National Metrology Technical Specification of the People's Republic of China JJF1024—1991
Reliability Analysis Principle for Measuring Instruments
Issued on March 4, 1991
Implementation on December 1, 1991
Issued by the State Bureau of Technical Supervision
JJF1024-1991
Reliability Analysis Principle for Measuring Instruments
Technical Norm of Reliability Analysis Principle for Measuring Instruments JJF1024—1991
This technical specification was approved by the State Bureau of Technical Supervision on March 4, 1991, and came into effect on December 1, 1991.
Responsible unit: China National Institute of Metrology Drafting unit: China National Institute of Metrology The drafting unit is responsible for interpreting the technical provisions of this regulation. The main drafter of this specification:
Liu Zhimin
JJF1024-1991
(China National Institute of Metrology)
Scope and purpose
Noun·
Reliability expression·
Reliability assessment
JJF1024—1991
(1)
1Scope and purpose
JJF1024—1991
Technical specification for reliability analysis principles of measuring instruments This specification proposes the basic concepts, characteristics and analysis methods of the reliability of measuring instruments to ensure the accuracy and reliability of the measurement value. 2 Noun
2.1 Reliability (reliability or reliability preformance) The ability of a measuring instrument to complete the specified function under specified conditions and within a specified time. 2.2 Reliability
(reliability)
The probability that a measuring instrument can complete its specified function under specified conditions and within a specified time. 2.3 Failure
(failure)
The loss of the specified function of a measuring instrument, which is manifested by its uncertainty exceeding the allowable value or malfunction (also called failure for repairable instruments).
2.4 Early failure
(early failure)
The failure of a measuring instrument due to factors such as unnecessary defects in design and manufacturing. Aging screening can eliminate early failures.
2.5 Random failure
(random failure)
The failure of a measuring instrument due to accidental factors. (wear-out failure)
2.6 Wear-out failure
The failure of a measuring instrument due to factors such as aging, wear, loss, and fatigue. 3 Reliability expression
3.1 Reliability function R(t)
The probability R(t) of completing the specified function within the specified time t under specified conditions is 0≤R()≤1
Example: There are 1000 instruments. One instrument fails within 500 hours of operation. After 1000 hours of operation, a total of 10 instruments fail. The estimated value of the reliability function is: R(500 h) = 1 000 -1
R(1 000 h)=1000 -10
3.2 Failure distribution function F(t)
and failure distribution density f(t）
F(t) = 1-R(t)
JJF1024—1991
f(t) =
Failure time t is a random variable, and f(t) is the failure (probability) distribution density of t. 3.3 Failure rate function π(t)
The probability of a measuring instrument that has not failed at a certain time t failing per unit time after that time a(t) =
-1.dR(t) -f(t)
Its estimated value is: the number of instruments that fail within [t, t+△t] divided by the product of △t and the number of instruments that have not failed at time t.
3.4 Average life 0
For repairable instruments, it is the mean time between failures MTBF, and for non-repairable instruments, it is the mean time before failure MTTF.
is the mathematical expectation of the failure time t9=Et.
tf(t)dt =R(t)dt
When the experimental life of n samples of measuring instruments is t1, t2,, tn, then the estimated value of θ is 6==17
3.5 Reliable life tR
Given the reliability R, the t that satisfies R(tr)=R is called the reliable life of reliability R. The reliable life to with R=0.5 is also called the median life.
3.6 Standard deviation of life.
Standard deviation of failure time t
(t -)f(t)dt
When the experimental life of n samples of measuring instruments is t1, tz,, tn, then. The estimated value is obtained by Bessel method (-)2
3.7 Failure distribution
The commonly used one is exponential distribution, and its reliable life under reliability R
k is shown in Table 1:
JJF1024—1991
f(t) = Ae-
R(t) =et
F(t) =1- e-α
（t)=入（constant）
It can be seen from the table that the median life to.50=0.69316, and the average life day is the reliable life with reliability R=e-1~0.368.
Measuring instruments are often used in the accidental failure period, when the failure rate is close to a constant, so this specification uses exponential distribution. Weibull distribution is also used in reliability.
3.8 Calculation of life expectancy in truncation experiment
3.8.1 No replacement: number of samples n, no replacement after failure, r failures before the end of the experiment, and failure time tro
Constant truncation: stop at one failure, then the estimated average life = 1/2t + (n -r)t,)
Timed truncation: the experiment ends at time t, then the estimated average life = 11
Ct + (nr)th
Example: for a certain instrument, under the condition of no replacement and timed t=20h truncation, the experiment n=88 sets, and the values in Table 2 are obtained. Table 2
Failure time/h
Number of failures
JJF1024—1991
#=10.32×2+0.76×2++19.60×18+(88-48)×20//48=27.33h
3.8.2 With replacement: The number of samples is n. Once a sample fails, there is a good sample to replace it. The failure is measured before the end of the experiment, and the failure time t≤t≤≤t is obtained. Constant truncation: When the experiment stops at t, the average life estimate is @=1
Timed truncation: When the experiment ends at time t, the average life estimate is 6
For example: A certain instrument is tested with 7 units under replacement timed t=70h truncation. Before the end of the experiment, 8 failures are measured. The average life estimate is
3.9 Accelerated test life calculation
X7×70=61.25h
In order to reduce the test time, an accelerated test can be performed. For the Arrhenius method, the reliable life is
t=104+b/T
Where: a, b—constants;
T——stress.
Igt=a+b/T
According to the existing (ti, T), a and b are obtained by algebraic method or least square method, and the life t under normal stress T is obtained.
Example: The average life t of a measuring instrument at temperature T is as follows: T = 463 K,
T2=493 K,
T3=513 K,
T4=533K,
Then the relationship between temperature T and average life t is
tj=5900h
t2=2600h
t3=1600h
t4=1040h
Igt = a + b
JJF1024—1991
The constant in it is calculated by the least square method as (n is the number of i) h
a =(gt: -6) =-1.98
4 Reliability Assessment
4.1 Verification Cycle
4.1.1 Based on Principles
On the premise that the reliability meets the requirements, the annual verification cost is the least. 4.1.2 Considerations
a. Required uncertainty;
b. Consequences caused by failure of measuring instruments; c.Use degree and severity;
d. Wear and drift trend;
e. Environmental conditions;
f. Manufacturer's recommendations;
g. Internal inspection and calibration frequency and method;
h. Frequency of comparison with other benchmarks;
i. Existing calibration data trend;
j. Maintenance and use record history.
4.1.3 Initial calibration
The initial calibration cycle shall be determined by experienced professionals based on the data of similar instruments at home and abroad, and then adjusted later. 4.1.4 Reliability function method
Determine the reliability function R (t) and the average life θ, and determine the working time t by t = 0t-lnR (t))
, and then determine the calibration cycle based on the average working time. For example: θ=5000h for a certain oscilloscope, R(t)=0.95, t=256h, the average working time of this oscilloscope is 50h/month, so the calibration period is 5 months. 4.1.5 Statistical calculation method
For a certain type of instrument, one of the following can be counted: a. T. year change of any one;
b. T. year change of several ones.
If the probability of exceeding the tolerance is equal to 0.05 (taking the reliability as 0.95), then T. year is the calibration period. JJF10241991
Example: The annual changes of a certain instrument are as follows (T.=1, 20 in total) 0.0, 0.8, 0.0, -0.2, 0.9, 0.0, 0.2, 0.9, 0.0, 0.1, 0.1, 0.5, 0.4, 0.1, 0.1, 0.3, 0.2, 0.1, 0.1, 1.1. The required tolerance is 1.0, then the probability of exceeding it in one year is 1/20, so the cycle is 1 year. 4.2 Reliability of combined equipment
4.2.1 Series connection: n equipment working independently Reliability
If each input is the same, then MTBF=MTBF/n
4.2.2 Parallel connection: n equipment working independently IR(t)
R(t) = 1-
Reliability
wwW.bzxz.NetIf each input is the same, then
(1- R;(t))
MTBF=MTBF(1+++
People's Republic of China
National Metrology Technical Specification
Principles of Reliability Analysis of Measuring Instruments
JJF1024—1991
Issued by the State Bureau of Technical Supervision
Published by China Metrology Press
No. 2, West Hepingli Street, Beijing
Postal Code 100013
Tel (010) 64275360
E
[email protected]Printed by Beijing Dixin Printing Factory
Published by Beijing Xinhua Bookstore
All rights reserved. No reproduction allowed
880mm×1230mm16-size
First edition in June 1991
Printing sheet 0.75 Word count 9,000 words
Second printing in July 2002
Print count 9001-10500
Standard book number 155026-1611
Tip: This standard content only shows part of the intercepted content of the complete standard. If you need the complete standard, please go to the top to download the complete standard document for free.