JP2005037370A - Method and program for reliability assessment - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a reliability assessment method and a reliability assessment program for quantitatively assessing the reliability of provided structural members. <P>SOLUTION: The reliability assessment method is for assessing the reliability of provided structural members. It comprises a process of measuring crack depth of a certain number of structural members among a plurality of structural members to be assessed, a process of calculating a plurality of stress ranges working on the structural members from the crack depth by using the crack progression rule, a process of statistically processing a plurality of stress ranges and calculating statistical distribution of the stress ranges, a process of calculating crack depth distribution at an arbitrary time point by using Monte-Carlo method from the statistical distribution, a process of calculating failure probability based on the fraction of crack depth exceeding a limit value from the crack depth distribution, and a process of calculating the assessment value of reliable section from the failure probability. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a reliability evaluation method and a reliability evaluation program for quantitatively evaluating the reliability of a plurality of structural members.

  Many large-scale facilities are used in large-scale facilities such as plant equipment, and the maintenance of these structural members is less affected by the failure of each structural member. Maintenance management by group management was frequently used due to reasons such as showing characteristics and large numbers that cannot be managed individually. Maintenance management by this group management grasps the overall deterioration tendency through sample surveys, etc., and when the overall deterioration degree is small, repairs are individually made as necessary, and the overall deterioration tendency is more than a certain degree. Then, the entire equipment was repaired or replaced.

  However, maintenance management by group management has the following problems. It was difficult to quantitatively evaluate the degree of deterioration of the entire facility because the overall deterioration tendency was grasped by individual sample surveys and the like. Therefore, it was not clear how many structural members should be sampled. In addition, there was no reliability prediction method for continuous operation of equipment, and it was not possible to clarify the criteria for repair and equipment renewal.

  Moreover, since the use method and use situation of the structural member are different, various factors of deterioration are assumed. For example, deterioration due to cracks in structural members in plant equipment is mainly used, and there are various generation factors such as fatigue, thermal fatigue, and corrosion fatigue. Therefore, it is necessary to carry out deterioration status and maintenance management according to the usage status of each structural member, and it is difficult to quantitatively evaluate the reliability of a plurality of structural members.

  The present invention is made in view of such a problem, and provides a reliability evaluation method and a reliability evaluation program capable of quantitatively evaluating the reliability of a plurality of structural members provided. Is a technical issue.

  That is, the present invention is a reliability evaluation method for quantitatively evaluating the reliability of a plurality of structural members provided, and measures the crack depth of a certain amount of structural members among the plurality of structural members to be evaluated. A step of calculating a plurality of stress ranges acting on the structural member using a crack propagation law from the crack depth, a step of statistically processing the plurality of stress ranges, and calculating a statistical distribution of the stress range, From the statistical distribution, the step of calculating the crack depth distribution at an arbitrary time by the Monte Carlo method, the step of calculating the failure probability based on the ratio of the crack depth exceeding the limit value from the crack depth distribution, and the failure probability And a step of calculating the confidence interval evaluation value.

Since the present invention quantitatively evaluates the reliability of a plurality of members provided by the above steps, can accurately grasp the degree of deterioration of structural members, and can predict the degree of future deterioration quantitatively, it can be rationally maintained. Management can be performed. By accurately grasping the deterioration state of each structural member, it is possible to perform only necessary maintenance on the necessary structural member, so it is possible to reduce unnecessary maintenance and contribute to reduction in maintenance cost. Here, examples of the statistical distribution include a normal distribution, a log normal distribution, and a Weibull distribution. Various statistical distributions can be used, and it is desirable to select and use them appropriately.

Further, in the step of statistically processing the plurality of stress ranges and calculating the statistical distribution of the stress ranges, inspection data without cracks can be evaluated from the measured crack depth by the following method. As a first evaluation method, the inspection data without cracks is handled assuming that the generated stress is equal to the threshold value Δσ _th (evaluation method A). In this evaluation method A, the stress range is set higher than the average value of the true stress range. As a second evaluation method, inspection data without a crack is treated as censored data, and the crack occurrence probability is determined to be equivalent to the measurement result (evaluation method B). This evaluation method B can evaluate the stress range average lower and the stress range dispersion larger than the evaluation method A described above. As a third evaluation method, inspection data without a crack is treated as a crack that does not occur when the generated stress range is equal to or less than a threshold value (evaluation method C). This evaluation method C can be suitably used when the ratio of inspection data without cracks is high in all inspection data. It should be noted that the present invention may be any evaluation method, and it is desirable to select and use various methods according to the characteristics of the inspection target data.

  Furthermore, in the step of calculating a plurality of stress ranges acting on the structural member from the crack depth using the crack propagation law, the present invention calculates the crack propagation law from the number of operations of the equipment including the structural member and the crack depth. It is desirable to use it to calculate the stress range.

  The crack propagation law is a law in which the progress of the crack depth of the structural member follows the Paris law. When calculating the stress range using this crack propagation law, the stress range is calculated in consideration of the measured crack depth and the time series leading to the crack depth. What is necessary is just to set this time series suitably according to the use of a structural member, a function, etc. For example, it can be set as the frequency | count of operation of the installation in which the structural member is used. It can be suitably used when the structural member deteriorates depending on the number of operations, and progresses.

  Furthermore, in the reliability evaluation method according to the present invention, it is desirable to set the number of structural members for measuring the crack depth according to the confidence interval evaluation value of the failure probability. The present invention measures the crack depth of a certain number of structural members among a plurality of structural members, and quantitatively evaluates the reliability of the structural members based on the measurement results. The number of structural members that perform this measurement is not particularly limited, and can be set to a certain ratio based on experience, for example. However, there are a wide variety of structural members that can be evaluated for reliability, and there are structural members that have very different deterioration conditions even if they are the same type of structural member. There is also. Therefore, the number of measurement members can be set rationally by setting the number of structural members to be measured based on the confidence interval evaluation value of the failure probability obtained in the present invention. Therefore, it is possible to reduce the number of structural members to be measured as compared with the case where a certain number of structural members are set, which can contribute to the reduction of maintenance costs, and the probability of damage and the reliability obtained. The accuracy of the interval evaluation value can be improved.

  Further, the present invention is a method for evaluating the reliability of a facility comprising a plurality of types of structural members, wherein the reliability interval evaluation value of the failure probability calculated for each structural member in the step is determined by the number of structural members provided in the facility. Weighting is performed, and the failure probability of the entire facility and its confidence interval evaluation value are calculated.

  ADVANTAGE OF THE INVENTION According to this invention, it becomes possible to perform the reliability evaluation of the whole installation of the installation provided with multiple types of structural member based on the reliability evaluation of each structural member. Therefore, it is possible to rationally perform maintenance management of the entire facility by grasping the deterioration state of each structural member. Therefore, it is possible to perform necessary maintenance only for necessary structural members based on the number of times of operation of the facility, and rational maintenance management can be performed.

  Further, the present invention may be a program for causing a computer to execute the above method in order to solve the problem.

  As described above, according to the reliability evaluation method and the reliability evaluation program according to the present invention, it is possible to quantitatively evaluate the reliability of a plurality of structural members provided in time series. In addition, by evaluating the reliability using the number of operations of the equipment in which the structural member is used, the reliability suitable for each structural member and the equipment that uses the structural member is considered in consideration of the use and usage status of each structural member. It becomes possible to perform sex evaluation. Furthermore, in facilities consisting of multiple types of structural members, it is possible to carry out necessary maintenance and plan only for the necessary structural members by quantitatively predicting the deterioration status of each structural member and its deterioration prediction. Become. Therefore, it is possible to provide an inspection method in which the maximum reliability can be obtained in accordance with the member configuration of equipment within a given maintenance cost. Moreover, since the number of structural members to be measured can be set based on the reliability evaluation value of each member, the number of necessary inspection members can be reduced and the inspection cost can be reduced.

  Hereinafter, the configuration of the present invention will be described in detail based on an example of an embodiment shown in the drawings. FIG. 13 is a schematic configuration diagram of a computer that executes the reliability evaluation method of the present invention. The computer 1 includes an arithmetic processing unit 12 including a CPU and a main memory in a main body 11, a storage device (hard disk) 13 for storing software (reliability evaluation program etc.) for arithmetic processing, and an input port for these data Alternatively, an input / output unit (I / O) 14 that is an output port is provided.

In the present embodiment, the reliability evaluation method of the present invention is executed by installing a reliability evaluation program in a general-purpose computer and causing the arithmetic processing unit 12 or the like to perform processing according to the program.
FIG. 14 is a flowchart of a reliability evaluation method executed by the computer. In the processing described below, what is described without particularly defining the subject is processing performed by the arithmetic processing unit of the computer.

  In the present embodiment, as an example of the evaluation method, an example is shown in which the reliability of the metal wall welded portion of the furnace wall and the thermal power generation facility in the thermal power generation facility is evaluated. In this embodiment, the reliability was evaluated based on the crack depth of the inner surface of the welded furnace wall tube of the skin bar, tie rod, and tension plate among the hardware welded to the furnace wall. FIG. 1 shows material data and measurement conditions of the measurement target member.

  First, the crack depth of each member is measured. In the present embodiment, the crack size is measured by an ultrasonic flaw detection test and input to the computer 1 from the input / output unit 14. The computer 1 receiving the input temporarily stores the measurement result in the memory. If there are a plurality of dimensions and materials in the same member, the crack depth in each dimension and material is measured. The crack depth distribution summarizing the measurement results is shown in FIG. 2 (step 1).

Ｃ，ｍ 亀裂進展を規定する定数
ΔＫ_th 亀裂発生を規定する限界応力拡大係数
いずれも材料特性であり、各部材の材料から導き出せる値である。
Δａ 亀裂深さ
ΔＫ 見掛けの応力拡大係数
ここで、見掛けの応力拡大係数ΔＫを以下のように表す。

Δσ 見掛けの応力範囲
亀裂のない部材の応力範囲は、以下のように表せる。

Next, based on the crack depth measurement data obtained in Step 1, the stress range acting on the inspection site is calculated using the crack propagation law (Step 2). In the present embodiment, the causes of cracks are fatigue, thermal fatigue, and corrosion fatigue. The following crack growth was used to represent the crack growth due to these degradations.

C, m Constants defining crack growth ΔK _th Critical stress intensity factor defining crack initiation Each is a material property and can be derived from the material of each member.
Δa Crack depth ΔK Apparent stress intensity factor Here, the apparent stress intensity factor ΔK is expressed as follows.

Δσ Apparent stress range The stress range of a crack-free member can be expressed as follows.

Ｃ，ｍ 亀裂進展を規定する定数
ΔＫ_th 亀裂進展しきい値
ａ₀ 初期亀裂寸法
ａ_j 計測亀裂寸法
ｔ₀ 初期亀裂に対応する起動停止繰り返し数（通常は０である）
ｔ 計測までの起動停止繰り返し数 Next, assuming that the cracks of each member follow the crack growth characteristics, the stress range Δσ is calculated backward from the crack depth measured based on Equation 4.

C, m Constants defining crack growth ΔK _th Crack growth threshold a ₀ Initial crack size a _j Measured crack size t ₀ Start / stop repetition number corresponding to initial crack (usually 0)
t Number of start / stop repetitions until measurement

The stress range obtained from the above is statistically processed to obtain a statistical distribution of the stress range of each member (step 3). First, as a result of measuring the crack depth, inspection data without a crack is evaluated by the following two methods. As a first evaluation method, from Equation 5, the inspection data without a crack is treated as the generated stress being equal to the threshold value Δσ _th (Evaluation Method A). In this evaluation method A, the stress range is set higher than the average value of the true distribution.

  As a second evaluation method, inspection data without a crack is treated as censored data, and the crack occurrence probability is determined to be equivalent to the measurement result (evaluation method B). This evaluation method B is the statistically best method, and can be evaluated with a lower stress range average and a larger stress range dispersion than the evaluation method A described above. Hereinafter, the evaluation method of this evaluation method B will be described in detail.

Δσの分布特性が正確に求めることが出来れば、上記亀裂進展しない確率は、図３に示す斜線部分の面積と等しい。すなわち、求める統計分布は、斜線の面積Ｓ_thが、Ｐ_ni＝Ｓ_thとなるように定めれば良い。以下、この算出方法の例として、２分法を用いた算出方法を説明する。 FIG. 3 shows the probability that the crack does not propagate in the stress range distribution function calculated from Equation 5. Regarding crack actual measurement data, out of N pieces of crack measurement data, when n pieces of cracks progress and (N−n) pieces of cracks do not progress, the probability that the crack does not progress is calculated as follows.

If the distribution characteristic of Δσ can be accurately obtained, the probability that the crack does not propagate is equal to the area of the hatched portion shown in FIG. In other words, the statistical distribution to be obtained may be determined so that the hatched area S _th is P _ni = S _th . Hereinafter, as an example of this calculation method, a calculation method using a bisection method will be described.

Δσ₀ Δσの最小下限値（入力値）
ｎ ：ステップ数、最初は１であり（Ｐ_ni−Ｓ_th）の符号が変わる度に１ずつ増やす。
そして、新しいΔσ_min ^*を用いて、再度Δσの分布特性を求める。平均値をμΔσ⁽¹⁾、標準偏差をσΔσ⁽¹⁾とする。また、この分布について、Δσ_thまでの面積Ｓ_th ⁽¹⁾を求める。 In order to determine the distribution characteristic of Δσ of the stress range by the bisection method, first, the initial value of the stress range Δσ _{min of the} member having no crack is set to Δσ _min = Δσ _th . Then, changing little by little the stress range Δσ _min from the following equation.

Δσ ₀ Minimum lower limit of Δσ (input value)
n: Number of steps, initially 1 and incremented by 1 each time the sign of (P _ni -S _th ) changes.
Then, the distribution characteristic of Δσ is obtained again using the new Δσ _min ^* . The average value is μΔσ ⁽¹⁾ , and the standard deviation is σΔσ ⁽¹⁾ . For this distribution, an area S _th ⁽¹⁾ up to Δσ _th is obtained.

Next, an improved value of Δσ _min and a distribution characteristic of Δσ are obtained again. The area up to Δσ _th is obtained again using the obtained new distribution characteristic. This calculation is repeated until the difference between S _th and P _ni falls within an allowable error. The obtained distribution characteristic is regarded as a true characteristic. FIG. 4 is a distribution schematic diagram of the nominal stress range in the case where censored data exists, and FIG. 5 is a distribution schematic diagram of the stress range in consideration of censored data using the distribution characteristics obtained by the calculation.

  The stress range distribution of each member obtained from step 3 is shown in FIGS. FIG. 9 shows the probability density function of each stress range distribution obtained by using this evaluation method B, and FIG. 10 shows the probability density function of the stress range obtained by using the evaluation method A. Comparing the probability density distribution of the stress range of each member, it can be seen that the probability density distribution of the tie rod is different from the probability density distribution of the skin bar and the tension plate, and the base is wide. This corresponds to the fact that many data with large crack depths were obtained from the measurement results of the tie rods.

ｔ_j 指定時刻あるいは指定運転回数までの応力サイクル数
ａ_j 指定時刻あるいは指定運転回数までの亀裂深さ計算値
尚、Δσが確率変数であるので、ａ_jも確率変数となる。 Next, from the stress range distribution obtained in step 3, the crack depth at an arbitrary time is calculated by using the Monte Carlo method using Equation 8, and the crack depth amount at an arbitrary future time for each member is calculated. (Step 4).

t _j Number of stress cycles up to specified time or specified number of operations _aj Calculated crack depth up to specified time or specified number of operations Note that since Δσ is a random variable, a _j is also a random variable.

ｐ_f 破損確率
ｐ_a 亀裂寸法の確率分布関数
ａ_j 式８から得られる亀裂寸法 Next, based on the crack depth distribution at an arbitrary time obtained in step 4, the failure probability and the reliability interval rating at each evaluation time are calculated for each member (step 5). The term “breakage” means that the crack depth has reached (or exceeded) the limit value. In this embodiment, the crack depth corresponding to 30% of the typical pipe thickness is set as the limit value. First, the damage probability is calculated. For example, the failure probability when the crack size at a certain point in time reaches (or exceeds) the critical crack size a _cr can be calculated by numerical integration of the following equation.

p _f damage probability p _a crack size which is obtained from the probability distribution function a _j-type 8 of crack size

A conceptual diagram of this mathematical formula is shown in FIG. On the other hand, as can be seen from equation 9, the evaluation formula of crack growth amount has become a complex shape, it is possible to find the probability distribution function p _a theoretical evaluation is considered to be difficult. Therefore, the failure probability is obtained using the Monte Carlo method. That is, a random value of the stress range according to the statistical distribution of the stress range is generated. Further, in the calculation of the breakage probability, the total number of crack actual measurement data for each member group is used as the scoring number. If the number of cracks that have reached (or exceeded) the critical crack a _cr is counted among the n initial cracks that have been scored, and the number of cracks is n _f , the failure probability p _f can be calculated by Equation 10. .

ｐ_f 破損確率の予測値
ｓ 標準偏差
ξ 信頼度に対応した正規分布の幅
９０％信頼区間を評価する場合は、ξ＝１．６４
９５％信頼区間を評価する場合は、ξ＝１．９６
９９％信頼区間を評価する場合は、ξ＝２．５８
また、式１１の信頼区間の評価には、破損確率の標準偏差ｓを用いる。ｎ個の亀裂が採点される場合の標準偏差ｓは以下の式１２から求める。

Next, a confidence interval evaluation value is obtained. The confidence interval when n cracks are scored is calculated as follows.

Predicted value of _pf failure probability s Standard deviation ξ Normal distribution width corresponding to reliability ξ = 1.64 when evaluating 90% confidence interval
When evaluating 95% confidence intervals, ξ = 1.96
When evaluating a 99% confidence interval, ξ = 2.58
Further, the standard deviation s of the failure probability is used for evaluation of the confidence interval of Equation 11. The standard deviation s when n cracks are scored is obtained from the following equation (12).

  Since the Monte Carlo method is a trial calculation, the failure probability calculation with the number of cracks actually measured as the score is repeated a sufficient number of times, and the average is used as the confidence interval evaluation value. The simulation is repeated until the convergence condition (tolerance ε%) designated for each member group is satisfied.

ｐｆ_k：各部材群の信頼区間評価値（ｐｆ₁，ｐｆ₂・・・，ｐｆ_k）
ｎ_k：各部材の亀裂実測データ数
Ｎ_k：各部材群の部材数 Next, the confidence interval evaluation value of each member obtained in step 5 is used in Expression 13, weighted by the number of structural members used in the entire facility, and the failure probability and reliability of the entire facility are evaluated (step 6). ).

pf _k : confidence interval evaluation value of each member group (pf ₁ , pf ₂ ... pf _k )
n _k : Number of cracks actually measured for each member N _k : Number of members in each member group

  FIG. 12 shows the relationship between the failure probability and the number of repetitions obtained in this way. Note that the overall failure probability is evaluated by weighting the failure probability of the three members by the number of members. As can be clearly seen from FIG. 12, the damage probability of the skin bar and the tension plate is almost zero. On the other hand, the tie rod is 0.01 when using the evaluation method A and 0.02 when using the evaluation method B in 2072 cycles, which is the number of crack depth measurements.

  Next, the failure probabilities of the structural members thus obtained are compared, and the priority order of the structural members to be maintained is determined. Further, when performing maintenance management of a plurality of facilities, the priority of facility maintenance is determined by comparing the probability of damage of each facility. And this evaluation result is output. For example, display on a display, output from a printer, transmission to another computer, etc. are performed (step 7). As described above, since the maintenance priority is determined based on the failure probability obtained in the above step, it is possible to perform rational maintenance management.

  In the above embodiment, when obtaining the statistical distribution of the stress range of each member in step 3, the crack-free inspection data is evaluated so that the crack generation probability is equivalent to the measurement result (evaluation). Method B). However, when the ratio of inspection data without cracks in the entire inspection data is high, an error may occur in the failure probability calculated based on this. As described above, when the ratio of the inspection data without cracks in the entire inspection data is high, the data without cracks is evaluated as a method in which the generated stress range is equal to or lower than the threshold value and cracks do not occur (evaluation method C). ) Can be suitably used. Example 1 is a reliability evaluation method using the evaluation method C. FIG. 15 is a flowchart of the reliability evaluation method according to the first embodiment. In the first embodiment, steps 1, 2, 4, 5, and 6 are the same as those in the above-described embodiment, and therefore the description of each step is omitted using the same formula. Step 3A in Example 1 corresponds to 3 in the above embodiment.

Example 1 is the reliability evaluation of the welded part of the furnace wall in the thermal power generation facility and the thermal power generation facility as in the above embodiment. In step 3A, the stress range obtained in step 2 is statistically processed to obtain a statistical distribution of the stress range of each member. If the number of data with cracks in the total number of inspection data N is n, the number of data without cracks is N−n. Data without cracks are treated as those in which the generated stress range is not more than a threshold value Δσ _th and cracks do not propagate (Evaluation Method C).

前記計測データ全体を母集団とした場合の確率密度ｐを式１５によって求める。

ｐ₀とは、亀裂のあるデータだけを用いて計算した応力範囲Δσの確率密度である。 The probability data are applied to the cracked data. As described above, in the first embodiment, since data without a crack is treated as not cracking below a threshold value Δσ _th , the data having a crack in the stress range obtained in step 2 is expressed by the following equation (14). To obtain a probability distribution for a new random variable Δσ ′. This distribution is used in the damage probability calculation by the Monte Carlo method.

The probability density p in the case where the entire measurement data is a population is obtained by Equation 15.

p ₀ is the probability density of the stress range Δσ calculated using only data with cracks.

  FIG. 16 is a diagram showing a Weibull distribution schematic diagram of the stress range Δσ ′ excluding crack-free data thus obtained and a normal distribution schematic diagram of the stress range Δσ including data without a crack. The base Weibull distribution schematic diagram is a diagram in which the probability distribution obtained by the method described in the paragraph 0038 is shifted to the original stress range Δσ and the probability density is converted into the entire measurement data by Expression 15 and displayed. The dotted line shown in the figure is the frequency distribution of data in the measurement, and it can be seen from this frequency distribution that the ratio of the number of data without cracks in the entire inspection data is high. Comparing and examining the normal distribution schematic diagram and the frequency distribution, it can be seen that the normal distribution schematic diagram shows almost no frequency of data with cracks. On the other hand, the tail Weibull distribution corresponds to the frequency of data with cracks, and it can be seen that it is preferable to treat the data without cracks as non-cracking and show it as the tail Weibull distribution. FIG. 17 shows an enlarged view of a part of the tail Weibull distribution.

  Next, from the stress range distribution obtained in Step 3A, the crack depth at an arbitrary time is calculated by using the Monte Carlo method using Equation 8, and the crack depth amount at an arbitrary future time for each member is calculated. (Step 4). In step 4, the amount of crack propagation up to the specified number of repetitions is calculated from the equation 8 using the probability density of the stress range calculated using only the data with cracks.

Similar to the above embodiment, the breakage probability is obtained using the Monte Carlo method. That is, a random value of the stress range according to the statistical distribution of the stress range is generated. The number of scoring here is not the total total number of crack actual measurement data for each member, but the number of data M for each member having a crack, in order to determine the reliability interval, unlike the above embodiment.
If the number of cracks that have reached (or exceeded) the limit crack a _cr is counted out of M scoring points in the Monte Carlo method, and the number is assumed to be m _f , the failure probability p _f0 can be calculated by Equation 16.

ｐ_f0：亀裂のある部材データから計算した破損確率
このようにして、求めた破損確率ｐ_fから、前記実施形態と同様にして信頼評価をすることができる。この結果得られた破損確率と繰り返し数との関係を図１８に示す。 Next, the failure probability p _f0 obtained in step 5 is corrected in accordance with the ratio between the number of cracked data and the total number of inspection data (step 5A). The failure probability p _f0 obtained in step 5 is a value calculated only from data with cracks, and the failure probability p _f for all inspection data is obtained from Equation 17.

p _f0 : Failure probability calculated from cracked member data In this manner, the reliability evaluation can be performed in the same manner as in the above-described embodiment from the determined failure probability p _f . FIG. 18 shows the relationship between the damage probability and the number of repetitions obtained as a result.

It is the material data and measurement conditions of the structural member to be measured in the reliability evaluation according to the present embodiment. It is the figure which showed the measurement result of the crack depth of each structural member. It is the figure which showed the probability that a crack does not progress in a stress range distribution function. It is a distribution schematic diagram of the nominal stress range when censored data exists. It is a distribution schematic diagram of the stress range in consideration of censored data. It is the figure which showed the stress range distribution of the skin bar. It is the figure which showed the stress range distribution of the tie rod. It is the figure which showed the stress range distribution of the tension plate. It is the figure which showed the probability density function of the stress range obtained using the evaluation method B for every member. It is the figure which showed the probability density function of the stress range obtained using the evaluation method A for every member. It is a conceptual diagram which shows a failure probability. It is the figure which showed the relationship between the failure probability of each member, and the number of repetitions. It is a schematic block diagram of the computer which performs the reliability evaluation method of this invention. It is a flowchart of the reliability evaluation method which concerns on embodiment. 3 is a flowchart of a reliability evaluation method according to the first embodiment. It is the figure which showed the distribution schematic diagram of stress range (DELTA) (sigma) which remove | excluded the data without a crack, and the various distribution schematic diagrams of stress range (DELTA) (sigma) including the data without a crack. FIG. 17 is an enlarged view of the bottom weibull distribution schematic diagram of FIG. 16. It is the figure which showed the relationship between the failure probability calculated | required in Example 1, and the number of repetitions.

Explanation of symbols

1 Computer 11 Computer Main Body 12 Arithmetic Processing Unit 13 Storage Device 14 Input / Output Unit

Claims (5)

A reliability evaluation method for quantitatively evaluating the reliability of a plurality of structural members provided,
Measuring a crack depth of a certain number of structural members among a plurality of structural members to be evaluated;
Calculating a plurality of stress ranges acting on the structural member using a crack propagation law from the crack depth;
Statistically processing the plurality of stress ranges and calculating a statistical distribution of the stress ranges;
Calculating a crack depth distribution at an arbitrary time point using the Monte Carlo method from the statistical distribution;
Calculating a failure probability based on a ratio of crack depth exceeding a limit value from the crack depth distribution;
And a step of calculating a confidence interval evaluation value from the failure probability. In the step of calculating a plurality of stress ranges acting on the structural member using the crack propagation law from the crack depth,
The reliability evaluation method according to claim 1, wherein the stress range is calculated from the number of operations of the equipment including the structural member and the crack depth using the crack propagation law. The reliability evaluation method according to claim 1, wherein the number of structural members for measuring the crack depth is set according to the reliability interval evaluation value. A method for evaluating the reliability of equipment comprising a plurality of types of structural members,
2. The reliability interval evaluation value calculated for each structural member in the step is weighted by the number of structural members provided in the facility, and the failure probability of the entire facility and its reliability interval evaluation value are calculated. The reliability evaluation method according to claim 3. A reliability evaluation program for quantitatively evaluating the reliability of a plurality of structural members,
Receiving a crack depth input of a certain number of structural members among a plurality of structural members to be evaluated;
Calculating a plurality of stress ranges acting on the structural member using the crack propagation law from the crack depth;
Statistically processing the plurality of stress ranges and calculating a statistical distribution of the stress ranges;
Calculating a crack depth distribution at an arbitrary time point using the Monte Carlo method from the statistical distribution;
Calculating a failure probability based on a ratio of crack depth exceeding a limit value from the crack depth distribution;
And a step of calculating a confidence interval evaluation value from the failure probability.

Images