JP2008128698A - Method and apparatus for estimating necessary life difference of significant difference determination in accelerated test - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for determining the presence of significant difference from the life between two lots obtained in an accelerated test and calculating life difference due to magnification. <P>SOLUTION: The random numbers corresponding to the test number of a lot of a level 1 are produced from a certain Weibull distribution, the random number corresponding to the test number of a lot of a level 2 are produced from the same Weibull distribution and the life calculated from the produced random numbers is operated to calculate one set of a life ratio. This processing is repeated the set number of times to calculate the life ration of the number of sets of the set number of times to form the probability distribution and cumulated probability distribution of this life ratio. The maximum and minimum life ratios in set reliable width are read to determine the presence of significant difference. When the presence of the significant difference is determined, the Weibull distribution is changed to a different Weibull distribution to repeat the analyzing step of the cumulated probability distribution. A graph showing the relationship between the set magnification obtained by the repetition of the analyzing step and the maximum life in the reliable width is created and the value of the set magnification corresponding to the life ratio of the lives of the levels 1 and 2 in input data is read to be set as the life difference capable of determining at least the read value. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  In the accelerated test for calculating the life from the time when the failure occurs, the test object consisting of a mechanical part such as a bearing or a test object is placed in a predetermined environmental condition that is severer than the use environment. The present invention relates to a method, apparatus, and program for determining whether there is a significant difference between lots, and for determining whether there is a significant difference in some cases, and for determining whether there is a significant difference or estimating a significant life difference.

  The life test is one of the tests indispensable for evaluating the performance of mechanical parts such as bearings. Life tests can be broadly divided into (1) actual machine tests that test under conditions close to the actual operating environment and (2) accelerated tests that perform life tests under relatively severe conditions. The former is a test that determines that there is no problem in the service life if the test continues without damage until a certain target time because the product is rarely damaged within a finite time (hereinafter referred to as " Called “censored test”). On the other hand, since the latter occurs in a relatively short time, the life can be calculated by Weibull plot (for example, Non-Patent Document 1), and is a test for determining the superiority or inferiority of the performance from the calculated life (hereinafter, such as this The test is called “accelerated test”).

Conventionally, the experienced life test has been conducted by experienced experts, and it is considered that he was able to make empirical judgments on the design of the life test and the interpretation of the life test results to determine the test conditions and the number of tests.
FIG. 14 shows a procedure for interpreting a life test design and a life test result, which has been conventionally performed, for each of an abort test and an acceleration test.
The details of what is currently determined empirically in the life test are shown in Table 1.

In addition, what uses Weibull distribution for the lifetime judgment of a machine component is proposed by various patent documents and nonpatent literature.
JP 2006-040203 A JP 2002-277382 A JP 2005-226829 A Author Makabe Satoshi, Introduction to Reliability Engineering 79, published in 1991

In the accelerated test, the life is calculated from the failure time as described above, and the superiority or inferiority of the performance is determined from the calculated life. However, there is a significant difference between the calculated life obtained from the life test results of 2 lots. There are many situations where you want to see if there is any.
Conventionally, the significant difference determination of the calculated life has been performed using the concept of confidence width. However, this determination method has several problems as follows.

First, the reliability range has a problem that a significant difference in lifetime between two levels cannot be determined. The reason is that it is impossible to quantitatively determine how much the reliability range overlaps and there is a difference in life. In order to determine a significant difference in life between two levels using the confidence width, the following procedure is required. The procedure will be described below with reference to FIG.
1) Obtain the median rank confidence level distribution (distribution A in the figure) in the L10 life. Here, this distribution is obtained using an F-number table and an approximate expression of Fisher.
2) The distribution of the distribution frequency of L10 (distribution B in the figure) is obtained from the distribution of the confidence level of the median rank in the L10 life and the Weibull slope.
3) Using this variation distribution f (x), the integral of the equation in the figure is taken, and the frequency at which the life multiplication factor n occurs from the same life distribution is examined. This is a variation distribution of the life ratio.
4) The cumulative probability distribution of the distribution of 3) is created, and the life multiplication factor that can be determined as having a significant difference with a probability of 90% or more is calculated from the graph.

  The above procedure makes it possible to test and design a significant difference in lifetime using the conventional confidence range. However, such a procedure is complicated and difficult to put into practical use. For this reason, it can be said that the test of significant difference in life using the confidence range has not been performed so far.

  LEONARD G. JOHNSON proposes a method to determine the significant difference between the average life and L10 life between two levels for each Weibull slope and number of specimens). However, the method is not actually used at present. This is due to the fact that there is no tool to perform the dominant difference test in an easy way.

  Conventionally, there has been no appropriate method for quantitatively indicating at least how much life difference can be said even if it is determined that there is a significant difference.

  The purpose of this invention is to determine the existence of a significant difference between two lots quantitatively and appropriately, and when it is determined that there is a significant difference, at least how much life difference can be said to be Can be easily and quickly calculated, can be quantitatively obtained and reliable, and can be used to appropriately calculate the required life difference even if not skilled. To provide a computer program.

In the accelerated test of the present invention, the presence / absence of significant difference / estimated life difference estimation method is determined by placing a test object consisting of a mechanical part such as a bearing or a test piece under predetermined environmental conditions that are severer than the operating environment, resulting in damage. In the accelerated test for calculating the life from time, a method for calculating the difference in life by magnification, which can be determined and at least determined from the life of two lots of the test object,
A process (M1) of inputting, as input information, the value of the Weibull slope of the Weibull distribution of the test object, the number of each test of the two lots of Level 1 and Level 2 to be compared, and the life of the test result as input information;
And a computer calculation process (M2) for causing the computer to determine whether or not there is a significant difference and to calculate at least a life difference that can be determined and to display a calculation result on a screen of a display device.
The lifetime is a lifetime with a predetermined reliability such as an L10 lifetime (90% reliability lifetime) or an L50 lifetime (50% reliability lifetime).

As the computer calculation process (M2),
A first random number generation procedure (N21) for generating a random number according to the Weibull distribution determined from the value of the Weibull slope of the input information of the test target product and the temporarily set life for the number of test pieces of the level 1 lot;
A second random number generation procedure (N22) for generating random numbers according to the same Weibull distribution as the first random number generation procedure (N21) for the test number of lots of level 2.
From the set number of random numbers generated in the first random number generation procedure (N21) by the predetermined lifetime calculation method and the set number of random numbers generated in the second random number generation procedure (N22), the predetermined lifetime calculation method. A life ratio calculation procedure (N23) for calculating a life ratio of a set of level 1 and level 2 lifespans from the obtained lifespan,
The first random number generation procedure (N21), the second random number generation procedure (N23), and the life ratio calculation procedure (N24) are repeated a set number of times to obtain the life ratio of the set number of sets, and the probability distribution of the life ratio A cumulative probability distribution analysis procedure (N24) for creating a cumulative probability distribution and reading the maximum life ratio within the set confidence range from the cumulative probability distribution;
It is judged that there is a significant difference when the life ratio of the life of Level 1 and Level 2 in the above input information exceeds the maximum life ratio obtained by the cumulative probability distribution analysis procedure (N24), and is valid when it is within the maximum life ratio. And a significant difference presence / absence determination procedure (N25) for determining that there is no difference.

In addition, as the computer calculation process (N2),
The cumulative probability distribution analysis is executed when it is determined that there is a significant difference in the significant difference presence / absence determination procedure (N25), and the Weibull distribution is sequentially changed to a Weibull distribution having a setting magnification with a different provisional set life for each repetition. A magnification change repetition procedure (N26) for repeating the procedure until the set magnification reaches the set maximum magnification from the set minimum magnification,
A life magnification relationship graph creation procedure (N27) for creating a life magnification relationship graph, which is a graph showing the relationship between the set magnification repeatedly changed and the maximum life ratio within the reliability range, obtained in this magnification change repetition procedure (N26). )When,
Reading the value of the set magnification corresponding to the life ratio of the life of Level 1 and Level 2 in the input information from the life magnification relationship graph and reading the significant life difference as a life difference that can be determined at least. (N28),
And a result output procedure (N29) for displaying on the display device at least a determinable life difference read in the significant difference presence / absence determination procedure (N25) and the significant life difference reading procedure (N28). .

In the first and second random number generation procedures (N21, N22), as the life for specifying the Weibull distribution used for random number generation, a life or actual value estimated as appropriate may be set.
In addition, in the life ratio calculation procedure (N23), the predetermined life calculation method for obtaining the life from the random number of the test pieces generated in the random number generation procedure is an appropriate life calculation method conventionally used in the accelerated test. Use it. In the accelerated test, the life is obtained by a Weibull plot from the time when the breakage occurs, and a calculation method for obtaining the life from the breakage time is used as the life calculation method in the life ratio calculation procedure.
The Weibull distribution is given by

Where m: Weibull slope, α: scale factor, γ: minimum life,
Specified by.

  The life of mechanical parts such as bearings is said to follow the Weibull distribution. The Weibull distribution has three parameters, a Weibull slope m, a scale factor α, and a minimum life γ, and is known as a universal distribution that can express an exponential distribution, a lognormal distribution, and a normal distribution by the Weibull slope m. In mass-produced bearings and the like, the actual value of the Weibull slope is often known, and in the method of the present invention, it is preferable to use the actual value of the product to be tested as the Weibull slope. If there is no actual value, a Weibull slope estimated by an appropriate method may be used. The calculation method of the minimum life γ is determined by various standards such as ISO, and it is preferable to use any one of the calculation methods determined as such. The scale factor α has an arithmetic expression that is uniquely determined from the value of the Weibull slope, the reliability of the required life, the value of the required life, and the minimum life γ, and may be specified using the arithmetic expression.

In this method, in the cumulative distribution creation procedure (N24), two lifespans are obtained as a result of performing a life test after extracting two sets of test target products from a lot having the same life distribution. It corresponds to checking the number of times set. Even in a test in which a test object is extracted from a lot having the same life distribution, the life varies. If the variation in the life ratio is defined as an appropriate confidence interval of the cumulative probability distribution, for example, a 5% and 95% interval (90% confidence interval), the life variation can be calculated from the cumulative probability distribution. This result shows that even if two sets of test target products are extracted from lots with the same life distribution, the life ratio is 90% within the confidence interval shown in the cumulative probability distribution. It shows that it will vary. In other words, if there is a maximum life difference in the confidence interval, there is a high possibility that the life distribution between the lots is different.
Therefore, as the significant difference presence / absence determination procedure (N25), when the life ratio of the level 1 and level 2 lives in the input information exceeds the maximum life ratio obtained by the cumulative probability distribution analysis procedure, there is a significant difference. By determining and determining that there is no significant difference within the maximum life ratio, the presence or absence of the significant difference can be appropriately determined quantitatively.

Also, the cumulative probability distribution similar to the above is calculated in combination with the Weibull distribution with different lifetimes. For example, when the calculation is performed with the combination of the Weibull distribution and the Weibull distribution with different lifetimes of 1.1, 1.2, ... Every time it changes 1.1, 1.2 ... 50 times, the life difference should vary around 1.1, 1.2 ... 50 times. The magnification change repetition procedure (N26) performs the calculation of the cumulative probability distribution and the probability distribution analysis procedure (N24) for the probability distribution.
A life magnification relationship graph, which is a graph showing the relationship between the set magnification repeatedly changed and the maximum life ratio within the reliability range, obtained in the magnification change repetition procedure (N26), is created, and from the graph, the input information When the value of the set magnification corresponding to the life ratio of the life of level 1 and level 2 is read, the read value is at least a life difference that can be determined.

  In the method of the present invention, the presence or absence of a significant difference is determined by random simulation based on the Weibull distribution, and at least the life difference that can be determined is obtained. You can estimate. Further, it can be quantitatively obtained and can be highly reliable, and can be calculated even if it is not an expert. The program used for this random number simulation is excellent in that it is a handy tool that can test the existence of a significant difference and at least determine the life difference that can be determined by inputting conditions.

In the accelerated test according to the present invention, the existence / non-existence of a significant difference / estimated life difference estimation device is caused by placing a test object consisting of a mechanical part such as a bearing or a test piece under a predetermined environmental condition that is severer than the use environment. In the accelerated test for calculating the life from time, it is a device for calculating the difference in life by magnification, which can be determined and at least determined from the life of two lots of the test object,
An arithmetic processing device (1), a display device (2) for displaying the output of the arithmetic processing device (1) on a screen, and an input means (3) for inputting to the arithmetic processing device (1) are provided.

  The arithmetic processing unit (1) displays, on the screen of the display unit (2), as input information, the value of the Weibull distribution of the test target product, each test of the two lots of level 1 and level 2 to be compared. Prompt screen output means (7M) that prompts the input of the number and test result life, and in response to the execution command, the presence / absence of the significant difference is calculated and at least the life difference that can be determined is calculated, and the calculation result is obtained. It is provided with a significant difference determination / significant life difference calculating means (52) for outputting to the screen of the display device (2).

  The significant difference determination / significant life difference calculating means (52) includes a first random number generating means (53), a second random number generating means (54), a life ratio calculating means (55), and a cumulative probability distribution analyzing means (56). ), Significant difference presence / absence determination means (57), magnification change repetition means (58), life magnification relation graph creation means (59), significant life difference reading means (60), and result output means (61).

The first random number generation means (53) is a means for generating random numbers according to the Weibull distribution determined from the value of the Weibull slope of the input information and the temporarily set lifetime for the number of test pieces of the level 1 lot.
The second random number generation means (54) is means for generating random numbers according to the same Weibull distribution as the first random number generation means (53) for the number of test pieces of the level 2 lot.
The life ratio calculation means (55) has a life determined by a predetermined life calculation method from a set number of random numbers generated by the first random number generation means (53) and a set number generated by the second random number generation means (54). Is a means for calculating the life ratio of a set of level 1 and level 2 lifespans from the lifespan obtained by the predetermined lifespan calculation method from the random number.

  The cumulative probability distribution analysis means (56) repeats the processes of the first random number generation means (53), the second random number generation means (54), and the life ratio calculation means (55) a set number of times, and the number of sets of the set number of times The life ratio is obtained, a probability distribution of the life ratio and a cumulative probability distribution are created, and the maximum life ratio within the set reliability range is read from the cumulative probability distribution.

  The significant difference presence / absence determining means (57) determines that there is a significant difference when the life ratio of level 1 and level 2 life in the input information exceeds the maximum life ratio obtained by the cumulative probability distribution analysis means (56). It is a means for determining that there is no significant difference when it is within the maximum life ratio.

The magnification change repetition means (58) is executed when there is a significant difference in the significant difference presence / absence determination means (57), and the Weibull distribution is sequentially changed to a Weibull distribution having a set magnification with a temporarily set lifetime different for each repetition. In other words, the cumulative probability distribution analysis means (56) repeats the processing until the set magnification becomes the set maximum magnification from the set minimum magnification.
The life magnification relationship graph creating means (59) is a life magnification relationship graph, which is a graph showing the relationship between the setting magnification repeatedly changed and the maximum life ratio within the reliability range, obtained by the magnification change repetition means (58). It is a means to create.
The significant life difference reading means (60) reads the value of the set magnification corresponding to the life ratio of the life of level 1 and level 2 in the input information from the life magnification relationship relation graph, and at least determines the read value. It is a means to make the life difference possible.

  The result output means (61) displays the result determined by the significant difference presence / absence determining means (57) and the life difference that can be determined at least determined by the significant life difference reading means (60) on the display device (2). It is a means to display.

  According to the significant difference presence / absence determination / significant life difference estimation apparatus in the accelerated test of this configuration, the presence / absence of significant difference is used in the significant difference existence determination / significant life difference estimation method of the present invention, and When it is determined that there is a significant difference, at least how much life difference can be said can be easily and quickly calculated, and quantitatively obtained and reliable. Even if it is not an expert, the life difference which can be determined at least can be estimated appropriately.

The significant difference presence / absence judgment / significant life difference estimation program (51) in the acceleration test of the present invention is a computer-executable program,
On the screen of the display device (2), as input information, the value of the Weibull slope of the Weibull distribution of the product under test, the number of each test in the two lots of Level 1 and Level 2 to be compared, and the life of the test result are entered. A prompt screen output procedure (N1) for prompting display;
In response to the execution command, the determination of the presence / absence of the significant difference and the calculation of the life difference that can be determined at least are performed, and the calculation result is output to the screen of the display device. including.

This significant difference judgment and significant life difference calculation procedure (N2)
A first random number generation procedure (N21) for generating a random number according to the Weibull distribution determined from the value of the Weibull slope of the input information and the temporarily set life, for the number of test pieces of the level 1 lot;
A second random number generation procedure (N22) for generating random numbers according to the same Weibull distribution as the first random number generation procedure (N21) for the test number of lots of level 2.
From the set number of random numbers generated in the first random number generation procedure (N21) by the predetermined lifetime calculation method and the set number of random numbers generated in the second random number generation procedure (N22), the predetermined lifetime calculation method. A life ratio calculation procedure (N23) for calculating a life ratio of a set of level 1 and level 2 lifespans from the obtained lifespan,
The first random number generation procedure (N21), the second random number generation procedure (N22), and the life ratio calculation procedure (N23) are repeated a set number of times to determine the life ratio of the set number of sets, and the probability distribution of the life ratio A cumulative probability distribution analysis procedure (N24) for creating a cumulative probability distribution and reading the maximum life ratio within the set confidence range from the cumulative probability distribution;
It is judged that there is a significant difference when the life ratio of the life of Level 1 and Level 2 in the above input information exceeds the maximum life ratio obtained by the cumulative probability distribution analysis procedure, and there is no significant difference when it is within the maximum life ratio. And a significant difference presence / absence determination procedure (N25).

The significant difference determination / significant life difference calculation procedure (N2) is further executed when there is a significant difference in the significant difference presence / absence determination procedure (N25). And a cumulative change distribution analysis procedure that repeats the cumulative probability distribution analysis procedure until the set magnification reaches the set maximum magnification from the set minimum magnification,
A life magnification relationship graph creation procedure (N27) for creating a life magnification relationship graph, which is a graph showing the relationship between the set magnification repeatedly changed and the maximum life ratio within the reliability range, obtained in this magnification change repetition procedure (N26). )When,
Reading the value of the set magnification corresponding to the life ratio of the life of Level 1 and Level 2 in the input information from the life magnification relationship graph and reading the significant life difference as a life difference that can be determined at least. (28) and
A result output procedure (29) for displaying on the display device the life difference that can be determined at least, the result determined in the significant difference presence / absence determination procedure and the significant life difference reading procedure;
including.

  The program for determining the presence / absence of significant difference / estimated life difference estimation in the accelerated test of this configuration is used to implement the existence / non-existence of significant difference / estimated life difference estimation of this invention. When it is determined that there is a difference, at least how much life difference can be estimated can be easily and quickly calculated, and it can be quantitatively obtained and reliable. Even if it is not a person, it can estimate the required life difference appropriately.

According to the method, apparatus, and program for determining whether there is a significant difference in the accelerated test of the present invention, the random number is generated from the Weibull distribution of the test object by computer simulation, and the random number is generated from the same Weibull distribution. The life obtained from the generated random number is calculated to calculate one set of life ratio, this process is repeated a set number of times to obtain the life ratio of the set number of sets, and the probability distribution of this life ratio and the cumulative probability distribution And read the maximum and minimum life ratio within the set confidence range from the cumulative probability distribution above, and determine the presence or absence of a significant difference, so the presence or absence of a significant difference can be easily and quickly determined, In addition, it can be obtained quantitatively and can be highly reliable, and can be determined even if it is not an expert.
In addition, when it is determined that there is a significant difference, the cumulative probability distribution analysis procedure is repeated by changing the Weibull distribution with different lifetimes every time, and the set ratio obtained by the iterations and the maximum lifetime ratio within the confidence range are repeated. From the graph, the value of the set magnification corresponding to the life ratio of the life of level 1 and level 2 in the input information is read, and the read value is at least In order to make a life difference that can be determined, if there is a significant difference, at least the life difference that can be determined can be determined easily and quickly, and it can be quantitatively determined and highly reliable. Even if it is not, a significant difference in life can be obtained.

  An embodiment of the present invention will be described. In this accelerated test, the existence of a significant difference and the estimation method of the significant life difference are determined by placing a test object consisting of a mechanical part such as a bearing or a test piece under specified environmental conditions that are severer than the operating environment, and starting from the time when the damage occurred. In the accelerated test for calculating the life, this is a method for calculating the life difference by magnification that can be determined and at least determined from the life of two lots of the test object.

Hereinafter, this embodiment will be described with reference to the drawings. In the acceleration test, the computer 1 shown in FIG. 1 executes a significant difference presence / absence determination / significant life difference estimation program 41 in the acceleration test, which is a random number simulation program. Do that. The computer 1 is composed of a personal computer or the like, has a central processing unit 4 and a memory 5, and operates by a predetermined operation system. The computer 1 is provided with a display device 2 that can be displayed on a screen such as a liquid crystal display device and an input device 3 such as a keyboard and a mouse. The computer 1, the display device 2, the input device 3, and the necessary life difference estimation program 41 constitute a significant difference presence / absence determination / significant life difference estimation device in which each function achievement means is shown as a block in FIG. 2. The configuration of the significant difference presence / absence determination / significant life difference estimation apparatus shown in FIG.
The significant difference presence / absence determination / significant life difference estimation program 51 includes the procedures shown in the flowcharts of FIGS. 4 and 5. The contents of this figure will be described later.

  As shown in FIG. 3, the required life difference estimation method includes an input process M1 for inputting predetermined information to the computer 1, and a computer calculation process M2 for performing calculation processing on the computer 1 and outputting a calculation result. Consists of.

In the input process M1, as shown in FIG. 6, an input screen 2a for prompting input of predetermined input information is displayed on the display device 2 by the output of the computer 1, and a display for prompting predetermined input is displayed on the input screen 2a. Is called.
The input screen 2a is displayed by switching to a plurality of screens in the example of FIG. Of the input screens, the first screen (FIG. 6A) displays whether to select a significant difference between the L10 life and the L50 life. By selecting the display portion consisting of, either the L10 life or the L50 life can be selected.
On the second screen (FIG. 6B), as input information, a display for prompting input of the value of the Weibull slope of the Weibull distribution of the test object and a display for filling in the input are performed. The operator enters a corresponding value in this entry display.
On the second screen (FIG. 6 (C)), the test numbers n1 and n2 of the two lots of level 1 and level 2 to be compared are input and a display for prompting the test result life is performed. A display for entry is entered. The operator inputs information corresponding to the entry display.

  It is preferable to input the actual value of the test as the value of the Weibull slope. As the actual value, it is desirable to use a result obtained by 10 or more tests, and more preferably 20 or more test results. As the number of tests and test results, the number and results of tests actually performed are input. Note that the number of tests and the lifespan are set in the Level 2 column as a procedure in the program for judging whether there is a significant difference or a significant lifespan difference estimation program 51 in which the long life side must be entered. . This procedure is provided as a part of the prompt screen output procedure N1, for example. With the above points in mind, calculation is started when a predetermined condition is input and the OK button on the input screen 2a is pressed.

  In the computer processing step M2 in FIG. 3, the life difference necessary for determining the existence of a significant difference is calculated from the input Weibull slope value, the test numbers n1, n2 of the two lots, and the life of the test result. Whether there is a significant difference or not, if there is a significant difference, calculate at least the life difference that can be determined.

When the calculation is completed, the result is displayed on the output screen shown in FIG. When there is no significant difference, as shown in FIG. 6A, a display indicating that there is no significant difference and a required life difference for determining that there is a significant difference are displayed.
When there is a significant difference, as shown in FIG. 5B, a display that there is a significant difference and a life difference (magnification) that can be determined at least are displayed.

  The significant difference presence / absence determination / significant life difference estimation program 51 of FIG. 1 is a program that can be executed by the computer 1 and includes the procedures shown in the flowcharts of FIGS. As shown in FIG. 4, the significant difference presence / absence determination / significant life difference estimation program 51 includes a prompt screen output procedure N1 and a required life difference calculation procedure N2. The prompt screen output procedure N1 is described above with reference to FIG. The input screen 2a is output. When the input information is input to the input screen 2a from the input means 3 and an execution command is input from the input means 3 by clicking the OK key on the input screen 2a, etc., it is determined whether there is a significant difference. Therefore, the life difference estimation procedure N2 is executed. The input process M1 in FIG. 3 is the input process M1 in FIG. 3, and the computer calculation process M2 in FIG. 4 executes the significant difference presence / absence / significant life difference estimation procedure N2 in FIG. It is a process.

The significant difference presence / absence determination / significant life difference estimation procedure N2 includes the respective steps shown in the flowchart of FIG. In this flowchart, a specific processing example for each of the procedures M21 to M29 is shown as an annotation.
For easy understanding, each of the procedures N21 to N29 will be described with reference to a specific processing example in FIG.

Assume that two lots of tests are performed in an accelerated test with a Weibull distribution having a certain L10 life (FIG. 8A) and a Weibull slope of 4. Assume that the number of tests n1 in the first lot (level 1) is 3, and the number of tests n2 in the second lot (level 2) is also 3.
First, according to the Weibull distribution, three random numbers corresponding to the test number n1 of the first lot are generated, and L10 life and L50 life are calculated from the three data (N21). As a calculation method, an appropriate method used for calculating the life from the result of the conventional acceleration test is used. Although only the L10 life or only the L50 life may be calculated, in this embodiment, in order to calculate the necessary life difference for each of the L10 life and the L50 life, both the L10 life and the L50 life are calculated. ing.
Next, from the Weibull distribution with the same L10 life, three random numbers that are the test number n2 of the second lot are generated, and the L10 life and L50 life are calculated from the three data by the same method as above. (N22).

  Next, the respective life ratios of the obtained L10 life and L50 life are calculated. That is, the life ratio between the L10 lives and the life ratio between the L50 lives are calculated (N23).

Next, these operations are repeated a set number of times (for example, 1000 times), and the probability distribution (FIG. 8B) and the cumulative probability distribution (FIG. 8C) of the life ratio of the L10 and L50 life of the set number of sets (1000 sets). Is created (N24).
These correspond to checking the set number of times (1000 times) to see how much difference in life results when two sets of three test pieces are extracted from the lot with the same life distribution and the life test is conducted. Yes.

From the frequency distribution diagram (FIG. 8B), it can be seen that the life varies even though the test pieces are extracted from the lots having the same life distribution. If the variation in the life ratio is defined as a predetermined reliability interval, for example, a 5% and 95% interval (90% confidence interval) of the cumulative probability distribution, the life variation is 0 as shown in FIG. It can be calculated to be 39 to 2.57 times (N243).
As a result, even when two sets of three test pieces were extracted from lots having the same life distribution, the life ratio varied between 0.39 and 2.57 times for 90%. It shows that it will end. In other words, if there is a life difference of 2.57 times or more, which is the maximum value within the variation, it is highly possible that the life distribution between the lots is different. When the number of tests is three, it can be said that the life difference that can determine the significant difference is 2.57 times or more.

  Therefore, the maximum value within the above-mentioned variation is 2.57 times and the life difference (magnification) between the two input lots (level 1 and level 2), that is, [(life of level 2) / (life of level 1). )] (N25), and if the difference in the lifespan of the two input lots is smaller, the difference in lifespan that can be used to judge the difference between the judgment result and the above-mentioned The value (2.57 in the above example) is displayed on the output screen (N29), and the processing by the computer 1 is terminated.

When the life difference between the two input lots is larger, the process proceeds to the next steps N26 to N28, and at least the life difference that can be determined is calculated.
In procedure N26, the same calculation as described above is performed in combination with a Weibull distribution having a different lifetime distribution. Specifically, the calculation is performed with a combination of the Weibull distribution and the Weibull distribution whose lifespan is 1.1, 1.2... 50 times different. When the calculation is performed in this way, every time the combination changes 1.1, 1.2... 50 times, the life difference should vary around 1.1, 1.2... 50 times.

  Finally, a graph (FIG. 8D) is created (N27) with the combination magnification as the horizontal axis and the 90% interval of each cumulative probability distribution as the vertical axis. This figure shows how the life difference varies when three sets of test pieces are extracted one by one from a group of two different test pieces at different magnifications and the life comparison is performed. For example, when two pairs of test pieces are extracted from a group of 2 lots of test specimens 4 times each and the life comparison is performed, the life difference is most often 4 times, but the life difference is 7.5 times. It can be seen from this figure that the case that becomes can occur with a probability of 10%.

Now, the difference in life between the two lots in question is three times. From the figure, it can be read that the life difference of 3 times is a life difference that can only occur between two lots having a life difference of 1.7 times or more, so at least the life difference that can be said can be 1.7 times. The value read in this way is determined as at least a life difference that can be determined (N28).
The determined life difference value is output to the output screen 2b (FIG. 7A) of the display device 2 together with a display that there is a significant difference (N29).
The above is the procedure for calculating at least the life difference that can be said from the life of two lots.

  The above description of the significant difference presence / absence determination / significant life difference estimation program 41 shown in FIGS. 4 and 5 has been described with reference to specific numerical values. This significant difference presence / absence determination / significant life difference estimation When organized, the program 41 is constituted by the following procedure.

The significant difference presence / absence judgment / significant life difference estimation program (51) in the acceleration test of the present invention is a computer-executable program,
On the screen of the display device (2), as input information, the value of the Weibull slope of the Weibull distribution of the product under test, the number of each test in the two lots of Level 1 and Level 2 to be compared, and the life of the test result are entered. A prompt screen output procedure (N1) for prompting display;
In response to the execution command, the determination of the presence / absence of the significant difference and the calculation of the life difference that can be determined at least are performed, and the calculation result is output to the screen of the display device. including.

This significant difference judgment and significant life difference calculation procedure (N2)
A first random number generation procedure (N21) for generating a random number according to the Weibull distribution determined from the value of the Weibull slope of the input information and the temporarily set life, for the number of test pieces of the level 1 lot;
A second random number generation procedure (N22) for generating random numbers according to the same Weibull distribution as the first random number generation procedure (N21) for the test number of lots of level 2.
From the set number of random numbers generated in the first random number generation procedure (N21) by the predetermined lifetime calculation method and the set number of random numbers generated in the second random number generation procedure (N22), the predetermined lifetime calculation method. A life ratio calculation procedure (N23) for calculating a life ratio of a set of level 1 and level 2 lifespans from the obtained lifespan,
The first random number generation procedure (N21), the second random number generation procedure (N22), and the life ratio calculation procedure (N23) are repeated a set number of times to determine the life ratio of the set number of sets, and the probability distribution of the life ratio A cumulative probability distribution analysis procedure (N24) for creating a cumulative probability distribution and reading the maximum life ratio within the set confidence range from the cumulative probability distribution;
It is determined that there is a significant difference when the life ratio of the life of Level 1 and Level 2 in the above input information exceeds the maximum life ratio obtained by the cumulative probability distribution analysis procedure, and there is no significant difference when it is within the maximum life ratio. And a significant difference presence / absence determination procedure (N25) for determination.

The significant difference determination / significant life difference calculation procedure (N2) is further executed when there is a significant difference in the significant difference presence / absence determination procedure (N25). And a cumulative change distribution analysis procedure that repeats the cumulative probability distribution analysis procedure until the set magnification reaches the set maximum magnification from the set minimum magnification,
A life magnification relationship graph creation procedure (N27) for creating a life magnification relationship graph, which is a graph showing the relationship between the set magnification repeatedly changed and the maximum life ratio within the reliability range, obtained in this magnification change repetition procedure (N26). )When,
Reading the value of the set magnification corresponding to the life ratio of the life of Level 1 and Level 2 in the input information from the life magnification relationship graph and reading the significant life difference as a life difference that can be determined at least. (28) and
And a result output procedure (29) for displaying on the display device at least a determinable life difference read by the significant difference presence / absence determination procedure (N28) and the significant life difference reading procedure.

  The cumulative probability distribution analysis procedure (N24) includes a procedure N241 for repeating steps N21 to N23 a set number of times, a procedure N242 for calculating a probability distribution and a cumulative probability distribution from the life ratio stored in the repetition process, And a procedure N243 for reading a long life ratio in the 90% range from the probability distribution.

  The magnification change repetition procedure (N26) includes a procedure N261 that is repeated until the set magnification is changed from the set minimum magnification to the set maximum magnification, and a magnification change procedure N231 that is changed to a different set magnification for each repetition.

Details of each random number generation procedure (N21, N22) will be described. These procedures N21 and N22 specify a Weibull distribution and generate a Weibull random number according to the specified Weibull distribution.
Generally, it is said that the bearing life distribution follows the Weibull distribution of the following equation 1).

Where m: Weibull slope, α: scale factor, γ: minimum life,
The Weibull distribution has three parameters and is known as a universal distribution that can express an exponential distribution, a lognormal distribution, and a normal distribution by the Weibull slope m. For reference, FIG. 10 shows changes in the Weibull distribution when various parameters are changed. The Weibull slope m is a parameter that governs the shape of the distribution, and it can be said that the smaller the value, the larger the variation. The scale factor α changes the scale of the horizontal axis (lifetime), and the larger the value, the longer the life. The minimum life γ simply shifts the horizontal axis (life) of the life distribution.

In this embodiment, a Weibull random number is generated. In order to generate this random number, it is necessary to determine three parameters of the Weibull distribution. The procedure for deciding is, for example, as follows.
1) The Weibull slope m is determined from the results.
2) Determine the reliability of the distribution for which random numbers are to be generated (for example, whether the life is L10 or L50).
3) From the Weibull slope m obtained from the reliability, the minimum life γ is determined using a predetermined formula. For example, from the scale factor α obtained from the L10 life or L50 life,
The minimum lifetime γ is determined using, for example, the following equation 2).
This equation is the minimum life of ISO established in 1990, and is a regression equation from experimental values.

This is an expression such that this value becomes 1 when R ≦ 10 and R = 0 (a1 with L10 life). In a formula that takes into account the minimum lifetime of ISO in the past, a lifetime that is less than or equal to the L10 lifetime is defined as a value that can write the L10 lifetime in this formula. R is a value corresponding to the reliability (a value at which 100-R is the reliability).
Note that the method of determining the minimum life may be changed with the times in various standards (for example, ISO), but a method according to the standard at the time of implementation may be adopted along with the change of the standard. In addition, there is a claim that it is better to describe the minimum life by a more general formula because it changes depending on the material test conditions, and an appropriate value may be used.

  The generation of Weibull random numbers will be described. Random numbers are qualitatively random sequences that have a uniform frequency of occurrence (equal probability) and no regularity in their generation (irregularity). Is impossible. Therefore, pseudorandom numbers that can be generated by a computer are used. A simple random number generation algorithm, for example, has a periodicity of about 20 digits in decimal notation, but some periodicity has a periodicity of 6,000 digits or more. It is preferable to use it.

In this embodiment, a Weibull random number that is not a uniform random number but a random number according to the Weibull distribution is generated. For this reason, a device is required for the generation method. When the probability density function is complicated, a method called a rejection method may be used to generate random numbers according to the distribution, and the rejection method is also used in this embodiment.
As shown in FIG. 11, it is assumed that the domain of the probability density function f (x) is in the range of 0 to X0, and the maximum value of f (x) in the domain is M. If RN is a uniform pseudorandom number in the interval [0,1], a uniform pseudorandom number xi in the interval [0, x0] can be generated by X0 · RN. Similarly, uniform pseudorandom numbers yi in the interval [0, M] can be generated by M · RN. Therefore, when the generated random numbers xi, yi satisfy the condition that f (xi)> yi, the random number xi is adopted as following the given probability density distribution. The random number xi is not adopted. A method of repeating this work, adopting random numbers xi with a probability according to the probability density distribution, and creating a sequence of random numbers according to the probability density distribution is called a rejection method. This method is not efficient as a random number generation method because random numbers that do not meet the conditions are discarded. However, in principle, if a good uniform random number is obtained, a correct number sequence can be obtained.

In the random number generation procedures N21 and N22 in FIG. 5, the life for specifying the Weibull distribution (L10 life) may be set as an appropriately assumed value in the calculation formula of these procedures N21 and N22, Further, it may be variable by input from the input means 3.
Further, in the life comparison calculation procedure N23, the calculation for obtaining the L10 life from the generated Weibull random number may use an appropriate life time calculation method conventionally used in the acceleration test. In the accelerated test, the life is obtained by a Weibull plot from the time when the breakage occurs, and a calculation method for obtaining the life from the breakage time is used as the life calculation method in the life ratio calculation procedure.

For this lifetime calculation, for example, the following method can be employed.
(1) Conduct a life test.
(2) Sort the obtained data (damaged time or number of broken loads) in ascending order.
(3) These data are plotted on the graph (Weibull probability paper) in FIG. 13 (vertical axis: cumulative failure probability, horizontal axis: life).
(4) The optimal straight line of the data plotted on the paper of FIG. 13 is drawn by the method of least squares. At this time, since there is a minimum life at a position below the L10 life, the value of the L10 life is divided into 10 (any number is possible, but the calculation time for fitting is set to be a reasonable time) A plot is added at the position where the cumulative probability is 0%. The optimum straight line that best fits the data is adopted among the 10 optimum curves.
(5) Then, the Weibull slope is the slope of this line, and the minimum life is based on the relationship between the L10 life value in any of 10 divisions, the L10 life (lifetime where the Weibull slope has a cumulative probability of 10%) and the scale factor α. α can be determined.

  The censoring time estimation apparatus program 51 obtains the relationship of the necessary life difference for determining that there is a life difference with a significant difference. FIG. 9 shows the relationship of the required life difference for determining that there is a life difference. Since this result changes depending on the Weibull slope and the number of tests input in advance, the figure shows the result of changing the Weibull slope and the number of tests. It can be seen that the inclination decreases as the number of tests increases or the Weibull slope increases. These results show that (1) the difference in life required to determine that there is a difference in life can be reduced as the number of tests increases, and (2) the difference in life required to determine that there is a difference in life for test conditions with a larger Weibull slope. It shows that it can be made smaller.

  Together with FIG. 2, a device for determining whether there is a significant difference in the acceleration test and estimating a significant life difference will be described. This significant difference presence / absence determination / significant life difference estimation device includes an arithmetic processing device 1, a display device 2 for displaying the output of the arithmetic processing device 1 on a screen, and an input means 3 for inputting to the arithmetic processing device 1. With.

  The arithmetic processing device 1 displays, on the screen of the display device 2, as input information, the value of the Weibull slope of the Weibull distribution of the test object, the number of tests in each of the two lots of level 1 and level 2 to be compared, and the test results. A prompt screen output means 7M for performing a display prompting the input of the life of the user, and in response to the execution command, the presence / absence of the significant difference is determined, and at least the life difference that can be determined is calculated, and the calculation result is displayed on the screen of the display device 2 And a significant life difference calculating / meaning life difference calculating means 51.

  The significant difference determination / significant life difference calculation means 52 includes a first random number generation means 53, a second random number generation means 54, a life ratio calculation means 55, a cumulative probability distribution analysis means 56, a significant difference presence / absence determination means 57, A magnification change repeating unit 58, a life magnification relationship graph creating unit 59, a significant life difference reading unit 60, and a result output unit 61 are provided.

The first random number generating means 53 is means for generating random numbers according to the Weibull distribution determined from the value of the Weibull slope of the input information and the temporarily set life for the number of test pieces of the level 1 lot, and will be described in the procedure N21 of FIG. Perform the process.
The second random number generation means 54 is means for generating random numbers according to the same Weibull distribution as the first random number generation means 53 for the test number of lots of level 2, and performs the processing described in step N22.
The life ratio calculating means 55 is configured to calculate the predetermined time from the life determined by a predetermined life calculation method from the set number of random numbers generated by the first random number generating means 53 and the set number of random numbers generated by the second random number generating means 54. It is a means for calculating the life ratio of a set of level 1 and level 2 lifespans from the lifespan determined by the lifespan calculation method, and performs the processing described in step N23.

  The cumulative probability distribution analysis means 56 repeats the processes of the first random number generation means 53, the second random number generation means 54, and the life ratio calculation means 55 a set number of times, obtains the life ratio of the set number of times, and this life This is means for creating a probability distribution of ratios and a cumulative probability distribution and reading the maximum life ratio within the set confidence width from the cumulative probability distribution, and performs the processing described in step N24.

  The significant difference presence / absence determination means 57 determines that there is a significant difference when the life ratio of the life of the level 1 and level 2 in the input information exceeds the maximum life ratio obtained by the cumulative probability distribution analysis means 56, and determines the maximum life. A means for determining that there is no significant difference when the ratio is within the ratio, and performs the processing described in step N25.

The magnification change repetition unit 58 is executed when the significant difference presence / absence determination unit 57 determines that there is a significant difference, and sequentially changes the Weibull distribution to a Weibull distribution having a set magnification with a different temporary setting life for each repetition. The process of the cumulative probability distribution analysis unit 56 is a unit that repeats the above-described setting magnification from the set minimum magnification to the set maximum magnification, and performs the processing described in step N26.
The life magnification relationship graph creating means 59 is a means for creating a life magnification relationship graph which is a graph showing the relationship between the set magnification repeatedly changed and the maximum life ratio within the reliability range obtained by the magnification change repetition means 58. Yes, the process described in step N27 is performed.
The significant life difference reading means 60 reads the value of the set magnification corresponding to the life ratio of the life of Level 1 and Level 2 in the input information from the life magnification relationship relation graph, and the life that can at least determine the read value. It is a means for making a difference, and the processing described in step N28 is performed.

  The result output means 61 is a means for causing the display device 2 to display the result determined by the significant difference presence / absence determining means 57 and the life difference that can be determined at least read by the significant life difference reading means 60, and the procedure N29. Perform the process described in.

1 is a schematic block diagram of a significant difference presence / absence determination / significant life difference estimation apparatus according to an embodiment of the present invention. FIG. It is a block diagram which shows the conceptual structure of the same significant difference existence determination / significant lifetime difference estimation apparatus. It is a general | schematic flowchart of the required lifetime difference estimation method using the same significant difference existence determination / significant lifetime difference estimation apparatus. It is a general | schematic flowchart of the significant difference existence determination / significant lifetime difference estimation program which implements the same significant difference presence determination / significant lifetime difference estimation method. It is a flowchart which shows the detail of the significant difference existence determination / significant lifetime difference estimation procedure in the program. It is explanatory drawing of the example of an input screen in the significant difference existence presence determination / significant lifetime difference estimation apparatus of FIG. It is explanatory drawing of the example of an output screen in the significant difference presence / absence determination / significant lifetime difference estimation apparatus of FIG. (A) is a graph of an example of the Weibull distribution, (B) is a graph of a probability distribution showing an example of the relationship between the frequency and the life ratio, (C) is a graph showing an example of the relation between the cumulative probability and the life ratio, and (D) is at least It is a graph which shows the relationship between the life difference (magnification) which can be determined, and a life difference. (A) is a graph showing the relationship between the life magnification for each Weibull slope and the required life difference, and (B) is a graph showing the relationship between the life magnification for each test number and the required life difference. It is a graph which shows the example of influence of each parameter of a Weibull distribution. It is a graph which shows how to define a Weibull distribution. It is distribution comparison figure of the lifetime dispersion | variation of the method of embodiment, and the method using the conventional reliability width. It is explanatory drawing of a Weibull probability paper. It is a flowchart which shows the procedure of the conventional truncation and acceleration test. It is explanatory drawing of the lifetime significant difference determination method using the conventional reliability range.

Explanation of symbols

1 Computer (arithmetic processing means)
2. Display device 3 ... Input device 7M ... Prompt screen output means 51 ... Significant difference presence / absence determination / significant life difference estimation program 52 ... Significant difference presence / absence determination / significant life difference estimation means 53 ... First random number generation means 54 ... second random number generating means 55 ... life ratio calculating means 56 ... cumulative distribution creating means 57 ... significant difference presence / absence judging means 58 ... magnification change repeating means 59 ... result output means

Claims (3)

In an accelerated test that calculates the life from the time at which damage occurred, placing the test object consisting of mechanical parts such as bearings or test pieces under specified environmental conditions that are severer than the operating environment, the life of the two lots of the test object is calculated. , A method for calculating the difference in life according to magnification, which can be determined whether there is a significant difference, and at least determined,
The process of inputting the value of the Weibull slope of the Weibull distribution of the test object, the number of tests in each of the two lots of Level 1 and Level 2 to be compared, and the life of the test results as input information to the computer,
A computer calculation process for causing the computer to determine the presence or absence of the significant difference and to calculate at least the life difference that can be determined and to display the calculation result on the screen of the display device;
As the above computer processing process,
A first random number generation procedure for generating a random number according to the Weibull distribution determined from the value of the Weibull slope of the input information of the test object and the temporarily set life for the number of test pieces of the level 1 lot;
A second random number generation procedure for generating random numbers according to the same Weibull distribution as the first random number generation procedure for the number of test pieces of the level 2 lot;
From the lifetime obtained by the predetermined lifetime calculation method from the set number of random numbers generated in the first random number generation procedure and the lifetime determined by the predetermined lifetime calculation method from the set number of random numbers generated in the second random number generation procedure, 1 A life ratio calculation procedure for calculating the life ratio of the life of the pair of level 1 and level 2,
The first random number generation procedure, the second random number generation procedure, and the life ratio calculation procedure are repeated a set number of times, the life ratio of the set number of times is obtained, and a probability distribution and a cumulative probability distribution of the life ratio are created and accumulated. The cumulative probability distribution analysis procedure that reads the maximum life ratio within the set confidence range from the probability distribution, and the life ratio of the level 1 and level 2 lives in the input information is the maximum life ratio obtained by the cumulative probability distribution analysis procedure. If it exceeds, it is determined that there is a significant difference, and if it is within the maximum life ratio, there is no significant difference.
This procedure is executed when there is a significant difference, and the Weibull distribution is changed to a Weibull distribution with a different set magnification with a temporary set life, and the cumulative probability distribution analysis procedure is set for the cumulative probability distribution analysis procedure. Repeat magnification change procedure that repeats from the minimum magnification to the set maximum magnification,
A life magnification relationship graph creation procedure for creating a life magnification relationship graph, which is a graph showing the relationship between the set magnification repeatedly changed and the maximum life ratio within the reliability range obtained in this magnification change repetition procedure,
Reading the value of the set magnification corresponding to the life ratio of the life of Level 1 and Level 2 in the input information from the life magnification relationship graph and reading the significant life difference as a life difference that can be determined at least. When,
A result output procedure for displaying on the display device a result determined in the significant difference presence / absence determination procedure and a life difference that can be determined at least read in the significant life difference reading procedure;
Including a significant difference in acceleration tests, and a method for estimating a significant life difference. In an accelerated test that calculates the life from the time at which damage occurred, placing the test object consisting of mechanical parts such as bearings or test pieces under specified environmental conditions that are severer than the operating environment, the life of the two lots of the test object is calculated. , A device for calculating the difference in life by magnification, which can be determined whether there is a significant difference, and at least determined,
An arithmetic processing device, a display device for displaying the output of the arithmetic processing device on a screen, and input means for inputting to the arithmetic processing device,
The arithmetic processing unit is
On the screen of the display device, as input information, a display prompting the user to input the value of the Weibull slope of the Weibull distribution of the product under test, the number of each test in the two lots of Level 1 and Level 2 to be compared, and the life of the test results Prompt screen output means to perform,
In response to the execution command, comprises the presence / absence of the significant difference and the calculation of the life difference that can be determined at least and outputs the calculation result to the screen of the display device.
The above significant difference judgment and significant life difference calculation means are:
First random number generating means for generating random numbers according to the Weibull distribution determined from the value of the Weibull slope of the input information and the temporarily set life for the number of test pieces of the level 1 lot;
Second random number generating means for generating random numbers according to the same Weibull distribution as the first random number generating means for the test number of lots of level 2;
From the lifetime determined by the predetermined lifetime calculation method from the set number of random numbers generated by the first random number generator and the lifetime determined by the predetermined lifetime calculation method from the set number of random numbers generated by the second random number generator. A life ratio calculating means for calculating the life ratio of the life of the pair of level 1 and level 2;
The processing of the first random number generation means, the second random number generation means, and the life ratio calculation means is repeated a set number of times, the life ratio of the set number of times is obtained, and a probability distribution and a cumulative probability distribution of the life ratio are created. A cumulative probability distribution analysis means for reading the maximum life ratio within the set reliability range from the cumulative probability distribution;
It is judged that there is a significant difference when the life ratio of the life of Level 1 and Level 2 in the input information exceeds the maximum life ratio obtained by the cumulative probability distribution analysis means, and there is no significant difference when it is within the maximum life ratio. Means for determining whether there is a significant difference, and
This means is executed when there is a significant difference, and the Weibull distribution is sequentially changed to a Weibull distribution having a different set magnification with a temporarily set life, and the process of the cumulative probability distribution analysis means is performed with the set magnification. Magnification change repetition means that repeats until the set minimum magnification reaches the set maximum magnification,
A life magnification relationship graph creating means for creating a life magnification relationship graph, which is a graph showing the relationship between the set magnification repeatedly changed and the maximum life ratio within the reliability range obtained by the magnification change repetition means,
Meaning life difference reading means for reading the value of the set magnification corresponding to the life ratio of the life of level 1 and level 2 in the input information from the life magnification relation graph and making the read value at least a life difference that can be determined When,
A result output means for displaying a result determined by the significant difference presence / absence determining means and a life difference that can be determined at least, read by the significant life difference reading means, on a display device;
Including significant difference in acceleration test, and useful life difference estimation device. A program executable on a computer,
On the screen of the display device, as input information, a display that prompts the user to input the Weibull slope value of the Weibull distribution of the test object, the number of each test in the two lots of Level 1 and Level 2 to be compared, and the life of the test results Prompt screen output procedure,
In response to the execution command, includes the determination of the presence or absence of a significant difference and the calculation of at least the life difference that can be determined, and the calculation result is output to the screen of the display device.
This significant difference judgment and significant life difference calculation procedure is:
A first random number generation procedure for generating a random number according to the Weibull distribution determined from the value of the Weibull slope of the input information and the temporarily set life for the number of test pieces of the level 1 lot;
A second random number generation procedure for generating random numbers according to the same Weibull distribution as the first random number generation procedure for the number of test pieces of the level 2 lot;
From the lifetime obtained by the predetermined lifetime calculation method from the set number of random numbers generated in the first random number generation procedure and the lifetime determined by the predetermined lifetime calculation method from the set number of random numbers generated in the second random number generation procedure, 1 A life ratio calculation procedure for calculating the life ratio of the life of the pair of level 1 and level 2,
The first random number generation procedure, the second random number generation procedure, and the life ratio calculation procedure are repeated a set number of times, the life ratio of the set number of times is obtained, and a probability distribution and a cumulative probability distribution of the life ratio are created and accumulated. The cumulative probability distribution analysis procedure that reads the maximum life ratio within the set confidence range from the probability distribution, and the life ratio of the level 1 and level 2 lives in the input information is the maximum life ratio obtained by the cumulative probability distribution analysis procedure. If it exceeds, it is determined that there is a significant difference, and if it is within the maximum life ratio, there is no significant difference.
This procedure is executed when there is a significant difference, and the Weibull distribution is changed to a Weibull distribution with a different set magnification with a temporary set life, and the cumulative probability distribution analysis procedure is set for the cumulative probability distribution analysis procedure. Repeat magnification change procedure that repeats from the minimum magnification to the set maximum magnification,
A life magnification relationship graph creation procedure for creating a life magnification relationship graph, which is a graph showing the relationship between the set magnification repeatedly changed and the maximum life ratio within the reliability range obtained in this magnification change repetition procedure,
Reading the value of the set magnification corresponding to the life ratio of the life of Level 1 and Level 2 in the input information from the life magnification relationship graph and reading the significant life difference as a life difference that can be determined at least. When,
A result output procedure for displaying on the display device a result determined in the significant difference presence / absence determination procedure and a life difference that can be determined at least read in the significant life difference reading procedure;
A program for determining the presence / absence of a significant difference and estimating a significant life difference in an accelerated test.

Images