Method for predicting the remaining life of a pressure vessel

Abstract

To provide a method for predicting the remaining lifespan of a pressure vessel that can quantitatively predict the remaining lifespan of the pressure vessel. [Solution] In one embodiment of the method for predicting the remaining life of a pressure vessel, stress-strain curves are obtained for each of a plurality of test pieces made of steel that have undergone accelerated hydrogen erosion testing. The value obtained by subtracting elastic strain and plastic strain from the total true strain obtained from the stress-strain curves is defined as hydrogen erosion strain. The hydrogen erosion strain is applied to the Ω method used for creep strain to numerically model the hydrogen erosion strain rate. The remaining life of the pressure vessel made of steel is predicted by computer simulation using the numerically modeled hydrogen erosion strain rate and the actual operating conditions of the pressure vessel.

Description

[Technical Field]

[0001] This disclosure relates to a method for predicting the remaining life of a pressure vessel. [Background technology]

[0002] For example, as disclosed in Patent Document 1, pressure vessels used in, for example, oil refining reactors are exposed to a high-temperature, high-pressure hydrogen environment. Pressure vessels made of steel materials such as Cr-Mo steel are known to be damaged by hydrogen erosion, in which carbon in the steel reacts with invading hydrogen to generate methane gas voids. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2005-024371 [Overview of the project] [Problems that the invention aims to solve]

[0004] For such pressure vessels, selecting steel materials within the permissible ranges of temperature and hydrogen partial pressure in the Nelson diagram and managing operating conditions should suppress the generation of methane gas, which causes hydrogen erosion.

[0005] However, hydrogen erosion can actually occur due to impurities and structural variations in the steel. Therefore, there is a need to predict the remaining lifespan of pressure vessels, but a quantitative method for predicting remaining lifespan has not yet been established. Other challenges and novel features will become apparent from the description and accompanying drawings in this specification. [Means for solving the problem]

[0006] A method for predicting the remaining lifespan of a pressure vessel according to one embodiment is: (a) A process of obtaining stress-strain curves for each of several test specimens made of steel that have undergone accelerated hydrogen erosion testing. (b) A step of defining hydrogen erosion strain as the value obtained by subtracting elastic strain and plastic strain from the total true strain obtained from the stress-strain curve, (c) A step of numerically modeling the hydrogen erosion strain rate by applying the hydrogen erosion strain to the Ω method used for creep strain, and (d) A step of predicting the remaining life of the pressure vessel made of steel by computer simulation using the numerically modeled hydrogen erosion strain rate and the actual operating conditions of the pressure vessel. [Effects of the Invention]

[0007] According to one embodiment, a method for predicting the remaining lifespan of a pressure vessel can be provided that allows for the quantitative prediction of the remaining lifespan of the pressure vessel. [Brief explanation of the drawing]

[0008] [Figure 1] This is a flowchart showing a method for predicting the remaining life of a pressure vessel according to the first embodiment. [Figure 2] This is a schematic graph showing the temperature pattern of the heat treatment performed on the steel plate according to the example. [Figure 3] This is a schematic perspective view of a welded test plate made from a steel plate according to the embodiment. [Figure 4] This is a SEM image of the microstructure in the HAZ of a welded joint after an accelerated hydrogen erosion test. [Figure 5] This is a schematic cross-sectional view of a welded joint, illustrating the location of the tensile test specimens to be taken. [Figure 6] This graph shows the definitions of hydrogen erosion strain and hydrogen erosion strain rate using true stress-true strain curves based on the hydrogen erosion accelerated test time. [Figure 7] This graph shows the relationship between hydrogen erosion strain εh (calculated by the Omega method) and the logarithm of remaining life log(tr-t) in welded joints of FSb material. [Figure 8] This graph shows the relationship between hydrogen erosion strain εh (calculated by the Omega method) and the logarithm of remaining life log(tr-t) in welded joints of 15Sb material. [Figure 9]This is a flowchart showing subroutines for implementation in numerical simulation. [Figure 10] This is a schematic diagram showing the analytical model of the reactor shell used to compare theoretical calculations with numerical analysis. [Figure 11] This is the initial true stress-true strain curve for the 1.25Cr-0.5Mo steel input for analysis. [Figure 12] This graph shows the initial principal stress distribution across the reactor shell thickness under actual operating conditions. [Figure 13] This graph shows a comparison between theoretical calculations and numerical analysis of the time-dependent changes in the distribution of hydrogen erosion damage Dh(t) across the reactor shell thickness in welded joints of FSb material. [Figure 14] This graph shows a comparison between theoretical calculations and numerical analysis of the time-dependent change in the distribution of hydrogen erosion strain εh(t) across the reactor shell thickness in welded joints made of FSb material. [Figure 15] This graph shows a comparison between theoretical calculations and numerical analysis of the time-dependent changes in the distribution of hydrogen erosion damage Dh(t) across the reactor shell thickness in welded joints made of 15Sb material. [Figure 16] This graph shows a comparison between theoretical calculations and numerical analysis of the time-dependent change in the distribution of hydrogen erosion strain εh(t) across the reactor shell thickness in welded joints made of 15Sb material. [Modes for carrying out the invention]

[0009] The following describes specific embodiments in detail with reference to the drawings. However, the embodiments are not limited to those described below. Also, for clarity, the following descriptions and drawings have been simplified as appropriate.

[0010] (First embodiment) <Method for predicting the remaining life of a pressure vessel> Referring to Figure 1, a method for predicting the remaining life of a pressure vessel according to the first embodiment will be described. Figure 1 is a flowchart showing the method for predicting the remaining life of a pressure vessel according to the first embodiment. Here, the pressure vessel whose remaining life is predicted by this manufacturing method is made of steel material such as Cr-Mo steel.

[0011] These pressure vessels, used, for example, in oil refining reactors, are exposed to high-temperature, high-pressure hydrogen environments. As a result, they are damaged by hydrogen erosion, where carbon in the steel reacts with invading hydrogen to generate methane gas voids. Therefore, a quantitative method for predicting remaining service life is needed.

[0012] First, as shown in Figure 1, stress-strain curves are obtained for each of several test specimens made of steel that have undergone accelerated hydrogen erosion testing (Step ST1). More specifically, the stress-strain curves are true stress-true strain curves. Here, the steel that has undergone accelerated hydrogen erosion testing is the same type of steel that constitutes the pressure vessel that is the subject of remaining life prediction.

[0013] In accelerated hydrogen erosion testing, exposure time is varied under higher temperatures and higher hydrogen pressures than those of the pressure vessel's actual operating conditions. While not particularly limited, the conditions for accelerated hydrogen erosion testing are typically around 500-600°C, 20 MPa, and 100-1000 hours of exposure time.

[0014] Next, as shown in Figure 1, the total true strain ε obtained from each stress-strain curve is... t From elastic strain ε e and plastic strain ε p The value obtained by subtracting from is the hydrogen erosion strain ε h Defined as (Step ST2). That is, hydrogen erosion strain ε h =ε t -ε e -ε p It is defined as follows.

[0015] Next, as shown in Figure 1, hydrogen erosion strain is applied to the Ω method used for creep strain to numerically model the hydrogen erosion strain rate (Step ST3). Finally, as shown in Fig. 1, the remaining life of the pressure vessel is predicted by computer simulation using the numerically modeled hydrogen erosion strain rate and the actual operating conditions of the pressure vessel (step ST4).

[0016] As described above, in the method for predicting the remaining life of the pressure vessel according to the present embodiment, by applying the hydrogen erosion strain to the Ω method used for creep strain and numerically modeling the hydrogen erosion strain rate, the remaining life of the pressure vessel can be quantitatively predicted.

[0017] <Details of Step ST3> Next, the details of step ST3 of numerically modeling the hydrogen erosion strain rate by applying the hydrogen erosion strain to the Ω method used for creep strain will be described.

[0018] Prager assumed that the acceleration of the creep strain rate consists of the following three independent factors. (1) Increase in stress (2) Increase in creep damage (3) Microstructural changes other than creep damage

[0019] By the way, in the hydrogen erosion acceleration test, due to the internal pressure of the methane gas generated by hydrogen erosion, in the heat affected zone (HAZ) of the welded joint, after grain boundary microcracks occur in the bainite coarse grain region, they connect along the grain boundaries. On the other hand, in the base material, voids occur between the ferrite / pearlite boundaries or pearlite colonies, and they connect and grow into fishers. Therefore, it is considered that the progress of damage due to such hydrogen erosion corresponds to the acceleration factor of the creep strain rate, and the hydrogen erosion strain ε h is applied to the Ω method used for creep strain.

[0020] Note that the remaining life can be predicted by numerical modeling using the Ω method in the HAZ of the welded joint and the base material as well. However, the progress of damage due to hydrogen erosion is much faster in the HAZ of the welded joint than in the base material. Therefore, for predicting the remaining life of the pressure vessel, it is sufficient to predict only the remaining life in the HAZ of the welded joint.

[0021] In the Omega method, hydrogen erosion strain ε h and hydrogen erosion strain rate ε' h The natural logarithm of lnε' h Between these, as shown in equation (1), the gradient is Ω and the intercept is lnε'. h0 This can be expressed as a linear relationship where Ω is the strain rate acceleration factor and ε' h0 This is the initial strain rate. lnε' h =Ωε h +lnε' h0 ...(1)

[0022] When the relationship in equation (1) holds, the hydrogen erosion curve corresponding to the creep curve is given by two parameters: the strain rate acceleration factor Ω and the initial strain rate ε'. h0 It can be expressed using only these two parameters. Therefore, the hydrogen erosion lifetime can also be expressed using these two parameters. By rearranging equation (1), we obtain equation (2). ε' h =ε' h0 exp(Ωε h )···(2)

[0023] Integrating equation (2) with respect to time t, we obtain ε at t=0. h Setting = 0 yields equation (3). ε h =-(1 / Ω)ln(1-ε' h0 Ωt)···(3) Here, if we define fracture as ε=∞, then the fracture lifetime t r This is expressed by equation (4). t r =1 / (ε' h0 Ω)...Equation (4)

[0024] On the other hand, differentiating equation (3) with respect to time t gives the hydrogen erosion strain rate ε' h Equation (5) is obtained for this. ε' h =ε' h0 / (1-ε' h0 Ωt)···(5) Furthermore, rearranging equation (5) for time t yields equation (6). t = 1 / (ε' h0 Ω)-1 / (ε' h Ω)···(6)

[0025] Then, from equations (4) and (6), the remaining life (t r Equation (7) for -t is obtained. t r -t=1 / (ε' h Ω)···(7) Equation (7) shows the remaining life (t r -t) and the hydrogen erosion strain rate ε' at time t h This means that the product of Ω and ε' is a constant value of 1 / Ω. In other words, the strain rate acceleration factor of the damaged material Ω and the hydrogen erosion strain rate ε' at a certain time t. h If we know that, we can determine the remaining lifespan (t) of the material. r -t) can be calculated.

[0026] Furthermore, by rearranging equations (2) and (5), we obtain the hydrogen erosion strain rate ε' increased by the damage. h Acceleration ε' h / ε' h0 Equation (8) is obtained for this. ε' h / ε' h0 =exp(Ωε h ) = 1 / (1-ε' h0 Ωt)···(8) Equation (8) shows the strain rate acceleration factor Ω of the damaged material and the hydrogen intrusion strain ε under operating conditions (temperature and stress). h If this is known, then the acceleration ε' of the hydrogen erosion strain rate increased by the damage under those operating conditions can be determined. h / ε' h0 This means that it is possible to calculate it.

[0027] Furthermore, after dividing equation (6) by equation (4), we use equation (8) to obtain the lifetime consumption rate t / t r Equation (9) is obtained for this. t / t r = 1 - ε h0 ' / ε h '=ε h0Ωt···(9)

[0028] In the remaining life assessment, the life consumption rate t / t is expressed by equation (9). r Let the sum of these be the cumulative damage D (≤ 1). Here, the cumulative damage D is the damage rate D' (= ε') obtained by differentiating equation (9) with respect to time t. h0 This is the calculated value obtained by integrating Ω, and is expressed by equation (10). D = Σt / t r =∫D'dt=∫ε' h0 Ωdt=ε' h0 Ωt···(10)

[0029] Substituting equation (10) into equation (5), we get the hydrogen erosion strain rate ε' h It is given by equation (11) using cumulative damage D. ε' h =ε' h0 / (1-D)···(11)

[0030] Initial strain rate ε' h0 The strain rate acceleration factor Ω is a function of temperature T and the initial true stress (hereinafter referred to as initial stress) σ0. The API-579-1 / ASME FFS-1 standard specifies the initial strain rate ε'. h0 This is where temperature T and stress parameter S are located. l (=Vibration 10 It is defined by equation (12) using [σ0]) and fitting parameters α1 to α4. Log 10 [ε' h0 ]=-(α1+α2S l +α3S l 2 +α4S l 3 ) / (273+T)···(12)

[0031] Here, the fitting parameters α1 to α4 in equation (12) are the stress parameters S obtained from the stress-strain curve. l (=Vibration 10 [σ0]) and initial strain rate ε' h0 The common logarithm of log 10 [ε' h0It is determined by fitting with respect to the relationship with

[0032] Also, in API-579-1 / ASME FFS-1 standard, the strain rate acceleration factor Ω is defined by Equation (13) using temperature T, stress parameter S l , and fitting parameters β1 to β4. log 10 [Ω]=(β1 + β2S l + β3S l 2 + β4S l 3 ) / (273 + T) ··· (13)

[0033] Here, the fitting parameters β1 to β4 in Equation (13) are determined by fitting with respect to the relationship between the stress parameter S l (= log 10 [σ0]) obtained from the stress-strain curve and the common logarithm log 10 [Ω] of the strain rate acceleration factor Ω. By the above numerical modeling, it becomes possible to predict the remaining life of the pressure vessel under various operating conditions (temperature T and initial stress σ0).

[0034] In addition, in step ST4 of predicting the remaining life of the pressure vessel by computer simulation, the effective stress σ e defined by the following Equation (14) is used as the initial stress σ0. σ e = 1 / √2[(σ1 - σ2) 2 +(σ1 - σ3) 2 +(σ2 - σ3) 2 (1 / 2) ··· (14)

[0035] Here, the stresses σ1, σ2, and σ3 in Equation (14) are, for example, in the case of a pressure vessel, the circumferential stress, meridional stress, and radial stress, respectively.

[0036] <Mathematical formulas used for numerical modeling> The following are the mathematical formulas used for numerical modeling. The formulas listed below are specified in API-579-1 / ASME FFS-1 standard. As described above, the hydrogen corrosion strain ε h is the total true strain ε t obtained from the stress-strain curve minus the elastic strain ε e and the plastic strain ε p and is expressed by Equation (15). ε h = ε t - ε e - ε p ···(15)

[0037] The elastic strain ε e in Equation (15) is, as shown in Equation (16), the true stress σ t divided by the Young's modulus E y at the target temperature. ε e = σ t / E y ···(16) ]

[0038] The plastic strain ε p in Equation (15) is, as shown in Equation (17), the sum of the true strain γ1 in the microscopic strain region of the stress-strain curve and the true strain γ2 in the macroscopic strain region of the stress-strain curve. <() ε<000()0111>= γ1 + γ2 ···(17)

[0039] The true strain γ1 in Equation (17) is expressed by Equation (18) using the true plastic strain ε1 in the microscopic strain region of the stress-strain curve and the Prager doctor coefficient H. γ1 = ε1(1.0 - tanh[H]) / 2 ···(18)

[0040] The true strain γ2 in Equation (17) is expressed by Equation (19) using the true plastic strain ε2 in the macroscopic strain region of the stress-strain curve and the Prager doctor coefficient H. γ2=ε2(1.0-tanh[H]) / 2···(19)

[0041] In equations (18) and (19), the Prager doctor coefficient H is the engineering yield stress σ evaluated at the target temperature. ys The engineering limit tensile stress σ evaluated at the target temperature uts Using the parameter K of the MPC stress-strain curve model, it can be expressed by equation (20). H=2[σ t -{σ ys +K(σ uts -σ ys )}] / {K(σ uts -σ ys )}···(20)

[0042] In equation (20), the parameter K is the engineering yield stress σ ys and engineering limit tensile stress σ uts Using this, it can be expressed in equation (21). K=1.5(σ ys / σ uts ) 1.5 +0.5(σ ys / σ uts ) 2.5 -(σ ys / σ uts ) 3.5 ···(twenty one)

[0043] In equation (18), the true plastic strain ε1 in the microscopic strain region is equal to the true stress σ t The stress-strain curve is expressed by equation (22), where A1 is the curve fitting constant for the elastic region of the stress-strain curve, and m1 is the curve fitting index of the stress-strain curve. Here, the curve fitting index m1 is equal to the true strain at the proportional limit and the strain hardening coefficient in the large strain region. ε1=(σ t / A1) (1 / m1) ···(twenty two)

[0044] In equation (22), the curve fitting constant A1 is equal to the engineering yield stress σ ys , 0.2% engineering offset strain ε ys(=2.0 × 10 -3 Using ), and the curve fitting index m1, it is expressed by equation (23). A1={σ ys (1+ε ys )} / (ln[1+ε ys ]) m1 ···(twenty three)

[0045] In equation (22), the curve fitting index m1 is equal to the engineering yield stress σ ys , engineering ultimate tensile stress σ uts , 0.2% engineering offset strain ε ys , and the 0.2% engineering offset strain ε for the proportional limit 0.2%p (=2.0 × 10 -5 Using ), it can be expressed in equation (24). m1={ln[σ ys / σ uts ]+(ε 0.2%p -ε ys )} / ln[ln[1+ε 0.2%p ] / ln[1+ε ys ]]···(twenty four)

[0046] In equation (19), the true plastic strain ε² in the macroscopic strain region of the stress-strain curve is equal to the true stress σ t The stress-strain curve is expressed by equation (25), where A2 is the curve fitting constant for the plastic region of the stress-strain curve, and m2 is the curve fitting index of the stress-strain curve. Here, the curve fitting index m2 is equal to the true strain at the true ultimate tensile stress. ε²=(σ t / A2) (1 / m2) ···(twenty five)

[0047] In equation (25), the curve fitting constant A2 is equal to the engineering limit tensile stress σ uts And using the curve fitting index m², it is expressed by equation (26). A2=(σ uts exp[m2]) / m2 m2 ...(26)

[0048] In equation (25), the curve fitting index m² is equal to the engineering yield stress σ. ys and engineering limit tensile stress σ uts Using this, it can be expressed in equation (27). m2 = 0.60(1.00 - σ) ys / σ uts )···(27) [Examples]

[0049] The following describes an example of the method for predicting the remaining lifespan of a pressure vessel according to the first embodiment. However, the method for predicting the remaining lifespan of a pressure vessel according to this embodiment is not limited to the following examples.

[0050] Table 1 shows the composition (mass%) and creep embrittlement factor (CEF) of the 1.25Cr-0.5Mo steel used in the examples. In Table 1, the composition of N and O is shown in mass ppm. Here, CEF is expressed by equation (28) using the concentrations (mass%) of As, Sn, and Sb. CEF=P+2.4As+3.6Sn+8.2Sb...(28)

[0051] As shown in Table 1, 50 kg square steel ingots of two types of 1.25Cr-0.5Mo steel, mainly differing in Sb concentration, were melted. As shown in Table 1, steel with an Sb concentration of 0.001 mass% is called FSb material, and steel with an Sb concentration of 0.015 mass% is called 15Sb material. 15Sb material is modeled after the CEF of 1.25Cr-0.5Mo steel manufactured in the 1960s, and FSb material is modeled after the CEF of 1.25Cr-0.5Mo steel manufactured in the 1990s. Both FSb and 15Sb materials are 1.25Cr-0.5Mo steel conforming to ASTM A387 Grade 11.

[0052] [Table 1]

[0053] <Test Conditions> Figure 2 is a schematic graph showing the temperature pattern of the heat treatment performed on the steel plates according to the example. For both FSb and 15Sb materials, a 50 kg square steel ingot was hot-rolled into a 20 mm thick steel plate, and the heat treatment shown in Figure 2 was performed in accordance with the requirements of ASME BPVC.Sec.II,Part A.

[0054] Specifically, as shown in Figure 2, the normalizing process involved first raising the temperature from 600°C to 920°C at a rate of 50°C / h, holding it at 920°C for 3 hours, and then furnace cooling. Here, the normalizing cooling rate was set to 1°C / min to obtain a pheto-pearlite structure. Next, the temperature was increased from 500°C to 650°C at a rate of 50°C / h, held at 650°C for 5 hours, and then subjected to a tempering process by furnace cooling.

[0055] Figure 3 is a schematic perspective view of a welded joint made from steel plates according to the embodiment. The two heat-treated steel plates were welded together in two passes using a φ4.0 mm covered arc welding rod to produce the welded joint shown in Figure 3. Although post-weld heat treatment is known to convert the supersaturated solid solution carbon in the as-welded state into stable carbides and reduce susceptibility to hydrogen erosion, it was omitted in this example because the steel plate thickness in this example was only 20 mm.

[0056] <Accelerated hydrogen erosion test> For both FSb and 15Sb materials, welded joints as shown in Figure 3 were placed in an autoclave and accelerated hydrogen erosion tests were conducted under hydrogen pressure. In the accelerated hydrogen erosion tests, only the exposure time was varied at a single temperature and hydrogen pressure, which were higher than the actual operating conditions of the pressure vessel. Specifically, the test temperature was 550°C, the hydrogen pressure was 20 MPa, and the exposure times were 96h, 168h, and 240h.

[0057] Here, Figure 4 shows SEM images of the microstructure in the HAZ of a welded joint after accelerated hydrogen erosion testing. Figure 4 shows the microstructure for FSb and 15Sb materials at exposure times of 96h, 168h, and 240h, respectively. As shown in Figure 4, for both FSb and 15Sb materials, grain boundary microcrack type damage in the coarse-grained bainite portion was reproduced at each exposure time. Furthermore, no damage was observed in the deposited metal.

[0058] <Tensile Test> Tensile tests were conducted on welded joints of FSb and 15Sb materials to investigate the deterioration of tensile properties, i.e., the stress-strain curve, as hydrogen erosion progresses. More specifically, tensile test specimens were prepared from welded joints of FSb and 15Sb materials at exposure times of 0h, 96h, 168h, and 240h in accelerated hydrogen erosion tests, and tensile tests were performed on these specimens. Here, an exposure time of 0h means before the accelerated hydrogen erosion test. This process corresponds to step ST1 in the flowchart shown in Figure 1.

[0059] Here, Figure 5 is a schematic cross-sectional view of the welded joint illustrating the location of the tensile test specimen to be taken. The tensile test specimen is a round bar-shaped specimen with a parallel section diameter of φ6.0 mm and a gauge length of 30 mm. As shown in Figure 5, the tensile test specimen was taken from the welded joint so that the HAZ (Heat-Absorbing Zone) was located in the longitudinal center of the tensile test specimen.

[0060] Table 2 summarizes the tensile test results for welded joints of FSb and 15Sb materials at various exposure times, namely yield stress (MPa), tensile strength (MPa), elongation at break (%), and reduction of area (%).

[0061] [Table 2]

[0062] As shown in Table 2, the deterioration of tensile properties associated with the progression of hydrogen erosion in the HAZ of welded joints manifests as a decrease in ductility (elongation at break and reduction of area in Table 2) and tensile strength. The degree of these decreases is more pronounced in 15Sb material than in FSb material. Therefore, it is considered that the progression of hydrogen erosion, such as the generation and coupling of grain boundary microcracks in the coarse-grained bainite portion of the HAZ of welded joints, is influenced by the impurity element Sb contained in the base material, similar to creep.

[0063] <Definition of hydrogen erosion strain> Each exposure time t i The nominal stress-nominal strain curves obtained from tensile test results of welded joints of FSb and 15Sb materials were converted to true stress-true strain curves. Each exposure time t i In the true stress-true strain curves for (=96h, 168h, 240h), for a certain stress value σ t Total true strain ε t_i From elastic strain ε e_i and plastic strain ε p_i The value obtained by subtracting from is the hydrogen erosion strain ε h_i This was defined as follows. This is the relationship shown in equation (15), and we will denote this relationship as equation (15)'. ε h_i =ε t_i -ε e_i -ε p_i ...(15) This process corresponds to step ST2 in the flowchart shown in Figure 1.

[0064] In equation (15)', the subscript i represents the exposure time, where i = 96, 168, and 240. Furthermore, the elastic strain ε in equation (15)' e_i This can be calculated using equation (16), and the plastic strain ε p_i This can be calculated using equations (17) to (27).

[0065] <Numerical modeling of hydrogen erosion strain using the Omega method> Next, the hydrogen erosion strain ε obtained by equation (15)' h_iApplying this to the Ω method, the hydrogen erosion strain rate ε' h This was modeled numerically. This process corresponds to step ST3 in the flowchart shown in Figure 1.

[0066] First, as shown in equation (29), hydrogen erosion strain ε h_i exposure time t i The value obtained by dividing by ε' is the hydrogen erosion strain rate. h_i (=∂ε _i / ∂t i ) was defined as follows. ∂ε h_i / ∂t i =ε h_i / t i ...(29) Furthermore, the stress value σ t Each exposure time t with initial stress σ0 i Hydrogen erosion strain value ε h_i The history is the hydrogen erosion strain ε over time under constant stress σ0. h It was assumed that this corresponds to an increase in [amount].

[0067] The hydrogen erosion strain ε using the true stress-true strain curve described above. h_i and hydrogen erosion strain rate ∂ε _i / ∂t i A concrete example of the definition is shown in Figure 6. Figure 6 is a graph showing the definitions of hydrogen erosion strain and hydrogen erosion strain rate using true stress-true strain curves based on the hydrogen erosion accelerated test time. As an example, Figure 6 shows the definitions of each exposure time t for 15Sb material. i Hydrogen erosion strain ε using true stress-true strain curves at (=96h, 168h, 240h) h_i and hydrogen erosion strain rate ∂ε _i / ∂t i This shows the definition. The same applies to FSb material.

[0068] Next, consider a certain stress value σ in the true stress-true strain curve described above. t (i.e., the hydrogen erosion strain ε obtained for the initial stress σ0) h_i and hydrogen erosion strain rate ∂ε _i / ∂ti Equation (1), i.e., the Ω method, is applied to the relationship. This gives the strain rate acceleration factor Ω and the initial strain rate ε' at the initial stress σ0. h0 The result is obtained. As a result, the fracture lifetime t can be used with equation (4). r (=1 / (ε' h0 Ω)) is obtained.

[0069] Table 3 shows the strain rate acceleration factor Ω and initial strain rate ε' at each initial stress σ0 (MPa) obtained from the true stress-true strain curves for welded joints of FSb and 15Sb materials. h0 (mm / mm / h), and fracture life t r The results of (h) are summarized below.

[0070] [Table 3]

[0071] Here, the validity of the Ω method is determined by the hydrogen erosion strain ε. h and the logarithm of remaining life log(t) r This can be verified by plotting the relationship with -t). Therefore, for FSb material and 15Sb material respectively, the hydrogen erosion strain ε at each initial stress σ0 shown in Table 3 h For this, the hydrogen erosion strain ε' defined in equation (29) h Substituting this into equation (7), we obtain the remaining lifetime (t r The logarithm of -t) was plotted.

[0072] Here, Figure 7 shows the hydrogen erosion strain ε by the Ω method in the welded joint of the aforementioned FSb material. h and the logarithm of remaining life log(t) r This graph shows the relationship with -t). The × marks indicate data points corresponding to exposure times of 96h, 168h, and 240h. The plots at each initial stress σ0 all show linearity, and it can be concluded that the application of the Ω method to the progression of hydrogen erosion strain in accelerated hydrogen erosion tests with varying exposure times of 96h, 168h, and 240h is appropriate.

[0073] Furthermore, Figure 8 shows the hydrogen erosion strain ε of the welded joint of the aforementioned 15Sb material, measured by the Ω method. h and the logarithm of remaining life log(t) r This graph shows the relationship with -t). The × marks indicate data points corresponding to exposure times of 96h, 168h, and 240h. The plots at each initial stress σ0 all show linearity, and it can be concluded that the application of the Ω method to the progression of hydrogen erosion strain in accelerated hydrogen erosion tests with varying exposure times of 96h, 168h, and 240h is appropriate.

[0074] Next, the initial strain rate ε' h0 The stress parameter S is defined in equation (12). l This was numerically modeled as a function of the target temperature T. Specifically, the initial strain rate ε' at the four initial stresses σ0 shown in Table 3. h0 Curve fitting was performed on the value of to determine the fitting parameters α1 to α4 in equation (12).

[0075] Similarly, the stress parameter S defines the strain rate acceleration factor Ω in equation (13). l This was then numerically modeled as a function of the target temperature T. Specifically, curve fitting was performed on the values ​​of the strain rate acceleration factor Ω at the four initial stresses σ0 shown in Table 3, and the fitting parameters β1 to β4 in equation (13) were determined.

[0076] The numerical model described above made it possible to predict the remaining life of welded joints made of FSb and 15Sb materials under various operating conditions (temperature T and initial stress σ0). More specifically, the initial strain rate ε' at the temperature T and initial stress σ0 to be evaluated. h0 And from the value of the strain rate acceleration factor Ω, the damage rate D'(=ε') can be calculated. h0 The value of Ω can be calculated. As a result, the hydrogen erosion strain rate ε' can be calculated using equation (11). h The value can be calculated.

[0077] <Implementation in computer simulations> Next, the numerically modeled hydrogen erosion strain rate ε' hThe remaining lifespan of the pressure vessel was predicted through computer simulations using the actual operating conditions of the pressure vessel. This process corresponds to step ST4 in the flowchart shown in Figure 1.

[0078] The strain rate acceleration factor Ω and initial strain rate ε' are calculated based on the actual operating conditions of initial stress σ0 and temperature T. h0 The strain rate ε' given by h A subroutine was created to implement this into a computer numerical simulation. The initial stress σ0 is the effective stress σ expressed by equation (14). e That is the case.

[0079] Here, Figure 9 is a flowchart showing the subroutines for implementation in numerical simulation. First, as shown in Figure 9, the effective stress σ e And the strain rate acceleration factor Ω from temperature T k and initial strain rate ε' h0,k Calculate the damage rate D' by multiplying the two together. h,k (=ε' h0,k Ω k Calculate (step ST41).

[0080] Next, as shown in Figure 9, the exposure time t is defined by equation (30). h Time step Δt k Using this, the hydrogen erosion damage rate D' is as shown in equation (31). h,k Numerically integrate the hydrogen erosion damage D h,k Calculate (step ST42). t h =ΣΔt k ...Equation (30) D h,k =D h,k-1 +D' h,k *Δt k =D h,k-1 +ε' h0,k *Ω k *Δt k ...Equation (31) Note that the asterisk (*) in equation (31), etc., signifies multiplication.

[0081] Next, as shown in Figure 9, the obtained hydrogen erosion damage D h,k Substitute this into equation (11) and get the strain rate ε' h,k Calculate (step ST43). ε' h,k =ε' h0,k / (1-D h,k )...Equation (11)

[0082] Next, as shown in Figure 9, the obtained strain rate ε' is given by equation (32). h,k time step Δt k Multiply by Δε to obtain the incremental value of hydrogen erosion strain. h,k Calculate (step ST44). Δε' h,k =ε' h,k *Δt k ...Equation (32)

[0083] Finally, as shown in Figure 9, the incremental value of hydrogen erosion strain Δε is obtained using equation (33). h,k The equivalent stress derivative is calculated. In equation (33), Δσ is a small equivalent stress value artificially given by inputting a command into the subroutine input data. Equation (33) shows a numerical approximation using the definition of the derivative that brings Δσ closer to 0. ∂Δε h / ∂σ e =lim{Δε h (σ e (+Δσ)-Δε h (σ e )} / Δσ···Equation (33) The above subroutine processing was implemented and analyzed using numerical simulation with the finite element method (FEA).

[0084] <Verification of computer simulation operation> Next, in order to verify the operation of the computer simulation using the above subroutine, hydrogen erosion damage D h (t) and hydrogen erosion strain ε hThe theoretical calculation results for (t) were compared with the numerical analysis results obtained from computer simulations.

[0085] Here, Figure 10 is a schematic diagram showing the analytical model of the reactor shell used for comparing theoretical calculations and numerical analysis. As shown in Figure 10, for the 1.25Cr-0.5Mo steel hydrocracking reactor, an internal pressure cylinder with actual shell dimensions (internal radius Ri = 1146.2 mm, external radius Ro = 1259.7 mm) was used as the analysis model. Furthermore, the actual operating conditions (pressure P = 10.93 MPa, temperature T = 427 °C) were used as the analysis conditions.

[0086] Here, Figure 11 shows the initial true stress-true strain curve of the 1.25Cr-0.5Mo steel input for analysis. Before the accelerated hydrogen erosion test, both the FSb and 15Sb materials showed no influence from the impurity elements in the base material, and their tensile properties were at the same level. Furthermore, in order to compare the analysis results using only the hydrogen erosion characteristics as the factor in damage progression, the true stress-true strain curve shown in Figure 11 was used as the initial true stress-true strain curve input for analysis for both the FSb and 15Sb materials.

[0087] Figure 12 is a graph showing the initial principal stress distribution across the reactor shell thickness under actual operating conditions. The operating conditions are pressure P = 10.93 MPa and temperature T = 427°C. The stresses σ1, σ2, and σ3 shown in Figure 12 are circumferential stress, meridional stress, and radial stress, respectively. Using equation (14), the effective stress σ e This can be calculated. Note that the meridional stress σ2 shown in Figure 12 is the end force F in the internal pressure cylinder used as the analysis model shown in Figure 10. end This is the same value (52.59 MPa).

[0088] Figure 13 shows hydrogen erosion damage D in welded joints of FSb material. h This graph shows a comparison between theoretical calculations and numerical analysis of the time-dependent changes in the distribution of (t) across the reactor shell thickness. Figure 14 also shows the hydrogen erosion strain ε in the welded joint of FSb material. h This graph shows a comparison between theoretical calculations and numerical analysis of the time-dependent changes in the distribution of (t) across the reactor shell thickness.

[0089] In Figures 13 and 14, theoretical calculation values ​​are shown by squares and circles. On the other hand, in Figures 13 and 14, numerical analysis values ​​obtained by computer simulation are shown by dotted lines. As shown in Figures 13 and 14, hydrogen erosion damage D h (t) and hydrogen erosion strain ε h The theoretical calculation value and the numerical analysis value for (t) were in agreement, which was extremely good.

[0090] Figure 15 shows hydrogen erosion damage D in a welded joint of 15Sb material. h This graph shows a comparison between theoretical calculations and numerical analysis of the time-dependent changes in the distribution of (t) across the reactor shell thickness. Figure 16 also shows the hydrogen erosion strain ε in a welded joint of 15Sb material. h This graph shows a comparison between theoretical calculations and numerical analysis of the time-dependent changes in the distribution of (t) across the reactor shell thickness.

[0091] In Figures 15 and 16, theoretical calculation values ​​are shown by squares and circles. On the other hand, in Figures 15 and 16, numerical analysis values ​​obtained by computer simulation are shown by solid lines. As shown in Figures 15 and 16, hydrogen erosion damage D h (t) and hydrogen erosion strain ε h The theoretical calculation value and the numerical analysis value for (t) were in close agreement, which was good.

[0092] Furthermore, when comparing FSb material and 15Sb material, the numerical analysis values ​​obtained by computer simulation for FSb material are in better agreement with the theoretical calculation values. Although not particularly limited, the method for predicting the remaining life of a pressure vessel according to this embodiment is more suitable for materials with an Sb concentration of 0.005 mass% or less, and even more so, 0.003 mass% or less.

[0093] The present invention has been described in detail above based on embodiments, but it goes without saying that the present invention is not limited to the embodiments already described, and various modifications are possible without departing from the spirit of the invention.

Claims

1. A method for predicting the remaining life of a pressure vessel made of steel, (a) A step of obtaining stress-strain curves for each of the multiple test pieces made of the steel material that has undergone hydrogen erosion acceleration testing, (b) A step of defining hydrogen erosion strain as the value obtained by subtracting elastic strain and plastic strain from the total true strain obtained from the stress-strain curve, (c) A step of numerically modeling the hydrogen erosion strain rate by applying the hydrogen erosion strain to the Ω method used for creep strain, and (d) A step of predicting the remaining life of the pressure vessel by computer simulation using the numerically modeled hydrogen erosion strain rate and the actual operating conditions of the pressure vessel, A method for predicting the remaining lifespan of a pressure vessel.

2. In step (a), each of the plurality of test pieces is a welded joint. The method for predicting the remaining lifespan of a pressure vessel according to claim 1.

3. In step (a), the heat-affected zone of the welded joint is located at the longitudinal center of each of the plurality of test pieces. The method for predicting the remaining lifespan of a pressure vessel according to claim 2.

4. The aforementioned steel material is Cr-Mo steel. A method for predicting the remaining life of a pressure vessel according to any one of claims 1 to 3.

5. The Sb concentration in the aforementioned steel material is 0.005% by mass or less. The method for predicting the remaining lifespan of a pressure vessel according to claim 4.

6. In the aforementioned accelerated hydrogen erosion test, only the exposure time is varied at a single temperature and hydrogen pressure that are higher and higher than the actual usage conditions. A method for predicting the remaining life of a pressure vessel according to any one of claims 1 to 3.

7. The pressure vessel is a reactor for petroleum refining. A method for predicting the remaining life of a pressure vessel according to any one of claims 1 to 3.

8. In step (c), the hydrogen erosion strain obtained from the stress-strain curve is divided by the exposure time in the accelerated hydrogen erosion test, and this value is defined as the hydrogen erosion strain rate. A method for predicting the remaining life of a pressure vessel according to any one of claims 1 to 3.

Images