DOI: https://doi.org/10.21203/rs.3.rs-2761781/v1
Repair welding is a common method to mitigation of damaged parts including creep voids and cracks in high temperature pressure pipelines. However, new cracks may be created due to the mechanical property degradation of aging material and excessive repair welding residual stress. In order to make a scientific and reasonable life extension or replacement decision, this paper carried out a feasibility study on the repair welding of an aged CrMo steel high temperature pressure pipe weldment. Firstly, the modified Kachanov-Rabotonov creep damage constitutive equation was used to predict the creep damage distribution of the weldment served for 26 years, and the most serious damage location for scarfing is determined. Then, the simulation study of welding residual stress was carried out with focus on the influence of material performance degradation and excavation methods. The results show that the maximum creep damage of the CrMo steel pipe weldment is concentrated in the heat affected zone (HAZ) and the adjacent base metal (BM).In total, the repair welding stress simulated with the degraded mechanical properties are lower than that with the initial mechanical properties. But, the stress discontinuity at the interface between the repair welding zone and BM is more severe due to the high level mismatch of mechanical properties, which is a key risk inducing repair welding cracks.The step repair method is recommended for engineering application based on the consideration from aspects of avoidance of the stress concentration inside the weld, and weaken the stress discontinuity of the weld boundary.
The weldment is the weak location of the high temperature pressure pipeline [1]. It is easy to arise defects such as creep voids and micro-cracks after long-term service, which threatens the subsequent service safety. In order to balance safety and economy, repair welding is a common engineering method to mitigate the above defects [2]. However, new welding cracks may be induced during the repair welding process, due to the combined effects of residual stress and degradation of material properties. Instead, it may lead to accelerated failure of the repaired welding joint.[3] Therefore, in order to evaluate the feasibility of repair welding reasonably, it is necessary to further understand the residual stress of repair welding and its influence on structural safety.
Many researchers have carried out a series of studies on repair welding residual stress form aspects of process parameters, groove shape and repair welding geometry [4, 5]. For example, Jiang et at. found that multi-layer welding and high heat input welding are beneficial to reduce residual stress [6]. The residual stress of deep repair weld is greater than that of shallow repair weld [7]. Yang K et al [8] found that there are proportionate relationships between the repair welding energy input and the HAS width. Sattari-Far I et al. [9] proposed that the shape of weld groove and the number of weld passes have a significant effect on the amplitudes and distribution of repair welding residual stress. Huang Y X [10] simulated the residual stress distribution in the repair welding zone of HP40 furnace tube, and found that the increase of length and width reduce the residual stress.
However, the above research mainly pay attention to the influence of repair welding process parameters on residual stress, the influence of material performance aging on repair welding residual stress is rarely considered. Moreover, the current research are only limited to the analysis of the distribution and evolution of residual stress in repair welding, the feasibility of repair welding was not studied. Therefore, the simulation study of repair welding residual stress was carried out for a CrMo steel weldment served for 26 years. The influence of material performance degradation and repair method on repair welding residual stress were mainly considered, and then the main factors affecting the feasibility of repair welding were analyzed, which provided a scientific reference for optimizing the repair welding process.
The geometric modeling of the cross-weld creep specimen was established based on the standard creep specimen geometry size [11], as shown in Fig. 1 (a). There are 30952 nodes and 159674 elements, and the element type is C3D4. One end of the model is set as a fixed constraint, and the other end is applied with a uniform tensile force of 46.87 MPa to ensure that the stress of the gauge section is 120 MPa. The validity of the constitutive model is verified by comparing the simulation results with the experimental results.
In order to predict the residual life and the maximum creep damage location of the CrMo steel pipe weldment served for 26 years, a 2D axisymmetric model was established, as shown in Fig. 1 (b). The outer diameter of the pipe is 274 mm, the wall thickness is 34 mm, and the welding groove is a V-groove. The chemical compositions of the weldment are listed in Table 1. The temperature field is calculated by DCAX4 element, and the stress field is calculated by CAX4R element. The axial displacement is constrained at one end of the pipe, and the radial displacement is constrained on the outer surface. The service temperature is 520°C and the internal pressure is 11.5 MPa.
In order to obtain the optimal repair welding method, the FEM model with partial excavation, full excavation and stepped full excavation (see Fig. 2) were established for repair welding residual stress simulation The geometric size of the model is the same as that in Fig. 1 (b). In order to reduce the residual stress at the edge of the repair weld and weaken the stress discontinuity at the weld boundary effectively, the repair welding is carried out from bottom to top and from both sides to the middle. The specific weld bead sequence is shown in the number of the repair area, and the geometric size of the repair area is shown in the Table 2. In order to study the influence of material performance aging on the residual stress of repair welding, the repair welding simulation under partial excavation conditions was carried out by using the initial performance and aging performance of CrMo steel respectively (In order to save space, this paper does not repeat other repair forms).
The temperature field is calculated by DCAX4 element, and the stress field is calculated by CAX4R element. The weld, HAZ and part of the BM were refined to ensure the calculation accuracy. In the analysis of welding residual stress, a fixed constraint is set at one end of the pipeline, and a constraint that can only be freely retracted in the axial direction is set at the other end.
|
Composion |
C |
Si |
Mn |
Cr |
Mo |
Ni |
S |
P |
|---|---|---|---|---|---|---|---|---|
|
Service steel |
0.118 |
0.733 |
0.531 |
1.152 |
0.373 |
0.117 |
0.0099 |
0.013 |
|
Filler material |
0.078 |
0.396 |
0.911 |
1.204 |
0.387 |
0.138 |
0.0093 |
0.014 |
|
Original steel |
0.119 |
0.729 |
0.507 |
1.185 |
0.418 |
0.113 |
0.0087 |
0.012 |
|
Method |
Width(mm) |
Depth(mm) |
Bevel angle |
|---|---|---|---|
|
Partial |
42 |
16 |
30° |
|
Full |
42 |
29 |
30° |
|
Stepped full |
46/29 |
16/10 |
30° |
In order to describe the primary, secondary and tertiary creep stage simultaneously, the Modified Kachanov-Rabotonov constitutive model [12] is used to predict creep damage and life, the detailed equations are as follows:
where \({\mathop \varepsilon \limits^{\cdot } _c}\) and \(\mathop \omega \limits^{\cdot }\) refer to creep strain rate and creep damage rate, repectively. \(\sigma\) is the stress. \(\omega\) is creep damage. A and n are material constant and stress exponent. The model parameters of service materials at 520℃ are listed in Table 3.
|
A |
n |
m |
B |
|
|---|---|---|---|---|
|
BM |
2.583e-27 |
10.914 |
0.3724 |
3.5447e-16 |
|
HAZ |
1.041e-25 |
11.093 |
0.2725 |
7.759e-10 |
|
WELD |
2.0575e-19 |
7.8689 |
0.2482 |
1.005e-14 |
|
n0 |
\(\chi\) |
\(\varphi\) |
M |
|
|
BM |
6.06445 |
11.40286 |
11.3047 |
2.0448e-26 |
|
HAZ |
4.079 |
11.1928 |
12.715 |
4.422e-28 |
|
WELD |
1.5262 |
9.0517 |
6.776 |
1.2167e-23 |
The thermal and mechanical properties of materials used in creep damage distribution simulation and thermal stress finite element simulation are shown in Figure.3.
Figure 4 shows the creep damage simulation results of the cross-weld specimen, in which the maximum damage located at HAZ with the fracture time of 1340 h. Obviously, both the calculated peak damage location and the rupture time is consistent with the experimental results (rupture time is 1332.4 h). It indicates that the creep constitutive model used in this paper can predict the maximum creep damage location and fracture time of the weldment accurately。
The creep damage contours of the CrMo steel pipe weldment at 11.12 kh and 12.45kh are shown in Fig. 5. It is found that the peak creep damage occurs in the HAZ (near the BM) at 11.12 k h, and then extends from the outer wall of the pipe to the inner wall along a path with 60°. Peak creep damage can be regarded as microcracks. At 12.45 k h, the length of microcrack is extended to 6mm and the depth is close to 5mm. At this time, the failure part needs to be removed and repaired.
Figure 6 shows that the repair welding residual stress contours of partial excavation. It is found that regardless of whether performance degradation is considered, the distribution of axial residual stress and hoop residual stress in the repair weld is similar. But, the distribution of residual stress in the weld edge, the inner and outer walls of the pipe weldment and the new HAZ is obviously different. The repair welding stress simulated with the degraded mechanical properties are lower than that with the initial mechanical properties. The stress discontinuity at the interface between the repair welding zone and BM is more severe due to the high level mismatch of mechanical properties. It is related with the degradation of the tensile properties and the decrease of the yield strength of the pipeline after long-term service. This shows that the repair welding of pipeline joints after service is more likely to cause stress corrosion on the inner wall.
As shown in Fig. 7, the influence of performance aging is mainly manifested in the obvious residual stress difference between the service steel and the original steel on the inner wall of the pipeline and the fusion line of the repair weld. Combined with Fig. 10, in the inner wall weld and HAZ, the axial residual tensile stress of the service steel is 180MPa-210MPa, which is 1.5 times that of the original steel. This is because the performance of the inner wall is seriously degraded, especially the yield strength and yield ratio, when the pipeline is in high temperature and pressure steam environment for a long time. Therefore, service steel will be more likely to produce stress corrosion cracking in the inner wall. In other words, using the data of the original steel for simulation will make the results more optimistic and underestimate the axial tensile stress on the inner wall, which will indirectly lead to early stress corrosion cracking of the pipeline after repair welding. In other words, using the data of the original steel for simulation will make the results more optimistic and underestimate the axial tensile stress on the inner wall, which will indirectly lead to early stress corrosion cracking of the pipeline after repair welding. Compared with the original steel, the residual stress value of the service steel on the BM side of the fusion line is significantly different, but it is smaller on the side of the weld, which increases the residual stress difference at the boundary of the repair weld and increases the discontinuous stress. Compared with the original steel, the residual stress value of the service steel on the BM side of the fusion line decreases significantly, but it does not change much on the weld side, resulting in an increase in the discontinuous stress at the boundary of the repair weld. As shown in Fig. 8, the tensile stress of the service steel in HAZ is only half of the original steel, and their hoop stress difference is close to 200 MPa. In addition, the interface mismatch of the service steel on the outer surface is larger than that of the original steel. Figure 9 shows that the interface mismatch is more manifested in the axial stress.
In summary, the mismatch between the repair welding material and the substrate after service is greater, and the performance mismatch increases. And it will produce a discontinuous stress distribution with a greater difference at the fusion line, and stress corrosion cracking is more likely to occur on the inner wall of the pipeline. For long-term service CrMo steel materials, when studying the residual stress of repair welding, if the original steel data is used for simulation, the results will be too optimistic, underestimating the influence of cross-section mismatch, and affecting the repair and life extension effect of the pipeline.
Figure 11 shows the axial stress distribution and hoop stress distribution under three kinds of excavation and repairing methods. It is found that the axial compressive stress of full excavation and stepped full excavation in the middle of the pipe thickness is 259.4MPa and 256MPa, which is slightly larger than 233MPa of partial excavation. The maximum axial tensile stress is all observed at the root of the last weld, indicating that the maximum axial tensile stress depends on the welding sequence and weld strength, and has little effect on the non-welded area. In the outer wall of the pipe, all the three methods are distributed with 12 similar weld beams, so the distributions of residual stress are almost the same.
As shown in Fig. 10, the ranges of the compressive stress obtained by the three methods are almost the same, and the maximum value of the hoop compressive stress of the stepped full excavation is the largest, which is 290.1 MPa, followed by 271.2 MPa of full excavation and 262.2 MPa of partial excavation. It could be seen that the stepped full excavation and full excavation are the preferred excavation methods.
Figure 12 is the axial stress and hoop stress distribution curve along the P1 path. For AXIAL STRESS, the stress discontinuity at the fusion line exists in all three methods, but not obvious. For hoop stress, there is almost no stress discontinuity at the fusion line between full excavation and stepped full excavation. Due to the multi-layer and multi-pass welding, the stress of full excavation and stepped full excavation gradually increases, avoiding the stress concentration inside the weld and the HAZ, which can effectively reduce the probability of internal cracking of the pipeline.
Figure 13 is the axial stress and hoop stress distribution curve along the P3 path. The residual stress induced by partial excavation, full excavation and stepped full excavation in HAZ did not exceed the yield strength (229.6MPa) after aging, so the residual life would not be affected by plastic deformation. However, the discontinuous stress in partial excavation situation is in the area with large creep damage (in the middle of the wall thickness), which will accelerate the creep damage rate and weaken the effect of repair welding. The full excavation and stepped full excavation not only removed the defects, but also removed the material with larger creep damage in the middle of wall thickness.
Figure 14 shows the axial stress and hoop stress distribution curve along the P4 path. In the original HAZ and the nearby base metal, the axial tensile stress in the full excavation weld is the largest among the three methods, followed by stepped full excavation, partial excavation. The repair weld seam of partial excavation is relatively far away from the inner wall of the pipeline, so the residual stress has little effect on the inner wall. The axial tensile stress in this area caused by full excavation has exceeded the yield strength of the service steel (229.6MPa), which will induce plastic deformation in the inner wall of the joint and will be more likely to produce stress corrosion cracking. It is contrary to the original intention of repair welding.
In summary, among the three repair welding methods of CrMo steel high-temperature pressure pipelines in extended service, the stepped full excavation has the best effect, which reduces the possibility of stress corrosion cracking of the inner wall of the joint, avoids the stress concentration inside the weld, and effectively alleviates the stress discontinuity of the weld boundary.
This work investigated the distribution of creep damage and welding residual stress in the weldments of a CrMo steel high temperature pressure pipeline by numerical and experimental study. The following conclusions can be drawn:
The maximum creep damage in the weldment is concentrated in the HAZ and the nearby base metal. The peak creep damage occurs in the HAZ (near the BM) at 11.12 k h. At 12.45 k h, the length of microcrack is extended to 6mm and the depth is close to 5mm.
The overall residual stress considering the degradation of mechanical properties is smaller than that without considering the degradation of mechanical properties. And more serious stress discontinuity will occur at the interface between the repair welding zone and the substrate, which is a potential threat to induce repair welding cracks. There is greater tensile stress on the inner wall of the pipeline, which is easy to produce stress corrosion cracking on the inner wall.
It is recommended to adopt the stepped full excavation method when repairing the weldment of CrMo steel high temperature pressure pipeline with extended service. The stepped full excavation can not only avoid the stress concentration inside the weld, but also effectively alleviate the stress discontinuity of the weld boundary.
Acknowledgements
Not applicable.
Authors’ contributions
Bin Yang: Methodology, Logical structure. Minghao Xiu: Data curation, Investigation, Writing- original draft preparation. Wenchun Jiang: Logical structure, Writing- Reviewing & Supervision. Wei Peng: Validation, Experiment design, Data processing.
Authors’ information
Bin Yang is currently an associate professor at China University of Petroleum (East China). His research interests include life assessment of energy equipment and Fracture mechanism of welded joint.
Minghao Xiu is currently a master candidate at China University of Petroleum (East China).
Wenchun Jiang is currently a professor at China University of Petroleum (East China). He received his PhD degree from Nanjing Tech University, China, in 2009. His research interests include simulation, measuring, and controling of welding residual stress in pressure vessels.
Wei Peng is currently a PhD student at China University of Petroleum (East China). His research interests include simulation and measuring of welding residual stress.
Funding
Supported by National Natural Science Foundation of China (Grant No. 52275168), National Natural Science Foundation of China (Grant No. 51805546), and Program of Shandong Province Natural Science Foundation (Grant No. ZR2019BEE050).
Competing interests
The authors declared that they have no conflicts of interest to this work.