welding heat treatment (PWHT) and repair welding of a large pressure vessel were investigated. A new strip cladding heat source model considering the magnetic pinch effect was proposed. The predicted welding pool size by the proposed heat source model has a good agreement with the experiment results. And the predicted residual stresses in every stage agree very well with those by indention measurement method. The residual stresses after strip clad welding along the through-thickness direction present tension-compression-tension distribution. The PWHT has little eliminating effect on tensile residual stress on cladding layer, and even worsens the discontinuity stress distributions across the interfaces between the clad metal and base metal. After repair welding, local detrimental tensile residual stresses were generated in the repaired zone. A new heat treatment method based on distinct temperature regulation is proposed, which has a good ability to eliminate the tensile stress of the cladding layer to compressive stress and has a good application in engineering.

Introduction

With the continuous improvement of modern energy technology, the pressure vessels gradually develop to the large scale, and its environment is facing high temperature, high pressure and severe corrosion [1]. Residual stress is an important factor affecting the fatigue performance and stress corrosion resistance of large pressure vessels. Therefore, it is of great significance to accurately evaluate the residual stress distribution in the manufacturing and service stages to ensure the structural integrity of large pressure vessels.

Strip clad welding is a key manufacturing process for pressure vessels such as hydrogenation reactor gasifier and nuclear reactor. The austenitic stainless steel or other corrosive alloys are often surfaced onto the inner wall made of low alloy steel, which evidently improves the corrosion resistance of pressure vessels [2,3]. The distribution of residual stresses induced by strip welding is very complex due to multi-layer, multi-pass, and the different thermal physical and mechanical properties between the clad metal and the base metal [4]. Although post-welding heat treatment (PWHT) is usually performed to eliminate the residual stress of strip welding, it is unclear how much the residual stresses are reduced. After long-term operation, the surfacing layer may be severely thinned or even fell off, as well as local corrosion cracking, which will seriously affect the service performance of the pressure vessels [5]. The repair welding is a commonly used technique to restore this defects or cracks to extend the service life of pressure vessels [6]. Generally, the effect of repair welding on the residual stress in the local repair zone is greater than that of initial weld residual stress [7]. A more complex residual stress distribution is generated in the repaired zone of strip welded joint due to complex constraint in repair zone [8]. Thus, a clear understanding on the evolution mechanism of residual stress during strip welding, heat treatment and repair process is crucial to guarantee the quality of strip cladded pressure vessels.

So far, very little work has been carried out on multilayer clad welding and repair welding though modern computational welding mechanics has been developed to predict the residual stress distribution more accurately [9,10]. For the strip welding on clad plate, the width of strip band is generally above 50Ā mm. The molten pool of strip clad welding becomes very flat and wide. The traditional Goldak double ellipsoid heat source model is no longer suitable for the simulation of wide strip welding [11]. On the other hand, the welding voltage of strip cladding is often very high. During strip welding, strong electromagnetic contraction force will be generated at both ends of the electrode, which will cause the molten pool metal to move toward the center and the redistribution of welding heat flux, thus directly affecting the shape of the molten pool [12]. Udagawa [13] carried out a simulation on the temperature field of submerged arc surfacing welding in which the heat source model was assumed that the heat flux at both ends of the strip electrode was consistent with the middle, and the effect of magnetic shrinkage was ignored. Amudha et al. [14] studied the distribution of the cladding residual stress on the surface of low carbon steel using traditional heat source model, and found that the residual stress of surfacing layer decreased with the increase of the number of surfacing layers. As the thickness of the base metal increases, the residual stress of the matrix and the surface layer increases, and the residual stress of the transition layer increases first and then decreases. Udagawa et al. [15] studied the residual stress distribution of strip surface cladding with single layer on the inner surface of nuclear reactor pressure vessels. It is found that the surfacing layer mainly produces tensile residual stress, and the residual stress near the interface between the surfacing layer and the matrix has a sudden change and a large stress gradient, which is the weak zone of the cladded structure. After PWHT, the residual stresses at the interface may be increased or decreased due to mismatch mechanical properties, thus a detailed analysis of PWHT should be taken to clarify the evolution of residual stresses in the cladded structure.

For the repair welding on cladded structure, residual stress is one of the important factor to affect the strength, fatigue life and corrosion resistance of the cladded structure. A series of studies have been conducted to analyze the repair residual stresses in homogeneous materials, and many useful results for the repair welding technology are obtained [16,17]. The residual stress distribution through-thickness within the repair is affected by various factors like repair parameters, materials and actual component configurations [18]. Jiang et al. [[19], [20], [21], [22], [23]] made a larger number of investigation on the repair welding residual stress of stainless steel/low carbon steel clad plate, and found that the repaired residual stresses can be decreased by changing heat input, welding layer, repair length, repair depth, repair width, clad metal thickness and base metal thickness. However, in their studies, the clad plate is formed by explosive welding, and the effects of original welding residual stresses are not considered, thus may affects the result accuracy. Song et al. [24] carried out an in-depth investigation of repair weld geometry effects on residual stress distributions of girth weld with considering original welding residual stresses. They concluded that a weld repair should be designed as long as possible, as narrow as possible, and as shallow as possible. The residual stress distribution of strip surface welding is different from that of girth weld because of specific welding molten pool shape and the material mismatching between clad and base metal. The repairing residual stress distribution and mechanism in multilayer clad plates remain enigmatic.

To sum up, the current heat source model is not appropriate to characterize the welding molten pool of strip surface cladding, leading to the cladding residual stresses is unclear, thus the evolution mechanisms of residual stresses on cladded structure during strip cladding, heat treatment and repair welding are also not revealed. Therefore, this paper first aims to establish an accurate heat source to simulate the welding temperature of strip clad welding. Then, the residual stress distributions and evolutions in strip cladding, PWHT, and repair welding were investigated systematically by numerical simulation. And the residual stresses in every condition were also measured by indentation strain-gage method, aiming to verify the accuracy of established numerical model. This study will provide better understanding in forming mechanism of residual stresses in strip cladded structures with multi-welding and multi-manufacturing process.

Section snippets

Specimen preparation

Cladded specimen is fabricated using thick plates made of SA387Gr11CL2 pearlitic heat-resistant steel (300LĀ Ć—Ā 300WĀ Ć—Ā 32Ā tĀ mm), as shown in Fig. 1. The strip clad welding was performed by a double layer submerged arc welding strip cladding (SAWSC) method. The welding parameters are listed in Table 1. In this case, the electrode of type 309Ā L stainless steel is used to make transition layer, then the electrode of type 316Ā L stainless steel is used to form surface layer subsequently. After the 6th

Finite element model

The full-scale finite element model was built by ABAQUS 6.10 for the welding and PWHT simulation is shown in Fig. 8, where the shapes and dimensions were defined based on the fabricated conditions mentioned above. The overlapping between passes in transition layer or surface layer overlaying was set to be 7.67Ā mm, and the thicknesses of surface layer and transition layer were 4Ā mm and 3Ā mm, respectively. The pass sequences during surface overlaying and repair welding are shown in Figs. 8 and 9.

Verification of heat source model

Fig. 13 shows the comparison of strip cladding temperature field variations at six measured points during the strip cladding process by experiment and simulation. It can be seen that the predicted cladding temperature field by proposed heat source model agreed very well with the experimental results. The predicted peak temperature by proposed heat source model is slightly lower than the experimental peak temperature. The maximum error of peak temperature is less than 9.1%. The first peak

Discussion

Based on above analysis, the residual stresses are redistributed after repair welding, especially the longitudinal stresses in the weld surface are increased. The repair welding is conducted immediately after PWHT on above analysis. However, the defects of pressure vessels may be generated in the early, middle or late stages of service. On one hand, some welding defects are generated during the strip cladding, such as, undercut, which is a groove or depression in weld toe. So, the repair

Conclusion

This paper presents a study on the evolution mechanism of residual stresses induced by strip clad welding, PWHT and repair welding in surfacing welded structure. The birth and death element technique has been used to simulate the deposition of claddings, as well as a sequential thermal mechanical analysis has been applied to discuss the residual stress evolution during PWHT and repair welding. Following conclusions can be summarized based on the obtained results:

  • (1)

    An improved moving heat source

Author statement

Wenchun Jiang: Methodology, Logical structure. Yun Luo: Data curation, Investigation, Writing-Original draft preparation. Qiang Zeng: Software, Validation, Experiment. Jinguang Wang & Shan-Tung Tu: Writing- Reviewing and Editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors gratefully acknowledge the support provided by National Natural Science Foundation of China (51905545), National Key R&D Program of China (2018YFC0808800), Fundamental Research Funds for the Central Universities (-20CX02219A; -18CX05002A) and the Natural Science Foundation of Shandong Province (ZR2019MEE108).

References (38)

Cited by (21)

  • Generation of compressive residual stress at the root of tube-to-tubesheet welded joints in a heat exchanger

    2022, International Journal of Pressure Vessels and Piping
    Citation Excerpt :

    In the process of simulation, the influence of element number on residual stress has been analyzed. The mesh sensitivity was checked until the difference of FE results for various element number was less than 5% [53]. A sequential coupling method was applied for the thermal and stress analysis.

View all citing articles on Scopus
View full text