JP2008145414A - Life estimation method and system from truncated life test - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method of a truncated life test capable of simply and quickly trial calculating the life level of the testing object even an unskilled person from the undamaged time while the testing object is still subjected to the test without being damaged with high reliability in the truncated life test. <P>SOLUTION: A computer generates the Weibull random numbers having established rate lives to the undamaged time by the testing number of pieces, and surveys the remaining random numbers removed in order of shorter of the random numbers by the number of testing pieces whether larger than the undamaged time or not. This treatment is repeated by the established number of times. A process for investigating the probability being ≥ investigated undamaged time is repeated from the prescribed shortest life to the longest life successively to the Weibull distribution having the lives successively changed in established rate for every repeated time. From the relation between the obtained life and the probability existing ≥ the undamaged time, the generation probability is determined as the lot life of the inspection object by reading out the value to be the life obtained by subtracting the prescribed reliability from 100%. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  This invention is a life test for bearings or other machine parts. In a censorship test that determines that the required life is satisfied if the test continues without damage until the censorship time, which is the target time, the test continues after the censorship time, etc. The present invention relates to a method, apparatus, and program for estimating the life of a lot of a test target product from an unbreakage time during which the test is continued without damaging at least some of the test target products.

  The life test is one of the tests indispensable for evaluating the performance of mechanical parts such as bearings. Life tests can be broadly classified into (1) actual machine tests that perform tests under conditions close to the actual operating environment and (2) accelerated tests that perform life tests under relatively harsh 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. 22 shows the procedure of the life test design and the interpretation of the life test results that have been conventionally performed for each of the abort test and the 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

The censorship test is a test in which the required life is satisfied after the censoring time has elapsed, but the test is continued after the censoring time has elapsed, such as when there is a margin for delivery, and the life level of the product under test is determined. You may want to check. This situation often occurs.
However, conventionally, there has been no appropriate method for estimating the life level of the product under test by continuing such a truncation test.
In addition, there is a case where it is desired to know not only the lower limit of the life of the target lot as the life level of the above test object but also the upper limit of the upper limit of the life of the lot. However, there has been no suitable method for estimating the upper limit of such life.

The object of the present invention is to easily and quickly calculate the life level of the lot of the test object from the non-breakage time in which the test is continued without damaging at least a part of the test object in the censoring test. It is possible to provide a method and an apparatus capable of being estimated and reliable, and capable of estimating the life level even if not skilled, and a computer program for carrying out the method.
Another object of the present invention is to easily and quickly calculate the upper limit of the life of a lot of the test object from the non-breakage time during which the test is continued without damaging at least a part of the test object. It is possible to provide a method, an apparatus, and a computer program for carrying out the method, which can be performed and have high reliability, and can estimate the upper limit of the lifetime without being an expert.

The life estimation method from the first life end test in the present invention is such that a test object consisting of a bearing or other machine part or test piece is placed in a predetermined use environment condition and is not damaged until the end time which is a target time. If the test continues, it is determined that the test target product will be tested from the non-breakage time during which the test is continued without damaging at least part of the test target product in the censorship test that determines that the required life for the specified reliability is satisfied. A method for estimating the life of a lot,
Input process to input the Weibull slope value of the test target product, the number of test target products, the number of undamaged or damaged items as the number of undamaged test target items, and the unbreakage time as input information to the computer (E1),
A computer calculation processing step (E2) for causing the computer to calculate a lifetime and display a calculation result on a screen of a display device.
The lifetime with the predetermined reliability is, for example, an L10 lifetime (90% reliability lifetime), an L50 lifetime (50% reliability lifetime), or the like. The non-breakage time and the life may not necessarily be a unit of time, and may be displayed, for example, by the number of loads such as the number of rotations when the test is performed by rotating the bearing.

As the computer calculation process (E2),
A random number generation procedure (G21) using a Weibull random number, which is a random number according to the Weibull distribution having a set life with respect to the unbreakage time, for the test number, and using the value of the Weibull slope of the input information for the Weibull distribution;
Random number analysis procedure (G22) for checking whether or not the remaining random numbers obtained by removing the random numbers corresponding to the damaged number in order from the shortest number among the Weibull random numbers corresponding to the generated test number are equal to or longer than the undamaged time;
A set percentage life satisfaction investigation procedure (G23) for repeating the random number generation procedure and the random number analysis procedure a set number of times, and examining the probability of being in the unbreakage time or more examined in the random number analysis procedure in each iteration of the repetition;
This set ratio life satisfaction check procedure (G23) is repeated for a Weibull distribution having a life in which the set ratio is sequentially changed from a predetermined shortest life shorter than the breakage time to a predetermined longest life that is gradually longer. Different life satisfaction investigation procedure (G24),
From the relationship between the life obtained by this different life satisfaction investigation procedure (G24) and the probability that it is longer than the unbreakage time, read the life when the probability of occurrence is a value obtained by subtracting the predetermined reliability from 100%, and then the product to be tested The life reading procedure (G25) determined as the life of the lot of
Execute.

  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 machine part 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.

According to this method, weibull random numbers, which are random numbers according to the Weibull distribution having a set life against unbreakage time, are generated for the number of tests, and among the generated number of Weibull random numbers, the random number for the number of breakage is short. Check whether the remaining random numbers removed in order from the one are longer than the unbreakage time. This procedure is repeated a set number of times, for example, thousands of times, and the probability of being over the unbreakage time is examined. This probability is obtained for a Weibull distribution having sequentially different lifetimes.
The relationship between the life thus obtained and the probability of being longer than the unbreakage time indicates what kind of life distribution the probability of being unbroken up to the unbreakage time is high. Therefore, the life of the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% (if the life is L10, 10% obtained by subtracting 90% from 100%) and determined as the life of the lot of the test object. be able to.
In this way, weibull random numbers of various lifetimes are generated for the number of tests, and the remaining random numbers excluding the number of breakage are used to determine what life distribution is likely to be unbroken until the unbreakage time. Since the probability distribution is obtained, the life level of the test object can be obtained with high reliability from the unbreakage time. In addition, the above processing is a computer simulation, and as input information to the computer, the value of the Weibull slope of the test object, the number of tests of the test object, the number of undamaged items or the number of breakage that is the number of undamaged test items Since the life level is output only by inputting the unbreakage time, the life level can be easily and quickly obtained from the unbreakage time without requiring skill.

The apparatus for estimating the life from the first life-end-of-life test in the present invention places a test object consisting of a bearing or other mechanical part or test piece in a predetermined use environment condition, and does not break until the end time, which is a target time. If the test continues, it is determined that the test target product will be tested from the non-breakage time during which the test is continued without damaging at least part of the test target product in the censorship test that determines that the required life for the specified reliability is satisfied. A device for estimating the life of a lot,
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 (2) are provided.

The arithmetic processing unit (1) displays, on the screen of the display unit (2), as input information, the value of the Weibull slope of the test object, the test number of the test object, and the number of undamaged test objects. A prompt screen output means (7E) for performing display for prompting input of the number of breakage or the number of breakage and the unbreakage time;
In response to the input of the execution command, a life estimation calculating means (22) for calculating the life and outputting it to the display screen is provided.

  The life estimation calculation means (22) includes a random number generation means (23), a random number analysis means (24), a set rate life satisfaction investigation means (25), a different life satisfaction investigation means (26), and a life reading means. (27) and reading result output means (28).

  The random number generation means (23) is means for generating Weibull random numbers, which are random numbers according to the Weibull distribution having a set ratio of lifetime with respect to the unbreakage time, for the number of tests, and the Weibull distribution includes a Weibull slope of the input information. Use the value.

  The random number analyzing means (24) is a means for examining whether or not the remaining random numbers obtained by removing the random numbers corresponding to the damaged number in order from the shortest number among the Weibull random numbers corresponding to the generated test number are equal to or longer than the undamaged time.

  The set ratio life satisfaction checking means (25) repeats the processes of the random number generating means (23) and the random number analyzing means (24) a set number of times, and the random number analyzing means (24) examined by the random number analyzing means (24) at each of the repetitions. It is a means of examining the probability of being over time.

  The different life satisfaction surveying means (26) repeats the processing of the set percentage life satisfaction surveying means (25) from the predetermined shortest life shorter than the breakage time to the predetermined longest life that is gradually longer. It is a means for repeating the Weibull distribution with sequentially changed lifetimes.

  The life reading means (27) is a value obtained by subtracting the predetermined reliability from 100% from the relationship between the life obtained by the processing of the different life satisfaction investigation means (26) and the probability that it is longer than the unbreakage time. Is a means for reading the life determined as the life of the lot of the test object.

  The reading result output means is means for outputting the life read by the life reading means to the display device.

  The apparatus for estimating the life from the life censoring test of this configuration can easily and quickly obtain the life level from the unbreakage time by carrying out the life estimation method of the present invention, and can be highly reliable. Even if it is not a person, the life level can be appropriately determined from the unbreakage time.

The life estimation program from the first life cutoff test in the present invention is:
A program executable on a computer,
On the screen of the display device, input the value of the Weibull slope of the product under test, the number of test products under test, the number of undamaged or damaged products, the number of undamaged products under test, and the time of unbreakage. A prompt screen output procedure (F1) for prompting display;
A life calculation procedure (F2) for calculating the life and displaying it on the screen of the display device in response to the execution command.

The above life calculation procedure (F2)
A random number generation procedure (G21) using a Weibull random number, which is a random number according to the Weibull distribution having a set life with respect to the unbreakage time, for the test number, and using the value of the Weibull slope of the input information for the Weibull distribution;
Random number analysis procedure (G22) for checking whether or not the remaining random numbers obtained by removing the random numbers corresponding to the damaged number in order from the shortest number among the Weibull random numbers corresponding to the generated test number are equal to or longer than the undamaged time;
A set percentage life satisfaction investigation procedure (G23) for repeating the random number generation procedure and the random number analysis procedure a set number of times, and examining the probability of being in the unbreakage time or more examined in the random number analysis procedure in each iteration of the repetition;
This set ratio life satisfaction check procedure (G23) is repeated for a Weibull distribution having a life in which the set ratio is sequentially changed from a predetermined shortest life shorter than the breakage time to a predetermined longest life that is gradually longer. Different life satisfaction investigation procedure (G24),
From the relationship between the life obtained by this different life satisfaction survey procedure and the probability that it is longer than the unbreakage time, read the life of which the probability of occurrence is a value obtained by subtracting the predetermined reliability from 100%. Life reading procedure (G25) determined as the life,
including.

  The life estimation program from the life end-of-life test of this configuration is used in the implementation of the life estimation method of the present invention, the life level can be easily and quickly determined from the unbreakage time, and can be highly reliable. Even if it is not an expert, a life level can be calculated | required appropriately from unbreakage time.

The life estimation method from the second life cutoff test in the present invention is a method for estimating the upper limit of the lifetime, that is, the upper limit that cannot be expected from the result of the lifetime test.
That is, the life estimation method from the second life cutoff test is:
If the test object consisting of a bearing or other mechanical part or test piece is placed in the specified operating environment conditions and the test continues without damage until the target time, which is the cutoff time, the required life for the life with the specified reliability can be obtained. In the censorship test that is judged to be satisfactory, the upper limit of the service life of the lot under test is estimated from the unbreakage time during which the test continues without causing damage to some of the test target products and the remaining target products. And
As input information to the computer, the value of the Weibull slope of the product under test, the number of tests of the product under test, the number of undamaged or broken items that are the number of untested products under test, and the unbreakage time (some An input process (E1A) for inputting a non-breakage time during which the test is continued without damaging the test target product and the remaining target product being damaged,
A computer calculation process (E2A) for causing the computer to calculate the upper limit of the lifetime and to display a calculation result on a screen of a display device.
The lifetime with the predetermined reliability is, for example, an L10 lifetime (90% reliability lifetime), an L50 lifetime (50% reliability lifetime), or the like. The non-breakage time and the life may not necessarily be a unit of time, and may be displayed, for example, by the number of loads such as the number of rotations when the test is performed by rotating the bearing.

As the computer calculation process (E2A),
A random number generation procedure (G21A) using a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to an unbreakage time, for the test number, and using the value of the Weibull slope of the input information for the Weibull distribution;
A random number analysis procedure (G22A) for checking whether or not the random number from the smallest to the number of breakage among the Weibull random numbers for the generated number of tests is equal to or less than the unbreakage time;
The above-mentioned random number generation procedure (G21A) and the above random number analysis procedure (G22A) are repeated a set number of times, and the set ratio life non-satisfaction investigation procedure (G23A) for examining the probability that the random number from the smallest one to the number of breakage will be less than the above-mentioned unbreakage time When,
This set ratio life non-satisfaction investigation procedure (G23A) is changed to a Weibull distribution with a life in which the set ratio is sequentially changed every time from a predetermined shortest life shorter than the unbreakage time to a predetermined longest long life. And the different life non-satisfaction investigation procedure (G24A),
From the relationship between the life obtained by this different life non-satisfaction investigation procedure (G24A) and the probability of being less than the unbreakage time, the life of which the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% is read. The life upper limit reading procedure (G25A) determined as the upper limit of the life of the product lot is executed.
The predetermined reliability to be subtracted from 100% is, for example, 90%.

  More specifically, the life upper limit reading procedure (G25A) is more specifically set to the probability that the failure is less than or equal to the unbreakage time for each set ratio life determined by the different life non-satisfaction investigation procedure (G24A). The cumulative probability distribution, which is the relationship of the probability of being less than or equal to the time, is read from this cumulative probability distribution, and the lifetime at which the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% is read as the upper limit of the lifetime of the lot under test. It is assumed that the procedure is determined.

According to this method, weibull random numbers, which are random numbers according to the Weibull distribution with a set life against unbreakage time, are generated for the number of tests, and from the smallest number of Weibull random numbers for the number of generated tests to the number of breakage. It is checked whether or not the random number is less than the unbreakage time. This procedure is repeated a set number of times, for example, thousands of times, and the probability of being over the unbreakage time is examined. This probability is obtained for a Weibull distribution having sequentially different lifetimes.
The relationship between the life thus obtained and the probability of being below the unbreakage time is that the life of the product under test from the shortest to the number of breakages is less than the above-mentioned unbreakage time for any life distribution. It indicates whether the probability is high. Therefore, by reading the life when the probability of occurrence is a value obtained by subtracting the predetermined reliability from 100% (if the life is L10, 10% obtained by subtracting 90% from 100%), the upper limit of the life of the lot of the test object is read. Can be determined.
In this way, weibull random numbers with various lifetimes are generated for the number of test pieces, and what kind of life distribution is there is a high probability that the life of the product under test from the shortest to the number of breakage will be less than the unbreakage time? Thus, the upper limit of the life of the test object can be obtained with high reliability from the unbreakage time. In addition, the above processing is a computer simulation, and as input information to the computer, the value of the Weibull slope of the test object, the number of tests of the test object, the number of undamaged items or the number of breakage that is the number of undamaged test items Since the life level is output only by inputting the unbreakage time, the upper limit of the life can be easily and quickly obtained from the unbreakage time without requiring skill.

The life estimation method from the second life end test in this invention may be used in combination with the life estimation method from the first life end test.
That is, in the life estimation method from the first life-end-of-life test, the lot of the test object is determined from the non-breakage time during which the test is continued without damaging a part of the test object and the remaining test object. As a method for estimating the upper limit, the computer includes a computer calculation process (E2A) in which the computer calculates the upper limit of the lifetime and displays the calculation result on the screen of the display device.

As the computer calculation process (E2A),
A random number generation procedure (G21A) using a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to an unbreakage time, for the test number, and using the value of the Weibull slope of the input information for the Weibull distribution;
A random number analysis procedure (G22A) for checking whether or not the random number from the smallest to the number of breakage among the Weibull random numbers for the generated number of tests is equal to or less than the unbreakage time;
A set ratio life non-satisfaction investigation procedure (G23A) that repeats the random number generation procedure (G1A) and the random number analysis procedure (G22A) a set number of times, and examines the probability that the random number from the smallest one to the number of breakage is below the unbreakage time; ,
This set ratio life non-satisfaction investigation procedure (G23A) is changed to a Weibull distribution with a life in which the set ratio is sequentially changed every time from a predetermined shortest life shorter than the unbreakage time to a predetermined longest long life. And the different life non-satisfaction investigation procedure (G24A),
From the relationship between the life obtained by this different life non-satisfaction investigation procedure (G24A) and the probability of being less than the unbreakage time, the life of which the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% is read. The life upper limit reading procedure (G25A) determined as the upper limit of the life of the product lot is executed.

  By the combined use of the first and second methods, both the lower limit and the upper limit of the life of the test object lot can be obtained. When the first and second methods are used in combination, common processing, for example, input processing, can be performed only once, without being repeated.

The lifetime estimation apparatus from the second lifetime termination test in the present invention is an apparatus that estimates the upper limit of the lifetime.
That is, the life estimation device from the second life cutoff test is
If the test object consisting of a bearing or other mechanical part or test piece is placed in the specified operating environment conditions and the test continues without damage until the target time, which is the cutoff time, the required life for the life with the specified reliability can be obtained. This is a device that estimates the upper limit of the life of a lot of test target products from the unbreakage time during which the test is continued and the part of the test target product is damaged and the remaining target products are not damaged. And
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)
On the screen of the above display device, as input information, the value of the Weibull slope of the product under test, the number of tests of the product under test, the number of undamaged or broken items that are the number of undamaged products under test, and the time of unbreakage Prompting screen output means (7EA) for performing display prompting,
Responsive to the input of the execution command, a life upper limit estimating calculating means (22A) is provided for performing an operation for estimating the upper limit of the life and outputting the operation result to the display screen.

The lifetime upper limit calculating means (22A)
A random number generator (23A) that generates a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to an unbreakage time, for the number of test pieces, and uses the value of the Weibull slope of the input information for the Weibull distribution;
Random number analysis means (24A) for checking whether or not the random number from the smallest one to the number of breakage among the Weibull random numbers for the generated test number is equal to or less than the unbreakage time;
The set ratio life non-satisfaction investigation means (in which the random number generation means (23A) and the random number analysis means (24A) are repeated a set number of times to determine the probability that the random number from the smallest one to the number of breakage will be less than the unbreakage time ( 25A)
This set ratio life non-satisfaction investigation means (25A) is processed into a Weibull distribution having a life in which the set ratio is sequentially changed every time from a predetermined shortest life shorter than the breakage time to a predetermined longest longevity. The different life non-satisfaction investigation means (26A) to be repeated,
From the relationship between the life obtained by the different life non-satisfaction investigation means (26A) and the probability of being less than the unbreakage time, the life of which the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% is read. Life upper limit reading means (29A) for determining the upper limit of the life of a product lot,
A reading result output means (28A) for causing the display device (2) to output the upper limit of the life read by the means (29A);
Is provided.

  The apparatus for estimating the life from the life censoring test of this configuration implements the second life estimation method of the present invention, and can easily and quickly obtain the upper limit of the life from the unbreakage time and is highly reliable. Even if it is not an expert, the upper limit of a lifetime can be calculated | required appropriately from unbreakage time.

The lifetime estimation program from the second lifetime termination test in the present invention is a program for estimating the upper limit of the lifetime.
That is, the lifetime estimation program from the second lifetime censoring test is
A program executable on a computer,
On the screen of the display device, input the value of the Weibull slope of the product under test, the number of test products under test, the number of undamaged or damaged products, the number of undamaged products under test, and the time of unbreakage. A prompt screen output procedure (F1A) for prompting display;
A lifetime upper limit calculation procedure (F2A) that calculates the upper limit of the lifetime and displays it on the screen of the display device in response to the execution command.

The above life upper limit calculation procedure (F1A)
A random number generation procedure (G21A) using a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to the unbreakage time, for a test number, and using the Weibull slope value of the input information for the Weibull distribution; ,
A random number analysis procedure (G22A) for checking whether or not the random number from the smallest to the number of breakage among the Weibull random numbers for the generated number of tests is equal to or less than the unbreakage time;
The above-mentioned random number generation procedure (G21A) and the above random number analysis procedure (G22A) are repeated a set number of times, and the set ratio life non-satisfaction investigation procedure (G23A) for examining the probability that the random number from the smallest one to the number of breakage will be less than the above-mentioned unbreakage time When,
For the Weibull distribution having a life in which the set ratio is sequentially changed from the predetermined shortest life shorter than the breakage time to the predetermined longest life that is gradually longer than the predetermined maximum life Repeat different life non-satisfaction investigation procedure (G24A),
From the relationship between the life obtained by this different life non-satisfaction investigation procedure (G24A) and the probability of being less than the unbreakage time, the life of which the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% is read. And a life upper limit reading procedure (G25A) determined as the upper limit of the life of the product lot.

  The lifetime estimation program from the lifetime censoring test of this configuration is used to implement the second lifetime estimation method of the present invention, and the upper limit of the lifetime can be easily and quickly determined from the unbreakage time and is highly reliable. Even if it is not an expert, a life level can be calculated | required appropriately from unbreakage time.

According to the life estimation method, apparatus, and program from the first life censoring test according to the present invention, weibull random numbers with various lifetimes are generated for the number of tests, and the remaining random numbers excluding the number of breakages are used to determine the life distribution. If there is, the probability distribution indicating whether the probability of non-breakage is high until the non-breakage time is obtained, so that the life level of the product to be tested can be obtained with high reliability from the non-breakage time. In addition, the computer inputs the Weibull slope value of the product under test, the number of tests of the product under test, the number of undamaged or broken items, which is the number of untested products under test, and the time of unbreakage. Thus, since the life level is output, the life level can be easily and quickly obtained from the unbreakage time without requiring skill.
According to the life estimation method, apparatus, and program from the second life censoring test of the present invention, weibull random numbers with various lifetimes are generated for the number of tests. Since the probability distribution that the probability that the life of the test object is less than the unbreakage time is high is obtained, the upper limit of the life of the test object can be obtained with high reliability from the unbreakage time. In addition, the computer inputs the Weibull slope value of the product under test, the number of tests of the product under test, the number of undamaged or broken items, which is the number of untested products under test, and the time of unbreakage. Thus, since the upper limit of the life is output, the upper limit of the life can be obtained from the unbreakage time easily and quickly without requiring skill.

A first embodiment of the present invention will be described. The service life estimation method based on the service life endurance test is based on the service endurance test that determines that the required service life will be satisfied if the bearing is placed in the specified operating environment conditions and the test continues without damage until the target end time. This is a method for estimating the life of a lot of a test target product from an unbreakage time during which the test is continued without damaging at least a part of the test target product when the test is continued after a certain period of time. Note that the present invention can also be applied to the case of estimating the cut-off time in a cut-off test of another mechanical part of a bearing or a test piece of the bearing or other mechanical part.
The life estimation method from the life cut-off test of this embodiment is applied to, for example, a test in which a part of the bearings in one lot is extracted to perform a cut-off life test and the life of the lot is confirmed. .

Hereinafter, this embodiment will be described with reference to the drawings. The life estimation method from the life cutoff test is performed by causing the computer 1 shown in FIG. 1 to execute the life estimation program 21 from the life cutoff test. 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 life estimation program 21 constitute a life estimation device in which each function achievement means is shown as a block in FIG. The configuration of the lifetime estimation apparatus shown in FIG.
The life estimation program 21 is a program that can be executed by the computer 1 and has the procedures shown in the flowcharts of FIGS. 4 and 5. The contents of this figure will be described later.

  As shown in FIG. 2, this life estimation method includes an input process E1 for inputting predetermined information to the computer 1 and a computer calculation process E2 for performing calculation processing on the computer 1 and outputting a calculation result. .

In the input process E1, as shown in FIG. 7, 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. On this screen, the display prompts the user to input the value of the Weibull slope of the bearing to be subjected to the abort test, the number of tests at the start of the test, the number of unbroken parts, and the unbreakage time as input information. As the Weibull slope value, the actual value of the bearing of the model number to be tested is input.
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.

  In the computer calculation processing step E2 of FIG. 3, the life which is the life level of the lot including the test object is calculated from the input unbreakage time, unbreakage number, etc., and the calculation result is output as shown in FIG. It is displayed on the screen 2b. That is, at least the life calculation result can be output as shown in the figure.

  The life estimation program 21 shown in FIG. 1 is a program that can be executed by the computer 1 and includes procedures shown in flowcharts in FIGS. 4 and 5. As shown in FIG. 4, the life estimation program 21 includes a prompt screen output procedure F1 and a life calculation procedure F2, and the prompt screen output procedure F1 outputs the input screen 2a described above together with FIG. When the input information is input from the input means 3 to the input screen 2a and an execution command is input from the input means 3 by clicking the OK key on the input screen 2a, the life calculation procedure F2 is executed. The The process of inputting to the input screen 2a in the figure is the input process A1 in FIG. 3, and the computer calculation process E2 in the figure is a process of executing the life calculation procedure F2 in FIG.

  The life calculation procedure F2 is configured by each procedure shown in the flowchart in FIG. In this flowchart, a specific processing example for each of the procedures G21 to G26 is also shown.

  For the sake of easy understanding, before explaining the flowchart of FIG. 5, the processing limited to the case where all of the processes in the flowchart of FIG.

FIG. 6 shows a procedure for calculating the lifetime from the intermediate results of the uncompleted abortion test that has been continued. Consider a situation where a test is being performed to guarantee 1000 hours with a target quality of L10 life. Since the delivery time was 3 months (2100 hours), 6 test pieces were prepared, and a censoring test with a total censoring time of 1860 hours was started.
The test was continued without causing any damage, and the result that the target quality could be guaranteed because the total cut-off time passed 1860 hours. Although the test may be terminated, it is not the situation that the testing machine is used immediately for other investigations, so the test was continued for the purpose of grasping the life level of the product.

  Now, the test is continued and 3000 hours have passed, and the life that can be said from the situation where all the six test pieces are not damaged is estimated. First, six random numbers are generated from a Weibull distribution (Weibull slope is set to 1.85 from the actual results) having an L10 life of 1/10 times 3000 (= 300 hours), which is the current total unbreakage time (procedure) G21 ′), it is checked whether all six random numbers obtained are over 3000 hours (procedure G22 ′).

  Next, this operation is repeated 5000 times, and it is examined how many times in 5000 times the situation exceeding 3000 hours occurs (G23 ′). These correspond to investigating the probability that all the specimens will be unbroken for 3000 hours when 6 specimens having a Weibull slope of 1.85 and an L10 life of 300 hours are tested. .

Further, this probability is investigated with a Weibull distribution having an L10 life of 1/10, 2/10, 3/10 ... 11/10 ... 200/10 times longer than 3000 hours (G24 '). With respect to the probabilities thus obtained, FIG. 9 (A) shows the probability that the horizontal axis is the L10 life and the vertical axis is the total number of unbroken 3000 hours. This figure shows what kind of life distribution has a high probability of non-peeling for 3000 hours. The life distribution in which the data for which no peeling is completed for 3000 hours can be obtained only at a frequency of 10% is the L10 life of 1530 hours. Therefore, when data that is not peeled off for 3000 hours is obtained, it can be said that the L10 life is 15630 hours or more with a probability of 90%. Thus, the time corresponding to the occurrence probability of 10% (the value obtained by subtracting the reliability of 90% of L10 life from 100%) is read, and the L10 life of the lot of the test object is read as 15630 hours or more. (G25 ').
The above is the method for calculating the lifetime from the intermediate result of the total number truncation test.

  Next, the life that can be said from the life cutoff test including the case where breakage occurs will be described. As shown in the item of interpretation (2) of the censoring test in Table 1 above, it is assumed that even if the crushing state of the censoring test is complicated, the life level of the test piece is to be confirmed. For example, a test of 10 test pieces was performed, and two of the test pieces were damaged, but the other eight test pieces were not damaged, so that the target quality could be satisfied. It is assumed that the user wants to confirm the life level of the test piece.

Here, a method for estimating the service life from a situation in which a small number of bearings are damaged in the total test bearing will be described. The flowchart of FIG. 5 shows the calculation procedure of the life that can be said from the life truncation test including the case where damage occurs.
In the flowchart of FIG. 5, description will be made with reference to an annotation portion showing a specific example. The basic procedure is the same as the procedure for estimating the lifetime from the result of the total number abort test described above with reference to FIG.

In order to simplify the description, specific examples will be given below. Consider a situation in which a test is performed to ensure that the target quality is 1000 hours with an L10 life. Since the delivery time was 5 months (3600 hours), six test pieces were prepared, and a censoring test with a total censoring time of 1860 hours was started. While continuing the test, one test piece was damaged in 1000 hours. 1000 hours is shorter than the set value 1860 hours of the total truncation time, which is the criterion for canceling the test, but exceeded the criterion for canceling the test (reference time for canceling the test if the required quality is not satisfied), which is 412 hours. The test was continued. In order to satisfy the target quality, the cut-off time of the remaining test piece is 2626 hours, but there is no problem in terms of delivery time. As a result of continuing the test, since the remaining test piece did not break in 2626 hours, the product could meet the target quality. Although the test may be terminated, it is not the situation that the testing machine is used immediately for other investigations, so the test was continued for the purpose of grasping the life level of the product.
Note that the abort time when all the parts are not damaged, the abort time when partly damaged, and the test stop reference time can be estimated by an appropriate method, but the description is omitted here.

Now, the test is continued for 3000 hours, and the life that can be said from the situation that 5 out of 6 test pieces are not damaged is estimated.
First, the Weibull distribution with an L10 life of an appropriate set ratio, for example, 1/10 times (= 300 hours), compared to the current total unbreakage time of 3000 hours (Weibull slope is 1.85 set from actual results) 6 random numbers are generated (step G21).
Next, the obtained six random numbers are rearranged in ascending order, and it is examined whether five data other than the smallest random number are 3000 hours or more (procedure G22).

Next, this operation is repeated a set number of times (5000 times in this example), and it is checked how many times in 5000 times a situation exceeding 3000 hours occurs (procedure G23).
When a test is performed on six test pieces having a Weibull slope of 1.85 and an L10 life of 300 hours, the probability that the five test specimens other than the test specimen having the shortest life will be 3000. Corresponds to investigating whether time is not damaged.

  Further, this probability is investigated with a Weibull distribution having an L10 life of 1/10, 2/10, 3/10 ... 11/10 ... 200/10 times 3000 hours (procedure G24). This is the calculation range when the shortest life that can be calculated by this method is set to 1/10 of 3000 hours and the longest life is set to 200/10 times 3000 hours.

  As shown in FIG. 10 (), the probability obtained in the procedure G23 is plotted with the abscissa indicating the L10 life and the ordinate indicating the probability of 5 non-damaged items in 3000 hours. B). This figure shows what kind of life distribution the five specimens other than the test bearing with the shortest life are likely to be unseparated for 3000 hours. The life distribution in which only 5 out of 3000 pieces of data can be obtained at a frequency of 10% (the value obtained by subtracting 90%, which is the reliability of the L10 life from 100%), is that the L10 life is 1099 hours. Is.

  Therefore, it can be said that the situation in which all five test pieces other than the test piece having the shortest life are not peeled for 3000 hours is a situation in which the L10 life is 1099 hours or more with a probability of 90%. As described above, from the relationship between the life distribution and the probability that the remaining random numbers are not peeled over the unbreakage time, the time of (100%)-(reliability) is read, and the life corresponding to the level of the bearing lot is obtained. (Procedure G25). The life thus read is displayed on the output screen 2b of the display device 2 (procedure G26).

  The above is the method for calculating the life from the situation in which a small number of test bearings in the total test bearing are damaged. Here, the test in this situation can be considered as a situation in which the total number of 1000 hours has not been damaged, or it can be considered that 5 out of 6 specimens have not been damaged for 3000 hours. The question remains as to whether or not In such a case, it may be set as appropriate which one is adopted. For example, it is possible to employ a method in which the lifetime is estimated in two situations and the lifetime is estimated to be long. Table 2 shows the life that can be estimated when 6 specimens have not been peeled for 1000, 2000, and 3000 hours, and the life that can be estimated when 5 out of 6 specimens have not been peeled for 1000, 2000, and 3000 hours. Indicates.

  It can be seen that the life that can be calculated when the total unbreakage time of 6 pieces is 1000 hours is shorter than the life that can be calculated when 5 pieces of 6 pieces have not been peeled for 3000 hours. In this case, for example, the one whose life is estimated to be long (the life that can be estimated when 5 out of 6 specimens have not been peeled for 3000 hours) is adopted.

  Table 3 shows another example of the test result.

  This test result can be interpreted in three ways: (1) 6 out of 6 are over 3050 hours, (2) 5 out of 6 are over 4000 hours, and (3) 4 out of 6 are over 6000 hours. is there. Table 4 shows the results of calculation with each interpretation.

  As a result of calculation, it is possible to determine the longest life side by interpreting that 4 out of 6 are 6000 hours or more. Therefore, it can be said from this test result that the L10 life is 1742 hours or more.

  The above description of the life estimation program 21 shown in FIG. 4 and FIG. 5 has been specifically described by taking numerical values as an example, but this life estimation program 21 is configured by the following procedure.

The life estimation program 21 from the life cutoff test of this embodiment is:
A program executable on a computer,
On the screen of the display device, input the value of the Weibull slope of the test object, the number of test objects, the number of undamaged or damaged items, and the unbreakage time as input information. Prompt screen output procedure (G1) for prompting display,
A life calculation procedure (G2) for calculating the life and displaying it on the screen of the display device in response to the execution command.

The life calculation procedure (G2) is
A random number generation procedure (G21) using a Weibull random number, which is a random number according to the Weibull distribution having a set life with respect to the unbreakage time, for the test number, and using the value of the Weibull slope of the input information for the Weibull distribution;
Random number analysis procedure (G22) for checking whether or not the remaining random numbers obtained by removing the random numbers corresponding to the damaged number in order from the shortest number among the Weibull random numbers corresponding to the generated test number are equal to or longer than the undamaged time;
The above-described random number generation procedure (G1) and the above random number analysis procedure (G22) are repeated a set number of times, and the set ratio life satisfaction investigation procedure (G23) for examining the probability of being in the unbreakage time or more examined in the above random number analysis procedure at each iteration. When,
This set ratio life satisfaction check procedure (G23) is repeated for a Weibull distribution having a life in which the set ratio is sequentially changed from a predetermined shortest life shorter than the breakage time to a predetermined longest life that is gradually longer. Different life satisfaction investigation procedure (G24),
From the relationship between the life obtained by this different life satisfaction investigation procedure (G24) and the probability that it is longer than the unbreakage time, read the life when the probability of occurrence is a value obtained by subtracting the predetermined reliability from 100%, and then the product to be tested The life reading procedure (G25) determined as the life of the lot of
Life output procedure (G26).

  The set proportion life satisfaction survey procedure (G23) includes the procedure (G231) of repeating the random number generation procedure (G1) and the random number analysis procedure (G22) a set number of times, and the random number analysis procedure in each of the repetitions. This is a probability investigation procedure (G232) for examining the probability of being at or above the breakage time.

  The different life satisfaction investigation procedure (G24) includes a repetition procedure (G241) for repeating the set percentage life satisfaction investigation procedure (G23) until a predetermined maximum life is reached, and a ratio change procedure for sequentially changing the set ratio for each repetition. (G242). Note that the initial value of the set ratio (1/10 in the above specific example) is determined in step G21.

Details of the random number generation procedure (procedure B21) will be described. This procedure B21 includes a Weibull distribution identification procedure (G211) and a procedure (G212) for generating a Weibull random number according to the identified Weibull distribution. In the Weibull distribution specifying procedure (G211), the next Weibull distribution is specified.
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. As a reference, FIG. 11 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 mathematical 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 the Weibull random number (procedure G212) 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. 12, 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) within 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.

With reference to FIG. 2, a life estimation apparatus based on the life cutoff test will be described. This life estimation device is a device that estimates the life of a lot of a test target product from an unbreakage time during which the test is continued without damaging at least a part of the test target product in a censoring test. A computer 1, a display device 2 that displays the output of the computer 1 on a screen, and an input device 3 that inputs the computer 1 are provided.
The computer 1 displays, on the screen of the display device 2, as input information, the value of the Weibull slope of the test object, the number of tests of the test object, the number of undamaged or broken objects, which is the number of undamaged test objects, Prompting screen output means 7E for performing display prompting for the input of breakage time;
Responsive to the input of the execution command, it includes life estimation calculation means 22 that performs a calculation for estimating the life and outputs the calculation result to the display screen.

  The life estimation calculation means 22 includes a random number generation means 23, a random number analysis means 24, a set rate life satisfaction investigation means 25, a different life satisfaction investigation means 26, a life reading means 27, and a reading result output means 28. .

The random number generation means 23 is a means for generating Weibull random numbers, which are random numbers according to the Weibull distribution having a set life with respect to the unbreakage time, for the number of tests, and performs the processing described in the procedure G21 in FIG. The Weibull distribution uses the value of the Weibull slope of the input information.
The random number generating means 23 includes a Weibull distribution specifying unit 23a and a random number generating unit 23b. The Weibull distribution specifying unit 23a performs the process described in the procedure G211 in FIG. 5, and the random number generation unit 23b performs the process described in the procedure G212D in FIG.

  The random number analyzing means 24 is a means for checking whether or not the remaining random numbers obtained by removing the random numbers corresponding to the damaged number in order from the shortest one among the Weibull random numbers corresponding to the generated test number is equal to or longer than the undamaged time. The process described in the procedure G22 in the flowchart of FIG.

  The set ratio life satisfaction check means 25 repeats the processes of the random number generation means 23 and the random number analysis means 24 for a set number of times, and checks the probability of being equal to or greater than the unbreakage time checked by the random number analysis means 24 in each repetition. The process described in the procedure G23 in the flowchart of FIG. 5 is performed.

  The different life satisfaction checking means 26 is a life in which the set ratio is sequentially changed every time the processing of the set percentage life satisfaction checking means 26 is repeated from a predetermined shortest life shorter than the breakage time to a predetermined longest life gradually longer. The process described in step G24 in the flowchart of FIG. 5 is performed.

  The life reading means 27 calculates a life whose occurrence probability is a value obtained by subtracting the predetermined reliability from 100%, based on the relationship between the life obtained by the processing of the different life satisfaction surveying means 26 and the probability that it is longer than the unbreakage time. This is a means for reading and determining the life of the lot of the test object, and performs the process described in the procedure G25 in the flowchart of FIG.

  The read result output means 28 is a means for causing the display device 2 to output the life read by the life read means 27, and performs the processing described in the procedure G26 in the flowchart of FIG.

A second embodiment of the lifetime estimation method, apparatus, and program according to the present invention will be described with reference to FIGS. In this embodiment, the matters other than those specifically described are the same as those in the first embodiment shown in FIGS.
This embodiment relates to a method of determining from the result of the life cutoff test that the upper limit of the life, that is, no longer life can be expected from the test result.

As with the first embodiment, the life estimation method from the life cutoff test of this embodiment is also required if the bearing is placed in a predetermined use environment condition and the test continues without being broken until the cutoff time, which is the target time. It is applied in the censoring test that judges that the life is satisfied.
However, this method is applied when a part of the test target product is damaged and the test is continued without damaging the remaining target product, and the upper limit of the life of the test target lot is estimated from the non-breakage time. Is the method.
The present invention can also be applied to the case of estimating the cut-off time in a cut-off test of another mechanical part of a bearing or a test piece of a bearing or other mechanical part.
The life estimation method from the life cut-off test of this embodiment is also applied to, for example, a test in which a part of bearings is extracted from a lot of bearings, a cut-off life test is performed, and the life of the lot is confirmed. .

Hereinafter, this embodiment will be described with reference to the drawings. The life estimation method from the life end test is performed by causing the computer 1 shown in FIG. 13 to execute the life estimation program 21A from the life end test. 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 life estimation program 21A constitute a life estimation device in which each function achievement means is shown as a block in FIG.
The life estimation program 21A is a program that can be executed by the computer 1, and has the procedures shown in the flowcharts of FIGS.
The life estimation program 21A in this embodiment is executed by the same computer 1 as the life estimation program 21 in the first embodiment. For example, after the life calculation in the first embodiment, the upper limit of the life in this embodiment is calculated. It is desirable to perform the calculation. That is, the life estimation method from the life end test according to the first embodiment and the life estimation method from the life end test according to the second embodiment can be used in combination. In this case, among the input items of the operator, duplicate input items can be performed at a time, and the input process can also be used. By using the first and second life estimation methods together, it is possible to know both the lower limit value and the upper limit value of the life of the lot of the test object.

  As shown in FIG. 15, this life estimation method includes an input process E1A for inputting predetermined information to the computer 1 and a computer arithmetic processing process E2A for performing arithmetic processing on the computer 1 and outputting the arithmetic result. .

  In the input process E1, as shown in FIG. 19, an input screen 2aA for prompting input of predetermined input information is displayed on the display device 2 by the output of the computer 1. Since this input screen 2aA and the input performed on this input screen are the same as those in the first embodiment (FIGS. 1 to 12), a duplicate description is omitted. However, in this embodiment, the estimated lifetime is the upper limit of the L10 lifetime and the L50 lifetime, and a message to that effect is displayed.

  In the computer calculation process E2A in FIG. 15, the upper limit of the life of the lot including the test object is calculated from the input unbreakage time, the number of unbroken pieces, etc. Then, the calculation result is displayed on the output screen 2bA as shown in FIG. That is, the upper limit value of the lot life is output as shown in FIG. This upper limit value is output for each of the L10 life and the L50 life.

  The life estimation program 21A shown in FIG. 13 is a program that can be executed by the computer 1, and includes procedures shown in flowcharts in FIGS. As shown in FIG. 16, the life estimation program 21A includes a prompt screen output procedure F1A and a life upper limit calculation procedure F2A, and the prompt screen output procedure F1 outputs the input screen 2aA described above with FIG. When the input information is input to the input screen 2aA from the input unit 3 and an execution command is input from the input unit 3 by clicking the OK key on the input screen 2aA, the life upper limit calculation procedure F2A is executed. Is done. The input process E1A in FIG. 15 is a process input to the input screen 2aA in FIG. 15, and the computer calculation process E2A in FIG. 15 is a process for executing the life upper limit calculation procedure F2A in FIG.

  The life upper limit calculation procedure F2A is configured by the respective procedures shown in the flowchart in FIG. In this flowchart, a specific processing example for each of the procedures G21A to G26A is also shown.

  For the sake of easy understanding, before describing the flowchart of FIG. 17, the processing limited to the case of one breakage in the processing of the flowchart of FIG.

FIG. 18 shows a procedure for calculating the upper limit of the life from the intermediate results of the unbroken abort test that has been continued.
Now, the life span which can be said from the situation in which the test of six test pieces has continued for 3000 hours and one of the test pieces has been damaged is estimated. According to the first embodiment, it is required that at least the life that can be said when all six test pieces are 3000 hours or longer (the lower limit L10 life that can be guaranteed with a probability of 90%) is 1530 hours or longer. Can do.
Hereinafter, the upper limit of the lifetime is obtained together with FIG. At present, one of the six specimens is damaged in 3000 hours, but the Weibull distribution has an L10 life of 0.5 times the failure time (= 1500 hours). .85), 6 random numbers are generated (G21A ′).
The obtained six random numbers are rearranged in ascending order, and it is checked whether the smallest random number is 3000 hours or less (G22A).
Next, this operation is repeated 5000 times, and the probability of how often the situation in which the smallest random number is 3000 hours or less occurs in 5000 times is examined (G23A ′).
These are investigations of the probability that data with less than 3000 hours will be obtained with the shortest life test piece when 6 test pieces with Weibull slope 1.85 and L10 life 1500 hours are tested. Corresponding to what you are doing.
Further, this probability is investigated with a Weibull distribution having an L0 life of 0.5, 1, 1.5... 25 times longer than 3000 hours (G24A ′).
The cumulative probability distribution is obtained from the result of this investigation, and the time when the probability distribution becomes 10% is read (G25A ′). That is, FIG. 21B shows the probability that the L10 life is plotted on the horizontal axis and the life of the shortest test piece is 3000 hours or less on the vertical axis. This figure shows what kind of life distribution the probability that the life of the shortest specimen is 3000 hours or less is high.
From this figure, it is a rare case of 10% that one of the six test pieces from the test piece having a life distribution of L316 life of 7316 hours is damaged for 3000 hours or less. Accordingly, it can be said that the life of the test piece group from which data of 3000 test pieces or less is obtained for one of the 6 test pieces is 90%, and the L10 life is 7316 hours or less. From this, it can be said that the range of lifetime is 1530 hours or more and 7316 hours or less.

The above is a method for calculating the upper limit of the lifetime when one or more pieces of damaged data are obtained. If this method is applied, the upper limit of the life can be calculated even when two pieces of damaged data are obtained. These calculations are performed using the life estimation programs 21 and 21A, and the input and output results are shown in FIGS. Here, as for the Weibull slope, it is desirable to input an actual value obtained from an actual test having 10 or more test pieces.
As is apparent from the above, according to the method for calculating the upper limit of the life, it is possible to judge from the result that the life beyond the expected value cannot be expected.

  Next, the details of the life upper limit calculation procedure F2A, that is, the procedure for calculating the upper limit of the life when one or more pieces of damaged data are obtained will be described with reference to FIG. Description will be made with reference to an annotation portion showing a specific example in the flowchart of FIG. The basic procedure is the same as that shown in FIG.

Although the description overlaps with FIG. 17, the example described with reference to FIG. 18 is used as a specific example. The numerical values in parentheses in the following description are specific examples.
Now, n (six) test pieces continue to be tested, a predetermined unbreakage time (3000 hours) has passed, and m (one) test piece can be said to have been damaged. Estimate the range.

  First, the process of step G21A (random number generation procedure) will be described. The above-mentioned unbreakage time, which is the breakage time when m pieces of n pieces (1 piece out of 6 pieces) have been broken within a predetermined unbreakage time T (3000 hours). Weibull distribution with L10 life (0.5 times = (1500 hours)) with respect to T (Weibull slope is set from results and generates n (6) random numbers from 1.85. The method of specifying the Weibull distribution (G211A) and random number generation (G212A) is the same as the method described in the first embodiment.

  Next, in the random number analysis procedure (G22A), the obtained random numbers n (6) are rearranged in ascending order, and random numbers from the smallest to the number m of damaged pieces are equal to or shorter than the unbreakage time T (3000 hours). Check whether or not. When the number of breakage m is 1, it is checked whether the smallest random number is 3000 hours or less.

Next, in the set ratio life non-satisfaction investigation procedure (G23A), the random number generation procedure (G21A) and the random number analysis procedure (G22A) are repeated a set number of times (5000 times) (G231A). It is examined whether the random number is equal to or less than the unbreakage time T (G232A). When the number of breakage m is 1, the probability of how often the situation where the smallest random number is 3000 hours or less occurs in 5000 times is examined.
These are the Weibull slope 1.85 and the L10 life is (set ratio) × T (1500 hours). This corresponds to the investigation of the probability that the data of the unbreakage time T (3000 hours) or less appears in the test specimens up to.

  Furthermore, in the different life satisfaction investigation procedure (G24A), the probability investigation (G23A) is performed by 0.5, 1 or more than the unbreakage time T (breakage time when m breakage occurs) (3000 hours). 1.5 ... Weibull distribution with 25 times the L0 life.

In the life upper limit reading procedure (G25A), the cumulative probability distribution is obtained from the result of the different life satisfaction survey procedure (G24A), and the occurrence probability is a value obtained by subtracting the predetermined reliability (90%) from 100% (10%). Read the lifetime which becomes a probability distribution (G25A).
That is, FIG. 21B shows the probability that the abscissa indicates the L10 life and the ordinate indicates the probability that the life of the shortest test piece is less than the unbreakage time T (3000 hours). This figure shows what kind of life distribution has a high probability that the life of the test piece from the shortest to the number m (1) of damage will be less than the unbreakage time T (3000 hours). ing.
From this figure, it is a rare case of 10% that one of the six test pieces from the test piece having a life distribution of L316 life of 7316 hours is damaged for 3000 hours or less. Accordingly, it can be said that the life of the test piece group from which data of 3000 test pieces or less is obtained for one of the 6 test pieces is 90%, and the L10 life is 7316 hours or less. From this, it can be said that the range of lifetime is 1530 hours or more and 7316 hours or less.

In the life upper limit output procedure (G26A), the upper limit of the life read in the life upper limit reading procedure (G25A) is displayed on the output screen 2b of the display device 2.
The above is the life upper limit calculation procedure F2A, that is, the procedure for calculating the upper limit of the life when one or more pieces of damaged data are obtained.

  In this way, weibull random numbers with various lifetimes are generated for the number of test pieces, and what kind of life distribution is there is a high probability that the life of the product under test from the shortest to the number of breakage will be less than the unbreakage time? Thus, the upper limit of the life of the test object can be obtained with high reliability from the unbreakage time. In addition, the above processing is a computer simulation, and as input information to the computer, the value of the Weibull slope of the test object, the number of tests of the test object, the number of undamaged items or the number of breakage that is the number of undamaged test items Since the life level is output only by inputting the unbreakage time, the upper limit of the life can be easily and quickly obtained from the unbreakage time without requiring skill.

  The above description of the life estimation program 21A shown in FIG. 17 and FIG. 18 has been specifically described by taking numerical values as an example, but this life estimation program 21A is configured by the following procedure.

The life estimation program 21A from the life end-of-life test of this embodiment is a program that can execute the upper limit of the life on a computer,
On the screen of the display device, input the value of the Weibull slope of the product under test, the number of test products under test, the number of undamaged or damaged products, the number of undamaged products under test, and the time of unbreakage. A prompt screen output procedure (G1A) for performing a prompt display;
And a life upper limit calculation procedure (G2A) for calculating the life and displaying it on the screen of the display device in response to the execution command.

The above life upper limit calculation procedure (G2A)
A random number generation procedure (G21A) using a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to an unbreakage time, for the test number, and using the value of the Weibull slope of the input information for the Weibull distribution;
A random number analysis procedure (G22A) for examining whether or not the random number from the smallest to the number of breakage among the Weibull random numbers for the generated number of tests is less than the unbreakage time;
A set ratio life non-satisfaction investigation procedure (G23A) for repeating the random number generation procedure and the random number analysis procedure a set number of times, and examining the probability that the random number from the smallest one to the number of breakage will be less than the unbreakage time;
This set ratio life non-satisfaction investigation procedure (G23A) is changed to a Weibull distribution with a life in which the set ratio is sequentially changed every time from a predetermined shortest life shorter than the unbreakage time to a predetermined longest long life. The different life non-satisfaction investigation procedure (24A)
From the relationship between the life obtained by this different life non-satisfaction investigation procedure and the probability of being less than the unbreakage time, the life of which the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% is read to determine the lot of the test object. Life upper limit reading procedure (G25A) determined as the upper limit of the life of
Life upper limit output procedure (G26A).

  The set ratio life non-satisfaction investigation procedure (G23A) was examined by the procedure (G231A) of repeating the random number generation procedure (G21A) and the random number analysis procedure (G22A) a set number of times, and the random number analysis procedure in each of these repetitions. This is a probability investigation procedure (G232A) for examining the probability that the random number from the smallest one to the number of breakage will be less than the unbreakage time.

  The different life non-satisfaction investigation procedure (G24A) includes a repetition procedure (G241A) in which the set rate life non-satisfaction investigation procedure (G23A) is repeated until a predetermined maximum life is reached, and a rate at which the set ratio is sequentially changed for each repetition. The change procedure (G242A) is performed. The initial value of the set ratio (5/10 in the above specific example) is determined by the random number generation procedure G21A.

  The life estimation program 21A from the life end-of-life test of this configuration is used for the execution of the second life estimation method of the present invention, and the upper limit of the life can be easily and quickly determined from the unbreakage time and is reliable. It can be high, and even if it is not an expert, a life level can be calculated | required appropriately from unbreakage time.

With reference to FIG. 14, a life estimation apparatus from the second life cutoff test will be described. The life estimation device from this life cutoff test is a device that estimates the upper limit of the life.
This life estimation device estimates the upper limit of the life of a lot of test target products from the unbreakage time during which the test is continued without causing damage to some of the test target products and the remaining target products in the censoring test. A computer 1 that is an arithmetic processing unit, a display device 2 that displays an output of the computer 1 on a screen, and an input unit 3 that inputs the computer 1 are provided.

The computer 1 displays, on the screen of the display device 2, as input information, the value of the Weibull slope of the test object, the number of tests of the test object, the number of undamaged or broken objects, which is the number of undamaged test objects, Prompt screen output means 7EA for performing display prompting for the input of breakage time;
Responsive to the input of the execution command, a life upper limit estimating calculating means 22A is provided for calculating the upper limit of the life and outputting the operation result to the display screen.

  The life upper limit estimation calculation means 22A includes a random number generation means 23A, a random number analysis means 24A, a set ratio life non-satisfaction investigation means 25A, a different life non-satisfaction investigation means 26A, a life reading means 27A, and a read result output means 28A. With.

The random number generation means 23A is means for generating Weibull random numbers, which are random numbers according to the Weibull distribution having a set life with respect to the unbreakage time, for the number of tests, and performs the processing described in the procedure G21A in FIG. The Weibull distribution uses the value of the Weibull slope of the input information.
The random number generating means 23 includes a Weibull distribution specifying unit 23Aa and a random number generating unit 23Ab. The Weibull distribution identification unit 23Aa performs the process described in the procedure G211A in FIG. 17, and the random number generation unit 23Ab performs the process described in the procedure G212A in FIG.

  The random number analysis means 24A is a means for checking whether or not the random number from the smallest to the number of breakage among the generated Weibull random numbers for the test number is less than the unbreakage time, and the procedure G22A in the flowchart of FIG. Perform the process described in.

  The set ratio life non-satisfaction investigation means 25A is a means for examining the probability that the random number from the smallest one to the number of breakage will be less than the unbreakage time by repeating the processes of the random number generation means 23A and the random number analysis means 24A a set number of times. Yes, the processing described in the procedure G23A in the flowchart of FIG. 17 is performed.

  The different life non-satisfaction investigation means 26A sequentially performs the above-described set ratios every time the processing of the set ratio life non-satisfaction investigation means 25A is repeated from a predetermined shortest life shorter than the unbreakage time to a predetermined longest life gradually longer. This is a means for repeating the Weibull distribution with the changed lifetime, and performs the processing described in the procedure G24A in the flowchart of FIG.

  In the life upper limit reading means 29A, the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% from the relationship between the life obtained by the processing of the different life non-satisfaction investigation means 26A and the probability of being less than the unbreakage time. A means for reading the life and determining the upper limit of the life of the lot of the test object, and performs the process described in the procedure G25A in the flowchart of FIG.

  The reading result output means A is means for outputting the upper limit of the life read by the life upper limit reading means 29A to the display device 2, and performs the processing described in the procedure G26A in the flowchart of FIG.

  According to the lifetime estimation apparatus from the lifetime censoring test of this configuration, the second lifetime estimation method in this embodiment can be implemented, and the upper limit of the lifetime can be easily and quickly determined from the unbreakage time, and the reliability Even if it is not an expert, the upper limit of a lifetime can be calculated | required appropriately from unbreakage time.

It is a schematic block diagram of the lifetime estimation apparatus which concerns on one Embodiment of this invention. It is a block diagram which shows the conceptual structure of the lifetime estimation apparatus. It is a schematic flowchart of the lifetime estimation method using the lifetime estimation apparatus. It is a general | schematic flowchart of the lifetime estimation program which implements the same lifetime estimation method. It is a flowchart which shows the detail of the lifetime calculation procedure in the program. It is a schematic flowchart which shows the example which applied the same flowchart to the case of a specific numerical value. It is explanatory drawing of the example of an input screen in the lifetime estimation apparatus of FIG. It is explanatory drawing of the example of an output screen in the lifetime estimation apparatus of FIG. (A) is a graph of an example of the Weibull distribution, and (B) is a graph showing the relationship between the life distribution and the probability that the whole is not damaged. (A) is a graph of an example of the Weibull distribution, and (B) is a graph showing the relationship between the life distribution and the probability that 5 out of 6 are not damaged. 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 a schematic block diagram of the lifetime estimation apparatus which concerns on 2nd Embodiment of this invention. It is a block diagram which shows the conceptual structure of the lifetime estimation apparatus. It is a schematic flowchart of the lifetime estimation method using the lifetime estimation apparatus. It is a general | schematic flowchart of the lifetime estimation program which implements the same lifetime estimation method. It is a flowchart which shows the detail of the lifetime calculation procedure in the program. It is a schematic flowchart which shows the example which applied the same flowchart to the case of a specific numerical value. It is explanatory drawing of the example of an input screen in the lifetime estimation apparatus of FIG. It is explanatory drawing of the example of an output screen in the lifetime estimation apparatus of FIG. The graph which shows the frequency by which a predetermined ratio lifetime becomes unsatisfactory, (B) is a graph which shows the relationship between a lifetime and a cumulative probability. It is a flowchart which shows the procedure of the conventional truncation and acceleration test.

Explanation of symbols

1 Computer (arithmetic processing means)
2 ... Display device 3 ... Input device 7E ... Prompt screen output means 21 ... Life estimation program 22 ... Life estimation calculation means 23 ... Random number generation means 24 ... Random number analysis means 25 ... Set ratio life investigation means 26 ... Different life fulfillment investigation means 27 ... Life reading means 28 ... Reading result output means 21A ... Life estimation program 7EA ... Prompt screen output means 22A ... Life upper limit estimation calculation means 23A ... Random number generation means 24A ... Random number analysis means 25A ... Set ratio life unsatisfaction investigation means 26A ... Unsatisfactory life fulfillment investigation means 27A ... Life reading means 28A ... Read result output means

Claims (8)

If the test object consisting of a bearing or other mechanical part or test piece is placed in the specified operating environment conditions and the test continues without damage until the target time, which is the cutoff time, the required life for the life with the specified reliability can be obtained. A method for estimating the life of a lot of a test object from an unbreakage time in which the test is continued without damaging at least a part of the test object in a censoring test judged to be satisfactory,
Input process to input the Weibull slope value of the test target product, the number of test target products, the number of undamaged or damaged items as the number of undamaged test target items, and the unbreakage time as input information to the computer When,
The computer includes a computer calculation process for calculating the life and displaying the calculation result on the screen of the display device,
As the above computer processing process,
A random number generation procedure using a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to an unbreakage time, for a test number, and using the value of the Weibull slope of the input information for the Weibull distribution,
Random number analysis procedure to check whether or not the remaining random numbers from the Weibull random number for the generated test number, in which the random number for the damaged number is removed in order from the shortest, is equal to or longer than the unbreakage time,
The above set random number generation procedure and the above random number analysis procedure are repeated a set number of times, and the set ratio life satisfaction survey procedure for examining the probability of being equal to or greater than the unbreakage time examined in the above random number analysis procedure in each of the repetitions;
This set percentage life satisfaction survey procedure is repeated for a Weibull distribution with a life in which the above set ratio is sequentially changed from a predetermined shortest life shorter than the failure time to a predetermined longest life that is gradually longer. Investigation procedure;
From the relationship between the life obtained by this different life satisfaction survey procedure and the probability that it is longer than the unbreakage time, read the life of which the probability of occurrence is a value obtained by subtracting the predetermined reliability from 100%. The life reading procedure determined as the life,
A lifetime estimation method from a lifetime censoring test. If the test object consisting of a bearing or other mechanical part or test piece is placed in the specified operating environment conditions and the test continues without damage until the target time, which is the cutoff time, the required life for the life with the specified reliability can be obtained. An apparatus for estimating the life of a lot of a test object from an unbreakage time in which the test is continued without damaging at least a part of the test object in a censoring test that is judged to be satisfactory,
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 above display device, as input information, the value of the Weibull slope of the product under test, the number of tests of the product under test, the number of undamaged or broken items that are the number of undamaged products under test, and the time of unbreakage Prompting screen output means for performing prompting display,
In response to the input of the execution instruction, the life estimation operation means for performing the operation of estimating the life and outputting the operation result to the display screen is provided.
The life estimation calculation means is
A random number generating means using a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to an unbreakage time, for a test number, and using the value of the Weibull slope of the input information for the Weibull distribution;
Random number analysis means for checking whether or not the remaining random numbers obtained by removing the random numbers corresponding to the damaged number in order from the shortest among the Weibull random numbers corresponding to the generated test number, are equal to or longer than the undamaged time,
A set ratio life satisfaction investigation means for repeating the process of the random number generation means and the random number analysis means for a set number of times, and examining the probability of being equal to or greater than the unbreakage time examined by the random number analysis means in each iteration of the repetition;
The process of the set ratio life satisfaction survey means is repeated for a Weibull distribution having a life in which the set ratio is sequentially changed from a predetermined shortest life shorter than the breakage time to a predetermined longest life that is gradually longer. A life satisfaction survey means;
From the relationship between the life obtained by this different life satisfaction survey means and the probability that it is longer than the unbreakage time, the occurrence probability is a value obtained by subtracting the predetermined reliability from 100%. A life reading means defined as a life;
Read result output means for outputting the life read by this means to the display device;
Life estimation device from life censoring test with A program executable on a computer,
On the screen of the display device, input the value of the Weibull slope of the product under test, the number of test products under test, the number of undamaged or damaged products, the number of undamaged products under test, and the time of unbreakage. Prompt screen output procedure for prompting display,
In response to the execution command, a life calculation procedure for calculating the life and displaying on the screen of the display device,
The above life calculation procedure is
A random number generation procedure using a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to an unbreakage time, for a test number, and using the value of the Weibull slope of the input information for the Weibull distribution,
Random number analysis procedure to check whether or not the remaining random numbers from the Weibull random number for the generated test number, in which the random number for the damaged number is removed in order from the shortest, is equal to or longer than the unbreakage time,
The above set random number generation procedure and the above random number analysis procedure are repeated a set number of times, and the set ratio life satisfaction survey procedure for examining the probability of being equal to or greater than the unbreakage time examined in the above random number analysis procedure in each of the repetitions;
This set percentage life satisfaction survey procedure is repeated for a Weibull distribution with a life in which the above set ratio is sequentially changed from a predetermined shortest life shorter than the failure time to a predetermined longest life that is gradually longer. Investigation procedure;
From the relationship between the life obtained by this different life satisfaction survey procedure and the probability that it is longer than the unbreakage time, read the life of which the probability of occurrence is a value obtained by subtracting the predetermined reliability from 100%. The life reading procedure determined as the life,
Life estimation program from the life censoring test including. If the test object consisting of a bearing or other mechanical part or test piece is placed in the specified operating environment conditions and the test continues without damage until the target time, which is the cutoff time, the required life for the life with the specified reliability can be obtained. In the censorship test that is judged to be satisfactory, a method is used to estimate the upper limit of the life of the lot under test from the unbreakage time during which the test continues without causing damage to some of the test subject and the rest of the test subject. There,
Inputs to the computer as input information the Weibull slope value of the product under test, the number of tests of the product under test, the number of undamaged or broken items that is the number of untested products under test, and the above-mentioned unbreakage time Process,
A computer operation process for causing the computer to calculate the upper limit of the lifetime and to display an operation result on a screen of a display device,
As the above computer processing process,
A random number generation procedure using a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to an unbreakage time, for a test number, and using the value of the Weibull slope of the input information for the Weibull distribution,
Random number analysis procedure to check whether the random number from the smallest to the number of breakage among the Weibull random number for the generated number of tests is less than the unbreakage time,
Repeat the above random number generation procedure and random number analysis procedure a set number of times, and set random life non-satisfaction investigation procedure to examine the probability that the random number from the smallest one to the number of breakage will be less than the above unbreakage time,
This set ratio life non-satisfaction investigation procedure is repeated for a Weibull distribution having a life in which the set ratio is sequentially changed from the predetermined shortest life shorter than the unbreakage time to the predetermined longest life gradually longer. A different life non-satisfaction investigation procedure,
From the relationship between the life obtained by this different life non-satisfaction investigation procedure and the probability of being less than the unbreakage time, the life of which the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% is read to determine the lot of the test object. The life upper limit reading procedure determined as the life upper limit of
A lifetime estimation method from a lifetime censoring test.   The life estimation method from the life truncation test according to claim 4, wherein the predetermined reliability subtracted from 100% is 90%. In Claim 1, including a method for estimating the upper limit of the life of a lot of the test object from an unbreakage time in which the test is continued without damaging a part of the test object and the remaining test object being damaged, As a method of estimating the upper limit, the computer includes a computer calculation process for calculating the upper limit of the lifetime and displaying the calculation result on the screen of the display device,
As the above computer processing process,
A random number generation procedure using a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to an unbreakage time, for a test number, and using the value of the Weibull slope of the input information for the Weibull distribution,
Random number analysis procedure to check whether the random number from the smallest to the number of breakage among the Weibull random number for the generated number of tests is less than the unbreakage time,
Repeat the above random number generation procedure and random number analysis procedure a set number of times, and set random life non-satisfaction investigation procedure to examine the probability that the random number from the smallest one to the number of breakage will be less than the above unbreakage time,
This set ratio life non-satisfaction investigation procedure is repeated for a Weibull distribution having a life in which the set ratio is sequentially changed from the predetermined shortest life shorter than the unbreakage time to the predetermined longest life gradually longer. A different life non-satisfaction investigation procedure,
From the relationship between the life obtained by this different life non-satisfaction investigation procedure and the probability of being less than the unbreakage time, the life of which the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% is read to determine the lot of the test object. The life upper limit reading procedure determined as the life upper limit of
A lifetime estimation method from a lifetime censoring test. If the test object consisting of a bearing or other mechanical part or test piece is placed in the specified operating environment conditions and the test continues without damage until the target time, which is the cutoff time, the required life for the life with the specified reliability can be obtained. A device that estimates the upper limit of the life of a lot under test from the non-breakage time during which the test is continued without causing damage to some of the test target products and the remaining test target products being damaged. There,
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, the value of the Weibull slope of the test object, the number of tests of the test object, the number of undamaged or broken objects, which is the number of undamaged test objects, and the unbreakage time A prompt screen output means for performing a display prompting input;
In response to the input of the execution command, a life upper limit estimation calculating means for performing an operation for estimating the upper limit of the life and outputting the operation result to the display screen is provided.
The above life upper limit estimation calculating means is:
A random number generating means using a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to an unbreakage time, for a test number, and using the value of the Weibull slope of the input information for the Weibull distribution;
Random number analysis means for checking whether or not the random number from the smallest to the number of breakage among the Weibull random numbers for the generated number of tests is equal to or less than the unbreakage time,
The set ratio life non-satisfaction investigation means for repeating the process of the random number generation means and the random number analysis means for a set number of times, and examining the probability that the random number from the smallest one to the number of breakage will be equal to or less than the unbreakage time;
The processing of the set ratio life non-satisfaction investigation means is repeated for a Weibull distribution having a life in which the set ratio is sequentially changed from a predetermined shortest life shorter than the breakage time to a predetermined longest life that is gradually longer. Different life expectancy survey means,
From the relationship between the life obtained by this different life non-satisfaction investigation means and the probability of being less than the unbreakage time, the life of which the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% is read to determine the lot of the test object. A life upper limit reading means that defines the upper limit of the life of
Reading result output means for causing the display device to output the upper limit of the life read by this means;
Life estimation device from life censoring test with A program executable on a computer,
On the screen of the display device, input the value of the Weibull slope of the product under test, the number of test products under test, the number of undamaged or damaged products, the number of undamaged products under test, and the time of unbreakage. Prompt screen output procedure for prompting display,
In response to the execution command, the upper limit of the lifetime is calculated and displayed on the screen of the display device.
The above life limit calculation procedure is as follows:
A random number generation procedure using a Weibull random number, which is a random number according to a Weibull distribution having a set life with respect to the unbreakage time, for a test number, and using the value of the Weibull slope of the input information for the Weibull distribution;
Random number analysis procedure to check whether the random number from the smallest to the number of breakage among the Weibull random number for the generated number of tests is less than the unbreakage time,
Repeat the above random number generation procedure and the above random number analysis procedure a set number of times, and set random life non-satisfaction investigation procedure to examine the probability that the random number from the smallest one to the number of breakage will be below the above unbreakage time,
This set percentage life non-satisfaction investigation procedure is repeated for a Weibull distribution with a life in which the above set percentage is sequentially changed from a predetermined shortest life shorter than the breakage time to a predetermined longest longest life. Unsatisfied investigation procedures;
From the relationship between the life obtained by this different life non-satisfaction investigation procedure and the probability of being less than the unbreakage time, the life of which the occurrence probability is a value obtained by subtracting the predetermined reliability from 100% is read to determine the lot of the test object. The life upper limit reading procedure determined as the life upper limit of
Life estimation program from the life censoring test including.

Images