JP4419802B2 - Method for analyzing residual stress in welded joints - Google Patents

Description

  The present invention relates to a method for analyzing residual stress in a welded joint portion of a large-diameter thick-walled tube by FEM analysis using an axisymmetric shell model.

  Conventionally, in thick pipe welds used in nuclear power plants, in order to prevent stress corrosion cracking (hereinafter referred to as SCC) due to welding residual stress, a high frequency induction heating residual stress improvement method ( Induction Heating Stress Improvement—hereinafter referred to as IHSI).

  IHSI is a heat treatment method using induction heating. A coil is wound around the outer surface of the object to be heated (the welded portion of the pipe), and the surface of the object to be heated is heated by induction heating by passing an alternating current through the heating coil. The principle of induction heating is that an alternating magnetic flux generated by an alternating current penetrates an object to be heated and induces a very high density current (eddy current) to enable heating in a short time. On the other hand, during IHSI construction, the inside of the pipe is cooled by water (for example, by flowing cooling water at a flow rate of 2 m / sec, about 300 ° K (Kelvin) or by stagnating the cooling water in the pipe), the inner surface of the pipe is cooled. A temperature difference can be generated in the pipe, and a compressive residual stress can be introduced to the inner surface side of the pipe (or a tensile residual stress can be relaxed) by thermal strain.

  In implementing IHSI, the dominant parameters of the residual stress relaxation effect are as follows: (1) Maximum heating temperature, (2) Necessary temperature difference between inner and outer surfaces, (3) Coil width, (4) Heating time, (5) There are five relative positions of the weld line and the coil width center (see Non-Patent Document 1). These values are referred to as EV (essential variable), and the construction conditions of IHSI are planned by confirming the relationship between EV and residual stress before construction.

  The relationship between the EV and the residual stress is derived by conducting an experiment with the set EV and actually measuring the residual stress. However, it is time / experimental to obtain effective IHSI construction conditions for all pipes. Difficult in terms of cost.

  Therefore, it is necessary to estimate the residual stress distribution in the vicinity of the weld joint where SCC is a concern by performing FEM analysis for the processes from welding to IHSI, and to clarify effective IHSI construction conditions by analysis.

  FEM analysis is a method in which the periphery of a welded portion of a pipe is divided into a plurality of elements, data for each element is input into a basic expression, an equation is assembled, and residual stress is analyzed based on the equation. Conventionally, in order to calculate the residual stress of a welded joint portion of a pipe, an analysis is performed as it is with a general analysis code using an axisymmetric shell model (see Non-Patent Document 2).

Iida et al., “Guidelines on Stress Relieving Method by High Frequency Induction Heating (SCC Countermeasure Method) TNS-G2804-1985”, Japan Thermal Power Technology Association, Nuclear Power Technology Committee, p. 5-26 "Finite Element Method Handbook I, II", published in 1992, Baifukan

  By the way, in an actual plant, it is expected to prevent damage due to SCC by implementing IHSI construction regardless of existing or new construction. Existing plants have a wide variety of pipe shapes (bore diameter and wall thickness), and it is necessary to clarify the IHSI construction conditions for effectively reducing the residual welding stress for each pipe joint. There is.

However, when the above-mentioned FEM analysis was performed, the difference between the analysis value and the actual measurement value was small in a small-diameter pipe having a diameter of about 50 mm, which was not a problem, but in a large-diameter and thick-walled pipe exceeding 500 mm in diameter, There is a problem that an error between the analysis value and the actual measurement value increases. In particular, the analysis value is calculated to be lower than the actual measurement value (it appears on the more advantageous side than the actual measurement value), so that it is difficult to use it for the evaluation of IHSI.

  When performing FEM analysis on large-diameter thick-walled pipes, if a moving heat source that simulates actual welding is analyzed using a 3D shell model or a 3D solid model, The error can be reduced. However, the error can be made very small in a three-dimensional model, especially a three-dimensional solid model, but the calculation becomes very complicated, so the analysis time is compared with the analysis time (for example, two hours) in the case of an axisymmetric shell model. There is a problem that a lot of time (for example, three months) is required.

  Therefore, the present invention has been devised to solve the above problems, and its purpose is to provide an analysis method capable of accurately calculating the residual stress of a welded joint portion of a large-diameter thick wall pipe in a short time. It is in.

  The invention of claim 1 divides the residual stress of the welded joint portion of the large-diameter thick wall pipe into a plurality of elements using an axisymmetric shell model,

In the analysis method for FEM analysis based on the above, the predetermined element of the axisymmetric shell model includes the radial displacement calculated by the FEM analysis at the time of all-around heating by the normal axisymmetric shell model and the local heating by the three-dimensional shell model. In addition to introducing a radial spring constraint having rigidity according to the difference from the radial displacement calculated by FEM analysis of the above, a predetermined element of the axisymmetric shell model is subjected to the entire circumferential heating by the normal axisymmetric shell model . FEM analysis is performed by introducing an axial spring constraint having rigidity corresponding to the difference between the axial displacement calculated by FEM analysis and the axial displacement calculated by FEM analysis at the time of local heating by a three-dimensional shell model. This is a method for analyzing residual stress in a welded joint.

According to the invention of claim 2, the introduction position of the radial spring constraint and the axial spring constraint, and the rigidity of the radial spring constraint and the axial spring constraint are the displacement behavior of the analysis by the axially symmetric shell model by FEM trial analysis. 2. A method for analyzing residual stress in a welded joint according to claim 1, wherein a value at which the displacement behavior of the analysis by the three-dimensional shell model coincides is calculated and set.

In the invention of claim 3, the radial direction and the axial displacement by the axisymmetric shell model are calculated by an analysis during the entire circumference heating of the large-diameter thick wall tube by the axisymmetric shell model, and the radial direction by the three-dimensional shell model is calculated. The axial displacement is a method for analyzing a residual stress in a welded joint according to claim 1 or 2, which is calculated by an analysis at the time of local heating of a large-diameter thick-walled pipe using a three-dimensional shell model.

In the invention of claim 4, the large-diameter thick wall pipe is made of stainless steel 600A piping, and the radial spring restraint is introduced at a position of ± about 30 mm in the axial direction from the weld center of the weld joint portion. Its rigidity is 2.581 × 10 ¹⁰ (N / m), and the axial spring restraint is introduced at a position of ± 100 mm in the axial direction from the welding center of the weld joint, and its rigidity is 11.621. × a ¹⁰ 10 (N / m) a method for analyzing the residual stress of the welded joint portion according to claims 1 to 3 or is.

  According to this invention, the outstanding effect that the residual stress of the welded joint part of a large diameter thick wall pipe can be calculated correctly in a short time is exhibited.

  High Frequency Induction Heating Residual Stress Improvement Method (IHSI) has been developed as one of the SCC countermeasures for austenitic stainless steel pipes in existing nuclear power plants, and improves residual tensile stress in the heat affected zone near the weld line. It is a construction method.

  IHSI is a heat treatment method using induction heating, in which a coil is wound around the outer surface of an object to be heated (piping weld), and the surface of the object to be heated is induced by induction heating by passing an alternating current through the heating coil. Can be heated. The principle of induction heating is that an alternating magnetic flux generated by an alternating current passes through an object to be heated and induces a very high-density current (eddy current) to enable heating in a short time. On the other hand, during IHSI construction, the inside of the pipe is water-cooled (for example, the pipe is filled or flushed with about 300 ° K (Kelvin) (approximately 27 ° C) cooling water), and the inside surface of the pipe is cooled with water cooling inside the pipe. A temperature difference is produced on the outer surface, and compressive residual stress can be introduced into the inner surface of the pipe by thermal strain (or relaxation of the tensile residual stress).

  About EV which becomes a dominant parameter of the residual stress relaxation effect in implementing IHSI, the construction condition of IHSI is planned by confirming the relationship between EV and residual stress before construction.

  In the present embodiment, the residual stress for the set EV is calculated by FEM analysis. Then, it is confirmed whether the calculated residual stress is a desired residual stress.

  In FEM analysis, the part to be analyzed (around the welded part of the pipe) is divided into a plurality of elements, data for each element is input to the computer, and an equation is constructed based on the basic formula by the computer. Analyze the residual stress.

  In this embodiment, it is divided into elements using an axisymmetric shell model. As shown in FIG. 3, the element division is finely divided near the boundary 61 between the welded portion 2 and the base material portion 1 of the pipe (large-diameter thick-walled tube) 62, which is particularly important in the analysis of stress relaxation by IHSI. is doing. Moreover, the welding part 2 is divided | segmented into the element E so that it may correspond to a welding pass.

  Then, each element E is analyzed based on the following basic formula.

  In the above equation, r represents the radial direction of the pipe, θ represents the circumferential direction of the pipe, z represents the axial direction of the pipe, σ represents stress, τ represents shear stress, ε represents direct strain, and γ represents shear strain. [D] indicates an elastic matrix, and a stress σ and a shear stress τ are expressed by applying a direct strain ε and a shear strain γ to the elastic matrix.

  In the present embodiment, the analysis code is, for example, ABAQUS Ver6.4, the model is an axisymmetric shell model, and the analysis method is performed using unsteady heat conduction temperature distribution analysis and thermal elastic-plastic stress analysis.

  In the above ABAQUS, by inputting the material physical properties such as Young's modulus E and Poisson's ratio ν at each temperature of the material, and by defining the node of each element E and the group definition of the welded portion and the piping portion, the above basic formula is obtained. Based on this, an equation is assembled for each element E, calculation is performed, and a stress for each element is calculated.

  By the way, the present invention is characterized in that FEM analysis is performed by introducing constraints by a spring in the axial direction and the radial direction in order to complement the difference between the main analysis by the axially symmetric shell model and the actual stress.

  This is because when welding pipes by welding, the heat source moves along the weld bead in actual welding, whereas in the axisymmetric shell model, the weld bead is warmed all around at the same time. This is because it expands freely in the axial direction and relatively easily deforms in the radial direction. Spring restraint conditions are introduced for the purpose of moderately restraining the free expansion in the axial direction and the ease of deformation in the radial direction, which do not actually occur.

  That is, when considering the phenomenon of circumferential welding in piping, the heat input associated with welding becomes a three-dimensional behavior because the point heat source moves. However, when analyzing the residual stress of a pipe having a welded part using an axisymmetric shell model, if a heat input simulating welding is applied, the annular ring is heated at the same time, and a point heat source like an actual machine Behaves differently from the movement of In the actual machine, the region heated locally by welding is constrained by the surrounding structure (because of the low temperature), but in the axisymmetric shell model, the radial displacement expands and contracts relatively freely. Therefore, it is considered that the result of analysis in the circumferential direction mainly differs from the actual machine. Focusing on the above points, in order to reproduce the actual deformation behavior in the axisymmetric shell model, a convenient constraint condition was given.

  The introduction position and rigidity of the spring restraint are determined by calculating a value in which the displacement behavior of the analysis based on the axially symmetric shell model and the displacement behavior actually obtained by the experiment coincide with each other by FEM trial analysis. In addition, the displacement behavior actually obtained by experiment is obtained by performing FEM analysis using a three-dimensional shell model.

  As described above, by introducing the spring restraint into the axisymmetric shell model, the residual stress of the pipe welded joint portion 63 (see FIG. 3) by welding can be accurately calculated in a short time.

  That is, the degree of increase in the yield point that has been work-hardened by heating during welding is obtained from the hardness distribution measured by the mock-up test on the actual joint cross section, and rolling is performed by the difference (%) from the welding residual stress analysis result. The analysis is performed by replacing the received material (with increased hardness).

  As described above, the residual stress of the pipe welded joint after IHSI is shortened by replacing the material of the part where the difference in strain occurs with the material that has undergone work hardening according to the difference and performing FEM analysis. It can be calculated accurately in time.

  Hereinafter, the determination process of the spring restraint of FEM analysis and the conditions of material replacement in this Embodiment is demonstrated concretely.

  First, the analysis target and the analysis model will be described.

  In this experiment, four mockups shown in FIG. 1 were formed and a mockup test was performed in order to confirm whether the analysis data by FEM analysis was consistent with the actual experimental data. No. Nos. 2, 3, and 4 are specimens that were subjected to multilayer circumferential welding and IHSI construction on pipes having different diameters. 1 is No. This is a test body having the same diameter as 2 and only welding. About each mockup, the temperature history during welding and IHSI construction was measured, respectively. Furthermore, after the construction was completed, the residual stress distribution was measured by a cutting method.

  On the other hand, the analysis was performed by an axisymmetric shell model simulating four mock-up tests. 2 and 3 show a model having a diameter of 600A as an example of an analysis model that simulates mockup. 2 is an overall cross-sectional view of the pipe, and FIG. 3 is an enlarged cross-sectional view of the main part in the vicinity of the welded portion of the pipe (A portion in FIG. 2).

  In the analysis, the analysis code is ABAQUS Ver6.4, and the analysis method is analysis by unsteady heat conduction temperature distribution analysis and thermal elastic-plastic stress analysis.

  And based on the temperature measurement result at the time of welding obtained from the mockup test, a heat conduction temperature analysis simulating heat input by welding is performed, and a temperature distribution is calculated. The thermal elastic-plastic stress analysis including the welded portion is performed using the thermal load obtained as a result, and then a thermal history simulating the IHSI construction is given to confirm the change in the residual stress around the welded portion.

  The physical properties of the materials used for the analysis (stress-strain relationship (yield stress) and specific heat, thermal expansion coefficient, Young's modulus, thermal conductivity, specific gravity) of the base metal and the weld metal are obtained from a separate “high-temperature tensile strength test”. Data was used. 4 to 6 show the true stress-plastic strain relationship of the materials used in the analysis.

  On the other hand, Mises' yield condition was adopted as the plastic yield condition, and the hardening rule was the isotropic hardening rule. Note that deformation mechanisms with speed dependency such as creep and transformation expansion are not considered.

  Next, a calculation method of temperature distribution (unsteady heat conduction temperature distribution analysis) will be described.

  In the analysis of temperature distribution by welding and IHSI, the conditions of heat removal by external atmosphere and cooling are set in the analysis model, and the elements corresponding to the welding pass are generated sequentially, and the heat input by welding is simulated to generate heat sequentially. Thus, the temperature distribution around the welded part and the welded part is calculated. At that time, the data of the mock-up test result is used for the welding conditions to be used.

  The temperature distribution analysis by welding will be described.

  The boundary conditions for temperature distribution analysis during welding and the outline of introduction of welding heat input are described below.

  (1) The boundary conditions considered heat transfer and radiant heat transfer to the external atmosphere on the inner and outer surfaces including the weld.

  (2) The setting of the welding path was modeled by element division simulating the welding path of an actual mock-up specimen in order to simulate the thermal effect of mock-up welding. An element corresponding to the weld metal was generated every time welding was performed, and a method of simulating the heat input by welding by generating heat from this element was used. FIG. 7 and FIG. 8 show the correspondence with the elements corresponding to the actual welding path in the analysis. 7 and 8 show examples of 600A piping. In FIG. 7, 1 indicates a base material part of the pipe, and 2 indicates a welding material part. The circled number indicates the analysis layer, and the number in the layer indicates the mock-up welding pass (see FIG. 8 for details).

  (3) When performing the initial setting of the heat history, the amount of heat input per unit length of the weld line can be obtained by current × voltage ÷ speed × heat input efficiency. These current, voltage, and speed were set based on welding construction records. As for the heat input efficiency of welding, the numbers 0.21 to 0.48 are introduced in “Distribution of temperatures in arc welding, Christensen, N.”, where 0.4 was used as the initial value in TIG welding. .

  In determining the amount of heat input per unit length of the weld line, the temporal change in the amount of heat input is assumed to be trapezoidal as shown in FIG. To do. Specifically, the amount of heat input increases at a predetermined rate of time relative to the first heat input time, then the amount of heat input is constant at a predetermined rate of time relative to the heat input time, and a predetermined rate of time relative to the last heat input time. The heat input decreases. As a result, even if the same amount of heat is input in total, the maximum temperature of the welded portion differs if the time width is different. In this analysis, the heat input time width was set so that the maximum temperature of the welded portion was 1300 ° C. while fixing the heat input amount per unit length, and was used as the initial value of the temperature fitting. Temperature fitting means bringing the temperature analysis value closer to the mock-up test measurement value.

  (4) When performing temperature fitting, the thermocouple temperature history measurement results obtained from the mock-up test results are compared with the previously set temperature history, and the heat input time and welding efficiency are matched to the measured temperature history. The initial setting temperature history was adjusted to a temperature history reflecting the actual measurement results. The adjustment of the heat input time and the welding efficiency is obtained by trial analysis based on a plurality of conditions and finding the most approximate condition. 10 and 11, the mock-up temperature measurement position 3 is indicated by black dots. A thermocouple is installed at each temperature measurement position 3. The temperature measurement positions 3 are provided on the inner surface and the outer surface of the pipe 1, respectively, and a plurality of temperature measurement positions 3 are provided at positions spaced apart from the welded portion by a predetermined distance.

  For temperature fitting, a standard is set so that the analysis result temperature is approximately ± 6% of the peak temperature at four measurement points (temperature measurement position 3) provided at 90 ° pitch on the pipe cross section. Evaluated. It has been confirmed that the peak temperature falls between the maximum value and the minimum value even when it does not fall within ± 6% (see FIGS. 12 to 14). By adjusting the time width shown in FIG. 9, the shape of the peak of the temperature curve was almost reproduced, and the adjustment of the heat input efficiency was also used to adjust the peak temperature at the measurement point. However, it was assumed that the heat input efficiency was between 0.2 and 0.8, and the time width was not less than 10 seconds. 12 to 14 show an example of the temperature fitting result of the 600A pipe. In FIGS. 12 to 14, reference numeral 4 denotes a reference temperature measurement point.

  In the actual analysis, the average value of the thermal efficiency and heating time obtained by this was calculated, and the analysis was carried out using this as the heat input conditions for all welding passes.

  As an example, no. The thermal efficiency and heating time in the analysis for 3 (550A piping) are as follows.

Thermal efficiency: 0.74
Heating time: 19.0 seconds Since the actual mock-up specimen contracted in the direction of the pipe axis, the points of the analysis mesh corresponding to the actually measured values were also offset and considered so as to follow the actual machine.

  As described above, by performing the temperature fitting, the temperature history of the welding process that moves three-dimensionally can be simulated with the axisymmetric shell model.

  Next, temperature distribution analysis by IHSI will be described.

  The boundary conditions for temperature distribution analysis during IHSI construction and the heat input by the coil are described below.

  The boundary conditions are as follows.

  (A) The outer surface of the pipe is a heat dissipation boundary with ambient air.

(B) The inner surface of the pipe is a forced convection boundary. (In order to simulate a state where cooling water of about 300 ° K (Kelvin) flows in the pipe at 2 m / sec, the inside of the pipe is set to forced convection heat transfer, and heat transfer is performed using the existing empirical formula (Dittus-Boelter formula). Find the rate.)
(C) Heat generated from the outer surface of the tube is internal heat generated by electromagnetic induction heating.

  Here, EV (essential variable) and temperature fitting will be described.

  As the dominant parameters of the residual stress relaxation effect in implementing IHSI, (1) maximum temperature, (2) required temperature difference, (3) coil width, (4) heating time, (5) weld line and coil width center There are five relative positions. These values are referred to as EV (Essential Variable), and IHSI construction conditions are planned by calculating and confirming before implementation.

  In this analysis, the heat generation position was adjusted so as to simulate the EV value of the mock-up construction conditions. Furthermore, temperature fitting was performed by adjusting the amount of heat input in the FEM analysis so as to coincide with the history and distribution of the pipe outer surface temperature obtained during the IHSI construction of the mock-up specimen.

  The temperature measurement position at the time of IHSI heating is shown in FIGS. 15 and 16 show an example of the temperature fitting result. In FIG. 15, reference numeral 5 denotes a temperature history group showing the temperature history at each temperature measurement position displayed on the right side of the drawing. Then, trial analysis was performed so as to match the measurement result of the lowest temperature among the measurement positions on the pipe outer surface side (the lowest temperature in the temperature history group 5), and the temperature fitting of the analysis was performed. As a result, as shown in FIG. 16, the temperature distribution in the axial direction obtained by the analysis is equivalent to the measurement result of MU (mock-up).

  As described above, by performing the temperature fitting, the temperature history of the IHSI process moving three-dimensionally can be simulated with the axisymmetric shell model.

  Next, a method for calculating residual stress (thermoelastic-plastic stress analysis) will be described.

  Based on the temperature distribution obtained by the above-mentioned unsteady heat conduction analysis, the thermal load was calculated, and the residual stress distribution inside and outside the pipe was calculated by thermal elastic-plastic large deformation analysis.

  Here, the residual stress analysis by welding will be described.

  First, the change of the time calendar of the temperature at each node calculated from the temperature distribution analysis by welding is input and converted into a thermal load through a linear expansion coefficient. The thermal load was considered from the temperature rising process at the beginning of heat input to the time when the temperature distribution around the welded portion hardly changed after the heat input was completed. In the stress analysis, the analysis proceeded while sequentially generating the elements of the heated area corresponding to the welding pass. The generation timing of this element was set to the time when the heating region became the highest temperature.

  By the way, the present invention is characterized in that, in order to complement the difference between the main analysis by the axially symmetric shell model and the actual three-dimensional welding phenomenon, the restraint by the spring is introduced in the axial direction and the radial direction during welding. .

  The timing of restraint is from the moment when heat input simulating molten metal is applied to after cooling, and thereafter, the restraint is once removed to obtain a state in which deformation due to self-equilibration of internal stress is applied. When the next pass was heated, the same restraint was performed again, and this operation was repeated for all the weld passes.

  The difference due to the difference in dimensions in FEM analysis is that the heat source moves along the weld bead in actual welding (a part of the three-dimensional body is warmed), whereas in the axially symmetric model, it extends over the entire circumference. This is an event in which the weld bead is heated at the same time, and is free to expand in the axial direction and is relatively easily deformed in the radial direction. Spring restraint conditions were introduced for the purpose of moderately restraining the free expansion in the axial direction and the ease of deformation in the radial direction, which do not actually occur.

  Since the rigidity of the peripheral region differs depending on the diameter and the plate thickness, it is necessary to carry out the study of this constraint condition for each pipe to be analyzed. The spring constraint condition is determined by paying attention to the following points.

  Regarding the axial direction and radial direction constraint calculation method at the time of welding, considering the phenomenon of circumferential welding in piping, the heat input accompanying welding moves three-dimensionally because the point heat source moves (FIG. 17). reference). Therefore, in order to analyze the residual stress of the pipe circumferential weld with high accuracy by FEM analysis, it is considered necessary to perform an analysis that reproduces the actual heat input state using a three-dimensional model. However, welding of actual machines becomes multi-pass (depending on the diameter and thickness of the pipe, but about 30-50 passes), and analysis with a three-dimensional model takes enormous time even in view of the remarkable improvement in the capabilities of recent computers. Therefore, it is common to analyze by replacing with an axisymmetric shell model when examining various conditions.

  However, when the residual stress of a pipe having a welded portion is analyzed using an axisymmetric shell model, when a heat input simulating welding is applied, the annular ring is heated at the same time (see FIG. 18). It shows a behavior different from the movement of a point heat source. As shown in FIG. 19, in the actual machine, the region heated locally by welding is constrained from deformation by the surrounding structure (because of the low temperature), but in the axially symmetric shell model, the radial displacement is relatively small. It will expand and contract freely, and it is thought that the analysis results in the circumferential direction will differ from the actual machine. Therefore, in order to reproduce the original three-dimensional deformation behavior with the axially symmetric shell model, it is considered necessary to give a convenient constraint condition as shown in FIG.

  By the way, when the strain in the cylindrical coordinate system is three-dimensional, the three-dimensional diameter and the axis are determined by considering the vertical three-component and the shear three-component, and bringing the diameter and axial displacement behavior of the axisymmetric shell model closer to the three-dimensional model. The strain behavior in the circumferential direction can be simulated.

  However, since the object is a cylinder, the displacement behavior is complex, and the same three-dimensional deformation behavior as an analytical solution is reproduced by an axially symmetric shell model, as in circumferential welding. It is considered difficult to determine such constraint conditions.

  Therefore, in the present invention, a trial analysis by FEM using an axisymmetric shell model and a three-dimensional shell model was performed, and the difference in deformation behavior between local heating and whole circumference heating was clarified. In addition, a case study using FEM was conducted to investigate a restraint method that matched the deformation behavior of local overheating and all-around heating.

  The method for determining the restraint method by FEM will be described below.

  First, a trial analysis method using FEM will be described.

  In order to clarify the difference in deformation behavior between the case where the entire circumference was welded simultaneously and the case where it was locally welded, an elastic analysis was performed using an axisymmetric shell model and a three-dimensional shell model. The following three types of analysis were used for comparison.

(1) Center heating using an axisymmetric shell model (equivalent to simultaneous heating around the entire circumference)
(2) All-around center heating using a 3D shell model (equivalent to all-around heating)
(3) Local heating at the center using a three-dimensional shell model (corresponding to the actual point heat source)
Although the same analysis result should be obtained for all-round heating using the axisymmetric shell model and the three-dimensional shell model, they are used for comparison in order to confirm the validity of the deformation of the axisymmetric shell model.

  FIG. 20 shows a 600A piping model as an example of the three-dimensional shell model used for the study. In FIG. 20 (a), the size of the local heating region indicated by 6 was set based on the temperature measurement result during welding performed with 600A piping. From the results of temperature measurement at the time of one-layer one-pass welding, the inner surface of the pipe is about 700 ° C. at a position of ± 8 mm from the groove center, about 390 ° C. at a position of ± 13 mm, and about 200 ° C. or less at a position of ± 23 mm.

  Therefore, in consideration of the analysis mesh, the heating region has a lower limit of 200 ° C. and a width of 20 mm. The length of the heating region was set to 128 mm from a welding speed of 90 mm / min and a high temperature duration of about 1.3 min at 200 ° C. or higher. In addition, as shown in FIG. 20B, the entire circumference heating region 7 (assuming heat input in an axially symmetric shell model) 7 was set to have a 20 mm width as a heat input range.

  Next, comparison of trial analysis results will be described.

  Three types of trial analysis using a three-dimensional shell model and an axially symmetric shell model were performed, and radial and axial displacements are shown in FIGS. In FIG. 21 and FIG. 22, 11 (square mark on the thin line) shows the displacement behavior in the case of 3D shell (three-dimensional shell model) and local heating, and 12 (thick line) in the case of 3D shell and all-around heating. The displacement behavior is shown, and 13 (circle mark on the thin line) shows the displacement behavior in the case of an axially symmetric shell (axially symmetric shell model) and all-around heating.

  From FIG. 21, it can be confirmed that the displacement behavior 12 when the entire circumference is heated by the three-dimensional shell model and the displacement behavior 13 when the circumference is heated by the axisymmetric shell model are substantially equal. On the other hand, it can be confirmed that the displacement behavior 11 in the radial direction when locally heated by the three-dimensional solid model generates considerably less displacement than the displacement behaviors 12 and 13 when heated all around. By correcting this difference, it is considered that the displacement behavior during welding is equivalent to that of the three-dimensional solid model even in the axisymmetric shell model, and as a result, the plastic strain and the stress balance approach the actual machine.

  As shown in FIG. 22, the displacement in the axial direction is similar to the radial displacement, the displacement behavior 12 when the entire circumference is heated with the three-dimensional shell model, and the displacement in the axial direction when the entire circumference is heated with the axisymmetric shell model. It can be seen that the behavior 13 is almost equal. On the other hand, it can be confirmed that the displacement behavior 11 when locally heated by the three-dimensional shell model generates considerably less displacement than the displacement behaviors 12 and 13 when heated all around. By correcting this difference, the displacement behavior during welding can be made equivalent to that of the three-dimensional solid model even in the axially symmetric shell model, and as a result, the plastic strain and the stress balance are considered to approach the actual machine.

  Next, the examination of the constraint conditions for matching the deformation behavior will be described.

  The stiffness in the peripheral region that does not appear in the axially symmetric shell model in both the radial and axial directions suppresses displacement due to heating. In order to reproduce the difference with an axisymmetric shell model, it is preferable to provide an appropriate rigidity (spring constraint) at an appropriate position and construct an analysis model. Since the object is a cylinder, the displacement behavior is complicated, and the introduction position and rigidity of the spring constraint need to be verified and obtained by FEM analysis. Therefore, a case study using FEM was conducted to confirm and determine the spring restraint setting position and rigidity.

  As a result of various case studies, we decided to introduce radial spring restraint at a position ± 30 mm from the center of the weld line and axial spring restraint at a position ± 100 mm from the weld center. The given spring stiffness was set by calculating from the trial analysis the stiffness with which the displacement behavior in the axisymmetric shell model and the three-dimensional shell model matched.

In other words, multiple conditions are input to the axisymmetric shell model and FEM analysis (parameter survey) is performed multiple times to find and set the stiffness (spring constant) that matches the displacement behavior of the axisymmetric shell model and the three-dimensional shell model. did. Specifically, in the case of 600A piping, the stiffness of the radial spring is 2.581 × 10 ¹⁰ (N / m), and the stiffness of the axial spring is 11.621 × 10 ¹⁰ (N / m). To do.

  When inputting conditions to ABAQUS, elements for introducing radial and axial spring restraints are specified, and the stiffness (spring constant) of each spring is input.

  FIG. 23 and FIG. 24 show a comparison between the displacement behavior of the axisymmetric shell model in which this constraint is introduced and the displacement behavior of the locally heated three-dimensional shell model. 23 and 24, 11 (square mark on thin line) indicates the displacement behavior in the case of 3D shell (three-dimensional shell model) and local heating, and 12 (thick line) indicates the case in the case of 3D shell and all-around heating. Shows displacement behavior, 13 (circle mark on thin line) shows displacement behavior in case of axially symmetric shell (axially symmetric shell model) and all-around heating, and 14 (thick line) shows shaft with spring restraint introduced in radial and axial directions The displacement behavior in the case of symmetrical shell (axisymmetric shell model) and all-around heating is shown.

  As shown in FIGS. 23 and 24, the displacement behavior 14 in the radial and axial directions when the entire circumference is heated by an axially symmetric shell (axially symmetric shell model) in which spring constraints are introduced in the radial and axial directions, and a three-dimensional shell model. It can be confirmed that the displacement behavior 11 in the diameter and axial directions when locally heated is approximately equal. That is, by introducing the above-described constraints, it becomes possible to simulate the deformation behavior of local heating in the three-dimensional shell model with the axisymmetric shell model.

  The effect of spring restraint in the axisymmetric shell model will be described.

  FIG. 25 and FIG. 26 show the results (As Weld analysis results) of the analysis by introducing the spring constraint obtained by the above method into the axisymmetric shell model for 600A piping. In FIGS. 25 and 26, 15 indicates the stress of the inner surface by FEM analysis (axisymmetric shell model) without spring constraint, and 16 indicates the stress of the inner surface by FEM analysis (axial symmetry shell model) with spring constraint. , 17 indicates the stress on the outer surface by FEM analysis (axisymmetric shell model) without spring constraint, and 18 indicates the stress on the outer surface by FEM analysis (axisymmetric shell model) with spring constraint. Reference numeral 19 denotes the stress at each temperature measurement position on the inner surface of the MU, and 20 denotes the stress at each temperature measurement position on the outer surface of the MU.

  As shown in FIG. 25, the As Weld analysis results (inner stress 16 and outer stress 18 by FEM analysis with spring restraint) are the results of conventional analysis (internal stress by FEM analysis without spring restraint) in the axial stress distribution inside the pipe. Compared with the stress 15 and the stress 17) on the outer surface, the inner and outer surfaces are slightly affected by radial restraint, but the tendency of the stress distribution is the same, and the stress level near the molten metal hardly changes. Further, as shown in FIG. 26, the stress distribution in the pipe inner circumferential direction of the stress 16 on the inner surface and the stress 18 on the outer surface by the FEM analysis with spring restraint decreases the stress level in the vicinity of the molten metal, and the actual measurement result (the inner surface of the MU). It has been confirmed that the stress 19 at each temperature measurement position and the stress 20) at each temperature measurement position on the outer surface are close to the tendency.

  From the above, it was found that the strain behavior due to the three-dimensional influence of the actual machine can be simulated even by FEM analysis using an axisymmetric shell model by giving boundary conditions for analysis.

  Next, residual stress analysis by IHSI will be described.

  The time calendar data (see Fig. 15) calculated by calculating the temperature distribution during IHSI construction is input as a time-varying temperature at each node in the same manner as the analysis during welding, and is converted into a thermal load via a linear expansion coefficient. . The thermal load was considered from the temperature rising process at the beginning of heat input to the time when the temperature distribution around the welded portion hardly changed after the heat input was completed.

  At the time of actual IHSI construction, there is almost no constraint in the axial direction. Therefore, in order to prevent rigid body deformation in the analysis, only one end of the pipe was constrained.

  The HAZ portion (boundary portion between the weld material and the base material) increases in hardness due to work hardening by welding, but this behavior cannot be completely reproduced in the analysis. After replacing the areas shown in A and B with physical properties of work-hardened materials obtained from the results of tests using separately performed rolled materials, the thermal load obtained by calculating the temperature distribution during IHSI construction was applied. The analysis was performed.

  That is, the degree of increase in the yield point that has been work-hardened by heating during welding is obtained from the hardness distribution measured on the actual joint cross section, and the analysis is performed by correcting the difference from the welding residual stress analysis result. Details will be described below.

  28 and 29 show stress-strain diagrams (true stress-plastic strain curves) obtained from the rolling material test.

  A method for correcting the amount of plastic strain near the weld will be described.

  Regarding the behavior in which residual stress in the vicinity of the weld is relaxed by the IHSI construction after the welding process, the results of the FEM analysis using the axisymmetric shell model with the spring constraint introduced agree well with the experimental results of the mock-up. It became clear to do. However, in the axial stress in the vicinity of the weld back surface portion (inner surface of the pipe), a perfect match is not obtained quantitatively. In particular, in the analysis, the residual stress after IHSI tends to be higher than the mock-up in the residual stress. Therefore, the evaluation of the residual stress of the actual machine is the opposite of the safety side in the analysis.

  It is considered that this part is subject to work hardening that is difficult to reproduce completely in the analysis because it is subjected to molten metal and nearby melting / solidification and thermal deformation. Therefore, in order to compare the degree of work hardening of analysis and mockup, the hardness distribution near the weld of As Weld was measured for a mockup specimen of 600A piping.

  FIG. 30 shows the measured hardness distribution. The hardness in the vicinity of the molten metal on the inner surface side of the pipe indicated by hatching A in FIG. 30 is about Hv 260, and the hardness in the vicinity of the molten metal at the center of the plate thickness is about Hv 220 (hatching B region). Generally, it is known that hardness and tensile strength are in a proportional relationship, and a proportional relationship is also expected with 0.2% proof stress. Therefore, a tensile test was performed on a SUSF316 (LC) material to which prestrain (rolling) was applied, and the relationship between hardness by prestrain and 0.2% proof stress was obtained.

  31 and 32 show a list of test results and the relationship between hardness and 0.2% proof stress. From these figures, it became clear that the hardness and the 0.2% proof stress are in a proportional relationship. From the hardness measured from the mock-up and FIG. 32, the 0.2% proof stress near the molten metal is estimated to be about 600 MPa. From this value and the table of FIG. 31, it is estimated that a strain history of about 15% was received in the hatched portion A of FIG. 30 and about 10% in the hatched B portion.

  The prestrain amount is estimated from the mock-up hardness distribution by such an estimation method, and the prestrain amount for each element of the corresponding FEM model is shown in a contour diagram as shown in FIGS. As shown in FIG. 33, it can be seen that the closer to the molten metal portion, the larger the pre-strain amount, and the inner surface has a larger pre-strain amount. In addition, as shown in FIG. 34, it can be seen that the amount of pre-strain increases as the distance from the center and the inner surface increases.

  On the other hand, the distribution of equivalent plastic strain obtained by FEM analysis is shown in FIGS. Difference between FIG. 33 and FIG. 35 (difference between the pre-strain amount of the pipe base material portion and the equivalent plastic strain amount), difference between FIG. 34 and FIG. 36 (difference between the pre-strain amount of the weld material portion and the equivalent plastic strain amount) The results of displaying are shown in FIGS. 37 and 38, respectively. In FIGS. 33 to 38, the numerical value (%) indicates plastic strain.

  As shown in the figure, the plastic strain is generally higher in the mock-up, and there is a difference in plastic strain of about 5 to 10% at the boundary between the molten metal and the SUS base material. In particular, the difference on the inner surface side is large. This difference is considered to be a deformation behavior that cannot be reproduced by FEM analysis.

  FIG. 39 shows the stress-strain relationship schematically representing the difference in plastic strain. Here, it is assumed that the stress-strain state of As Weld is at A in FIG. 39 in the FEM analysis result. On the other hand, from the difference between the amount of plastic strain estimated from actual measurement and the amount of plastic strain obtained by analysis, it is expected that the mock-up is at A '(stress is the same, but the amount of plastic strain is different).

  Next, FIG. 40 shows the stress-strain behavior of the inner surface of the pipe when subjected to the IHSI thermal cycle after welding. In the analysis, a path such as A → B → C with cooling and C → D with cooling is followed, and high compressive residual stress is introduced after IHSI construction. However, when the plastic strain is A ′, the heating process follows A ′ → B ′ → C ′ and the cooling process follows C ′ → D ′ by the IHSI construction. That is, in the case of A ′ having a large pre-strain history, since the stress that yields again at B ′ is higher than that of B, as a result, the compressive residual stress introduced by IHSI is a relatively low value. In a material with remarkable work hardening such as SUS material, it is considered that an increase in yield stress accompanying work hardening significantly affects the residual stress distribution after IHSI.

  Therefore, in order to obtain the stress at the time D ′ from the state A in the analysis by the IHSI thermal cycle, the stress that causes the behavior of A → B ″ → C ″ → D ″ as shown in FIG. 41 in the process of IHSI− It is considered that the simulation can be performed by using the strain relationship, that is, it is necessary to replace the material property value having the stress strain relationship as shown by the broken line passing through B "C" immediately before the IHSI analysis.

  In order to obtain the stress-strain relationship indicated by the broken line, FIG. 42 shows a true stress-plastic strain curve of the rolled material obtained from the experiment. In FIG. 42, a line indicated by 21 is a true stress-plastic strain curve of a material that has not undergone prestrain. If the plastic strain at mock-up As Weld is 20%, it is considered that 22 loops in FIG. 42 are followed during IHSI construction. On the other hand, in order to follow the same loop as mock-up in the region where the equivalent plastic strain is 15% in FEM, follow the true stress-plastic strain curve of 21 on the line shifted to the left by 5% (one-dot chain line 23 in the figure). There is a need. This 5% shifted true stress-plastic strain curve 23 is almost equivalent to the true stress-plastic strain curve (24 curves in the figure) of the material subjected to 5% rolling. In order to follow the same loop as the mock-up in the region where the equivalent plastic strain is 10% in FEM, follow the true stress-plastic strain curve of 21 on the line shifted to the left by 10% (dashed line 25 in the figure). There is a need. It can be seen that this diagram 25 is close to the true stress-plastic strain curve (26 curve in the figure) of the material subjected to 10% rolling.

  In order to make the mock-up and the stress after IHSI of FEM analysis equivalent from the above study, it can be reproduced by using the stress-strain curve of the material pre-strained by the difference in plastic strain after As Weld in FEM analysis. it is conceivable that. Therefore, a trial analysis using an axisymmetric shell model was conducted for 600A piping, and the effect of material replacement on the residual stress distribution after IHSI was examined. In the region shown in FIG. 27 where the plastic strain difference between FIGS. 37 and 38 is simplified, the stress strain behavior of the material subjected to 5% and 10% rolling, respectively, was replaced by the calculation after the end of the welding analysis.

  The result of the above-mentioned FEM analysis is shown in FIGS. FIG. 43 shows the effect of material replacement on the axial residual stress distribution, and FIG. 44 shows the effect of material replacement on the circumferential residual stress distribution. In FIG. 43, a thick line 27 indicates an axial stress distribution on the inner surface of the pipe as an analysis result obtained by replacing the material. In FIG. 44, a thick line 27 indicates a circumferential stress distribution on the inner surface of the pipe as an analysis result obtained by replacing the material. On the other hand, an analysis result without replacing the material is indicated by a thin line 28. From these comparisons, it can be seen that the FEM analysis value and the mock-up stress distribution almost coincide with each other by replacing the materials. In addition, the thick line 29 has shown the axial direction stress distribution (FIG. 43) and the circumferential direction stress distribution (FIG. 44) in the pipe outer surface of the analysis result which performed material replacement. On the other hand, an analysis result without replacing the material is indicated by a thin line 30. Further, 31 indicates the stress at each temperature measurement position on the inner surface of the MU, and 32 indicates the stress at each temperature measurement position on the outer surface of the MU.

  From the above results, it was found that the axial stress after IHSI can be estimated equivalent to mock-up by replacing the material reflecting the hardness distribution after the end of the welding process in the FEM analysis.

  Here, the analysis result and the experimental result are compared.

  The mock-up test result in the state of As Weld is compared with the analysis result.

  No. FIG. 45 and FIG. 46 show the results of comparing the experimental results of one mock-up specimen (600A piping) with the FEM analysis results simulating it. 45 shows the stress distribution of the axial stress on the outer surface and inner surface of the pipe, and FIG. 46 shows the stress distribution of the circumferential stress on the outer surface and inner surface of the pipe. 45 and 46, a line 33 indicates the stress distribution on the inner surface of the pipe by FEM analysis, and a line 34 indicates the stress distribution on the outer surface of the pipe by FEM analysis. 35 indicates the stress at each temperature measurement position on the inner surface of the MU, and 36 indicates the stress at each temperature measurement position on the outer surface of the MU.

  45 and 46, the residual stress distribution in the vicinity of the welded portion (within ± 50 mm from the weld center) in the As Weld state is analyzed (stress distribution 33 on the pipe inner surface by FEM analysis and stress distribution 34 on the pipe outer surface by FEM analysis). ) And the experiment (the stress 35 at each temperature measurement position on the inner surface of the MU and the stress 36 at each temperature measurement position on the outer surface of the MU) show good agreement (substantially the same tendency). Moreover, the result that the maximum stress in the axial direction reaches about 400 MPa was obtained.

  The mock-up test results after IHSI construction and the analysis results will be compared.

  No. 47 and 48 show the results of comparing the experimental results of the two mock-up specimen (600A piping) and the FEM analysis results simulating it. 47 shows the stress distribution of the axial stress on the outer surface and inner surface of the pipe, and FIG. 48 shows the stress distribution of the circumferential stress on the outer surface and inner surface of the pipe. 47 and 48, a line 37 indicates the stress distribution on the inner surface of the pipe by FEM analysis, and a line 38 indicates the stress distribution on the outer surface of the pipe by FEM analysis. 39 indicates the stress at each temperature measurement position on the inner surface of the MU, and 40 indicates the stress at each temperature measurement position on the outer surface of the MU.

  47 and 48, the residual stress distribution in the vicinity of the welded portion (within ± 50 mm from the weld center) after IHSI is analyzed (the stress distribution 37 on the pipe inner surface by FEM analysis and the stress distribution 38 on the pipe outer surface by FEM analysis). It can be seen that the experiments (stress 39 at each temperature measurement position on the inner surface of the MU and stress 40 at each temperature measurement position on the outer surface of the MU) show good agreement (substantially the same tendency). It can be seen that the analysis of changing the hardness of the material agrees well with the result of the mockup.

  No. FIG. 49 and FIG. 50 show the results of comparing the experimental results of the three mock-up specimen (550A piping) with the FEM analysis results simulating it. FIG. 49 shows the stress distribution of axial stress on the outer surface and inner surface of the pipe, and FIG. 50 shows the stress distribution of circumferential stress on the outer surface and inner surface of the pipe. 49 and 50, a line 43 indicates the stress distribution on the inner surface of the pipe by FEM analysis, and a line 44 indicates the stress distribution on the outer surface of the pipe by FEM analysis. 45 indicates the stress at each temperature measurement position on the inner surface of the MU, and 46 indicates the stress at each temperature measurement position on the outer surface of the MU.

  The trend is No. Similar to the 2 mock-up test, it can be seen that the stress distribution in the axial direction agrees well with the mock-up experiment result by changing the hardness.

  No. 51 and 52 show the results of comparing the experimental results of the 4-mock-up specimen (300A piping) and the FEM analysis results simulating it. 51 shows the stress distribution of axial stress on the outer surface and inner surface of the pipe, and FIG. 52 shows the stress distribution of circumferential stress on the outer surface and inner surface of the pipe. 51 and 52, a line 47 indicates the stress distribution on the inner surface of the pipe by FEM analysis, and a line 48 indicates the stress distribution on the outer surface of the pipe by FEM analysis. Further, 49 indicates the stress at each temperature measurement position on the inner surface of the MU, and 50 indicates the stress at each temperature measurement position on the outer surface of the MU.

  As described above, the following results were obtained by conducting an analysis simulating multilayer welding and IHSI of a thick-walled pipe using an axisymmetric shell model.

  Transient temperature analysis simulating mock-up multi-layer welding path was performed, and residual welding stress analysis was performed by thermo-elasto-plastic analysis using this as a loading condition. As a result, in large diameter pipes, it was found that the residual stress distribution in the vicinity of the mock-up weld was in good agreement by introducing appropriate spring restraints in the axial and radial directions and replacement of material properties in the vicinity of the molten metal. On the other hand, it was found that the medium diameter pipe of about 300A has a larger heat input range and less influence of spring restraint than the large diameter pipe.

  In short, by calculating the restraint of the spring and changing the hardness of the material, a value equivalent to the actual residual stress is calculated with an axisymmetric shell model with a short analysis time, even for welded joints of large-diameter thick-walled pipes. It is possible to accurately calculate the residual stress of the welded joint portion of the large-diameter thick-walled pipe in a short time.

  The present invention has been described by taking the analysis of the residual stress of the welded joint in IHSI as an example. However, the present invention is not limited to this, and the residual stress can be analyzed over the entire welded portion of the pipe. is there.

It is the table | surface which showed the structure of the mockup. It is a whole sectional view showing a model with a diameter of 600A as an example of an analysis model simulating mockup. It is a principal part expanded sectional view which showed the model of the aperture diameter 600A in an example of the analysis model which simulated mockup. It is the graph which showed the stress-strain characteristic of the base material. It is the graph which showed the stress-strain characteristic of the base material. It is the graph which showed the stress-strain characteristic of the base material. It is sectional drawing which showed the welding pass. It is the table | surface which showed the relationship between a weld layer and a welding path. It is the graph which showed the time change of the heat input of welding. It is sectional drawing which showed the temperature measurement position of the mockup. It is sectional drawing which showed the temperature measurement position of the mockup. It is the graph which showed the temperature fitting example at the time of 600A piping welding. It is the graph which showed the temperature fitting example at the time of 600A piping welding. It is the graph which showed the temperature fitting example at the time of 600A piping welding. It is the graph which showed the time calendar temperature fitting of piping outer surface side temperature. It is the graph which showed the temperature distribution fitting of a pipe axis direction. It is the perspective view which showed the behavior of the circumference welding. It is the figure which showed the axial symmetry shell model and its heating range. It is the perspective view which showed the restraint state of the welding part periphery area | region at the time of local heating. (A) is the figure which showed the heating area | region of the local heating of a three-dimensional shell model, (b) is the figure which showed the heating area | region of the perimeter heating of a three-dimensional shell model. It is the graph which showed the comparison of the radial direction displacement of an axisymmetric shell and a three-dimensional shell model. It is the graph which showed the comparison of the axial direction displacement of an axisymmetric shell and a three-dimensional shell model. It is the graph which showed the comparison of the radial direction displacement of the axisymmetric shell and the three-dimensional shell model in the model which put the spring restraint. It is the graph which showed the comparison of the axial direction displacement of the axisymmetric shell and the three-dimensional shell model in the model which put the spring restraint. It is the graph which showed the effect (axial stress distribution) of the spring restraint by an axisymmetric shell model. It is the graph which showed the effect (circumferential stress distribution) of the spring restraint by an axisymmetric shell model. It is the distribution contour figure which showed the area | region which received work hardening. It is a stress-strain diagram obtained from the rolled material which received 5% work hardening. It is the stress-strain diagram obtained from the rolling material which received 10% work hardening. It is the figure which showed the welding part hardness distribution measurement result of 600A piping. It is the table | surface which showed the tension test result. It is the graph which showed the relationship between the hardness of SUS316L, and 0.2% yield strength. It is the distribution contour figure which showed the pre-strain (plastic strain) distribution of the SUS base material part. It is the distribution contour figure which showed the pre-strain (plastic strain) distribution of the molten metal part. It is the distribution contour figure which showed the equivalent plastic strain distribution of the SUS base material part. It is the distribution contour figure which showed the equivalent plastic strain distribution of the molten metal part. It is the distribution contour figure which showed the difference of the plastic strain of the measurement and FEM analysis in a SUS base material part. It is the distribution contour figure which showed the difference of the plastic strain of the measurement and FEM analysis in a molten metal part. It is the figure showing the analysis in an As Weld state, and the comparison of the stress-strain state of a mockup. It is the figure which showed the comparison of the stress-strain behavior during IHSI construction. It is the figure which showed the stress-strain curve required in order to make the mockup after IHSI and the stress of FEM analysis equivalent. It is the graph which showed the true stress-strain behavior of the rolling material, and the true stress-strain behavior during IHSI construction. It is the graph which showed the influence which material replacement has on axial direction residual stress distribution. It is the graph which showed the influence which material exchange has on the circumferential direction residual stress distribution. Mock-up No. It is the graph which showed the axial direction stress distribution of 1 (600A piping As Weld). Mock-up No. It is the graph which showed the circumferential direction stress distribution of 1 (600A piping As Weld). Mock-up No. It is the graph which showed the axial direction stress distribution of 2 (after 600A piping IHSI). Mock-up No. It is the graph which showed the circumferential direction stress distribution of 2 (after 600A piping IHSI). Mock-up No. 3 is a graph showing an axial stress distribution of 3 (after 550A piping IHSI). Mock-up No. It is the graph which showed the circumferential direction stress distribution of 3 (550A piping after IHSI). Mock-up No. It is the graph which showed the axial direction stress distribution of 4 (300A piping IHSI after). Mock-up No. It is the graph which showed the circumferential direction stress distribution of 4 (after 300A piping IHSI).

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Pipe base material part 2 Welding material part 62 Large diameter thick wall pipe 63 Welded joint part E Element

Claims (4)

The residual stress of the welded joint of large diameter thick wall pipe is divided into multiple elements using an axisymmetric shell model,

In the analysis method for FEM analysis based on
The predetermined element of the axisymmetric shell model includes a radial displacement calculated by FEM analysis at the time of all-round heating by a normal axisymmetric shell model and a radial displacement calculated by FEM analysis at the time of local heating by a three-dimensional shell model. A radial spring constraint having rigidity corresponding to the difference between the axial displacement and the axial displacement calculated by FEM analysis at the time of all-around heating by the normal axisymmetric shell model is added to a predetermined element of the axisymmetric shell model and 3 Residual stress analysis of welded joints characterized by FEM analysis by introducing axial spring restraint having rigidity according to the difference from axial displacement calculated by FEM analysis during local heating with a three-dimensional shell model Method.   The introduction position of the radial spring constraint and the axial spring constraint, and the rigidity of the radial spring constraint and the axial spring constraint are analyzed by the displacement analysis of the analysis by the axially symmetric shell model by the FEM trial analysis and the analysis by the three-dimensional shell model. The method for analyzing residual stress in a welded joint according to claim 1, wherein a value in which the displacement behaviors of the welds coincide is calculated and set. The radial direction and axial displacement by the above-mentioned axially symmetric shell model are calculated by analysis at the time of all-around heating of the large-diameter thick wall tube by the axially symmetric shell model, and the radial direction and axial direction displacement by the above three-dimensional shell model are 3 The method for analyzing residual stress in a welded joint according to claim 1 or 2, wherein the residual stress is calculated by analysis during local heating of a large-diameter thick-walled pipe using a three-dimensional shell model. Stainless steel 600A piping is used for the large-diameter thick-walled pipe, and the radial spring restraint is introduced at a position of ± 30 mm in the axial direction from the weld center of the weld joint, and its rigidity is 2.581 ×. 10 ¹⁰ (N / m), and the axial spring restraint is introduced at a position of ± about 100 mm in the axial direction from the weld center of the weld joint, and its rigidity is 11.621 × 10 ¹⁰ (N / m). The method for analyzing residual stress in a welded joint according to any one of claims 1 to 3.

Images