ISO 19973-1:2007
(Main)Pneumatic fluid power — Assessment of component reliability by testing — Part 1: General procedures
Pneumatic fluid power — Assessment of component reliability by testing — Part 1: General procedures
ISO 19973-1:2007 provides general procedures, the calculation method for assessing the reliability of pneumatic fluid power components and the methods of reporting. These procedures are independent of the kinds of components and of their design. ISO 19973-1:2007 also provides general test conditions and a method for data evaluation. Because the service life of any component is subject to accidental variations, a statistical evaluation assists the interpretation of the test results. The methods specified in ISO 19973-1:2007 apply to the first failure without repairs, but exclude outliers; however, because outliers can be highly significant, information about how to deal with them is given.
Transmissions pneumatiques — Évaluation par essais de la fiabilité des composants — Partie 1: Procédures générales
General Information
Relations
Standards Content (Sample)
INTERNATIONAL ISO
STANDARD 19973-1
First edition
2007-08-01
Pneumatic fluid power — Assessment of
component reliability by testing —
Part 1:
General procedures
Transmissions pneumatiques — Évaluation par essais de la fiabilité des
composants —
Partie 1: Procédures générales
Reference number
ISO 19973-1:2007(E)
©
ISO 2007
---------------------- Page: 1 ----------------------
ISO 19973-1:2007(E)
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but
shall not be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In
downloading this file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat
accepts no liability in this area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation
parameters were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In
the unlikely event that a problem relating to it is found, please inform the Central Secretariat at the address given below.
COPYRIGHT PROTECTED DOCUMENT
© ISO 2007
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or
ISO's member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Published in Switzerland
ii © ISO 2007 – All rights reserved
---------------------- Page: 2 ----------------------
ISO 19973-1:2007(E)
Contents Page
Foreword. iv
Introduction . v
1 Scope . 1
2 Normative references . 1
3 Terms and definitions. 1
4 Symbols and units of measurement. 3
5 Concept of reliability . 3
6 Strategies for conducting testing . 4
6.1 Accelerated life testing. 4
6.2 Test stand and measurement of parameters . 4
6.3 Test planning. 4
7 Statistical analysis. 4
8 Test conditions . 4
9 Sample size and selection criteria . 5
10 End of test . 5
10.1 Minimum number of failures required . 5
10.2 Termination cycle count . 6
10.3 Suspended test unit. 6
10.4 Censored test . 6
11 Evaluation of reliability characteristics from the test data . 6
12 Test report . 8
13 Identification statement (reference to this part of ISO 19973). 8
Annex A (informative) Calculation procedures for censored data without suspensions. 9
Annex B (informative) Calculation procedures for censored data with suspensions . 14
Annex C (informative) Verification of minimum life at a specified reliability and one-sided
confidence level. 21
Annex D (informative) Dealing with outliers in test data. 26
Bibliography . 31
© ISO 2007 – All rights reserved iii
---------------------- Page: 3 ----------------------
ISO 19973-1:2007(E)
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO 19973-1 was prepared by Technical Committee ISO/TC 131, Fluid power systems.
ISO 19973 consists of the following parts, under the general title Pneumatic fluid power — Assessment of
component reliability by testing:
⎯ Part 1: General procedures
⎯ Part 2: Directional control valves
⎯ Part 3: Cylinders with piston rod
⎯ Part 4: Pressure regulators
iv © ISO 2007 – All rights reserved
---------------------- Page: 4 ----------------------
ISO 19973-1:2007(E)
Introduction
In pneumatic fluid power systems, power is transmitted and controlled through a gas under pressure within a
circuit. Pneumatic fluid power systems are composed of components and are an integral part of various types
of machines and equipment. Efficient and economical production requires highly reliable machines and
equipment.
It is necessary that machine producers know the reliability of the components that make up their machine’s
pneumatic fluid power system. Knowing the reliability characteristic of the component, which can be
determined from laboratory testing, the producers can model the system and make decisions on service
intervals, spare parts inventory and areas for future improvements.
There are three primary levels in the determination of component reliability:
a) preliminary design analysis: finite element analysis (FEA), failure mode and effect
analysis (FMEA);
b) laboratory testing and reliability modelling: physics of failure, reliability prediction, pre-production
evaluation;
c) collection of field data: maintenance reports, warranty analysis.
Each level has its application during the life of a component. A preliminary design analysis is useful to identify
possible failure modes and eliminate them or reduce their effect on reliability. When prototypes are available,
in-house laboratory reliability tests are run and initial reliability can be determined. Reliability testing is often
continued into the initial production run and throughout the production lifetime as a continuing evaluation of
the component. Collection of field data is possible when products are operating and data on their failures are
available.
Specific component test procedures and exclusions are provided in ISO 19973-2, ISO 19973-3 and
ISO 19973-4.
© ISO 2007 – All rights reserved v
---------------------- Page: 5 ----------------------
INTERNATIONAL STANDARD ISO 19973-1:2007(E)
Pneumatic fluid power — Assessment of component reliability
by testing —
Part 1:
General procedures
1 Scope
This part of ISO 19973 provides general procedures, the calculation method for assessing the reliability of
pneumatic fluid power components and the methods of reporting. These procedures are independent of the
kinds of components and of their design.
This part of ISO 19973 also provides general test conditions and a method for data evaluation.
NOTE Because the service life of any component is subject to variations, a statistical evaluation assists the
interpretation of the test results.
[4]
The methods specified in this part of ISO 19973 apply to the first failure without repairs (see IEC 60300-3-5) ,
but exclude outliers; however, because outliers can be highly significant, information about how to deal with
them is given in Annex D.
2 Normative references
The following referenced documents are indispensable for the application of this document. For dated
references, only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments) applies.
ISO 1000, SI units and recommendations for the use of their multiples and of certain other units
ISO 3534-1, Statistics — Vocabulary and symbols — Part 1: General statistical terms and terms used in
probability
ISO 5598, Fluid power systems and components — Vocabulary
IEC 60050-191, International Electrotechnical Vocabulary, chapter 191: Dependability and quality of service
IEC 61649, Goodness-of-fit tests, confidence intervals and lower confidence limits for Weibull distributed data
3 Terms and definitions
For the purposes of this document, the terms and definitions given in ISO 3534-1, ISO 5598 and
IEC 60050-191 and the following apply.
3.1
catastrophic failure
sudden failure of an item that results in its complete inability to perform all required functions
[IEC 60050-191]
© ISO 2007 – All rights reserved 1
---------------------- Page: 6 ----------------------
ISO 19973-1:2007(E)
3.2
confidence coefficient
confidence level
value (1 − α) of the probability associated with a confidence interval or a statistical coverage interval
NOTE 1 See also 3.6.
NOTE 2 See ISO 3534-1 for notes related to this term and definition.
3.3
confidence limit
either of the limits, T or T , of the two-sided confidence interval, or the limit, T, of the one-sided confidence
1 2
interval
NOTE See ISO 3534-1 for notes related to this term and definition.
3.4
failure
termination of the ability of an item to perform a required function
[IEC 60050-191]
NOTE In the ISO 19973 (all parts), the reaching of a threshold level for statistical calculation is also considered a
failure.
3.5
mean time to failure
MTTF
expectation of the time to failure
[IEC 60050-191]
NOTE In ISO 19973-2 to ISO 19973-4, the concept of time is expressed as either cycles or distance.
3.6
one-sided confidence interval
T
interval estimator for a population, Θ, comprised of the interval from the smallest possible value of the
population, Θ, up to T or the interval from T up to the largest possible value of Θ, where the probability
p(T W Θ ) or p(T u Θ ) is at least equal to (1 − α), where (1 − α) is a fixed number, positive and less than 1
NOTE See ISO 3534-1 for notes related to this term and definition.
3.7
relevant failure
failure that should be included in interpreting test or operational results or in calculating the value of a
reliability performance measure
[IEC 60050-191]
3.8
reliability
probability that an item can perform a required function under given conditions for a given time interval
[IEC 60050-191]
3.9
sample
one or more test units taken from a population and intended to provide information on the population
NOTE A sample can serve as a basis for a decision on the population or on the process that produced it.
2 © ISO 2007 – All rights reserved
---------------------- Page: 7 ----------------------
ISO 19973-1:2007(E)
3.10
sample size
number of test units in the sample
NOTE In a multi-stage sample, the sample size is the total number of test units at the conclusion of the final stage of
sampling.
3.11
threshold level
value of a performance characteristic (for example, leakage, flow rate, current, etc.) against which the
component’s test data is compared
NOTE This is an arbitrary value defined by the experts as the critical value for performance comparisons, but is not
necessarily indicative of a component failure.
4 Symbols and units of measurement
4.1 The symbols used in this part of ISO 19973 are given in Table 1.
Table 1 — Symbol list
a
Symbol Definition
B Expected time at which 10 % of the population fails (10 % fractile of the lifetime)
10
η Scale parameter (characteristic life) of the Weibull distribution
F(t) Probability of failure, expressed in percent
β Shape parameter (slope) of the Weibull distribution
R(t) Reliability value of the entire sample; cumulative reliability
a
Other symbols can be used in other documents and software.
4.2 Units of measurement are in accordance with ISO 1000.
5 Concept of reliability
For the purposes of this part of ISO 19973, reliability is the probability that a component does not have a
relevant failure for a specified interval of time, number of cycles or distance when it operates under stated
conditions.
A relevant failure occurs when a component
⎯ exceeds any of its defined threshold levels, or
⎯ experiences a catastrophic failure (burst, fatigue or functional failure, etc.).
Threshold levels of the components covered by ISO 19973 (all parts) are specified in the component-specific
parts of this International Standard.
This probability can be determined by analysing the results of a series of tests and describing the population
failure by statistical methods. There are many different statistical distributions that describe the population of
failures that result from testing.
NOTE See Annex C for information about verifying the minimum life of a component at a specified reliability and one-
sided confidence level.
© ISO 2007 – All rights reserved 3
---------------------- Page: 8 ----------------------
ISO 19973-1:2007(E)
6 Strategies for conducting testing
6.1 Accelerated life testing
One of the major difficulties encountered in specifying a reliability test is the time it takes to reach the
threshold levels without accelerating the test. Accelerated testing, in which a component is subjected to
environmental conditions more severe than those for which the component is rated, is sometimes necessary
in order to keep the test time at a reasonable length. The degree and amount of accelerated testing specified
in ISO 19973 (all parts) reflect the best judgement of, and agreement among, those who developed these
International Standards. The primary criterion for determining test acceleration factors is that the failure mode
or failure mechanism should not change or be different from that expected from a non-accelerated test.
6.2 Test stand and measurement of parameters
Two other important factors are the test stand and measurement of parameters. The test stand shall be
designed to operate reliably within the planned environmental conditions. Its configuration shall not affect the
results of the test being run on the component. Evaluation and maintenance of the test stand during the
reliability test program is critical. The accuracy of parameter measurement and control of parameter values
shall be within the specified tolerances to assure accurate and repeatable test results.
6.3 Test planning
Proper test planning is essential in order to produce results that accurately predict the component’s reliability
under specified conditions. The goals and objectives of the test program shall be clearly defined if a supplier
and user agree to apply ISO 19973 (all parts).
7 Statistical analysis
The resulting test data shall be evaluated for calculating an assessment of the reliability. One of the most
commonly used methods is the Weibull analysis because of its versatility in modelling various statistical
distributions. This method shall be used for the analysis of the test data to ensure comparability of the results.
Examples of applying Weibull analysis are given in Annex A.
NOTE Commercial software can be helpful for this purpose.
8 Test conditions
8.1 Testing shall be carried out in accordance with the provisions defined in the part of ISO 19973 that
relates to the component tested, including the test parameters that are measured and threshold levels
specified for each test parameter.
8.2 No repairs are permitted on the samples during the reliability test.
8.3 Unless otherwise specified in the relevant part of ISO 19973 that relates to the component being tested,
or when agreed between the user and supplier, all tests shall be carried out under the conditions specified in
Table 2.
4 © ISO 2007 – All rights reserved
---------------------- Page: 9 ----------------------
ISO 19973-1:2007(E)
Table 2 — General test conditions
Parameter Value
Working pressure 630 kPa ± 30 kPa (6,3 bar ± 0,3 bar)
Ambient temperature
23 °C ±10 °C
Temperature of the medium
23 °C ±10 °C
Air quality Filtration: nominal filtration rating 5 µm
Dryer: maximum pressure dew point +3 °C
Lubrication None
8.4 During the endurance test, the test operator shall determine the measuring intervals in relation to the
total number of cycles expected. Perform measurements more frequently at the beginning of the test until
confidence with the operation is assured. It is up to the experience and judgement of the people in the test
laboratory to determine effective measuring intervals; test intervals that are too long can reduce the
statistically-determined lifetime of the test unit (see 10.2); measuring intervals that are too short increase the
cost of testing. The first measurement shall be done within 10 % of the expected lifetime of the test unit. Test
units shall be operated continuously and checked periodically for proper functioning.
9 Sample size and selection criteria
9.1 The samples shall be representative of the population and shall be selected randomly.
9.2 The minimum sample size shall be seven units.
NOTE It is important to have at least seven samples in order that the first data point on the Weibull graph is below
the 10 % cumulative-failure point. This allows a more accurate projection of the lower confidence limit lines to intersect the
10 % cumulative-failure point and determine a B life.
10
9.3 For a product series with the same design principle, it is not necessary to test all types or sizes.
However, the test program shall include the type with the most critical conditions, for example, highest velocity.
10 End of test
10.1 Minimum number of failures required
The minimum number of test units that are required to fail (e.g. reach a threshold level) is described in Table 3.
This number does not include suspensions, which are not considered failures.
NOTE It is desirable to achieve at least 10 failures in accordance with IEC 61649. Fewer failures result in a wider
confidence interval and a shorter B life at the lower confidence limit.
10
Table 3 — Minimum number of failures for evaluation of the characteristic life
Sample size 7 8 9 10 > 10
a
Minimum numbers of failures 5 6 7 7
70 % of the sample size, truncated
a
For example, if the sample size is 11, the minimum number of failures is 7.
© ISO 2007 – All rights reserved 5
---------------------- Page: 10 ----------------------
ISO 19973-1:2007(E)
10.2 Termination cycle count
When a test unit fails between consecutive observations, the data collected is referred to as left-censored
interval data; in this case, both the last cycle count at which the test unit was operating properly and the cycle
count at which the test unit failed shall be recorded. In some cases, when a more precise determination of a
failure point is desired, it can be necessary to continuously monitor the performance of a test unit by using
limit switches or other suitable means to detect failures when they occur.
10.3 Suspended test unit
Testing on an individual test unit may be stopped before a relevant failure occurs. This is known as a
suspension. Some examples of suspensions include
⎯ a unit which has been disassembled for inspection, or
⎯ a unit which has been accidentally crushed.
However, these units had achieved a number of cycles until the point of suspension and these data have a
positive influence on the calculation of the statistical parameters.
10.4 Censored test
If the test is stopped after the minimum number of failures specified in Table 3 is reached but the remaining
samples are still operating, the test shall be considered censored. If the censored test does not include any
suspensions, the method specified in Annex A shall be used to calculate the statistical parameters. If the
censored test includes one or more suspensions, the method specified in Annex B shall be used to calculate
the statistical parameters.
11 Evaluation of reliability characteristics from the test data
11.1 To improve the interpretation of the calculation results, the failure mode shall be specified and recorded.
11.2 Calculations shall be made from the test data to determine
⎯ characteristic life, η : relative location of the straight line in the Weibull plot relative to the x-axis
(time or scale parameter);
⎯ Weibull shape parameter, β : slope of the straight line in the Weibull plot.
11.3 Calculate B life at the median ranks (see Figure 1, footnote a).
10
NOTE See Annex B for information on how to deal with censored data with suspensions.
11.4 Calculate the confidence limit of the B life at the one-sided 95 % confidence level using Fisher Matrix
10
(see Figure 1, footnote a).
6 © ISO 2007 – All rights reserved
---------------------- Page: 11 ----------------------
ISO 19973-1:2007(E)
Key
X number of cycles until failure, t
Y probability of failure, F(t), expressed in percent
1 best fit line
2 confidence limit, one-sided 95 %, obtained by Fisher Matrix
3 10 % fractile line
4 line at which 63,2 % of test units failed
a
B life at the one-sided 95 % confidence level.
10
b
B life.
10
c
Characteristic life, η.
NOTE 1 The MTTF value can be calculated using the characteristic life.
NOTE 2 Commercial software can be useful in constructing the graphs.
Figure 1 — Example of how B life value is determined
10
© ISO 2007 – All rights reserved 7
---------------------- Page: 12 ----------------------
ISO 19973-1:2007(E)
12 Test report
The test report shall include at least the following data:
a) number of the relevant part of ISO 19973, including the component-specific part number (for example,
ISO 19973-2 for valves);
b) date of the test report;
c) component description (manufacturer, type designation, series number);
d) sample size;
e) test conditions (working pressure, temperature, air quality, frequency, load, etc.);
f) threshold levels;
g) type of failure for each test unit;
h) B life at the median rank and confidence limit of B life at one-sided 95 % confidence level;
10 10
i) characteristic life, η ;
j) number of failures considered;
k) method used to calculate the Weibull data (for example, maximum-likelihood, rank regression,
Fisher Matrix);
l) other remarks, as necessary.
13 Identification statement (reference to this part of ISO 19973)
It is strongly recommended to manufacturers who have chosen to conform to this part of ISO 19973 that the
following statement be used in test reports, catalogues and sales literature:
“General procedures for assessing pneumatic component reliability by testing performed in accordance with
ISO 19973-1, Pneumatic fluid power — Assessment of component reliability by testing — Part 1: General
procedures.”
8 © ISO 2007 – All rights reserved
---------------------- Page: 13 ----------------------
ISO 19973-1:2007(E)
Annex A
(informative)
Calculation procedures for censored data without suspensions
A.1 Example using maximum-likelihood estimates
A.1.1 Consider a test run on a sample of seven test units of a component, and parameters related to three
failure modes (1, 2, 3) are measured during a reliability test. Raw data from each parameter is collected as the
test progresses. When a failure occurs (either by no longer being able to perform a required function or by
reaching a threshold level; see 3.4), the cycle count at which the test unit was last observed in satisfactory
condition is recorded, along with the cycle count at which it was first noticed that the failure had occurred. In
this case, the termination cycle count is somewhere between the two observed cycle counts.
For the example shown in Table A.1, test unit number 5 experiences a mode 3 failure somewhere between
the two cycle counts shown. The data for other test units are likewise determined from the observations when
a test unit experiences any failure mode for first time. These are shown as shaded cells in Table A.1. The test
is concluded when the minimum number of test units specified in Table 3 fail, in this case, five test units.
If the manufacturer desires to know more information about any test unit that has failed by reaching a
6
threshold level, testing on that unit may continue. However, further data, such as observed at 36,6 × 10 and
6
43,1 × 10 cycle counts for test unit 5, shall not be considered in the reliability analysis.
Table A.1 — Example of test unit cycle counts and failure modes
Cycle count at
observation of last Cycle count at Failure mode 1 Failure mode 2 Failure mode 3
satisfactory observation of failure (leakage – seal A) (leakage – seal B) (shifting time)
operation
— —
6 6
Test unit number 5
10,8 × 10 12,8 × 10
— —
6 6
Test unit number 1
19,5 × 10 23,5 × 10
— —
6 6
Test unit number 2
28,2 × 10 32,2 × 10
—
6 6
Test unit number 2 Test unit number 5
34,2 × 10 36,6 × 10
—
6 6
Test unit number 3 Test unit number 1
37,8 × 10 41,8 × 10
— —
6 6
Test unit number 5
42,9 × 10 43,9 × 10
— —
6 6
Test unit number 6
44,9 × 10 44,9 × 10
—
6
Test ended – test units 4 and 7 removed from test
44,9 × 10
NOTE The example illustrates how test units can reach several threshold levels if they continue to be tested beyond the point at
which they experience their first failure mode. The shaded cells indicate which test units experienced a failure mode for the first time.
Note also that some test units did not experience a failure mode and were censored at the end of the test. This example shows
termination-cycle counts for test units that were continuously monitored.
A.1.2 In this example, the Weibull parameters are then determined from a maximum-likelihood estimation,
using both cycle counts for each test unit. Results are graphed on a Weibull plot, as shown in Figure A.1. The
failure points use the cycle-count range in Table A.1, and the plot line and Weibull parameters are based on
the maximum-likelihood calculation. The confidence limit is based on a Fisher Matrix calculation.
© ISO 2007 – All rights reserved 9
---------------------- Page: 14 ----------------------
ISO 19973-1:2007(E)
Key
X1 number of cycles to failure, t
X2 slope of the Weibull distribution, β, equal in this case to 2,382 3
Y probability of failure, F(t), expressed in percent
1 best fit line
2 95 % one-sided confidence limit
3 10 % fractile line
4 line at which 63,2 % of test units fail
Weibull data point
●
a
B life at the one-sided 95 % confidence level.
10
b
B life.
10
c 6
Characteristic life, η, is equal to 42,365 × 10 .
Figure A.1 — Weibull plot for maximum-likelihood estimation
(data from Table A.1)
A.1.3 Results for the median conditions are
6
⎯ characteristic life: 42,37 × 10 cycles;
⎯ slope: 2,38.
10 © ISO 2007 – All rights reserved
---------------------- Page: 15 --------------------
...
Questions, Comments and Discussion
Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.