JP6439498B2 - Fatigue life evaluation method, fatigue life evaluation system, and fatigue life evaluation program - Google Patents

Description

  The present invention relates to a fatigue life evaluation method, a fatigue life evaluation system, and a fatigue life evaluation system for evaluating the fatigue life of a measurement object when a predetermined load waveform is regularly and repeatedly applied to the measurement object, using a cumulative fatigue damage law, and The present invention relates to a fatigue life evaluation program.

  Generally, the cumulative fatigue damage law (corrected minor law) is used as a method for evaluating the fatigue life of a measurement object. In the cumulative fatigue damage law, a state in which loads such as stresses with various amplitudes are randomly generated is regarded as the sum of repeated loads with different amplitudes, and the fatigue life is estimated from this sum. Specifically, as a result of analyzing the load waveform, assuming that amplitudes of σ1, σ2,..., Σi are generated, the measurement target breaks when a load with these amplitudes is applied to the measurement target. The number of repetitions up to is read from the SN curve, and are defined as N1, N2,. When these amplitudes are repeated n1, n2,..., Ni times, respectively, the degree of damage is considered as n1 / N1, n2 / N2,. The sum of these individual damage levels is defined as the total damage level D. It is assumed that fatigue failure occurs when the overall damage degree D is greater than 1.

  In the cumulative fatigue damage law, when the damage degree D is obtained, it is necessary to calculate the number of repetitions of amplitude n1, n2,. For this reason, conventionally, a technique using a rainflow method is known as a method for calculating the frequency of amplitude (see, for example, Patent Document 1). In the conventional configuration, only the amplitude frequency is recorded using the rainflow method for the difference between the peak value and the valley value, and the amount of data to be recorded can be reduced.

JP-A-9-264706

  By the way, in a fatigue test of an object to be measured, a predetermined test load waveform is repeatedly given over a plurality of periods. The rainflow method has an advantage that the amplitude can be obtained in a random waveform. However, in a regular waveform in which the load waveform is periodically repeated, such as in a fatigue test, amplitude is generated at the boundary between the end point of the load waveform of one cycle and the start point of the load waveform of the next cycle. Since the loss may occur, there is a drawback that the regularity cannot be used. In particular, when the number of repetitions of the load waveform is large, there is a possibility that the influence of the boundary described above appears strongly. For this reason, in the structure which calculates the frequency of the amplitude for all periods by multiplying the frequency of the amplitude for one period of the load waveform by the number of repetitions of the load waveform, there is a problem that an actual value and an error occur.

  For this reason, in the case of the regular waveform described above, it is necessary to perform the calculation by the rainflow method on the data of the entire period as the test period in order to accurately calculate the frequency of the amplitude for the entire period. There is a problem that the calculation work becomes very large and complicated.

  The present invention has been made in view of the above, and a fatigue life evaluation method and a fatigue life evaluation system capable of easily and accurately evaluating the fatigue life of a measurement object when a load waveform is regularly and repeatedly applied. And to provide a fatigue life evaluation program.

  In order to solve the above-described problems and achieve the object, the present invention provides the fatigue life of a measurement object when a predetermined test load waveform is repeatedly applied to the measurement object over M cycles. A fatigue life evaluation method for evaluating using a cumulative fatigue damage law, which is based on a step of creating waveform data of a predetermined m period (3 ≦ m <M) from test load waveform data, and a rainflow method The step of extracting the amplitude of m period, the step of obtaining the number Na of amplitudes of m period, the remainder Na% m when the number Na of amplitudes is divided by the number m of periods, and the remainder Na% m Based on the step of obtaining a range of amplitude with periodicity in m periods, the frequency of amplitude of m periods and the frequency of amplitude of periodicity among m periods, and the amplitude of M periods from each frequency The step of determining the frequency of M and the amplitude of the M period Based on the frequency, characterized by comprising a step of evaluating the fatigue life from the cumulative fatigue damage rule, the.

  According to this configuration, it is possible to easily and accurately calculate the frequency of the amplitude of the M cycle serving as the test period from the data of the test load waveform for one cycle, and based on the frequency of the amplitude of the M cycle. Thus, the fatigue life of the measurement object can be easily and accurately evaluated.

  In this configuration, the test load waveform may be formed by combining a plurality of waveform patterns having different frequencies, amplitudes, and phases. According to this structure, even if it is a complicated waveform pattern, the fatigue life of a measuring object can be evaluated easily and correctly.

  In addition, the number m of m periods preferably satisfies the range of 3 ≦ m <M / 2. According to this configuration, it is possible to reduce the amount of data to be processed for calculation as compared with the case where calculation is performed by the rainflow method on all M cycles of data that are the test period. In particular, when m = 3, the amount of data can be minimized.

  Further, the present invention provides a fatigue for evaluating a fatigue life of a measurement object when a predetermined test load waveform is repeatedly applied to the measurement object over M cycles using a cumulative fatigue damage law. A life evaluation system that extracts waveform data from a test load waveform data and generates waveform data of a predetermined m cycle (3 ≦ m <M) and the amplitude of the m cycle based on the rainflow method. An amplitude extraction unit that performs the calculation, obtains the number Na of amplitudes in the m period, and the remainder Na% m obtained by dividing the number Na of the amplitudes by the number m of the periods. Amplitude range extraction unit for obtaining a characteristic amplitude range, an amplitude frequency for m periods and an amplitude frequency having a periodicity out of m periods, and an amplitude for obtaining an M period amplitude frequency from each frequency Accumulated based on frequency calculation unit and frequency of amplitude of M period Characterized by comprising the lifetime evaluation calculation unit for evaluating the fatigue life from the labor damage law, the.

  The present invention also allows a computer to evaluate the fatigue life of a measurement object when a predetermined test load waveform is repeatedly applied to the measurement object over M cycles using the cumulative fatigue damage law. A fatigue life evaluation program, comprising: generating waveform data of a predetermined m cycle (3 ≦ m <M) from test load waveform data; obtaining an m cycle amplitude based on a rainflow method; The step of obtaining the number Na of amplitudes in m periods and the residual Na% m when the number Na of amplitudes is divided by the number m of periods, and having periodicity in the m periods based on the residual Na% m A step of obtaining an amplitude range, a step of obtaining an amplitude frequency of m cycles and a frequency of an amplitude having periodicity among the m cycles, obtaining a frequency of amplitude of M cycles from each frequency, and an amplitude of M cycles Based on the frequency of Possible to execute a step of evaluating the fatigue life from fatigue damage law to the computer and it said.

  According to the present invention, it is possible to easily and accurately calculate the frequency of the amplitude of the M cycle serving as the test period from the data of the test load waveform for one cycle, and based on the frequency of the amplitude of the M cycle. Thus, the fatigue life of the measurement object can be easily and accurately evaluated.

FIG. 1 is a block diagram showing a functional configuration of the fatigue life evaluation system according to the present embodiment. FIG. 2 is a flowchart showing an operation procedure for evaluating the fatigue life. FIG. 3 is a diagram for explaining the rainflow method. FIG. 4 shows the relationship between the characteristics of the test load waveform and the remainder. FIG. 5 shows the relationship between the remainder and the periodic amplitude. FIG. 6A to FIG. 6F are diagrams for explaining a series of procedures for obtaining the frequency of the amplitude of the M period according to the present embodiment. FIG. 7A to FIG. 7D are diagrams for explaining a series of procedures for obtaining the frequency of the amplitude of the M period according to Reference Example 1. 8A to 8D are diagrams for explaining a series of procedures for obtaining the frequency of the amplitude of the M cycle according to Reference Example 2. FIG. FIG. 9 shows the relationship between the first number Nb and the last number Ne and the mathematical expression.

  EMBODIMENT OF THE INVENTION The form (embodiment) for implementing this invention is demonstrated in detail, referring drawings. The present invention is not limited by the contents described in the present embodiment. The constituent elements described below include those that can be easily assumed by those skilled in the art and those that are substantially the same. Furthermore, the constituent elements described below can be appropriately combined.

  FIG. 1 is a block diagram showing a functional configuration of the fatigue life evaluation system according to the present embodiment. The fatigue life evaluation system 10 virtually evaluates the fatigue life of a test target component (measurement object) on a computer. For example, when a predetermined test load waveform is repeatedly applied to a test target part over M cycles (for example, 200,000 cycles), it is evaluated whether the test target part is subject to fatigue failure, A predetermined test load waveform is repeatedly applied to the part, and the time (number of times) until the test target part undergoes fatigue failure is evaluated.

  The fatigue life evaluation system 10 includes an arithmetic unit 11 as shown in FIG. The arithmetic device 11 is composed of a CPU of a computer, and an input device 21 such as a keyboard and a mouse and an output device 22 such as a printer and a display are connected to the arithmetic device 11. Further, a storage device 31 composed of a hard disk, a semiconductor memory, or the like is connected to the arithmetic device 11. The storage device 31 stores a computer program (fatigue life evaluation program) 32 that causes the fatigue life evaluation system 10 to execute a procedure for evaluating the fatigue life. Although not shown in the drawing, the storage device 31 is necessary when conducting fatigue life evaluation tests such as SN curve data corresponding to various test target parts and various test load waveform data. Various data are stored. The test load waveform is preferably formed by combining a plurality of waveform patterns having different frequencies, amplitudes, and phases.

  The calculation device 11 includes a waveform data creation unit 12, an amplitude extraction unit 13, an amplitude range extraction unit 14, an amplitude frequency calculation unit 15, and a life evaluation calculation unit 16.

  The waveform data creation unit 12 creates waveform data that repeats the test load waveform for m cycles from the test load waveform data read from the storage device 31. This m cycle can be arbitrarily set within the range of 3 ≦ m <M, but is preferably a smaller range (3 ≦ m <M / 2) from the viewpoint of speeding up the arithmetic processing. Three cycles are used.

  The amplitude extraction unit 13 extracts the amplitude of m-cycle waveform data created by the waveform data creation unit 12 based on the rainflow method. The amplitude range extraction unit 14 divides the amplitude Na from the amplitude of the m-period waveform data extracted by the amplitude extraction unit 13 and the remainder Na when dividing the amplitude Na by the number m of periods. % M is obtained. Here, “%” is an operator for obtaining a remainder when Na is divided by m. Further, the amplitude range extraction unit 14 obtains a periodic amplitude range among the amplitudes of the m-cycle waveform data based on the remainder Na% m.

  The amplitude frequency calculation unit 15 obtains the amplitude frequency (number of repetitions) of the m-cycle waveform data and the frequency of the periodic amplitude (number of repetitions), and sets the test load waveform to M periods from each frequency. The frequency of the amplitude when it is repeated over is obtained. The life evaluation calculation unit 16 evaluates the fatigue life of the test target component using the cumulative fatigue damage law based on the frequency of the amplitude of the M cycle.

  Next, a procedure for fatigue life evaluation according to the present embodiment will be described. FIG. 2 is a flowchart showing an operation procedure for evaluating the fatigue life, and FIG. 6 is a diagram for explaining a series of procedures for obtaining the frequency of the amplitude of the M cycle according to the present embodiment. The operation for evaluating the fatigue life is executed by the arithmetic unit 11 of the fatigue life evaluation system 10 based on the computer program 32. In general, when the fatigue life evaluation test is performed, the cycle (number of times) M for repeatedly applying the test load waveform is large (for example, 200,000 cycles). However, in this embodiment, M is set to 5 for convenience of explanation. A case where the period is set will be described.

  First, the waveform data creation unit 12 creates waveform data that repeats the test load waveform for 3 (m = 3) cycles from the test load waveform data read from the storage device 31 (step S1). Specifically, as shown in FIG. 6 (A), waveform data for testing is created by simply multiplying the test load waveform for one cycle in which the stress (load) fluctuates with the passage of time and repeating the cycle three times. In the present embodiment, for the sake of convenience of explanation, the test load waveform is the one that shows the temporal change of the stress at 5 points. However, as long as the period is determined, any number of the stress points that change with time may be used.

  Next, the amplitude extraction unit 13 extracts the left and right flow amplitude data (also simply referred to as amplitude) of the three cycles of waveform data created by the waveform data creation unit 12 based on the rainflow method (step S2). . Here, the rainflow method will be briefly described. FIG. 3 is a diagram for explaining the rainflow method. FIG. 3 shows the relationship between time and stress. The rainflow method is a method of finding the largest possible amplitude from a random waveform and counting the frequency of the amplitude.

  As shown in FIG. 3, in the rainflow method, time is plotted on the vertical axis and stress is plotted on the horizontal axis, and the temporal change of the stress is represented as a waveform graph. The corrugation is likened to the roof, and water drops flow from the base of the roof toward the eaves. The water droplet stops when the following stop conditions are satisfied, and the abscissa (distance from the root to the eaves) that flows until the water droplet stops is the amplitude. In the rain flow method, the number of times (frequency) of one amplitude is set to 0.5, and it is measured as one round trip.

  Stopping conditions are as follows: (1) When water droplets flow from the base of the roof toward the eaves, water droplets located above are already flowing, (2) When water droplets fall from the last roof, (3) Water droplets When a root that is the same as or at the beginning of the flow of the water appears deeper than the start of the flow, the water droplet stops when any of the conditions is satisfied. When the left and right flows are obtained from FIG. 3 in accordance with the above stop conditions, the water droplet L1 flowing from time 1 (start point ST1 of the first cycle) falls at time 2 and stops at time 4 for the left flow. To do. In this case, since the stress is changed from −1 to −2, the amplitude is −1. Here, the sign of the amplitude is originally positive (plus), but a negative (minus) sign is used for the amplitude of the left flow as a symbol for distinguishing the flow direction (left-right direction in FIG. 3). ing. When the amplitude frequency is obtained, the absolute value of the amplitude is used. Further, the water droplet L2 flowing from time 4 falls at time 5 (end point EN1 of the first cycle) and stops at time 9. In this case, since the stress is changed from 0 to -3, the amplitude is -3. As for the flow on the right side, the water droplet R1 that has flowed from time 3 falls at time 4 and stops at time 5. In this case, since the stress is changed from −2 to 0, the amplitude is 2. Similarly, as shown in FIGS. 3 and 6B, the positions at which the water droplets stop are obtained for the left and right flows of the three-cycle waveform data, and the amplitudes of the left and right flows are extracted. FIG. 6C summarizes the amplitudes corresponding to the water droplets divided into left and right flows.

  Returning again to the flowchart of FIG. Once the left and right flow amplitudes are extracted, data processing is performed for each left and right flow (step S3). This step S3 has a subroutine of steps Sa11 to Sa14, and executes steps Sa11 to Sa14 for the left and right flows, respectively.

  The amplitude range extraction unit 14 obtains the number of amplitudes Na from the amplitude of the three periods of waveform data extracted by the amplitude extraction unit 13 (step Sa11), and also divides the number of amplitudes Na by the number of cycles. The remainder Na% 3 is obtained (step Sa12). As shown in FIG. 3, in the regular waveform in which the test load waveform is periodically repeated as in the present embodiment, the end point EN1 of the load waveform of one cycle (first cycle) and the next cycle ( In some cases, an amplitude is generated at the boundary 40 between the load waveform start point ST2 of the second period). The applicant considered the influence of the amplitude at the boundary 40 in the rainflow method, and found the relationship between the characteristics of the waveform data due to the amplitude generated / disappeared at the boundary 40 and the number of amplitudes Na and the remainder Na% m. FIG. 4 shows the relationship between the characteristics of the test load waveform and the remainder, and FIG. 5 shows the relationship between the remainder and the periodic amplitude.

  In the present embodiment, as shown in FIG. 3, an amplitude that does not exist in the first period from the start point ST1 to the end point EN1 occurs at the boundary 40. Although not occurring in the present embodiment, the amplitude may disappear at the boundary 40. In some cases, the amplitude does not occur or disappear at the boundary 40. As shown in FIG. 4, when the characteristics of the amplitude at the boundary 40 when the amplitude is generated, the following characteristics are obtained.

  (1) The direction of the flow in contact with the boundary 40 is, for example, leftward and coincides with each other as indicated by the water drops L2 and L3. (2) The end point and start point (for example, end point EN1 and start point ST2) of each cycle are inconsistent because the stress is -3 and -1. (3) The direction of the flow generated at the boundary 40 is, for example, rightward as indicated by the water droplet R2, and is inconsistent with the direction of flow (leftward) in contact with the boundary 40 shown in (1). In examining the occurrence / disappearance of the amplitude, examining the characteristics of such a waveform is a complicated operation. Therefore, the generation / disappearance of the amplitude and the waveform data were examined, and it was found that there was a relationship between the generation / disappearance of the amplitude and the remainder Na% 3 when the number Na of the amplitude was divided by the number of periods (for example, 3). . That is, as shown in FIG. 4, the generation / disappearance of the amplitude can be obtained for any waveform having any characteristic only by obtaining the remainder Na% 3. This means that a range of amplitude having periodicity can be obtained out of the three periods based on the remainder Na% 3. The remainder value is three values of 0, 1, m−1 (2 in the present embodiment) shown in FIG. Therefore, by referring to FIG. 4, the generation / disappearance of the amplitude can be understood from the remainder value.

  In this embodiment, the amplitude number Na in the left flow is 6, as shown in FIG. 6C, and the remainder Na% 3 is 6% 3 = 0. Similarly, the number Na of amplitudes in the flow on the right side is 5 as shown in FIG. 6C, and the remainder Na% 3 is 5% 3 = 2.

  Returning again to the flowchart of FIG. Next, the amplitude range extraction unit 14 obtains a periodic amplitude range among the amplitudes of the waveform data of three periods based on the remainder Na% 3. Specifically, referring to FIG. 5, based on the value of the remainder Na% 3, the first number Nb of the periodic amplitude is obtained (Step Sa13), and the value of the remainder Na% 3 is used. Then, the last number Ne of the amplitude having periodicity is obtained (step Sa14). A period between the first number Nb and the last number Ne is a periodic amplitude range.

In the present embodiment, since the remainder is 0 in the left flow, as shown in FIG.
Nb = Na / m + 1 (a)
This is expressed by the following equation (a), and 6/3 + 1 = 3. The last number Ne is
Ne = Na / m × 2 (b)
(6) × 6 = 2 = 4. Thus, in the flow on the left side, the range of the amplitude with periodicity is between L3 and L4 in FIG. 6C.

Similarly, the remainder is 2 in the flow on the right. Here, the handling of data is divided into a case where 1 is reduced at the beginning and a case where 1 is reduced at the end. This is an agreement on whether data (for example, water droplets L2 and L4) straddling each cycle is handled as data of the previous cycle or as subsequent data. As long as you follow this convention, you can do either. In the present embodiment, data that straddles a cycle is handled as data of the previous cycle, and therefore corresponds to a case where the last is reduced by one. For this reason, the first number Nb is
Nb = Int (Na / m) +2 (c)
It is represented by the formula (c). Here, Int is an operator that truncates the decimal part. From this equation (c), Int (5/3) + 2 = 3. The last number Ne is
Ne = Int (Na / m) × 2 + 2 (d)
Int (5/3) × 2 + 2 = 4. Thus, in the flow on the right side, the range of the amplitude with periodicity is between R3 and R4 in FIG. 6C.

  Next, the amplitude frequency calculation unit 15 obtains the frequency Fmi of the i-th amplitude in three cycles (step S4). Specifically, as shown in FIG. 6D, the amplitude magnitude (absolute value) frequencies are obtained on the left and right sides, and the left and right frequencies are added. When the amplitude is 1, the frequency on the left side is 3 and the frequency on the right side is 2, so the total is 5. Similarly, when the amplitude is 2, the frequency on the left side is 0 and the frequency on the right side is 1. Therefore, the total is 1, and when the amplitude is 3, the frequency on the left side is 3. Since the frequency on the right side is 2, the total is 5.

  Next, the amplitude frequency calculation unit 15 obtains the frequency Fci of the i-th amplitude in the periodical range among the three periods (step S5). Specifically, as shown in FIG. 6E, the frequency of the amplitude (absolute value) in the periodic range is obtained on the left and right sides, and the left and right frequencies are added. When the amplitude is 1, the frequency on the left side is 1 and the frequency on the right side is 1, so the total is 2. Similarly, when the amplitude is 2, the frequency on the left is 0 and the frequency on the right is 0, so the total is 0. When the amplitude is 3, the frequency on the left is 1, Since the frequency on the right side is 1, the total is 2.

Subsequently, the amplitude frequency calculation unit 15 obtains the frequency FMi of the i-th amplitude in the M period from the frequency Fmi of the i-th amplitude and the frequency Fci of the i-th amplitude in the periodic range (step S6). ). Specifically, the frequency FMi is
FMi = Fmi + Fci × (M−m) (e)
And is calculated for each amplitude (absolute value). When the amplitude is 1, 5 + 2 × (5-3) = 9. Similarly, when the amplitude is 2, 1 + 0 × (5-3) = 1, and when the amplitude is 3, 5 + 2 × (5-3) = 9. Thus, the frequency FMi of the i-th amplitude in the M period can be easily obtained. Here, as described above, in the rainflow method, since the number of times (frequency) of one amplitude is generally 0.5 and is measured as one round trip, the calculated frequency FMi is finally 1 as well. / Then, it is doubled to move to the next step.

  Subsequently, the life evaluation calculation unit 16 evaluates the fatigue life from the cumulative fatigue damage law (for example, the modified minor law) based on the obtained frequency FMi of the i-th amplitude of the M period (step S7). In the cumulative fatigue damage rule, the frequency FMi of the i-th amplitude of the obtained M period is put into n1, n2,..., Ni of the formula 1, and it is determined whether or not the overall damage degree D is greater than 1. To do.

  Next, as a reference example 1, a method for calculating the frequency FMi of the i-th amplitude in the M period by the rainflow method for the data in the entire period will be described. FIG. 7A to FIG. 7D are diagrams for explaining a series of procedures for obtaining the frequency of the i-th amplitude of the M period according to Reference Example 1.

  In this reference example 1, M is also 5. As shown in FIG. 7A, waveform data that repeats the test load waveform for five cycles is created, and the left and right flow amplitudes of the waveform data are extracted based on the rainflow method. In Reference Example 1, as shown in FIG. 7B, water drops L1 to L10 flow on the left side, and water drops R1 to R9 flow on the right side.

  Next, as shown in FIG. 7C, each amplitude corresponding to the water droplet is divided into left and right flows, and the frequency of the amplitude is obtained. Specifically, as shown in FIG. 7D, the amplitude magnitude (absolute value) frequencies are obtained on the left and right sides, and the left and right frequencies are added. When the amplitude is 1, the frequency on the left side is 5 and the frequency on the right side is 4, so the total is 9. Similarly, when the amplitude is 2, the frequency on the left is 0 and the frequency on the right is 1, so that the total is 1. When the amplitude is 3, the frequency on the left is 5, Since the frequency on the right side is 4, the total is 9. Even in this case, the calculated frequency FMi is finally multiplied by 1/2.

  In the method of the present embodiment, the same calculation result as that obtained when the frequency FMi of the i-th amplitude in the M period is calculated by the rainflow method on the data in all periods can be obtained. Furthermore, although the case where M = 5 has been described in the present embodiment, since this M is normally set to a large number (for example, 200,000), the calculation in the present embodiment becomes larger as M becomes larger. Compared with the method according to the reference example 1, the amount of data processed for obtaining the amplitude is reduced to m / M times, and the amount of data processed for obtaining the frequency is approximately ( It has a very large effect of being reduced to m + 1) / M times.

  Next, as Reference Example 2, a method for calculating the frequency FMi of the i-th amplitude in the M period by multiplying the waveform data in one period by M will be described. 8A to 8D are diagrams for explaining a series of procedures for obtaining the frequency of the amplitude of the M cycle according to Reference Example 2. FIG.

  In this reference example 2, M is also 5. As shown in FIG. 8A, waveform data having a test load waveform as one cycle is created, and the left and right flow amplitudes of the waveform data are extracted based on the rainflow method. In Reference Example 2, as shown in FIG. 8B, water droplets L1 and L2 flow on the left side, and water droplet R1 flows on the right side.

  Next, as shown in FIG. 8C, each amplitude corresponding to the water droplet is divided into left and right flows, and the frequency of the amplitude is obtained. Specifically, as shown in FIG. 8D, the frequency of the amplitude (absolute value) in one cycle is obtained on the left side and the right side, respectively, and these are multiplied by 5 to obtain the i-th amplitude in the M cycle. The frequency FMi is obtained. When the amplitude is 1, the frequency on the left side is 1 and the frequency on the right side is 0, so (1 + 0) × 5 = 5. Similarly, when the amplitude is 2, the frequency on the left is 1 and the frequency on the right is 0, so (1 + 0) × 5 = 5, and when the amplitude is 3, Is 1 and the frequency on the right is 0, so (1 + 0) × 5 = 5. Even in this case, the calculated frequency FMi is finally multiplied by 1/2. In this way, the method of calculating the frequency FMi of the i-th amplitude of the M period by multiplying the waveform data of one period by M does not consider the influence of the boundary that may occur during each period. An erroneous calculation result different from the actual frequency FMi is obtained. For this reason, the frequency FMi of the i-th amplitude of the M cycle cannot be calculated simply by multiplying the waveform data of one cycle by M.

  As described above, according to the present embodiment, waveform data for m cycles is created from the test load waveform data (S1), and the amplitude of m cycles is extracted based on the rainflow method (S2). The number Na of amplitudes and the remainder Na% m obtained by dividing the number Na of amplitudes by the number m of periods are obtained (Sa11, Sa12). Based on the remainder Na% m, there is periodicity among the m periods. The amplitude range is determined (Sa13, Sa14), the frequency Fmi of the i-th amplitude in m periods and the frequency Fci of the i-th amplitude having periodicity in m periods (S4, S5), Since the frequency FMi of the i-th amplitude of the M cycle is obtained from the frequency (S6), the frequency FMi of the i-th amplitude of the M cycle serving as the test period is simply obtained from the data of the test load waveform for one cycle, and Can be calculated accurately, i-th of this M period Based on the width of the frequency FMi, the fatigue life of the test target part easily and can be evaluated accurately.

  Further, according to the present embodiment, the test load waveform is formed by combining a plurality of waveform patterns having different frequencies, amplitudes, and phases. Lifetime can be evaluated easily and accurately.

  Further, according to the present embodiment, the number m of m periods preferably satisfies the range of 3 ≦ m <M / 2. According to this configuration, it is possible to reduce the amount of data to be processed for calculation as compared with the case where calculation is performed by the rainflow method on all M cycles of data that are the test period. In particular, when m = 3, the amount of data can be minimized.

  As mentioned above, although embodiment of this invention was described, this invention is not limited to the said embodiment. For example, in the above embodiment, it is determined whether the value of the remainder Na% m (m = 3) is 0, 1, or m−1, and the periodic amplitude is based on the remainder Na% m. The first number Nb and the last number Ne are obtained, but the first number Nb and the last number Ne can be obtained by applying the remainder of Na% m to a predetermined mathematical expression. In this case, the process for determining the value of the remaining Na% m can be omitted, so that the processing speed can be increased.

  Next, a method for calculating the first number Nb and the last number Ne by applying the value of the remainder Na% m to the formula will be described. In the above embodiment, as shown in FIG. 5, the first number Nb and the last number Ne are handled as data across cycles (whether handled as data of the previous cycle or data of the later cycle), and Depending on the value of the remainder Na% m, the definition formula is different. As shown in FIG. 9, the inventor can obtain the first number Nb and the last number Ne without determining whether the value of the remainder Na% m is 0, 1, or m-1. The formula was derived.

Here, a process of deriving equations for obtaining the first number Nb and the last number Ne will be described in the case where data that crosses the period is handled as the previous period (incorporated into the previous period). First, as shown in FIG. 5, the definition formula of the first number Nb differs depending on the value of the remainder Na% m, and when data that crosses the cycle is treated as the previous cycle (incorporated into the previous cycle), Expressions (A) to (C) are established.
When the remainder Na% m = 0, Nb = Na / m + 1 (A)
When the remainder Na% m = 1 Nb = Int (Na / m) +1 (B)
In case of excess Na% m = m−1 Nb = Int (Na / m) +2 (C)
Here, as described above, Int is an operator that rounds off the decimal part.

  In the formula (A), the fact that the remainder Na% m is 0 is divisible when the number Na is divided by the number m of periods. That is, Na / m is represented only by an integer, and there is no number after the decimal point. Further, since the remainder Na% m is 0, adding the remainder Na% m does not change the result.

For this reason, when formula (A) is modified, formula (D) is obtained.
Nb = (Na / m) +1 (A)
= {Int (Na / m)} + 1
= {Int (Na / m) + Na% m} +1
= {Int (Na / m) + Sgn (Na% m- (m-1)) + 1} +1 (D)
Here, Sgn is a sign function (operator) that obtains −1 in the case of a negative value, 0 in the case of 0, and +1 in the case of a positive value. In this case, the remainder Na% m is 0, and the number m of cycles is 3 ≦ m <M (m = 3), so the value of Sgn (Na% m− (m−1)) + 1 Becomes 0. Therefore, it turns out that Formula (D) is equivalent to Formula (A).

Next, consider equation (B). In the formula (B), the remainder Na% m is 1, and the number of periods m is 3 ≦ m <M (m = 3). For this reason, the value of Sgn (Na% m- (m-1)) + 1 is 0, and even if this value is added, the result does not change. Therefore, equation (B) can be transformed into equation (D) as follows.
Nb = Int (Na / m) +1 (B)
= {Int (Na / m) + Sgn (Na% m- (m-1)) + 1} +1 (D)

In the case of Expression (C), 2 is first divided into 1 + 1. Further, since the remainder Na% m is m-1, the value of Sgn (Na% m- (m-1)) is 0. For this reason, adding this value does not change the result. Therefore, equation (C) can be transformed into equation (D) as follows.
Nb = Int (Na / m) +2 (C)
= {Int (Na / m) +1} +1
= {Int (Na / m) + Sgn (Na% m- (m-1)) + 1} +1 (D)

  As described above, any of the above formulas (A) to (C) can be transformed into the same formula (D). Therefore, regardless of the value of the remainder Na% m, the first number Nb can be obtained from one equation (D).

On the other hand, the definition formula of the last number Ne also differs depending on the value of the remainder Na% m, and when data that crosses the cycle is handled as the previous cycle (incorporated into the previous cycle), the following formulas (E) to (E G).
When the remainder Na% m = 0, Ne = Na / m × 2 (E)
When the remainder Na% m = 1 Ne = Int (Na / m) × 2 + 0 (F)
In case of excess Na% m = m−1 Ne = Int (Na / m) × 2 + 2 (G)

Regarding the formula (E), since the remainder Na% m is 0 and the number m of cycles is 3 ≦ m <M (m = 3), Sgn (Na% m− (m−1)) The value of +1 is 0. Therefore, equation (E) can be transformed into equation (H) as follows.
Ne = (Na / m) × 2 (E)
= {Int (Na / m)} × 2
= {Int (Na / m) + Na% m} × 2
= {Int (Na / m) + Sgn (Na% m)} × 2
= {Int (Na / m) + Sgn (Na% m- (m-1)) + 1} * 2 (H)

Similarly, for the formula (F), since the remainder Na% m is 1, and the number m of cycles is 3 ≦ m <M (m = 3), Sgn (Na% m− (m− 1) The value of +1 is 0. Therefore, Formula (F) can be transformed into Formula (H) as follows.
Ne = Int (Na / m) × 2 + 0 (F)
= {Int (Na / m) + Sgn (Na% m- (m-1)) + 1} * 2 (H)

Similarly, in the formula (G), since the remainder Na% m is m−1, the value of Sgn (Na% m− (m−1)) is 0. Therefore, equation (G) can be transformed into equation (H) as follows.
Ne = Int (Na / m) × 2 + 2 (G)
= {Int (Na / m) +1} × 2
= {Int (Na / m) + Sgn (Na% m- (m-1)) + 1} * 2 (H)

  Thus, any of the above-described formulas (E) to (G) can be transformed into the same formula (H). Therefore, whatever the value of the remainder Na% m, the last number Ne can be obtained from one equation (H). Therefore, by substituting (applying) the values of the number of amplitudes Na, the number of periods m, and the remainder Na% m into these formulas (D) and (H), respectively, the first number Nb and the last number Ne. Can be obtained. Thereby, the process which determines the result of remainder Na% m can be abbreviate | omitted, and the speed of a calculation process can be improved by that much.

In the above description, the case where data crossing a cycle is handled as the previous cycle (incorporated in the previous cycle) has been described, but data crossing the cycle can be handled as the subsequent cycle (incorporated in the subsequent cycle). . In this case, although details are omitted, the value of the first number Nb can be obtained from the equation (I), and the value of the last number Ne can be obtained from the equation (J). Also in this case, the process of determining the result of the remaining Na% m can be omitted, and the calculation processing speed can be improved accordingly.
Nb = {Int (Na / m) + Sgn (Na% m- (m-1)) + 1} + ((Na + 1)% m-1) +1 (I)
Ne = {Int (Na / m) + Sgn (Na% m− (m−1)) + 1} × 2 + ((Na + 1)% m−1) (J)

  The above formulas (D), (H) and formulas (I), (J) are examples of mathematical formulas for obtaining the values of the first number Nb and the last number Ne, and for the input (remainder Na% m). Other equations may be used as long as the output (Nb, Ne) matches that shown in FIG.

DESCRIPTION OF SYMBOLS 10 Fatigue life evaluation system 11 Computation device 12 Waveform data creation part 13 Amplitude extraction part 14 Amplitude range extraction part 15 Amplitude frequency computation part 16 Life evaluation computation part 21 Input device 22 Output device 31 Storage device 32 Computer program 40 Boundary D Damage degree FMi I-th amplitude of M period Fci frequency of i-th amplitude in a periodic range Fmi frequency of i-th amplitude ST1 to ST3 start point EN1 to EN3 end point L1 to L10 water droplet R1 to R9 water droplet

Claims (5)

A fatigue life evaluation method for evaluating a fatigue life of a measurement object when a predetermined test load waveform is repeatedly applied to the measurement object over M cycles using a cumulative fatigue damage law. ,
Creating waveform data of a predetermined m period (3 ≦ m <M) from the test load waveform data;
Extracting the m-cycle amplitude based on a rainflow method;
Obtaining the number Na of amplitudes in the m period and the residual Na% m when the number Na of amplitudes is divided by the number m of periods;
Obtaining a periodic amplitude range of the m periods based on the remainder Na% m;
Obtaining a frequency of the amplitude of the m period and a frequency of the periodic amplitude of the m periods, and obtaining the frequency of the amplitude of the M period from each of the frequencies;
And evaluating the fatigue life from the cumulative fatigue damage law based on the frequency of the amplitude of the M period.   The fatigue life evaluation method according to claim 1, wherein the test load waveform is formed by combining a plurality of waveform patterns having different frequencies, amplitudes, and phases.   The fatigue life evaluation method according to claim 1, wherein the number m of the m periods satisfies a range of 3 ≦ m <M / 2. A fatigue life evaluation system for evaluating a fatigue life of a measurement object when a predetermined test load waveform is repeatedly applied to the measurement object over M cycles using a cumulative fatigue damage law. ,
A waveform data creation unit for creating waveform data of a predetermined m period (3 ≦ m <M) from the test load waveform data;
An amplitude extraction unit that extracts the amplitude of the m period based on a rainflow method;
The number Na of amplitudes in the m period and the residual Na% m obtained by dividing the number Na of amplitudes by the number m of periods are obtained, and the periodicity of the m periods is determined based on the residual Na% m. An amplitude range extraction unit for obtaining a certain amplitude range;
An amplitude frequency calculation unit for determining the frequency of the amplitude of the m period and the frequency of the periodic amplitude of the m periods, and determining the frequency of the amplitude of the M period from each frequency;
A fatigue life evaluation system comprising: a life evaluation calculation unit that evaluates the fatigue life from the cumulative fatigue damage law based on the frequency of the amplitude of the M cycle. A fatigue life evaluation program that causes a computer to evaluate the fatigue life of the measurement object when a predetermined test load waveform is repeatedly applied to the measurement object over M cycles using a cumulative fatigue damage law. There,
Creating waveform data of a predetermined m period (3 ≦ m <M) from the test load waveform data;
Obtaining the m-cycle amplitude based on a rainflow method;
Obtaining the number Na of amplitudes in the m period and the residual Na% m when the number Na of amplitudes is divided by the number m of periods;
Obtaining a periodic amplitude range of the m periods based on the remainder Na% m;
Obtaining a frequency of the amplitude of the m period and a frequency of the periodic amplitude of the m periods, and obtaining the frequency of the amplitude of the M period from each of the frequencies;
A fatigue life evaluation program that causes the computer to execute the step of evaluating the fatigue life from the cumulative fatigue damage law based on the frequency of the amplitude of the M period.

Images