JPH10160646A - Method for anticipating fatigue life of structure member - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for anticipating fatigue life of a structure member wherein future usable life is appropriately anticipated, by measuring crack length based on low-cycle thermal fatigue occurring with the structure member, for anticipating crack length at operation thereafter. SOLUTION: With stress distribution analysis based on data for use state (1), stress distribution of structure member is divided (2) to set stress-division region (3). From an image input of crack length in the set division region (4), crack length is measured to decide maximum crack length in division region (5 and 6). Damage repetition times ratio is decided as fatigue life reference value of structure member obtained from maser curve generated in advance with the maximum crack length in the division region (7). With the value, while future operation.stop operation times are added (8), future development of crack length of structure is anticipated (9), to be compared to a threshold crack length which leads to damage (10), and based on the difference, start-up.stop operation times for structure is judged (11).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for estimating the fatigue life of a structural member, and more particularly to a method of estimating a usable life by measuring low-cycle thermal fatigue damage occurring in a structural member used over a long period of time at a high temperature. The present invention relates to a method for predicting the fatigue life of a structural member that accurately predicts the fatigue life.

[0002]

2. Description of the Related Art For example, in a very large prime mover such as a gas turbine, thermal stress is applied to structural members such as a gas turbine combustor, a gas turbine stationary blade, and a gas turbine rotor due to a temperature change caused by repeated start / stop operations. This thermal stress accumulates over a long period of operation, and as a result, it is one factor of causing cracks in structural members as so-called low cycle thermal fatigue damage.

[0003] It has been reported in the literature (eg, Fujiyama et al., JSME [No.940-34] Symposium Proceedings and Materials, Structural Strength and Fracture, "Estimation of High-Temperature Low-Cycle Fatigue Crack Initiation and Growth Process of Superalloy Using Simple Simulation Analysis Model").

However, when a crack occurs in a structural member,
Even if the development of cracks has been elucidated, it is difficult to predict the life of the structural member over how long it can be used.

[0005]

In order to accurately evaluate the remaining life of a structural member due to low-cycle thermal fatigue damage and to set an inspection / repair interval, it is necessary to establish an accurate evaluation and prediction method. However, conventionally, the number of failure cycles Nf of a test piece having a diameter of about 10 mm was used as a reference for the life, and it was considered that the number of failure cycles Nf of the test piece corresponded to the occurrence of a crack in a target portion of a structural member applied to an actual machine. I was stopped. In this case, when the load level of the test piece falls to 75% of the steady state, the number of failure cycles N
f (JIS standard Z2279), and if the cross-sectional area of the semi-circular crack in the cross-sectional area at this time is 25%, the crack length on the surface becomes about 8 mm, and the crack length becomes longer. Is not guaranteed.

However, the structural member applied to the actual machine often has a damaged area much larger than that of the test piece, and the observable crack length is more than two orders of magnitude in comparison with the test piece. In addition, the surface distribution pattern of a large number of cracks generated in a region subjected to the same number of failure cycles Nf in the same strain range is considered to be the same for both the structural member and the test piece applied to the actual machine. Since the point at which the peak load, which is the number of failure repetitions Nf, falls by 25% is set as failure, the applicable range is extremely narrow. For this reason, even when the structural member applied to the actual machine is exposed to a high temperature state, the structural member is restrained by the coupling relationship with other structural members, and if the rigidity is maintained by the restraining force, further damage is caused in the damage area. Even if it proceeded, it was not damaged, and a crack longer than the crack of the test piece was sometimes observed.

[0007] Therefore, in the conventional crack reference method, there are inconveniences and disadvantages in directly applying the method to structural members applicable to an actual machine.

Further, as a method of evaluating fatigue life based on cracks of a structural member which has already been proposed, for example, Japanese Unexamined Patent Publication No.
This evaluation method estimates the crack length of a structural member applied to an actual machine by an extreme value statistical method.

However, in this evaluation method, the crack length of the structural member is limited to the range at the test piece level, and it is impossible to apply the crack length to a structural member applicable to an actual machine.

The present invention has been made in view of such circumstances, and measures a crack length based on low cycle thermal fatigue generated in a structural member, and based on the measured data, determines a crack in a subsequent operation. An object of the present invention is to provide a fatigue life prediction method for a structural member that predicts a length and accurately predicts a future usable life.

[0011]

According to a first aspect of the present invention, there is provided a method for predicting the fatigue life of a structural member, comprising the steps of: Set the stress division area by dividing the stress distribution area of the member, input the crack length of this set division area as an image, measure the crack length from this image and determine the maximum crack length in the division area Then, based on the maximum crack length in the sectioned area, a failure cycle ratio as a fatigue life reference value of the structural member obtained from a master curve created in advance is determined, and the determined failure cycle number Predicting the future growth of the structure's crack length taking into account the number of shutdown operations, comparing the predicted structure's crack length with the critical crack length that can lead to structure damage, and calculating the difference from the difference How many times the object will be activated in the future It is to determine whether it is possible to stop operating frequency.

According to a second aspect of the present invention, there is provided a method for predicting the fatigue life of a structural member, comprising the steps of: From the thickness and operating time, the operating temperature of the structural member was estimated using the data of the oxide film thickness, temperature, and time obtained separately in the high-temperature oxidation test, and from the estimated temperature, repeated damage to the structural member as a fatigue life standard value The life ratio of the structural member is predicted by correcting the numerical ratio.

According to a third aspect of the present invention, there is provided a method for predicting the fatigue life of a structural member, wherein the thickness of the oxide film of the structural member is measured using an oxide film size measuring instrument. Is what you do.

According to a fourth aspect of the present invention, there is provided a method for predicting the fatigue life of a structural member, wherein the thickness of the oxide film of the structural member is non-destructively measured by an ultrasonic method. It is something to be done.

According to a fifth aspect of the present invention, there is provided a method for predicting the fatigue life of a structural member, wherein the thickness of the oxide film on the structural member is non-destructively measured by an electromagnetic method. What to do.

According to a sixth aspect of the present invention, there is provided a method for predicting the fatigue life of a structural member, the method comprising the steps of: The crack length density represented by the sum of the crack lengths per unit area is measured, and the stress division region of the structural member is set from the crack length density distribution.

In the method for predicting the fatigue life of a structural member according to the present invention, in order to achieve the above object, the crack length density of the structural member is measured by a replica as described in claim 7. .

In order to achieve the above object, the method for predicting the fatigue life of a structural member according to the present invention, as described in claim 8, determines the ratio of the number of failure cycles as the reference value of the fatigue life of the structural member to the crack length. An evaluation formula determined from the density and the number of start / stop operations is used.

In order to achieve the above object, the method for predicting the fatigue life of a structural member according to the present invention, when determining the number of repeated failures as a reference value for the fatigue life of a structural member, as described in claim 9, An evaluation formula that can be used to identify the number of fractures determined from the number of cracks and crack length during the previous inspection and the number of fractures determined from the number of cracks and crack length during the current evaluation for each inspection It is.

According to a tenth aspect of the present invention, there is provided a method for predicting the fatigue life of a structural member according to the present invention. The stress / strain acting on the region is determined, the stress / strain distribution in the thickness direction of the structural member is corrected from the obtained stress / strain value, and the amount of crack propagation in the thickness direction of the structural member is predicted.

In order to achieve the above object, the method for predicting the fatigue life of a structural member according to the present invention is characterized in that the stress / strain value is calculated by using the equation for the repetitive stress / strain relationship of the structural member. It is obtained from the intersection with the equation of the Neubar stress-strain relationship.

[0022]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a method for predicting fatigue life of a structural member according to the present invention will be described below with reference to the accompanying drawings.

FIG. 1 is a block diagram conceptually showing a first embodiment of a method for predicting the fatigue life of a structural member according to the present invention.

In the present embodiment, for example, a stress analysis means 1 for analyzing stress generated in a structural member based on operating condition data of a gas turbine, a stress distribution is obtained from stress data calculated by stress analysis, and a low cycle thermal fatigue damage region Means for setting a stress distribution setting means for determining the fatigue life from the low-cycle thermal fatigue damage area, means for setting a section area for performing the fatigue life evaluation, surface image input means for inputting a surface crack image of a structural member, cracks from the input image. Image processing means 5 for generating coordinate data of the above, a crack length measuring means 6 for measuring a crack length from the coordinate data of the crack and obtaining a maximum crack length, and a data for the obtained maximum crack length and the number of start / stop operations. , Fatigue life reference value setting means 7 for determining the standard failure cycle number ratio, operating condition setting means 8 for setting future operating conditions, maximum crack length from future operating conditions Critical crack setting means 10, critical crack setting means 10, and critical crack setting means 10 for determining whether or not a crack is to be coalesced based on information on the predicted crack length predicting means 9, segmented area setting means 3, and crack length measuring means 6. Each step is provided with a remaining life estimating means 11 for determining an allowable damage repetition ratio from the information of the estimating means 9.

As shown in FIG. 2, the stress analysis means 1 operates, for example, the operating time such as the temperature, pressure, flow rate, flow rate, and rotation speed of the combustion gas for driving the gas turbine 12, which is input to the operation control panel 13. A stress analysis that performs temperature analysis for each part of structural members such as gas turbine combustors, gas turbine vanes, and gas turbine rotor blades based on data for changes, and obtains thermal stress and the like generated in each part from the analysis temperature of these structural members Is what you do.

Further, as shown in FIG. 3, the stress distribution setting means 2 applies the stress value calculated by the stress analysis means 1 to the gas turbine stationary blade in the order of the stress value in the order of higher stress values as shown in FIG. Since the distribution area B can be expressed as the stress distribution area C, for example, when the stress distribution area B exceeds the yield stress of the structural member and becomes a low-cycle thermal fatigue condition, as shown in FIG. B is set as a low-cycle thermal fatigue damage area D.

Further, as shown in FIG. 5, the section area setting means 3 sets a section area to be divided into a rectangular area having the same width as the low cycle thermal fatigue damage area D.

As shown in FIG. 6, the surface image input means may be used to inspect the structural members
The guide 15 corresponding to the curvature is set, and an image input head 17 such as a CCD camera mounted on a mount 16 that moves along the guide 15 is scanned into a predetermined divided area 18 and input as digital image data. is there. This image is subjected to black-and-white binarization processing by the image processing means 5, noise elimination is performed, black lines are thinned, and the coordinates of cracks are converted into data.

Further, as shown in FIG. 7, the crack length measuring means 6 creates a circumscribed rectangle 20 of the crack 19, determines the length of the long side or diagonal line as the crack length, and determines the length of the crack. Sort the data in ascending numerical order to get the maximum value a
Determine max.

In addition, the fatigue life reference value setting means 7 determines the ratio of the number of failure cycles N to the maximum value amax of the crack determined above.
/ Nf is obtained from the master curve shown in FIG.

In this master curve, the maximum crack length amax and the number of failure cycles N
/ Nf is obtained from the following equation.

[0032]

(Equation 1) Here, N indicates the number of crack occurrences of the test piece (however, the number of start / stop operations in the actual machine), Nf indicates the ratio of the number of times of failure of the test piece, and A and B indicate constants, respectively.

Conventionally, test piece breakage and repetition rate Nf
Shows that when the test piece is subjected to a low cycle thermal fatigue test under strain control, the peak stress at each repetition is 75% of the steady state.
% Is defined as the number N of occurrences of cracks when the number of cracks decreases.
For this reason, the existing surface crack length of the test piece is only about 8 mm when 25% of the cross-sectional area is occupied by a semicircular crack, and its application range extends only to the broken line shown in FIG. Not very narrow.

Therefore, in the above equation (1), the maximum crack length is no longer applicable when the maximum crack length exceeds the crack length corresponding to the number Nf of repeated failures of the test piece.

However, in this embodiment, as shown in the evaluation line indicated by the solid line in FIG. 8, even if the maximum crack length amax of the actual structural member is greatly increased, the failure repetition ratio N /
Nf can be applied.

That is, the maximum crack is generated by coalescence of frequently occurring cracks. In this case, the cumulative crack coalescing number density Σnc, which is an index of the crack coalescence rate, is obtained by accumulating the observed crack coalescing number and dividing by the observed area. Given by

[0037]

(Equation 2) Here, Δεf indicates the entire strain range, I0 indicates the saturation value (material constant) of the crack length density, and C and d indicate the constants.

The above equation (2) is normalized by the saturation value Io of the crack length density based on the material difference, and the difference in strain range is expressed by Δ
Corrected by εf. When the failure repetition ratio N / Nf is obtained from the test piece data using the above equation (2), the evaluation line monotonically increases as shown by the solid line in FIG. This monotonic increase is (N / Bf)>
Even if it becomes 1, it has not changed. Even when (N / Nf)> 1, it is considered that the reason why the evaluation line is monotonically increasing is that the crack propagating machine due to the coalescence of the cracks is continuously generated in the plastic strain distribution region. For this reason, the master curve shown in FIG. 8 shows the maximum crack length ama obtained with the test piece.
Even if the relationship between x and the failure repetition rate ratio N / Nf is extended as it is, it can be applied to actual structural members.

Actually, the actual structural member has cracks over a much larger area than the test piece, and the length of the cracks is long due to the coalescence of the cracks. By the way, when taking a gas turbine stationary blade as an example, the maximum crack length amax
May exceed 200 mm depending on the number N of start / stop operations as shown in FIG.

Even in such a case, the number of failure cycles determined by substituting the maximum crack length amax of the actual structural member into the above equation (1) and the number of failure cycles determined from the master curve shown in FIG. This is consistent with the ratio N / Nf. If the number of start / stop operations up to the evaluation time is known, the number of failure cycles Nf as a fatigue life reference value at the evaluation time is determined.
Is determined, and it is possible to predict how long a future crack will grow on the basis of the number Nf of breakage cycles.

As described above, the fatigue life reference value setting means 7 according to the present embodiment calculates the maximum value amax of the crack at the time of the current evaluation.
The ratio N / Nf of the number of times of failure can be easily obtained from the master curve shown in FIG.

Next, the crack length estimating means 9 sets the future number of start / stop operations N 'based on the information of the number of start / stop operations at the time of the current evaluation from the operation condition setting means 8, (N + N ') / Nf is calculated from the set number of start / stop times N', and the value of (N + N ') / Nf is calculated based on the number of start / stop operations N' set through the master curve evaluation line shown in FIG. This is to predict the maximum crack length amax value.

On the other hand, as shown in FIG. 11, the limit crack setting means 10 determines whether or not cracks occurring in the divided areas 1 and 2 defined on the gas turbine stationary blade are combined. is there. That is, when a crack is present in each of the adjacent divided areas 1 and 2, the cracks are combined so as to reach the area width of each of the two divided areas, and the root portion of the gas turbine vane is damaged. In this state, it is determined that the limit crack at this time is the sum of the crack lengths of the two divided areas is the sum of the widths of the two divided areas.

As described above, the limit crack setting means 10 according to the present embodiment determines the length of the crack at the time of the current evaluation until the structural member is broken by the coalescence of the cracks.

Finally, the remaining life judging means repeatedly uses the master curve again from the information on the crack length at the current evaluation pointed out by the above-described limit crack setting means 10 and passes through the evaluation line shown in FIG. While the number ratio N / Nf is determined, the failure repetition ratio N / Nf is determined from the maximum crack length amax predicted by the crack length prediction means 9 using a master curve, and both the failure repetition ratios N / Nf are obtained. In comparison with the above, the number of start / stop operations until the structural member is damaged is determined.

As described above, in the present embodiment, conventionally, the master curve of the ratio of the number of failure cycles to the maximum crack length, which is extremely narrow in the applicable range, is used to calculate the number of failure cycles even when the maximum crack length of the actual structural member is greatly increased. Since the master curve from which the numerical ratio can be obtained is created, a simple and accurate fatigue life prediction can be performed, and a highly reliable fatigue life prediction method for a structural member can be realized.

FIG. 12 is a block diagram conceptually showing a second embodiment of the method for predicting the fatigue life of a structural member according to the present invention. In the present embodiment, an oxide film thickness measuring unit 21 and a temperature estimating unit 22 are added to the configuration of the first embodiment. The same components as those of the first embodiment are denoted by the same reference numerals, and only different components will be described.

In recent gas turbines, the combustion gas temperature has reached an extremely high temperature of 1500 ° C. or higher in order to achieve high thermal efficiency and high output. In order to cope with this extremely high temperature, the thermal conductivity of the structural members of the gas turbine has been increased. Is covered with a heat shielding layer made of a relatively low ceramic material. For this reason, in the present embodiment, it is necessary to remove the heat shielding layer when predicting the fatigue life of the structural member.

As shown in FIG. 13, the oxide film thickness measuring means 21 grinds the oxide film 23 by grinder and buff polishing until the metal surface 24 completely appears and a metallic luster appears to remove the oxide film 23. The oxide film thickness δ is measured by using the oxide film size measuring instrument 25 on the step with respect to the portion not having the part.

The temperature estimating means 22 estimates the metal temperature of the structural member. That is, the oxide film 2
The temperature of 3 is a function of the operating time, as shown in FIG. For this reason, since both the operation time and the oxide film thickness are known, even if the structural member is covered with the oxide film 23, the metal temperature of the structural member can be easily estimated from FIG.

On the other hand, as shown in FIG. 15, the damage repetition number Nf is a function of the metal temperature T (° K), and is expressed by the following equation.

[0052]

(Equation 3) Here, No indicates the number of repetitions of the damage at the reference temperature T0, K indicates the Boltzmann constant, and Q indicates the constant.

The number Nf of failure cycles determined from the above equation (3)
Is replaced with Nf in the equation (1), it is possible to perform a more accurate pre-life evaluation.

The oxide film thickness measuring means 21 according to the present embodiment has been described to measure the thickness δ of the oxide film 23 by using the oxide film dimension measuring instrument 25. The eddy current method may be used instead.

As shown in FIG. 16, the surface ultrasonic measurement method uses an ultrasonic transmission probe 27 and an ultrasonic reception probe 28.
, The ultrasonic wave 29 may be transmitted to and received from the measured section 26. Further, in the surface ultrasonic measurement method, as shown in FIG. 17, there is no need to remove the oxide film 23, but the oxide film 2
There is a difference between the sound speeds of the ultrasonic waves propagating in each of the measuring part 3 and the measured part 26. Therefore, an accurate value of the thickness δ of the oxide film 23 can be obtained from the surface supersonic calibration curve 30 shown in FIG. Note that the surface supersonic speed 31 of the part to be measured is
Is measured in advance by a test piece.

On the other hand, in the eddy current method, as shown in FIG.
An eddy current probe 32 is attached to the measurement section 26 to measure the magnetic permeability. As shown in FIG. 20, the eddy current probe 32 is attached to the oxide film 23 of the measurement section 26 and the magnetic permeability of the oxide film 23 is measured. Was measured, and the difference between the two voltages was measured as shown in FIG.
Is calibrated by the calibration curve 34 of the output voltage indicated by
Is to measure the thickness δ.

As described above, the surface ultrasonic measurement method and the eddy current, both of which can automatically measure the thickness δ of the oxide film 23, can further improve the measurement accuracy.

In this embodiment, the oxide film thickness measuring means 21
The film thickness of the structural member is measured by the mechanical or electric / magnetic method described above, and the metal temperature of the structural member covered with the oxide film 23 is estimated by the temperature estimating means 22. Since the input is made to the reference value setting means 7, the metal temperature of the damaged structural member can be accurately evaluated, and the number Nf of failure cycles as the fatigue life reference value in consideration of the metal temperature effect can be selected. A fatigue life prediction method can be realized.

FIG. 22 is a block diagram conceptually showing a third embodiment of the method for predicting the fatigue life of a structural member according to the present invention. In the present embodiment, a replica sampling unit 35 and a replica image input unit 36 are provided instead of the stress analysis unit 1 and the stress distribution setting unit 2 in the configuration of the first embodiment. The same parts as those of the first embodiment are denoted by the same reference numerals.

The replica collecting means 35 according to the present embodiment
As shown in FIG. 22, for example, after removing an oxide film coated on the structural member 14 of the gas turbine stationary blade, the polishing is performed, and a replica 37 of the sectioned area 18 is collected. The image of the crack of the structural member 14 transferred by the replica 37 is input by a replica image input unit 36 through a low magnification microscope. When the cracks in the structural member 14 are generated by low cycle thermal fatigue, as shown in FIG. 24, a large number of cracks 38 are densely packed. From the crack image data formed in the dense state, the crack length measuring means divides the sum of the crack lengths by the observation visual field area to obtain the crack length density I. In this case, in the same part, in the same region where the crack length density I is the same, it is considered that the strain applied thereto is also equal. Therefore, by using the crack length density I as the evaluation target region setting criterion, the divided region 18 to be evaluated can be determined without performing the stress analysis.

As described above, the segmented area 18 to be evaluated
Is set, the prediction of fatigue life of the structural member 14 is performed in the same manner as in the first embodiment.

Further, the fatigue life reference value setting means 7 according to the present embodiment is for obtaining the failure repetition ratio N / Nf with respect to the maximum crack value amax, similarly to the first embodiment. There is a method using crack length density.

In the test piece, the relationship between the crack length density I and the failure repetition ratio N / Nf is an evaluation line indicated by a solid line in FIG. 25, and this evaluation line is obtained by the following equation.

[0064]

(Equation 4) Where Io is the saturation value (material constant) of crack length density, μL
And σL indicate constants, respectively.

In the above equation (4), by replacing I / Io with the cumulative probability, the damage repetition rate ratio Nf / N can be expressed as a lognormal distribution. That is, assuming that the crack initiation length is a constant value, the left side of the above equation (4) is the crack occurrence cumulative probability, the right side of the above equation (4) is the crack initiation life distribution equation,
Each can be considered.

The above equation (4) has a large change in the region of N / Nf <0.5 where cracks occur relatively frequently, and is suitable for grasping relatively early damage.

The replica sampling means 35 can only observe cracks in a relatively small area and cannot measure the maximum crack length in some cases. However, the crack length density can be uniformly distributed in the same strain range. , Stable evaluation can be performed.

After the failure repetition rate ratio Nf / N is determined and the fatigue life reference value is set, the same calculation processing as in the first embodiment is performed to predict the fatigue life of the structural member 14.

In this embodiment, the replica collection means 35,
Since the replica image input means 36 is used, a simple fatigue life prediction can be performed by omitting the stress analysis of the structural member,
Further, the fatigue life of the initial fatigue damage can be accurately predicted using the crack length density.

In estimating the fatigue life of a structural member, the structural member is exposed to a severe condition due to the heat load of the high-temperature combustion gas during operation, and has been used for a long period of time. For this reason, it is expected that the degree of crack growth occurring in the structural member will be measured differently for each periodic inspection.

Therefore, the fatigue life criterion setting means 7 in the first, second, or third embodiment needs to identify the number of repetitions of damage Nf represented by the equation (1) for each periodic inspection. is there. The identification of the damage repetition number Nf is applied to inspection data for each periodic inspection, and if each data is appended with a suffix 1 or 2, the damage repetition number Nf for each periodic inspection is calculated by the following equation.

[0072]

(Equation 5) Here, N1 and N2 are the number of crack occurrences for each periodic inspection, a1
, A2 indicate the crack length for each periodic inspection, and B indicates a constant.

As described above, in the present embodiment, the above equation (5)
Can be used to identify the number of repetitions of failure Nf of the structural member for each periodic inspection, so that the degree of crack propagation of the structural member for each periodic inspection can be easily recognized, and highly accurate fatigue life prediction can be realized. it can.

FIG. 26 is a block diagram conceptually showing a fifth embodiment of the method for predicting the fatigue life of a structural member according to the present invention. This embodiment is obtained by adding a stress / strain distribution correcting means 39 and a crack depth predicting means 40 to the first embodiment. The same components as those of the first embodiment are denoted by the same reference numerals, and only different components will be described.

The stress / strain distribution correcting means 39 according to this embodiment corrects the stress / strain based on the information of the fatigue life reference value setting means 7 and the information of the stress distribution setting means 2,
The data is more accurate.

Since the fatigue life reference value setting means 7 determines the number Nf of failure cycles as the fatigue life reference value from the test piece, the strain range Δεt of the structural member is determined as shown in FIG. Life evaluation line 41
Can be estimated from

The stress distribution setting means 2 determines the stress distribution in the thickness direction of the structural member. When determining the relationship between stress and strain, for example, a diagram shown in FIG. Can be obtained from Here, the Neubar's equation is as follows, assuming an elasto-plastic stress range Δσ, an elasto-plastic total strain range Δεt, a stress range Δσo by elastic analysis, and a total strain range Δεo by elastic analysis.

[0078]

(Equation 6) Here, Kt indicates an elastic stress concentration coefficient (constant).

On the other hand, the relationship between the repetitive stress and strain of the structural member is as follows:

(Equation 7) It has become.

Therefore, since the entire strain range εt is set by the fatigue life reference value setting means 7, the stress range Δσ
Can be obtained from the intersection shown in FIG.

The crack depth predicting means 40 sets the total strain range distribution in the thickness direction with the structural member set by the stress distribution setting means 2 on the surface of the structural member by the fatigue life reference value setting means 7. The correction is made to match the entire strain range .epsilon.t.

Number of failure cycles Nf as reference value for fatigue life
Is determined by the fatigue life reference value setting means 7, and the evaluation of the progress of the crack depth Cmax is performed using the following equation.

[0083]

(Equation 8) Here, A 'and B' are constants.

In evaluating the progress of the crack depth using the above equation (8), the crack depth of the structural member is calculated by using the future start / stop operation number Ne set by the operating condition setting means 8. As shown in FIG. 29, the total strain range distribution curve shown by the solid line can be corrected to the total strain range distribution curve shown by the broken line, and the development of the crack depth with high accuracy can be predicted.

On the other hand, the remaining life judging means 11 is based on the information on the evaluation of the crack depth of the structural member by the above-mentioned crack depth estimating means 40 and the information on the evaluation of the crack length of the structural member by the limit crack setting means 10. Since the allowable number of start / stop operations is determined, the evaluation of the determination can be further improved.

Therefore, according to the present embodiment, the future allowable number of starts / stops is predicted based on both the information on the evaluation of the crack length and the information on the evaluation of the crack depth. It can be carried out.

[0087]

As described above, according to the method for predicting the fatigue life of a structural member according to the present invention, the stress values of the structural members are obtained from the operation data, and the obtained stress values are arranged for each of the divided areas and image-processed. From the image processing, the crack length was converted to data to determine the maximum crack length, and based on the maximum crack length, the ratio of the number of repeated failures of the structural member was determined by the master curve based on the maximum crack length. Determine the failure repeat ratio,
The crack length of the future structural member is predicted again by the master curve, taking into account the number of start and stop of the future structural member in addition to the determined number of cycles of failure of the structural member. The number of failure cycles calculated from the master curve and the ratio of the number of failure cycles based on the crack length of the future structural member are compared with each other. Since it is determined whether the number of start / stop operations is possible, it is possible to easily and accurately predict the fatigue life of structural members, and to make accurate maintenance management decisions such as repair and replacement of structural members. Excellent effects, such as possible.

In the method for predicting the fatigue life of a structural member according to the present invention, the thickness of an oxide film coated on the structural member is measured simply and accurately using a mechanical or electromagnetic method. Can easily predict the fatigue life.

Further, the method for predicting the fatigue life of a structural member according to the present invention employs a replica sampling method and integrates the density of the transferred crack length into a single stable crack as a reference for setting the evaluation target region of the structural member. Since the length is measured, the fatigue life of the structural member can be predicted with high accuracy. In particular, it is possible to accurately predict the fatigue life of the structural member in the initial state.

Further, the method for predicting the fatigue life of a structural member according to the present invention uses a method capable of identifying the fatigue life of a structural member by the ratio of the number of failure cycles at each periodic inspection. It is possible to easily make management judgments such as replacement and repair while comparing with the damage repetition ratio.

Further, the method for predicting the fatigue life of a structural member according to the present invention uses a method of correcting stress / strain of a crack generated in a structural member and predicting a crack depth based on the correction of the stress / strain. Therefore, in predicting the fatigue life of a structural member, highly reliable and accurate information can be obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram conceptually showing a first embodiment of a fatigue life prediction method for a structural member according to the present invention.

FIG. 2 is a view for explaining collection of data as a basis for stress analysis of a structural member when predicting a fatigue life of the structural member according to the present invention.

FIG. 3 is a diagram showing an example in which the stress distribution analysis of the method for predicting the fatigue life of a structural member according to the present invention is applied to a gas turbine stationary blade.

FIG. 4 is a diagram showing an example in which a low-cycle thermal fatigue region is set from a stress distribution analysis result of the fatigue life prediction method for a structural member according to the present invention.

FIG. 5 is a diagram showing a setting example of a segmented area to be evaluated by the method for predicting fatigue life of a structural member according to the present invention.

FIG. 6 is a view for explaining image input of a structural member in the method for predicting fatigue life of a structural member according to the present invention.

FIG. 7 is a view showing a measurement example of a crack length in the fatigue life prediction method for a structural member according to the present invention.

FIG. 8 is a diagram summarizing the relationship between the maximum crack length of the structural member used for predicting the fatigue life of the structural member according to the present invention and the ratio of the number of failure cycles by a master curve.

FIG. 9 is a diagram showing the relationship between the cumulative crack coalescing rate and the ratio of the number of failure cycles, which is the basis of the master curve used in the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 10 is a diagram comparing the results of evaluation of the maximum crack length by the method for estimating the fatigue life of a structural member according to the present invention with the measurement results of actual structural members.

FIG. 11 is an explanatory diagram used to determine a limit crack length in the method for predicting fatigue life of a structural member according to the present invention.

FIG. 12 is a block diagram conceptually showing a second embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 13 is an explanatory view of mechanically measuring an oxide film thickness in a second embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 14 is a view showing a relationship among an oxide film thickness, temperature, and time in a second embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 15 is a diagram showing the relationship between the number of repeated failures of the test piece and the temperature in the second embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 16 is a schematic diagram showing that the thickness of a structural member is measured by an ultrasonic method in the second embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 17 is a schematic diagram showing that an oxide film thickness of a structural member is measured by an ultrasonic method in a second embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 18 is a diagram showing the relationship between ultrasonic sound speed and oxide film thickness in a second embodiment of the method for predicting fatigue life of a structural member according to the present invention.

FIG. 19 is a schematic diagram showing that the thickness of a structural member is measured by an eddy current method in a second embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 20 is a schematic view showing that an oxide film thickness of a structural member is measured by an eddy current method in a second embodiment of the method for predicting fatigue life of a structural member according to the present invention.

FIG. 21 is a view showing the relationship between the output voltage of eddy current and the thickness of an oxide film in the second embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 22 is a block diagram conceptually showing a third embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 23 is a view showing that a crack of a structural member is collected by a replica in the third embodiment of the method for predicting fatigue life of a structural member according to the present invention.

FIG. 24 is a view showing a crack form of a structural member obtained by a replica in the third embodiment of the method for predicting fatigue life of a structural member according to the present invention.

FIG. 25 is a view showing the relationship between the crack length density and the ratio of the number of failure cycles in the third embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 26 is a block diagram conceptually showing a fifth embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 27 is a view showing the relationship between the total strain range of the test piece and the ratio of the number of failure cycles in the fifth embodiment of the method for predicting the fatigue life of a structural member according to the present invention.

FIG. 28 is a diagram for determining a stress range from a total strain range using the Neubar method in the fifth embodiment of the fatigue life prediction method for a structural member according to the present invention.

FIG. 29 is a schematic diagram showing an example of correcting a strain distribution in a thickness direction of a structural member in a fifth embodiment of the method for predicting fatigue life of a structural member according to the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Stress analysis means 2 Stress distribution setting means 3 Sectional area setting means 4 Surface image input means 5 Image processing means 6 Crack length measuring means 7 Fatigue life reference value setting means 8 Operating condition setting means 9 Crack length prediction means 10 Limit Crack setting means 11 remaining life determining means 12 gas turbine 13 operation control panel 14 structural member 15 guide 16 mount 17 image input head 21 oxide film thickness measuring means 22 temperature estimating means 23 oxide film 25 oxide film size measuring instrument 35 replica sampling means 36 Replica image input means 37 Replica 39 Stress / strain distribution correction means 40 Crack depth prediction means

Continued on the front page (72) Inventor Hiroaki Yoshioka 2-4 Suehirocho, Tsurumi-ku, Yokohama, Kanagawa Prefecture Inside the Toshiba Keihin Office (72) Inventor Daizo Saito 2-4 Suehirocho, Tsurumi-ku, Yokohama-shi, Kanagawa Prefecture Toshiba Keihin Corporation In business office

Claims (11)

[Claims] 1. A stress distribution region of a structural member is divided by stress distribution analysis from data on the state of use to set a stress division region, and a crack length of the set division region is input as an image. Measure the length to determine the maximum crack length in the sectioned area, and the failure cycle ratio as the fatigue life standard value of the structural member obtained from the master curve created in advance based on the maximum crack length in the sectioned area Is determined, and the future crack length of the structure is predicted by considering the determined number of failure cycles and the number of future operation and shutdown operations, and the predicted crack length of the structure and the structure A method for predicting the fatigue life of a structural member, comprising: comparing a critical crack length that leads to breakage with a critical crack length; and determining from the difference the number of times the structure can be started and stopped in the future. 2. A structural member using data of an oxide film thickness, a temperature, and a time of a high-temperature oxidation test separately obtained from an oxide film thickness measured by removing an oxide film at a crack measurement site of a structural member and an operation time. 2. The fatigue life of a structural member according to claim 1, wherein a service life of the structural member is estimated by estimating a service temperature of the structural member, and correcting the ratio of the number of cycles of failure of the structural member as a fatigue life reference value from the estimated temperature. Forecasting method. 3. The method for predicting the fatigue life of a structural member according to claim 2, wherein the thickness of the oxide film on the structural member is measured with an oxide film size measuring instrument. 4. The method for predicting the fatigue life of a structural member according to claim 2, wherein the thickness of the oxide film on the structural member is determined non-destructively by using an ultrasonic method. 5. The method for predicting the fatigue life of a structural member according to claim 2, wherein the thickness of the oxide film on the structural member is determined non-destructively by using an electromagnetic method. 6. Image processing of a crack length in a stress division region of a structural member is performed by measuring a crack length density represented by a sum of crack lengths per unit area, and calculating the crack length density distribution from the crack length density distribution. The fatigue life prediction method for a structural member according to claim 1, wherein a stress division region is set. 7. The method for predicting the fatigue life of a structural member according to claim 6, wherein the measurement of the crack length density of the structural member is performed using a replica. 8. The structural member according to claim 6, wherein an evaluation formula for determining the ratio of the number of failure cycles as a fatigue life reference value of the structural member from the crack length density and the number of start / stop operations is used. Fatigue life prediction method. 9. In determining the number of failure cycles as a fatigue life reference value of a structural member, the number of crack occurrences determined from the number of crack occurrences and crack length in the previous inspection, the number of crack occurrences in the current evaluation, and 2. The fatigue life prediction method for a structural member according to claim 1, wherein an evaluation formula is used which can identify the number of breakage cycles determined from the crack length for each inspection. 10. A stress / strain acting on a stress division region is determined from a failure repetition ratio as a fatigue life reference value of a structural member, and a stress / strain distribution in a thickness direction of the structural member is determined from the obtained stress / strain value. And correcting the amount of crack propagation in the thickness direction of the structural member.
The method for predicting the fatigue life of a structural member as described in the above. 11. The fatigue life of a structural member according to claim 1, wherein the stress / strain value is obtained from an intersection of an equation of a repeated stress / strain relation of the structural member and an equation of a stress / strain relation of Neubar. Forecasting method.