ISO/TR 13086-4:2024

Technical
Report
ISO/TR 13086-4
Second edition
Gas cylinders — Guidance for
2024-11
design of composite cylinders —
Part 4:
Cyclic fatigue of fibres and liners
Bouteilles à gaz — Recommandations pour la conception des
bouteilles en matière composite —
Partie 4: Fatigue cyclique des fibres et liners
Reference number
© ISO 2024
All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may
be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting on
the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address below
or ISO’s member body in the country of the requester.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: +41 22 749 01 11
Email: copyright@iso.org
Website: www.iso.org
Published in Switzerland
ii
Contents Page
Foreword .iv
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 Background . 1
5 Cyclic fatigue evaluation . 2
6 Elements of cyclic fatigue . 2
6.1 Service conditions and requirements .2
6.1.1 Temperature and moisture .2
6.1.2 Pressure .3
6.1.3 Pressure cycles .3
6.2 Test conditions and specimens .4
6.3 Fibre materials and their fatigue properties .6
6.3.1 Materials .6
6.3.2 Material properties and data .6
6.3.3 Hybrid construction .7
6.4 Liner materials and their fatigue properties .8
6.4.1 Materials used .8
6.4.2 Material properties and data .8
6.4.3 Issues with localized strain differences .8
6.5 Resin materials and their fatigue properties .9
6.6 Composite/liner load sharing .9
6.7 Autofrettage .9
6.8 Analysis methods .10
6.9 Leak before burst (LBB) .10
6.10 Damage tolerance.11
6.11 Aging and environment .11
6.12 Counting cycles .11
6.13 Combining cycles . 13
6.14 Qualification testing . 13
7 Summary and conclusions .13
Annex A (informative) Equivalent pressure cycling . 14
Bibliography .20

iii
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.
The procedures used to develop this document and those intended for its further maintenance are described
in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the different types
of ISO document should be noted. This document was drafted in accordance with the editorial rules of the
ISO/IEC Directives, Part 2 (see www.iso.org/directives).
ISO draws attention to the possibility that the implementation of this document may involve the use of (a)
patent(s). ISO takes no position concerning the evidence, validity or applicability of any claimed patent
rights in respect thereof. As of the date of publication of this document, ISO had not received notice of (a)
patent(s) which may be required to implement this document. However, implementers are cautioned that
this may not represent the latest information, which may be obtained from the patent database available at
www.iso.org/patents. ISO shall not be held responsible for identifying any or all such patent rights.
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and expressions
related to conformity assessment, as well as information about ISO's adherence to the World Trade
Organization (WTO) principles in the Technical Barriers to Trade (TBT), see www.iso.org/iso/foreword.html.
This document was prepared by Technical Committee ISO/TC 58, Gas cylinders, Subcommittee SC 3,
Cylinder design.
This second edition of ISO/TR 13086-4 cancels and replaces the first edition (ISO/TR 13086-4:2019) which
was technicall revised.
The main changes are as follows:
— editorial and technical changes throughout the document.
A list of all parts in the ISO 13086 series can be found on the ISO website.
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www.iso.org/members.html.

iv
Technical Report ISO/TR 13086-4:2024(en)
Gas cylinders — Guidance for design of composite
cylinders —
Part 4:
Cyclic fatigue of fibres and liners
1 Scope
This document addresses the topic of cyclic fatigue of structural reinforcing fibres as used in composite
cylinders, and cyclic fatigue of structural and non-structural liners in these cylinders. This document
provides a basic level of understanding of these topics.
2 Normative references
There are no normative references in this document.
3 Terms and definitions
No terms and definitions are listed in this document.
NOTE Terms and definitions related to gas cylinders can be found in ISO 10286.
ISO and IEC maintain terminological databases for use in standardization at the following addresses:
— IEC Electropedia: available at https:// www .electropedia .org/
— ISO Online browsing platform: available at https:// www .iso .org/ obp
4 Background
Composite cylinders began service in the 1950s, initially as rocket motor cases with glass fibre reinforcement.
This soon led to glass fibre pressure vessels with rubber liners, and then to glass fibre pressure vessels with
metal liners. Metal liners were typically either aluminium alloy or steel. Eventually, new structural fibres,
such as aramid and carbon, came into use for reinforcing pressure vessels. Today, typical reinforcements
for composite gas cylinders are glass and carbon, either individually or together as a hybrid. Typical
liner materials are steel, aluminium alloy or polymers, for example, high-density polyethylene (HDPE) or
polyamide (PA); other materials could be acceptable.
Each of these materials is subject to cyclic fatigue based on the type of service and the construction of the
cylinder. Cylinders used in transport service generally see full range cycles, with a limited number of cycles
per year. Cylinders used as fuel containers typically see up to three pressure cycles per day for fleet vehicles,
and less for private vehicles. Cylinders used in stationary applications such as refuelling cascades could see
a very large number of partial cycles in a year. Some cylinders could see a combination of these conditions.
Stationary cylinders used for fuel cells or emergency breathing applications could see a very limited number
of cycles. Design working pressures for high pressure cylinders are typically in the range of 20 bar to
1 100 bar. Cylinders for liquified gases such as propane can operate at pressures up to 20 bar, and normally
see fewer pressure cycles.
The different reinforcing fibres have different fatigue lives for a given stress or strain range. Liner materials
can also have different fatigue lives for a given stress or strain range. The load-sharing characteristics
of a liner material with a given reinforcement will affect their fatigue lives. An autofrettage cycle is used

with metal lined cylinders to improve fatigue life. The low modulus of elasticity of polymer liner materials
often results in the liner being in compression when the cylinder is pressurized, so their fatigue life could
be very high. Welds in a liner, whether it is metal or polymer, can affect the fatigue life due to the different
mechanical properties in a weld and in heat affected zones.
Surface quality and conditions such as roughness will affect cyclic fatigue, particularly crack initiation in
Type 2 or Type 3 cylinders. Autofrettage generally blunts cracks, and adds surface compression, which will
improve fatigue life.
Evaluation and understanding of cyclic fatigue will lead to improved designs and reduce the risk of cyclic
fatigue failures without the need to overdesign the cylinders or conduct extensive qualification testing on
each new design.
5 Cyclic fatigue evaluation
Cyclic fatigue of composite cylinders can be addressed with an understanding of
— service conditions and requirements,
— test conditions and specimens,
— fibre materials and their fatigue properties,
— liner materials and their fatigue properties,
— resin materials and their fatigue properties,
— composite/liner load sharing,
— autofrettage,
— analysis methods,
— leak before burst (LBB),
— damage tolerance,
— aging and environment,
— counting and combining different cycles, and
— qualification testing.
6 Elements of cyclic fatigue
6.1 Service conditions and requirements
6.1.1 Temperature and moisture
Service conditions depend largely on location and usage of the cylinder. If the cylinders are located and used
outdoors, they must be able to withstand ambient conditions. Common conditions include temperature
ranges from −40 °C to +85 °C (−40 °F to +185 °F), which include higher temperature exposure due to solar
input and storage in confined spaces. This could include use in a vehicle or shipment in a rail car where direct
sunlight will raise temperatures within the storage compartment. Surface absorptivity and emissivity of the
cylinder can affect solar input to the cylinder and its equilibrium temperature. It is less common to require
operation in temperatures to -55 °C (−67 °F), and in some cases to even lower temperatures.
Moisture levels in outdoor locations can range from very high to very low depending on ambient conditions.
Some cylinders are actually located in a water bath. Moisture itself generally does not affect fatigue of
structural materials used in cylinders, but can cause corrosion, which could affect fatigue life. Moisture can
also be absorbed into polymer liners, and resulting property changes need to be understood by the cylinder

designer. Moisture can also bring in chemicals that could affect material strength and fatigue properties,
particularly those glass fibres that are not corrosion resistant.
Some cylinders are maintained in a controlled environment, such that temperature and moisture are
monitored and controlled. However, conditions must be guaranteed if a controlled environment is needed
to meet fatigue requirements. Otherwise, it is best to assume the cylinders will be exposed to worst-case
conditions.
Temperature and moisture changes from a reference point can cause dimensional changes in the cylinder
components, which can likely result in stresses within the cylinder. These stresses can result from either
[2]
transient or steady-state conditions of temperature and moisture, as shown by Newhouse .
6.1.2 Pressure
Working pressures typically range from 5 bar to 20 bar for liquified gas applications, and up to 1 100 bar
for compressed gas applications, with allowance for pressure increases due to temperature increases. The
maximum allowable working pressure in stationary applications, more commonly known as the design
pressure or maximum service pressure for this application, is the maximum pressure the cylinder can be
exposed to. The pressure could be at the design limit regardless of the service temperature.
In transportable and vehicle fuel container applications, the working pressure is the settled pressure
at 15 °C, and can increase up to about 130 % of the rated working pressure during extreme temperature
conditions. Operating pressures will be below the rated working pressure when ambient temperature drops
below 15 °C. Note that in North America, the reference temperature is usually 21,1 °C (70 °F).
Cylinders in various applications can also be subject to test pressures that are generally 150 % of the rated
working pressure, but can range from 125 % to 167 % of the rated working pressure, with generally not more
than 50 such cycles over a lifetime. Although some cylinder standards or regulations allow pressurizing to test
pressure during fill, cylinders must not be filled with more gas than can settle to working pressure at 15 °C.
6.1.3 Pressure cycles
Some applications require only a limited number of cycles in a lifetime, so fatigue evaluation is not a
significant concern. Such applications include emergency breathing cylinders, and fuel containers for fuel
cells providing power when primary power is out of service. It can also include applications where the
cylinder is in limited use, and could only experience one or two pressure cycles in a month.
Transportable cylinders are generally designed for a specified lifetime, either limited or non-limited, and
qualified by conducting a specified number of cycles. A typical qualification test requirement is 12,000 cycles
to test pressure, or in dedicated gas service 24,000 cycles to maximum developed pressure, for a non-limited
life. For a limited life, cycling 250 times to test pressure, or in dedicated gas service 500 times to maximum
developed pressure, per year of design life is a common requirement. Specific standards could require more
or less cycles. Transportable cylinders are generally not expected to be filled more than once a day, and
cycling to the test pressure provides a margin of safety.
Vehicle fuel tanks, containing either natural gas or hydrogen, could be filled two to three times a day in
fleet use, such as in buses or medium- and heavy-duty trucks. This is the basis for qualification testing of
750 cycles to 1 000 cycles per year used in fuel container standards.
Stationary cylinders, generally referred to as pressure vessels, could be subject to a high number of pressure
cycles. One such application is use as a refuelling cascade for natural gas and hydrogen powered vehicles.
These cylinders could be in use continually as vehicles are brought in for refuelling, resulting in a high
number of cycles per day. In some cases, the cylinders could be refuelled from another fuel reservoir, such
as from a pipeline, as soon as the pressure begins to drop. These cylinders can see a very high number of
partial cycles. Some cylinders can see a high number of partial cycles, combined with a given number of full
cycles, in the course of a day.

6.2 Test conditions and specimens
Testing is generally conducted at ambient temperature. Care is taken to avoid testing at temperatures that
can affect test results. Consideration is given to actual conditions, and how that can affect fatigue results.
Low temperatures can increase strength of the material being tested, but can also cause embrittlement
that decreases the fatigue life. High temperatures can decrease the strength of material being tested.
Extreme temperatures will also affect load share between liner and overwrap materials due to differences
in thermal coefficient of expansion, and will also affect stress distribution if hybrid construction is used
for the composite overwrap. For example, as temperature decreases, an aluminium alloy liner tends to
decrease interface pressure with the composite overwrap, causing the liner to carry a larger percentage
of the pressure load. Analysis needs to be conducted to evaluate the effect of temperature on stresses and
strains within an actual cylinder.
Testing with liquid versus gas to pressurize a cylinder results in the same pressure on the inside of the
cylinder, and therefore the same stress in the cylinder. However, there can be temperature differences
resulting from the use of different fluids, depending on energy to compress the fluid. This could also be
a consideration for the service conditions, although filling and discharge are generally over a longer time
period in service compared with testing.
Fibre strength in the helical and hoop directions is the basic design criteria for design of the composite
overwrap for the cylinder. As cylinder design pressure increases, laminate thickness is increased in order
to maintain the stress and strain at the same level vs. their design stress and strain in the helical and hoop
directions. Although the peak fibre stresses generally remain the same, the radial compressive stress
increases in the inner part of the laminate. This change in stress conditions can have a significant effect on
the fatigue life of the composite and of a metal liner. Therefore, consideration is given to test pressure versus
service pressure when evaluating fatigue life.
Options for test specimens to evaluate laminate strength and fatigue resistance include flat coupons, tube
sections, and cylinders. Each option has advantages and disadvantages. As the test specimen gets closer to
the actual product configuration, the results will be more valid, but more difficult to obtain.
Flat coupons can include unidirectional specimens and cross-plied laminates. These specimens could be
suitable for comparisons between fibres as to strength and fatigue properties, but are generally not suitable
from which to predict cylinder performance directly. Loading is only in the principle direction, unlike the
three-dimensional loading of a pressure cylinder. If loaded in tension, consider the stress concentrations
caused by the grips, and the geometry of the specimen, including edge effects. If loaded in bending, consider
that the specimen loading is further removed from the type of loading seen in a cylinder. Nevertheless, the
ability to quickly test comparative specimens can have some value.
A flat coupon is not suitable for evaluating the interaction between a metal liner and a composite overwrap.
[3] [4]
Tube specimens can include unidirectional specimens and cross-plied laminates. NOL or ASTM rings
are one option for unidirectional tubular specimens. Tube specimens can also be wound with helical and/or
hoop layers over a longer cylindrical mandrel. Cross-plied tube specimens could be tested in axial tension
using end grips that interface with tube ends that have additional reinforcement to avoid grip failures. That
is, the tube has similarity to a flat tensile specimen with wider or thicker ends (i.e. a “dog-bone” specimen).
Tubular specimens can also be tested using internal pressure. The resultant will be hoop stress if the
pressure source was contained within a double ended piston, so that axial load was contained within the
piston. Alternatively, the tube will experience both hoop load and axial tension if the tube ends were closed
such that the end closures apply tension to the tube, such as when doing an axial tension test, but using the
internal pressure for loading.
A tubular specimen loaded in either axial tension or in hoop loading has advantages over a flat specimen
given that it is testing of a curved specimen, but the single direction loading has limitations. As with a flat
specimen, it is suitable for comparisons between fibres as to strength and fatigue properties, and gives more
representative results, but is still not as accurate as an actual specimen. A tubular specimen loaded in both
axial tension and hoop loading will have an even greater fidelity, with consideration to the level the laminate
reflects the construction of an actual cylinder.

Cylinder specimens give the best fidelity when assessing strength and fatigue life. However, the relative cost
can make them less attractive for a study involving many specimens. Subscale cylinder specimens offer an
option for good fidelity at a lower cost than full scale cylinders.
Figure 1 shows how fatigue results can vary with the choice of test specimen. The upper line, with data from
[5]
Mandell , reflects use of a unidirectional carbon fibre reinforced specimen loaded in tension. The middle
[6]
line, with data from Liber and Daniel, reflects use of a flattened tube with a symmetric laminate having
longitudinal fibre layers and ±45° layers loaded in tension. This construction results in a more complex
laminate, with more complex loading within the laminate. The data from this specimen shows a reduction in
fatigue life compared with the test specimen using unidirectional fibre.
The lower line reflects cyclic pressure testing of high pressure gas cylinders. These cylinders have a more
complex laminate and loading within the laminate than the test specimens of Mandell and of Liber and
Daniel. The lower line reflects a reduction in life compared with the other two specimens, but it does contain
some conservativism. The points plotted reflect test cycles conducted, but not necessarily with a resulting
test failure. It therefore represents a lower limit on fatigue life, rather than an average life.
Key
x
X log cycles (10 ) 1 mandell
Y fraction of ultimate strength 2 liber and daniel
3 pressure vessel
Figure 1 — Fatigue results using different configuration test specimens
The data presented in Figure 1 reflects what was stated above, that cyclic fatigue performance depends
on laminate construction, method of loading and other factors. Resin selection and laminate construction
can also affect results, as load transfer through the wall is dependent on radial laminate properties. It is
therefore accepted to demonstrate cyclic fatigue performance on a representative gas cylinder to properly
address fatigue performance in service due to pressure cycling. Note that the lines from Mandell and Liber
and Daniel reflect mean values, while the pressure vessel line reflects the methodology of this report.
The data presented in Figure 1 also indicates that cyclic fatigue testing can be accelerated by increasing
the upper pressure limit. Sufficient cyclic fatigue data, over a range of stress levels, is needed to get
representative results.
6.3 Fibre materials and their fatigue properties
6.3.1 Materials
Common composite reinforcing materials include glass, aramid and carbon fibres, generally filament wound
with an epoxy or vinyl ester resin matrix. Other resin matrix materials could be suitable. Other reinforcing
fibres could be available, but none have developed as being viable alternatives to glass, aramid and carbon
fibre at this time.
Glass fibre was the first to be developed and was in use in the 1950s and 1960s. The most commonly used
grade for gas cylinders is ECR-glass. This is fundamentally an E-glass, but has enhanced corrosion resistance
resulting from removal of boron from the glass formulation. Other grades of glass fibre are suitable, but
are less widely used. Glass fibre is essentially a super-cooled liquid, and is subject to creep flow and surface
cracking. It has the least resistance to fatigue failure of the three commonly used fibre types.
Aramid fibre (aromatic polyamide) was developed in the 1960s and came into use in gas cylinders in the
1970s. It has greater strength, lower density and improved fatigue resistance compared with glass fibre. It
has a long-chain molecular structure, with very high strength in the longitudinal direction, but relatively
weak transverse properties.
Carbon fibre suitable for use in gas cylinders was developed in the 1960s and 1970s. It came into widespread
use in commercial gas cylinders in the 1990s. Carbon fibre is more of a crystalline structure, and is generally
processed from a PAN precursor. It has higher tensile strength and modulus than glass and aramid fibre.
Carbon fibre has the best fatigue resistance of the commonly used fibre reinforcements, but is more sensitive
to mechanical impacts.
6.3.2 Material properties and data
Table 1 provides typical properties for glass, aramid and carbon fibres. Actual fibres used could have higher
or lower values, particularly for strength and modulus, depending on the characteristics of the fibre.
Table 1 — Typical fibre properties
Property ECR-Glass Aramid Carbon
a
Tensile strength, MPa (ksi) 1 500 (220) 2 500 (360) 4 500 (650)
b
Working strength, MPa (ksi) 430 (63) 830 (120) 2 000 (290)
Tensile modulus, GPa (msi) 72 (10,5) 131 (19) 220 (32)
Density, g/cc (pounds per cubic inch) 2,55 (0,092) 1,44 (0,052) 1,80 (0,065)
a
Nominal design fibre strength in the hoop direction of a pressure vessel at minimum burst
pressure.
b
Nominal design fibre strength in the hoop direction of a pressure vessel at service pressure.
NOTE ECR refers to corrosion resistant E-glass, from which boron has been removed as a constituent.
[5]
Figure 2 compares nominal cyclic fatigue for glass, aramid and carbon fibre.

Key
X log (cycles to fail, N) 1 carbon
Y maximum stress/static strength 2 aramid
3 glass
Figure 2 — S/N data for carbon, aramid and glass reinforcement
6.3.3 Hybrid construction
Some gas cylinders are manufactured using hybrid construction. That is, using two or more different
reinforcing fibres in the gas cylinder. This could be a combination of a glass and carbon fibre, or it could be
a combination of two different carbon fibres. This can be in the form of intraply hybrids, where there are
different fibres within a single winding band, or it can be alternating layers of fibres.
Hybrid construction is discussed here in terms of structural reinforcement. However, in some cases,
materials such as glass fibre can be wound on the outside surface as a protective layer. This layer is generally
not considered structural, but there is an awareness that it does have a contribution to the structure and
could share load.
The evaluation and analysis of hybrid construction can be accomplished by considering the basic elements
being evaluated. With a layered hybrid, each layer can be modelled using the properties of the single
material, with consideration for orientation. With an intraply hybrid, consider the number of fibre tows of
each material in the strand, the cross-sectional area of each tow, and the mechanical properties of each fibre,
in order to calculate the equivalent properties.
The concept of generalized plane strain applies to calculation of mechanical properties and strain within the
band and laminate. That is, all tows within the band have the same axial strain. The mechanical properties
in the fibre direction are based on the effective area and modulus of each material. Once the strains for the
laminate have been calculated, the strain in each fibre, along with the elastic modulus of the fibre, determine
the stress in the fibre.
In-plane transverse property calculation is a bit more involved, as the materials are in series rather than in
parallel. However, computer software is available that will evaluate the properties of a hybrid band. Also,
the in-plane transverse stiffness of the laminate is less significant than the properties in the direction of the
band, so the properties in the direction of the band dominate the laminate response to loading.

6.4 Liner materials and their fatigue properties
6.4.1 Materials used
Liners are generally made of metals or polymers. Aluminium alloys, carbon steel, CrMo steel or stainless
steel are the most commonly used metal liner materials. Polymer liner materials are often high-density
polyethylene (HPDE), or polyamide (PA) materials.
Metal liners generally carry part of the structural load, with consideration for liner thickness, as a result
of their relatively high elastic modulus. Given that the composite material is the primary reinforcement
material, it is generally the dominant factor in the strains seen at working pressure, and could result in the
metal liner being used at strain levels in the liner that are greater than is used in an all-metal cylinder of the
same material. High strain levels will have an influence on the liner fatigue life.
Polymer liner materials are generally non-load sharing, regardless of thickness. The strain in the liner is
fully dependent on the composite. Given that the polymer elastic modulus is low, it is generally in a state of
compressive stress at working pressure, as the Poisson’s effects from radial pressure overcome the in-plane
tensile strain.
Polymer liners generally have a metal end boss. Materials suitable for liners are generally suitable for end
bosses. Many other metals or alloys are also suitable for use as an end boss on a polymer liner, as the end
boss can be designed with more options for geometry than can the ends of one-piece (seamless) liners.
6.4.2 Material properties and data
Properties of some commonly used metal and polymer liner materials are given in Table 2 and Table 3. These
properties are provided for general information, and can be confirmed before use in analysis. Other metal
alloys or polymers are also suitable for use as liners.
Table 2 — Typical metal liner properties
Property Aluminium alloy Carbon steel Stainless steel
Alloy-heat treat 6 061-T6 4 340 SS 316
a
Tensile strength, MPa 260 690 580
Elongation, % 10 21 50
Tensile modulus, GPa 68 200 190
Poisson’s ratio 0,33 0,30 0,29
Density, g/cc 2,71 7,85 8,03
a
Nominal design strength in a pressure cylinder at minimum burst pressure.
Table 3 — Typical polymer liner properties
Property HDPE PA
Tensile strength, MPa 27 72
Elongation, % 600 100
Tensile modulus, GPa 0,97 1,45
Poisson’s ratio 0,40 0,40
Density, g/cc 0,96 1,13
6.4.3 Issues with localized strain differences
The cylindrical portion of a gas cylinder is generally at a constant thickness, and generally has equal strain
at all locations. Therefore, the fatigue characteristics of the composite and liner are nearly constant over

this region. However, there are localized conditions that require more detailed evaluation for fatigue. Some
examples are:
— The liner could have stress concentrations if it is welded.
— There could be bending stresses in the dome region that add to the extensional stresses.
— There are generally stress/strain discontinuities at the juncture of the cylinder and dome.
— The boss flange and neck could be different materials (as is the case with a polymer liner) and thicknesses
than the liner membrane.
— The transition from the flange to the membrane thickness of the liner generally results in strain
concentrations.
6.5 Resin materials and their fatigue properties
Resin matrix materials used in the composite construction are generally thermoset, but can also be
thermoplastic. Thermoset materials are typically epoxy, vinyl ester or polyester. The resin matrix is
primarily used to hold the fibre materials in place, and to facilitate transfer of radial compressive loads with
the laminate.
The resin matrix does not contribute significantly to the in-plane strength of the laminate. It will generally
micro-crack, or craze, when taken to test pressure, and this micro-cracking could progress further as the
cylinder is pressure cycled. This micro-cracking could have some effect on the fatigue life of the laminate.
6.6 Composite/liner load sharing
The load sharing between the laminate and the liner, and between fibres within a laminate, will have an
influence on fatigue. The in-plane strains, in particular, are affected by load sharing. Load sharing will
depend on relative fibre stiffness within hybrid composites, and will depend on the stiffness of the laminate
compared with the stiffness of the liner.
Load sharing could depend on manufacturing processes. One factor is the fibre tension during winding. High
tension will apply a compressive pre-load to the liner. However, resin can flow out of the laminate during
winding, reducing this compressive pre-load somewhat. Thermosetting resins can be more subject to such
flow than thermoplastic resins.
A second factor in load sharing is the effective “stress-free” temperature of the composite and liner. The
temperature during the winding process, and during the curing process, will affect the “stress-free”
temperature. The difference between the “stress-free” temperature and the ambient temperature can result
in tensile or compressive stresses in the liner or composite. Note that “stress-free” can be relative, as some
thermally induced stresses could always be present in the laminate.
A third factor in load sharing involved the autofrettage process, which is discussed in 6.7. Autofrettage will
result in residual compressive stress in the liner, and residual tensile stress in the composite.
6.7 Autofrettage
Autofrettage is one method of improving cyclic fatigue life of a metal liner. The cylinder is pressurized to
achieve liner yielding. Autofrettage pressure can be held for a sufficient length of time that the liner does not
yield further. A typical time is 1 min. This preload results in a net compression in the liner at zero pressure,
thereby reducing the average stress in the liner while pressurized. A reduced average stress during pressure
cycling increases the fatigue life of a metal liner. The yielding of the liner during autofrettage also blunts any
stress concentrations (e.g. cracks), which will increase the fatigue life of the liner.
However, the compressive preload in the liner results in a tensile preload in the composite reinforcement.
Care is taken to avoid putting too much preload into the composite, as it could result in excessive stress,
[7]
resulting in violation of the stress ratio requirements, as explained further in ISO/TR 13086-3 .

Autofrettage will not improve the cyclic fatigue life of a polymer liner, given that it is non-structural, and
generally will not yield during pressurization.
6.8 Analysis methods
Fatigue performance of the composite reinforcement and the liner are generally confirmed by means of
performance testing. However, analysis of the cylinder to determine stresses and strains, particularly where
there could be localized stresses and strain peaks, is of great value in understanding margins of safety, and
evaluate means to improve on the fatigue performance of the cylinder reinforcement and liner.
Finite element analysis is a generally accepted method of determining stresses and strains throughout the
cylinder. There are closed form methods that can also be used, particularly in the cylindrical portion of the
gas cylinder, where the stresses and strains are not dependent on axial or circumferential location.
Regardless of the analysis method, it is appropriate to validate the analysis results by means of strain gages,
displacement gages, or other mechanical means to measure the actual strains and/or displacements.
Knowing the stresses and strains during cycling, the fatigue life of the composite reinforcement can be
predicted by, and compared with, data such as that presented in Figures 1 and 2. The stresses and strains
in a polymer can similarly be used to predict liner cyclic fatigue life. However, it is noted that, depending on
the modulus of the polymer, the liner will likely be in compressive stress when pressurized, so there is little
tendency for crack formation and growth.
For a metal liner, knowing the stresses and strains, and the effect of contained gases (such as hydrogen), will
facilitate prediction of liner cyclic fatigue life. Fracture mechanics is one method to predict life of the metal
[8]
liner. A second method is to use fatigue curves generated by testing, such as those in MIL-HDBK-5J . A third
method is to use empirical methods based on evaluation of data.
[9]
One such empirical method is the low cycle fatigue equation developed by S.S. Manson :
V
σ
 1 
V V
ult
3 4
Δε =⋅V ⋅+N ln ⋅N
1  
ff
E 1−R
 A 
where
N is the fatigue life;
f
Δε
is the strain range;
σ is the ultimate strength;
ult
E is the Young’s modulus;
R is the reduction of area for the metal material;
A
V = 3,5;
V = 0,6;
V = 0,12;
V = 0,6.
6.9 Leak before burst (LBB)
Leak before burst (LBB) is a term used to describe a preferred failure mode when a cylinder is pressure
cycled to failure. LBB is not applicable to the failure mode in a burst test. A metal cylinder generally meets
this requirement by use of a material that was ductile to a point where a through crack is not likely to
propagate rapidly. It could also use an intentional flaw to initiate fatigue failure in a region of the cylinder
that is resistant to rapid flaw growth.

Typically, the fatigue life of a metal liner is lower than that of the composite reinforcement. With proper
material selection and design, liner fatigue failure results in leakage rather than rupture. However, if the
liner carries a large enough fraction of the load, and a through crack results in rapid failure of the liner, it is
possible to dynamically transfer the liner load into the composite reinforcement, resulting in a rupture.
It is generally accepted that if the margin against failure by pressure cycling is sufficiently high, such that
the likelihood of fatigue failure is low, the cylinder is safe for use. The safety factor on cycling is generally
2 or 3 times the maximum number of cycles expected in the lifetime of the cylinder. Some standards have
addressed LBB by pressure cycling to test pressure. Alternatively, a weak point could be designed into a
component, such as an end boss, that results in failure by leakage before burst.
6.10 Damage tolerance
Damage tolerance of the laminate will affect its cyclic fatigue life. Impact damage, cuts and abrasions could
directly cut or remove fibre in the laminate. This can reduce laminate strength at the points surrounding the
damage, resulting in higher stresses locally, and propagate d
...