ISO/TR 12391-4:2002

TECHNICAL ISO/TR
REPORT 12391-4
First edition
2002-12-15
Gas cylinders — Refillable seamless
steel — Performance tests —
Part 4:
Flawed-cylinder cycle test
Bouteilles à gaz — Rechargeables en acier sans soudure — Essais de
performance —
Partie 4: Cycle d'essai pour bouteilles défectueuses

Reference number
©
ISO 2002
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.

©  ISO 2002
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 2002 — All rights reserved

Contents Page
Foreword. iv
Introduction . v
1 Scope. 1
2 References. 1
3 Terms and definitions. 2
4 Symbols. 2
5 Background. 3
6 Experimental test programme . 4
6.1 Types of cylinder tested. 4
6.2 Material properties tests. 5
6.3 Description of the flawed-cylinder cycle test. 6
7 Flawed-cylinder cycle test results. 8
7.1 Flawed-cylinder cycle test procedure. 8
7.2 Flawed-cylinder cycle test results for group F-B materials. 8
7.3 Flawed-cylinder cycle test results for group F-C materials. 8
7.4 Flawed-cylinder cycle test results for group F-D materials. 9
7.5 Flawed-cylinder cycle test results for group F-E materials. 10
8 Discussion. 10
8.1 Background. 10
8.2 Flawed-cylinder cycle test procedures and acceptance criteria (ISO 9809-2). 10
8.3 Pressure cycling tests (ISO 9809-2) . 11
9 Conclusions. 11

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.
In exceptional circumstances, when a technical committee has collected data of a different kind from that
which is normally published as an International Standard (“state of the art”, for example), it may decide by a
simple majority vote of its participating members to publish a Technical Report. A Technical Report is entirely
informative in nature and does not have to be reviewed until the data it provides are considered to be no
longer valid or useful.
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/TR 12391-4 was prepared by Technical Committee ISO/TC 58, Gas cylinders, Subcommittee SC 3,
Cylinder design.
ISO/TR 12391 consists of the following parts, under the general title Gas cylinders — Refillable seamless
steel — Performance tests:
 Part 1: Philosophy, background and conclusions
 Part 2: Fracture performance tests — Monotonic burst tests
 Part 3: Fracture performance tests — Cyclical burst tests
 Part 4: Flawed-cylinder cycle test
iv © ISO 2002 — All rights reserved

Introduction
Gas cylinders as specified in ISO 9809-1 have been constructed of steel with a maximum tensile strength of
less than 1 100 MPa. With the technical changes in steel-making using a two-stage process, referred to as
ladle metallurgy or secondary refining, significant improvement in mechanical properties have been achieved.
These improved mechanical properties provide the opportunity of producing gas cylinders with higher tensile
strength, which achieve a lower ratio of steel to gas weight. The major concern in using steels of higher tensile
strength with correspondingly higher design wall stress is safety throughout the life of the gas cylinder.
When ISO/TC 58/SC 3 began drafting ISO 9809-2, Working Group 14 was formed to study the need for
additional controls for the manufacture of steel gas cylinders having a tensile strength greater than 1 100 MPa.
This part of ISO/TR 12391 presents all of the specific test results of the flawed-cylinder cycle tests that were
conducted to evaluate the fatigue performance of cylinders ranging in tensile strength from less 800 MPa to
greater than 1 350 MPa.
TECHNICAL REPORT ISO/TR 12391-4:2002(E)

Gas cylinders — Refillable seamless steel — Performance
tests —
Part 4:
Flawed-cylinder cycle test
1 Scope
This part of ISO/TR 12391 applies to seamless steel refillable cylinders of all sizes from 0,5 l up to and including
150 l water capacity produced of steel with tensile strength, R , greater than 1 100 MPa.
m
It can also be applied to cylinders produced from steels used at lower tensile strengths. In particular, it provides
the technical rationale and background to guide future alterations of existing ISO standards or for developing
advanced design standards.
This part of ISO/TR 12391 is a summary and compilation of the test results obtained during the development
of the “flawed-cylinder cycle test”. The “flawed-cylinder cycle test” was developed as part of a co-operative
project under the direction of ISO/TC 58/SC 3/WG 14. The “flawed-cylinder cycle test” is a test method to
evaluate the fatigue performance of steel cylinders that are used to transport high pressure, compressed
gases.
The concept and development of the flawed-cylinder cycle test is described in ISO/TR 12391-1. The details of
the test method and the criteria for acceptable fatigue performance of steel cylinders are given in 9.2.6 of
ISO 9809-2:2000, “flawed-cylinder cycle test”. In this part of ISO/TR 12391, test results are reported for more
than a hundred flawed-cylinder cycle tests that were conducted on seamless steel cylinders ranging in
measured tensile strength from less than 800 MPa to greater than 1 350 MPa. The test method is intended to
be used for the selection of materials and design parameters in the development of new cylinder designs.
2 References
ISO 148:1983, Steel — Charpy impact test (V-notch)
1)
ISO 6406:— , Seamless steel gas cylinders — Periodic inspection and testing
ISO 9809-1:1999, Gas cylinders — Refillable seamless steel gas cylinders — Design, construction and
testing — Part 1: Quenched and tempered steel cylinders with tensile strength less than 1 100 MPa
ISO 9809-2:2000, Gas cylinders — Refillable seamless steel gas cylinders — Design, construction and
testing — Part 2: Quenched and tempered steel cylinders with tensile strength greater than or equal to
1 100 MPa
ISO/TR 12391-1, Gas cylinders — Refillable seamless steel — Performance tests — Part 1: Philosophy,
background and conclusions
1) To be published. (Revision of ISO 6406:1992)
ISO/TR 12391-2, Gas cylinders — Refillable seamless steel — Performance tests — Part 2: Fracture
performance test — Monotonic burst tests
ISO/TR 12391-3, Gas cylinders — Refillable seamless steel — Performance tests — Part 3: Fracture
performance tests — Cyclical burst tests
3 Terms and definitions
3.1
flawed-cylinder cycle test
test conducted on a finished gas cylinder having a shallow prescribed flaw of 10 % of the cylindrical wall
thickness machined into the exterior sidewall and failed by cyclical internal pressurization that is normally
hydraulic
3.2
flawed-cylinder burst test
test conducted on a finished gas cylinder having a deep prescribed flaw in the range of 75 % of the cylindrical
wall thickness machined into the exterior sidewall and failed by internal pressurization that may be hydraulic,
and is applied either monotonically or cyclically
3.3
pressure cycling test
test conducted on a finished gas cylinder that does not have a flaw machined into the exterior sidewall and
failed by cyclical internal pressurization that is normally hydraulic
4 Symbols
d is the flaw depth, expressed in millimetres as a percentage of t ;
d
D is the outside diameter of the cylinder, expressed in millimetres;
l is the flaw length, expressed in millimetres (= n × t );
o d
n represents multiples of t (= l /t );
d o d
P is the calculated design test pressure for the cylinder, expressed in bar;
h
P is the calculated design service pressure for the cylinder, expressed in bar;
s
R is the guaranteed minimum yield strength;
e
R is the actual measured value of yield strength, expressed in megapascals;
ea
R is the maximum value of tensile strength guaranteed by the manufacturer, expressed in megapascals;
g, max
R is the minimum value of tensile strength guaranteed by the manufacturer, expressed in megapascals;
g, min
R is the actual measured value of tensile strength, expressed in megapascals;
m
t is the actual measured wall thickness at the location of the flaw, expressed in millimetres;
a
t is the calculated minimum design wall thickness, expressed in millimetres.
d
2 © ISO 2002 — All rights reserved

5 Background
High-pressure industrial gases (such as oxygen, nitrogen, argon, hydrogen, helium, etc.) are stored and
transported in portable steel cylinders. These cylinders are designed, manufactured, and maintained in
accordance with ISO 9809-1 and ISO 9809-2. The cylinders are constructed from specified alloy steels that
generally contain chromium and molybdenum as the principal alloying elements. The cylinders are of
seamless construction and are manufactured by either a forging process, a tube-drawing process or by a
plate-drawing process. The required mechanical properties are obtained by using an austenitizing, quenching,
and tempering heat treatment. Typical sizes of these cylinders are 100 mm to 250 mm in diameter, 500 mm to
2 000 mm in length, and 3 mm to 20 mm in wall thickness. Typical working pressure ranges are 100 bar to
400 bar.
Until recently, the tensile strength of the steels used in the construction of such cylinders has been limited to a
maximum of about 1 100 MPa. This limitation for the maximum tensile strength occurs because the fracture
toughness and ductility of the steels decreases with increase in the tensile strength and above a tensile
strength of about 1 100 MPa the fracture toughness and ductility were not adequate to prevent fracture of the
cylinders. Recently developed new steel alloys that have both high tensile strength and high fracture
toughness and ductility make it possible to construct lighter cylinders with higher tensile strength steels. This
permits the use of cylinder designs with higher permissible stresses in the cylinder wall increased for a
constant wall thickness. The use of higher strength steels therefore leads to a lower ratio of steel weight to
gas weight that reduces shipping and handling costs.
A major concern in using higher strength steels for cylinder construction and correspondingly higher design
wall stress is the ability to maintain the same level of safety throughout the life of the cylinder. In particular,
increasing the tensile strength of the steels and increasing the stress in the wall of the cylinders could make
the cylinders less fracture resistant and more subject to fatigue failure than cylinders made from steels with
the traditionally used lower tensile strength levels. In order to use steels with strength levels higher than
1 100 MPa, it was decided that new requirements were needed to assure adequate fracture and fatigue
resistance of the cylinders.
To develop these requirements, WG 14 was formed under ISO/TC 58/SC 3. WG 14 was assigned the task:
“develop a suitable test method and specifications to assure adequate fracture resistance for gas cylinders
made from steels with tensile strengths greater than 1 100 MPa”. The results of the test programme to
develop suitable test methods and acceptance criteria to ensure adequate fracture performance are described
in ISO/TR 12391-1, ISO/TR 12391-2 and ISO/TR 12391-3.
The original scope of the WG 14 work was amended to also include the development of a suitable test method
and acceptance criteria to ensure adequate fatigue resistance for gas cylinders made from steels with tensile
strengths greater than 1 100 MPa. This was required because the fatigue crack growth rate is controlled by
the wall stress in the cylinder, so that by increasing the tensile strength of the steels and increasing the stress
in the wall of the cylinders the cylinders may become less fatigue resistant and more subject to fatigue failure
than cylinders made from steels with the traditionally used lower tensile strength levels.
WG 14 decided that the test method and acceptance criteria that were developed to evaluate the fatigue
performance of the cylinders should demonstrate that the overall “fatigue resistance” of cylinders made from
higher strength steels was equivalent to that of cylinders made from lower strength steels. It was decided that
the test method that was developed should measure the total fatigue resistance of the cylinder and not just the
fatigue crack growth rate of the steel used in the cylinder. Therefore, the test method that was developed to
evaluate the total fatigue performance of cylinders was the “flawed-cylinder cycle test”. The concept of the
flawed-cylinder cycle test and the development conducted under WG 14 is described in the ISO/TR 12391-1.
In the “flawed-cylinder cycle test”, the fatigue test is performed on an actual, full size, cylinder rather than by
measuring the fatigue properties of the material alone by taking small-scale test specimens. This test method
requires the testing of cylinders in which flaws of specified sizes are machined into the external surface of the
cylinders. The cylinders are cyclically pressurized to a specified maximum pressure until failure occurs either
by leaking or by fracturing or for a defined maximum number of pressure cycles without failure. The maximum
and minimum cycling pressure and the number of pressurization cycles is recorded. If the cylinder fails either
by leaking or by fracture, the failure mode and number of pressurization cycles to failure are recorded as the
test results. If the maximum number of pressurization cycles is reached without the cylinder failing, the
cylinder is confirmed as having adequate fatigue resistance.
In the development of the test method and acceptance criteria for the flawed-cylinder cycle test, it was
decided that the fatigue resistance of newer higher-strength steel cylinders should be essentially the same as
that of the lower strength existing cylinders because the existing cylinders have provided adequate fatigue
performance during their many years of service. Therefore, flawed-cylinder cycle tests were conducted on
cylinders with strength levels covering the full range of strength levels currently being produced in the world.
Tests were conducted on cylinders made from steels ranging in tensile strength from less than 800 MPa to
greater than 1 350 MPa. During the development of the flawed-cylinder cycle test, more than one hundred
flawed-cylinder cycle tests, were conducted.
The acceptance criteria for the flawed-cylinder cycle test was based on the maximum pressurization cycles
and the maximum pressure that a cylinder is likely to experience in service. The maximum number of
pressurization cycles was established based on a cylinder being filled rather frequently (e.g. once per day).
The cycle life of a cylinder having an acceptable defect at the time of manufacture or at the time of retesting
should therefore withstand an average 3 500 cycles within a 10 year re-testing period (i.e. 350 d × 10 years).
In addition, for the purpose of testing, it was assumed that the absolute maximum developed pressure in a
cylinder could be up to the design test pressure of the cylinder. Therefore, this pressure level was chosen for
the flawed-cylinder cycle test.
The size of the standard flaw that was machined in the test cylinders was based on the size of flaws that can
occur during manufacturing of the cylinder or that can be developed in service. For cylinders manufactured in
accordance with ISO 9809-2, an ultrasonic inspection was required for each cylinder at the time of
manufacture. The flaw detection sensitivity for this inspection is limited to 5 % of wall thickness. Therefore,
flaws developed by service abuse would not be of concern unless the flaws are deeper than 5 % of the wall
thickness. Furthermore, according to ISO 6406, during periodic inspection, flaws such as "cuts and gouges"
are acceptable provided the depth of the flaw does not exceed 10 % of the wall thickness. Therefore, WG 14
established that a standard flaw type for the flawed-cylinder cycle test that is similar to the flaw type used in
the flawed-cylinder burst test (ISO/TR 12391-2 and ISO/TR 12391-3) but with a smaller depth of 10 % of the
wall thickness would be appropriate for evaluation of the effect of service induced flaws on fatigue cycle life.
The standard flaw has a length of approximately 10 × wall thickness.
The flawed-cylinder cycle test is included in ISO 9809-2 as a design approval test. The test is used for the
design approval of all newly designed cylinders. The details of the test method and the acceptance criteria are
given in 9.2.6 of ISO 9809-2:2000.
This part of ISO/TR 12391 is limited to a summary and compilation of the results of the flawed-cylinder cycle
tests that were conducted by WG 14 during the development of the flawed-cylinder cycle test method. This
part of ISO/TR 12391 is in the form of a data-base of the test results and is intended to be used for further
analysis of the fatigue performance of steel cylinders and to define acceptable sizes of flaws for use at the
time of periodic inspection.
6 Experimental test programme
6.1 Types of cylinder tested
Flawed-cylinder cycle tests were conducted on cylinders that represented most of the currently used and
proposed new types of seamless steel cylinders. A brief description of all the cylinders that were tested is
shown in Tables 1 to 4.
The cylinders are classified in material groups based on strength level that is consistent with the classification
of the cylinder materials used in the WG 14 report on the flawed-cylinder burst test described in
ISO/TR 12391-2 and ISO/TR 12391-3. For this study, the cylinders were classified into material groups
(designated Group B to E) based on the actual measured tensile strength, R , of the cylinders that were
m
tested. No flawed-cylinder tests were conducted on cylinders of material group A strength levels (tensile
strength less than 750 MPa). The actual measured tensile strength, R , for each group of cylinders that was
m
tested is shown in Tables 1 to 4. The general description of the cylinders in each material group is shown
4 © ISO 2002 — All rights reserved

below. Cylinders made from materials in groups B to D are currently being produced and used throughout the
world. Cylinders made from material group E, are experimental and are not currently authorized for use.
Tensile strength R
Material Group Description of cylinder
m
B Cylinders made from alloy steel (Cr-Mo steels) heat 750 MPa < R u 950 MPa
m
treated by quenching and tempering; these cylinders
may generally be used for all gases.
C Cylinders made from alloy steel (Cr-Mo steels) heat 950 MPa < R u 1 080 MPa
m
treated by quenching and tempering; these cylinders are
restricted to use with non-corrosive gases made in
accordance with ISO 9809-1.
D Cylinders made from alloy steel (Cr-Mo steels) heat 1 080 MPa < R u 1 210 MPa
m
treated by quenching and tempering, high strength and
high toughness steel cylinders: these cylinders are
restricted to use with non-corrosive gases made in
accordance with ISO 9809-2.
E Experimental cylinders; extra high strength; not currently R > 1 210 MPa
m
authorized for use.
Within each main material group (FB to FE) shown in Tables 1 to 4, material subgroups are designated; e.g.,
material subgroup F-B-1, F-B-2. The material group coding, e.g. F-B-1 indicates that the test was a fatigue
cycle test (F) and the material strength was in the B group range (R = 750 MPa to 950 MPa). All the
m
cylinders within a given material subgroup were made to the same specification, of the same dimensions
(diameter, thickness and volume), the same material, the same specified tensile strength range, the same
designated service pressure and test pressure, and were made by the same manufacturing process. The
cylinders in a specific material subgroup (e.g. material subgroup F-B-2) may be of a different alloy, size,
design specification or manufacturing process than cylinders in a different materials subgroup (e.g. F-B-3) in
the same main material group (group F-B). However, the actual measured tensile strength for all cylinders in a
material group will be in the same range (e.g., 750 MPa to 950 MPa for all cylinders in group F-B).
In Tables 1 to 4, each flawed-cylinder cycle test is assigned a number in sequence, as shown in the first
column, for the purpose of tracking each test. The same number is then used to identify the cylinders in the
tables for the cycle test results (Tables 5 to 8). In addition, each individual cylinder tested is assigned a
number, such as F-B-1, as shown in the second column of the tables.
Additional information to fully describe each cylinder is shown in Tables 1 to 4. This information includes the
outside diameter of the cylinder, D, the minimum design wall thickness of the cylinder, t , the maximum design
d
test pressure, P , the maximum design service pressure, P , the actual wall thickness, t and the cylinder
h s a
volume (in litres).
It should be noted that in a few cases, the actual measured tensile strength (R ) for one or more cylinders in a
m
particular material subgroup is slightly outside the designated range for the tensile strength of the particular
material subgroup in which the cylinder is included. However, the measured tensile strength of the rest of the
cylinders from that material subgroup that were tested is within the appropriate tensile strength range for that
material subgroup.
6.2 Material properties tests
Conventional mechanical properties tests, such as tensile tests and Charpy-V-notch tests, were conducted on
each set of cylinders on which flawed-cylinder cycle tests were performed. The results of these tests are
shown in Tables 1 to 4 for each group of materials.
The tensile test results shown in Tables 1 to 4 are the actual measured yield strength, R , and the actual
ea
measured tensile strength, R . These materials properties are required to be measured by all of the existing
m
ISO cylinder design standards. The actual measured tensile strength, R value is used to determine whether
m
the cylinder meets the standard to which it is manufactured and is used in this test programme to determine in
which material group the tested cylinder should be placed. The actual measured yield strength, R , is used to
ea
determine whether the cylinder meets the requirement for the yield strength to tensile strength ratio when this
ratio is a part of the standard.
The Charpy-V-notch tests were conducted in accordance with the test method described in ISO 148:1983.
The Charpy-V-notch tests were conducted either at ambient temperature (20 °C) or at low temperature
(− 20 °C or – 50 °C ), as shown in Tables 1 to 4. The Charpy-V-notch test specimens were all oriented with
the longitudinal axis perpendicular to the longitudinal axis of the cylinder (designated transverse specimens).
The total energy absorbed in breaking the Charpy-V-notch test specimens was measured in joules (J). All
Charpy-V-notch test results are reported as J/cm , where the total energy absorbed is divided by the area of
the specimen ligament below the specimen notch. The Charpy-V-notch energy test results are not used to
evaluate the results of the flawed-cylinder cycle test. However, the Charpy-V-notch energy test results are
reported here because these results may be used to evaluate the fatigue and fracture performance of the
cylinders using alternate analysis procedures to the flawed-cylinder cycle test.
6.3 Description of the flawed-cylinder cycle test
The flawed-cylinder cycle test is used to evaluate the overall fatigue performance of the entire cylinder and is
used only as a “design approval test”. The full details of the test and the criteria for acceptable fatigue
performance of steel cylinders are given in 9.2.6 of ISO 9809-2:2000.
In the flawed-cylinder cycle test, the fatigue performance of the cylinder is evaluated by cyclically pressurizing
a cylinder with a designated type (shape and sharpness) and dimension (length and depth) of surface flaw,
until failure. The cylinder to be tested has a flaw machined into the exterior surface of the cylinder wall. The
flaw is machined in the location of probable maximum stress under pressurized loading, i.e. a longitudinal
surface flaw at mid-length and at thinnest place in the cylinder wall. To make the tests adequately uniform and
reproducible, a surface flaw with a standard geometry is required.
All tests carried out for this project were conducted in accordance with the requirements specified in
ISO 9809-2. These requirements are as follows.
A standard Charpy-V-notch milling cutter is used to machine the flaw to the designated length and depth. The
milling cutter is required to meet the following specifications:
 thickness of the cutter = 12,5 mm ± 0,2 mm;
 angle of the cutter = 45° ± 1°;
 tip radius u 0,25 mm ± 0,025 mm;
 for cylinders u 140 mm in diameter, cutter diameter = 50 mm ± 0,5 mm;
 for cylinders > 140 mm in diameter, cutter diameter = 65 mm to 80 mm.
The cycling frequency shall not exceed 5 cycles/min.
0,5
The flaw length l shall be 1,6 × (D × t )
o d
NOTE 1 For the specific test conducted here, the flaw length is approximately expressed as multiples of the design wall
thickness, i.e. n × t and is approximately 10 × t for all tests.
d d
The depth, d of the flaw shall be not less than 10 % of the wall thickness, t .
d
When measuring the actual flaw depth, a deviation not exceeding 0,1 mm is acceptable (e.g. for an actual wall
thickness of 7 mm the flaw depth shall in no case be less than 0,6 mm).
6 © ISO 2002 — All rights reserved

The “standard surface flaw” geometry is shown in Figure 1. The flaw length, l , is normally expressed in
o
multiples, n, of the cylinder design minimum wall thickness, t , (l = n x t ). The flaw depth is expressed as a
d o d
percentage of the cylinder design minimum wall thickness, t , i.e. flaw depth = d/t × 100.
d d
Pressurization is carried out hydrostatically. The requirements of ISO 9809-2 are that the maximum cyclical
pressure be equal to at least the design test pressure, P , and that the minimum cyclical pressures be 10 % of
h
the maximum cyclical pressure.
NOTE 2 Some of the tests conducted in this study used slightly different pressure ranges than specified by ISO 9809-2.
During the test, each cylinder is filled with water at room temperature and the pressure is cycled continuously
until the cylinder reaches the required number of cycles or fails by leaking or fracturing. The minimum and
maximum cyclical pressure, the number of cycles and the failure mode (if the cylinder fails during the test) are
reported in the test results.
The acceptance criterion for the flawed-cylinder cycle test specified by ISO 9809-2 is that the cylinder shall
have passed the test if the number of cycles attained without failure exceeds 3 500 as a mean value of the
two cylinders tested but with an absolute minimum of 3 000;
If the test is continued to failure, then the mode of failure shall be reported (i.e. leak or fracture).
9.2.4 of ISO 9809-2:2000 specifies a “pressure cycling test”. The pressure cycling test is conducted in the
same manner as the flawed-cylinder cycle test except that a cylinder without an external flawed is tested. The
acceptance test for the pressure cycling test is that the cylinder shall withstand 12 000 pressurization cycles
without failure. If the flawed-cylinder cycle test withstands a minimum of 12 000 cycles without failure (by
leaking or fracturing) then the pressure cycling test does not have to be carried out. Many of the tests
conducted in this programme satisfied this requirement.

a
Cutter angle
b
Cutter profile radius
c
Ligament
d
Cutter diameter
Figure 1 — Standard flaw geometry for the flawed-cylinder cycle test
7 Flawed-cylinder cycle test results
7.1 Flawed-cylinder cycle test procedure
The results of all of the flawed-cylinder cycle tests that were conducted are shown in Tables 5 to 8. For each
cylinder tested, the nominal flaw length, l , in terms of a multiple of the actual measured cylinder wall
o
thickness, t , is given as l = n × t (e.g. l = 10 t ). In a few cases, the actual measured wall thickness, t , was
a o a o a a
not given and the design wall thickness, t , was used (e.g. l = 10 t ) to calculate the nominal flaw length. The
d o d
nominal flaw length was used as a common reference to compare cylinders with different wall thicknesses.
The flaw depth, d, is given as a percentage of the design minimum cylinder wall thickness, t ,
d
(e.g. 100 × d/t = 10 %).
d
A flaw of the required size, generally 10 × t long and 10 % t deep is machined in the cylinder wall and the

a d
cylinder is cyclically pressurized to a maximum pressure equal to the test pressure for the required number of
cycles (usually 3 500 cycles). The test results reported in Tables 5 to 8 show the flaw length, the flaw depth,
the minimum and maximum cycling pressure, the total number of cycles and the failure mode. If the test was
continued until failure occurred, the failure mode was either leaking or fracture (defined as an extension of the
original flaw by at least 10 %). For some tests, no failure occurred after a large number of pressurization
cycles. These tests are considered to be “run-outs” and the failure mode is shown in Tables 5 to 8 as “none”.
7.2 Flawed-cylinder cycle test results for group F-B materials
The cylinders made from group F-B materials have the lowest tensile strength of the cylinders tested in this
programme. These cylinders have measured tensile strengths of less than 950 MPa and are representative of
a number of cylinders that have been in worldwide use for about 60 years. Cylinders of this type normally
have a service pressure rating of less than 200 bar.
All of the cylinders tested had an initial flaw depth of at least 10 % t and a nominal flaw length of 10 × t . As
d a
shown in Table 5, the fatigue cycle life for the all of the cylinders in material group F-B exceeded 3 500 cycles
at a maximum cyclical pressure equal to the cylinder test pressure that is required by ISO 9809-2. All of the
cylinders in material subgroups F-B-1 and F-B-2 failed by leaking after at least 20 000 cycles. The cylinders in
material subgroups F-B-3 and F-B-4 failed by both leaking and fracture after at least 7 500 cycles.
The fatigue cycle life of all but one of the cylinders in material group F-B exceeded 12 000 cycles and
therefore satisfied the requirements given in 9.4.2 of ISO 9809-2:2002 for the “Pressure cycling test” for an
unflawed-cylinder.
7.3 Flawed-cylinder cycle test results for group F-C materials
The cylinders made from group F-C materials have measured tensile strengths ranging from 950 MPa to
about 980 MPa. Cylinders of this type have a normal service pressure rating ranging from about 200 bar to
about 300 bar. These cylinders are manufactured according to the requirements of ISO 9809-1.
All cylinders had an initial flaw depth of at least 10 % t and a nominal length of 10 × t . As shown in Table 6,
d a
the fatigue cycle life for the all of the cylinders in material subgroups F-C-1, F-C-2 and F-C-4 had an average
of 3 500 cycles when the maximum cyclical pressure was equal to the cylinder test pressure and the minimum
cyclical pressure was 10 % of the maximum cyclical pressure. Therefore, all of these cylinders satisfied the
requirements of ISO 9809-2. All of these cylinders failed by leaking after at least 6 000 cycles.
As shown in Table 6, the cylinders in material subgroup F-C-3 were tested with the maximum cyclical
pressure approximately equal to the design service pressure of the cylinders instead of equal to the test
pressure of the cylinders. Therefore, these tests were not conducted in full compliance with the requirements
of ISO 9809-2 which requires that the maximum cyclical pressure to be equal to the test pressure rather than
the design service pressure. However, because the fatigue cycle life of these cylinders exceeded
14 000 cycles when the maximum cyclical pressure was equal to the design service pressure, it is likely that
they would have exceeded the requirement of 3 500 cycles when tested at a maximum cyclical pressure equal
to the test pressure which is 3/2 times the design service pressure.
8 © ISO 2002 — All rights reserved

The fatigue cycle life of cylinders in material subgroups F-C-1 and F-C-4 exceeded 12 000 cycles when tested
at a maximum cyclical pressure equal to the test pressure and therefore satisfied the requirements given in
9.4.2 of ISO 9809-2:2000. The fatigue cycle life of cylinders in material subgroup F-C-2 did not satisfy the
12 000 cycles required by 9.4.2 of ISO 9809-2:2000. The cylinders in material subgroup F-C-3 were tested at
a maximum cyclical pressure equal to the design service pressure rather than the test pressure. Although the
fatigue cycle life of cylinders in material subgroup F-C-3 exceeded 12 000 cycles when tested to maximum
cyclical pressure equal to design service pressure, it is unlikely that the fatigue cycle lift would have exceeded
12 000 cycles if the tests had been conducted at the test pressure and therefore, these cylinders would have
to be tested with no flaw in order to determine if they meet the requirements given in 9.4.2 of ISO 9809-2:2000.
7.4 Flawed-cylinder cycle test results for group F-D materials
The cylinders made from group F-D materials are the highest strength steel cylinders currently being
manufactured. These cylinders are manufactured to ISO 9809-2 requirements. They are restricted to use for
shipping non-corrosive (non-hydrogen bearing) gases. These cylinders are generally made from modified
chromium-molybdenum alloy steels that have a good combination of tensile strength and fracture toughness.
These cylinders normally have a design service pressures at or above 300 bar.
Most of the cylinders used in these flawed-cylinder cycle tests had an initial flaw depth of at least 10 % t and

d
a nominal flaw length of 10 × t as required by ISO 9809-2. However, a few of the cylinders had initial flaw
a
depths greater than the required 10 % t (e.g. 12 %, 14 %, 15 %, 50 % and 60 %) and a few had initial flaw
d
depths less than the required 10 % t (e.g. 5 % and 6 %). When tested at the same pressure range, the effect
d
of the deeper initial flaws is to reduce the total number of cycles to failure and the effect of the shallower flaws
is to increase the number of cycles to failure. Most of the cylinders used in these flawed-cylinder cycle tests
were tested to a pressure cycle range with the maximum pressure equal to the design test pressure of the
cylinder and the minimum pressure equal to 10 % of the maximum pressure as required by ISO 9809-2.
However, a few tests were conducted with the maximum pressure equal to the design service pressure
(e.g. material groups F-D-4 and F-D-6) or to a maximum pressure of twice the design service pressure
(e.g. material group F-D-5). Although most of the tests were conducted with the minimum cyclical pressure
equal to 10 % of the maximum cyclical pressure, a few were conducted with the minimum cyclical pressures
set at about 1 % of the maximum cyclical pressure or at zero pressure (e.g. material groups F-D-3, F-D-7,
F-D-8, F-D-9, F-D-10, F-D-11 and F-D-18). The effect of this larger pressure cycling range, in which the
minimum cyclical pressure range is less than 10 % of the maximum cyclical pressure, is to reduce the number
of cycles to failure.
As shown in Table 7, the fatigue cycle life for the all of the cylinders in material group F-D had an average of
3 500 cycles when the maximum cyclical pressure equal to the cylinder test pressure and the minimum
cyclical pressure was 10 % of the maximum cyclical pressure and the initial flaw depth was 10 % t . Therefore,
d
all of the cylinders tested satisfied the requirements of ISO 9809-2.
For the tests conducted according to the requirements of ISO 9809-2, most of the cylinders failed by leaking
after exceeding the required 3 500 pressurization cycles. The exceptions were some cylinders in material
groups F-D-3, F-D-4, F-D-12, F-D-18 and F-D-20 that failed by fracture.
The cylinders that were tested under conditions that differed from those required by ISO 9809-2 did not meet
the requirement that the fatigue cycle life exceed 3 500 cycles without failure. The cylinders in material group
F-D-4 that had initial flaw depths of 50 % and 60 % of the wall thickness failed by leaking at less than
700 pressurization cycles. The cylinders in material group F-D-5 that were tested at a maximum cyclical
pressure equal to more than twice the design service pressure failed by fracturing at 2 217 and
4 200 pressurization cycles. Two of the cylinders in material subgroup F-D-6 were tested with the maximum
cyclical pressure at the design service pressure instead of the design test pressure. The fatigue cycle life of
these cylinders greatly exceeded the cycle life of 3 500 cycles required by ISO 9809-2. However, the one
cylinder from material subgroup F-D-6 that was tested with the maximum cyclical pressure at the design test
pressure had a fatigue cycle life of 7 589 cycles and therefore fully satisfied the requirements of ISO 9809-2
for the flawed-cylinder cycle test.
For the tests conducted according to the requirements of ISO 9809-2, most of the cylinders in material groups
F-D-1, F-D-3, F-D-14, F-D-15, F-D-17, F-D-18 and F-D-20 had a fatigue cycle life of greater than
12 000 cycles and therefore satisfied the requirements given in 9.4.2 of ISO 9809-2:2000. Cylinders in
material groups F-D-2, F-D-4, F-D-7, F-D-8, F-D-9, F-D-10, F-D-11, F-D-12, F-D-13 and F-D-19 had a fatigue
cycle life of less than 12 000 cycles and therefore did not satisfy the requirements given in 9.4.2 of
ISO 9809-2:2000. Therefore, additional tests using unflawed-cylinders would need to be conducted in order to
fully satisfy the fatigue requirements of ISO 9809-2.
7.5 Flawed-cylinder cycle test results for group F-E materials
The cylinders made from group F-E materials have the highest strengths of any cylinders tested in this
programme. These cylinders have measured tensile strengths of greater than 1 300 MPa. This is higher than
currently permitted by any safety regulations in the world. These cylinders are experimental cylinders for
evaluating the feasibility of using higher strength steels in cylinders without risking failure by fatigue or fracture
in service.
All cylinders had an initial flaw depth of at least 10 % t and a nominal len
...