Some standard content:
Mechanical Industry Standard of the People's Republic of China
JB/T6881-93
Pump Reliability Test
Published on July 13, 1993
Ministry of Machinery Industry of the People's Republic of China
Implementation on July 1, 1994
Mechanical Industry Standard of the People's Republic of China
Pump Reliability Test
1 Subject Content and Scope of Application
JB/T6881-93
This standard specifies the index selection, sampling regulations, fault judgment criteria, test procedures and test data processing methods of pump reliability test.
Reliability test is a test to determine the reliability characteristic value of the system. The purpose of the day is to determine the actual reliability level achieved by the pump so that these data can be recorded in the technical documents.
This standard applies to pump products that need to understand the reliability index value, and is applicable to both laboratory and field reliability test methods. 2 Reference standards
GB3187
Basic terms and definitions of reliability
GB5080.4 Equipment reliability test Point estimation and interval estimation methods for reliability measurement tests (exponential distribution) 3 Reliability indicators and definitions
3.1 Mean life (mean life)
a. Mean time between failures (mean time between failures - MTBF) The mean working time between two adjacent failures during the operation of the pump. b. Mean time to failure (mean time to failure - MTTF) The mean working time of the pump from the start of operation to the occurrence of a failure that is inconvenient to repair. c. Mean overhaul life (mean overhaul life - MOL) The mean working time when the pump needs to be disassembled for inspection and repair. 3.2 Mean time to repair (mean time to repair - MTTR) The mean time required for the pump to recover its specified function from the discovery of the fault. 3.3 Availability (availability - A(t)) The probability that the pump has or maintains its specified function within a specified time interval. 3.4 Reliability (reliability-R(t)): The probability that a pump can complete a specified function under specified conditions and within a specified time. The observed values of the above indicators shall be calculated in accordance with the provisions of GB3187. 4 Selection of reliability indicators
4.1 The reliability indicators of repairable pump products should be the mean time between failures (MTBF) and the mean maintenance life (MOL). 4.2 Some pumps that are repairable in principle but are inconvenient, difficult or even impossible to repair can be treated as non-repairable products (such as deep well pumps and certain marine pumps, etc.), and their reliability indicators are the mean time to failure (MTTF). 4.3 When conditions permit to collect pump maintenance data, in addition to 4.1 or 4.2, the reliability indicators can also be the mean time to repair (MTTR) and effectiveness (A(t)).
In engineering, effectiveness is usually called availability. When the failure interval and maintenance interval both obey exponential distribution, effectiveness is: Approved by the Ministry of Machinery Industry on July 13, 1993
Implemented on July 1, 1994
JB/T 6881-93
A(t)=MTBF+MTTR
4.4When assessing the parts and components of the pump (such as impeller, mechanical seal, etc.), in addition to the above indicators, the reliability indicator can also use the reliability R(t).
5 Sampling regulations
5.1 The purpose of sampling is to reduce the number of test benches and save costs, but it must be ensured that the test results of the samples can represent the level of the entire batch of pumps. 5.2 The reliability test of the pump is used in the following situations: a. Developed models or prototypes;
b. Mass production.
The population must be essentially the same, that is, the pumps of the same population are produced and assembled in the same way and under stable process conditions to ensure that the reliability test is representative. If it can be proved that the reliability level of the pump is independent of the pump specifications, the pumps of the same model but different specifications can be treated as a population.
The tested pumps must be randomly counted and sampled from the represented population. 5.3 Attention should be paid to the selection of sites for on-site reliability tests: the user's use conditions should be consistent with the design conditions; a.
b. Users with a large amount of use of the product being tested should be given priority as much as possible; e.
If multiple users are selected, attention should be paid to the examination of the same or basically the same use conditions. 5.4 The laboratory test should select no less than 3 pumps of the same specification and model produced in the same year. The on-site test should be sampled from the pumps of the same specification and model that have been running for a period of time, with a sample size of no less than 5 units, and for the same specification and model pumps with an annual output of more than 150 units, the sample size should be no less than 15 units. 5.5 For the pump products tested on site, in order to expose as many faults as possible, the sample size does not have to be large, but sufficient test time must be guaranteed, and at the same time, it must be ensured that the commissioning time of the tested pumps should not differ too much. 5.6 During the on-site test, in general, full sample tests should be given priority, that is, the test should be carried out until the maintenance life of each sampled tested pump. For pumps with a long average maintenance life (MOL ≥ 3 years), incomplete sample tests can be selected. 6 Fault judgment criteria
6.1 Fault
Any phenomenon that cannot complete its specified function according to the operating conditions required by the order data sheet, or its performance indicators deteriorate to outside the allowable range is a fault.
6.2 Manifestation
Under the rated head, the flow rate does not reach the specified value or cannot meet the user's usage value, abnormal external leakage, high vibration intensity, poor lubrication, local overheating, and frequent replacement of wearing parts without proper belts are all faults. 6.3 Classification
- Failures that endanger personal and equipment safety and cause major economic losses. Such as pumping failure, explosion, etc. Class I failure-
Class 1 failures--Failures where the main parts and components are seriously damaged and need to be disassembled for inspection and repair. Such as impeller cracks, replacement of mechanical seals and bearings, etc. Class ■ failures
-Failures where the pump function is reduced and can be easily repaired and eliminated in a short time by replacing wearing parts and accompanying tools. Such as changing packing, etc.
Class N failures-Failures that do not affect the function and use requirements of the pump and can be easily eliminated without stopping the machine. Such as label shedding, slight leakage, etc.
6.4 Judgment criteria
6.4.1 Unplanned downtime due to Class I, Class II, and Class III faults is considered as one fault. 6.4.2 Planned downtime for maintenance is not counted in the number of faults, and replacement of wearing parts according to plan is not counted in the number of faults. 6.4.3 Class N faults and other subordinate faults caused by reasons other than the pump itself, such as instrument failure, misoperation, etc., must be recorded but are not counted in the number of faults.
JB/T6881-93
6.4.4 The determination of the fault category should be based on the severity of the final consequences. 6.4.5 Due to the different structures of various pumps, the consequences and difficulty of eliminating the same fault may vary greatly. Therefore, the fault category should be determined according to the specific situation.
6.4.6 If it is found that several parameters of the pump deviate from the specified range of use during field use, and it cannot be proved that they are caused by the same reason, then each deviation of the parameter is considered as a fault of the tested pump. If there are several reasons, it is considered as several faults. If it is caused by the same reason, it is considered as one fault.
7 Requirements for the tested product
7.1 The tested product shall be tested in accordance with Section 5. 7.2 The tested product must be a qualified product that has passed the inspection, and it should be disassembled after completing the performance parameter test. Reliability testing is allowed only when no abnormal signs of parts or components are found. 7.3 Reliability testing can be carried out on a special test bench or on site. However, the operating conditions should be within the working range specified in the design, and the operating environment should comply with the relevant provisions of the instruction manual of the tested pump. 7.4 Laboratory tests should be carried out under specified conditions and be equipped with a timer. Accelerated test methods are allowed, but it must be proved that the method used does not change the failure rate. Mode and failure mechanism and the acceleration coefficient is accurate enough. 7.5 When testing under the actual working load at the site, the equipment should meet the following conditions: a.
It can measure the pump head, flow rate, power, and the measurement accuracy does not require too high, and the method is not limited; it can measure the quantities that must be measured to determine whether the pump is operating normally, such as bearing temperature, external leakage, pressure in certain parts or pressure difference, etc. b.
;
Accident protection or monitoring devices are complete.
7.6 Plans should be made for components and parts that are allowed to be replaced. 8 Test procedures
8.1 Test record
Randomly select a specified number of samples from the given pump product population and put them into the specified test process, record and accumulate the relevant test time, failure interval time and relevant failure number of all tested pumps one by one, and record them in Table 1. 8.2 Relevant test time
The relevant test time of the tested pump shall be determined in accordance with the provisions of GB5080.4 (see Figure 1). Number of tested products
★Failurewww.bzxz.net
Failure number table-1
Figure 1 Related test time of test pump
Related test time
.i+te,
Test process
JB/T6881-93
Each and wages
8.3 Calculation of cumulative related test time
JB/T6881-93
When the relevant test time of each tested pump can be measured, the cumulative relevant test time T for the occurrence of the and failures is: or T=
Where: n is the number of tested pumps;
t.—Among the tested pumps, the relevant test time for the pump with serial number m until the occurrence of the th failure; r-number of related failures (number of failures);
tthe:th failure interval time.
In the timed test, the cumulative relevant test time T is: T
Where: t—the test time of the test product with serial number m until the specified point. 8.4 Test monitoring time
(2)
During the entire test period, the manufacturer must assign a special person to observe the relevant test time of the corresponding test product and observe the relevant failure number at the end of the test.
If continuous monitoring is not possible, the time interval between monitoring must be specified so that it must be short enough not to substantially affect the test results.
Generally, the monitoring interval should be less than 0.2m. Where m. is the specified acceptable mean time between failures. 8.5 Test record
A test record should be established for each test pump, and data should be recorded in sequence at the specified time and after each failure. 8.5.1 The general maintenance personnel on site shall collect data according to the maintenance form used on site. 8.5.2 The on-site operator shall collect data according to Table 1. 8.5.3
The test personnel shall collect data regularly according to the monitoring interval and record them in Table 1.8.5.4
All events that the test personnel consider to be related to the test shall be recorded. 9 Parameter estimation and distribution assumptions
A point estimate is a single value used to represent the unknown true value of a statistical parameter. For example, the mean trouble-free working time, the validity 9.1
confidence limit specifies the confidence interval around the estimated value, which contains the true value of the estimated parameter with a certain probability (i.e., confidence level).
The one-sided confidence level 7=0.7 should be taken.
9.3 When the pump life is not verified to belong to a certain distribution, the exponential distribution is selected. The Weibull distribution is selected for the life of zero and components. When there is enough data to fit the distribution curve, the distribution fitted by the data shall prevail. 10 Test Scheme
Reliability test includes two methods: laboratory and field network. You can choose one method to conduct the test according to specific conditions. 10.1 Complete sample test-the test is carried out until each test pump reaches the maintenance life period. 10.2 Incomplete sample test:
a. Timed truncation test-a test that stops when the test time T\ specified before the test is reached. When the sample size is large, especially the laboratory test, the timed truncation test scheme can be selected. b. Fixed number truncation test-a test that stops when the test reaches the failure number r specified before the test. When the user limits the number of pump failures, the fixed number truncation test scheme can be selected. 11 Data processing
11.1 Data processing of reliability test includes two methods: graphical analysis and numerical analysis. For small samples (n≤20), the numerical analysis 5
JB/T6881-93
method should be used. For large samples (n>20), in addition to numerical analysis methods, graphical analysis methods can also be used. In addition, whether it is a large sample or a small sample, if there is computer programming ability, the random truncation method should be used. Graphical analysis method - a method of using various probability papers (such as normal, Weibull, log-normal probability papers) to plot points, estimate distribution parameters and reliability characteristic values for the data obtained from the reliability test; - a mathematical method of estimating reliability indicators using probability models and statistical models; numerical analysis method
Random truncation method - a special processing method for reliability test data obtained from the timing or fixed number truncation test plan that does not meet the complete sample or incomplete sample (such as the case of product withdrawal from the test midway or field test data with both faults and no faults).
11.2 The data obtained through the test are distinguished according to the model specifications of the pump and different environmental conditions, and recorded in Table 1. The original data for collation are:
number of test tickets n;
total number of failures r;
working time T of the i-th failure of each pump
failure interval time t,=TT:-1:
cumulative related test time T=
where T; and T:-1, are the total working time when the i-th and 1-1 failures occurred, respectively. 11.3 Exponential distribution -
When the life of the pump product obeys the exponential distribution, after obtaining the MTBF index, the point estimate of other indicators (such as failure rate, reliability R, etc.) can be obtained accordingly: T
R(t)exp(-
Interval estimation (confidence = 1α=0.7). Constant truncation one-sided confidence lower limit:
MTBF≥x(2r)
txi(2r))≤R(t)≤1
Timed truncation one-sided confidence lower limit:
MTBF≥(2r+2)
exp(二xi(+ 2)≤R(1)≤1
where the distribution of X2() is shown in Table 2.
.·(5)
（7）
·（8）
·（9）
6881-93
920-12
D=(iX<:X)d
JB/T 688193
503020
z50\ts
280*68
281°21
JB/T6881-93
Example: The life of a certain type of pump is considered to obey an exponential distribution. Five pumps are selected for life test. When 10 failures occur, the test is stopped. The failure intervals are 330, 480, 905, 1260, 1910, 2498, 2704, 3614, 4100, and 5311h respectively. Try to make a point estimate of the average life Calculation and one-sided confidence interval estimation: a.
According to formula (1): T:
According to formula (3): MTBF
According to formula (4): =
=330+480+905+1260+1910+2498+2704+3614+4100+5311=23112hT
=0.433×10-1/h
MTBF=2311.2
From =0.7, then 4=1-0.7=0.3. Look up Table 2 to get X(2r)=X(2×1 0)=22.7752×23112
Calculate the one-sided confidence lower limit according to formula (6): MTBF≥222.775
When the ticket runs for t=1500h, try to make a point estimate and one-sided confidence interval estimate of the reliability: b.
According to formula (5): R(1500)=exp(
-1500×22.775)=0.477≤R(1500)Calculate the one-sided confidence lower limit according to formula (7): exp(11.4Normal distribution and Weibull distribution
2×23112| |tt||11.4.1 Calculate the reliability characteristic value according to the original data using the following formula: Observation value of mean trouble-free working time and mean maintenance life: MTBF
Experimental root mean square difference of mean trouble-free working time and mean maintenance life: MTBF)2
(OL,-MOL)
Experimental coefficient of variation of mean trouble-free working time and mean maintenance life: Xu
11.4.2 According to the different coefficients of variation, use normal distribution (when v≤0.35) or Weibull distribution (when>0.35) respectively. 11.4.3 Calculate the mean trouble-free working time and mean maintenance life of the pump according to the following formula. When using normal distribution:
MTBF-MTBF
MOL=MOL
Where:
and t.
Select from Table 3.
When using Weibull distribution:
.(10)
·(11)
·(13)
（15)
JB/T6881-93
MTBF=K,
MOL=K,
Where: rs is selected from Table 3; K, and 6 are selected from Table 4 or Figure 2. Table 3
(n or r)
Normal distribution coefficient
and
(n or r)
Weibull distribution coefficient
..(18)
..(19)
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.