2019

Composite Pressure Vessels

A Term Paper

Submitted By
Neel Nadpara

MSE 5364: Advanced Composites

Instructor: Professor Rajiv Naik

Department of Materials Science and Engineering
University of Connecticut

December 1, 2019
Table of Contents
I. Introduction .................................................................................................. 2
II. Material Description ..................................................................................... 6
III. Mechanical Properties .................................................................................. 9
IV. Fabrication process ..................................................................................... 15
V. Composite analysis ..................................................................................... 20
VI. Fatigue and toughness ................................................................................ 27
VII. Recommendations ...................................................................................... 30
VIII. Summary ................................................................................................. 32
IX. References .................................................................................................. 33

1
 I. Introduction
Composite materials provide some unique advantages over traditional engineering materials
such as metals. While it may be surprising, natural composite materials have been utilized for centuries.
In Mesopotamia, one of the first known civilizations, mud brick was reinforced with straw and dried in
the sun to craft building materials around 4900 B.C as seen in Figure 1. However, the creation of high-
strength man-made fibers and developments in polymer chemistry in the 20th century have permitted
the creation of composite materials that can rival and even surpass traditional materials such as metals.
Composite materials combine two or more materials to produce a material with improved or more
desirable properties than the individual materials alone. This is achieved by using a matrix material to
secure a reinforcement material which works to supplement the material properties of both materials in
an advantageous way (Campbell, 2010).

Figure 1: Mud Brick Reinforced with Straw

One potential application of composite material are pressure vessels. Pressure vessels are
containers that are designed to hold liquids, vapors, or gases at pressures that are considerably different
from ambient conditions. Pressure vessels are some of the most prevalent engineering structures and
can be found in nearly any complex machine (Vasiliev, 2009). Pressure vessels typically consist of a shell
with a general spherical shape or cylindrical shape (with hemispherical or dished heads) as seen in
Figure 2. Pressure vessels are used extensively in different industries such as chemical processing, power
plants (both conventional and nuclear), shipbuilding, and aerospace. Examples of pressure vessels
include storage tanks, heat exchangers, boilers, reactor vessels, steam generators, and rocket motor
cases as seen in Figure 3.

Figure 2: Left: Cylindrical Pressure Vessel, Right: Spherical Pressure Vessel

2
 Figure 3: Left: Heat Exchanger, Center: Steam Generator, Right: Rocket Motor Case

Pressure vessels typically hold operational pressures greater than 15 psig and can be subject to
different temperatures and conditions depending on the application (Pressure Vessel, 2019). To add to
the complexity, the pressure and temperature within the vessel often fluctuates during operation.
Moreover, pressure vessels can contain deleterious material including radioactive, corrosive, or toxic
materials. Pressure vessels are usually critical to the operation of an overall system and failure to
perform can cause serious problems.

Since pressure vessels are expected to perform under enormous pressures, a pressure vessel
failure can be immensely dangerous. A pressure vessel bursting can lead to explosions, poison gas leaks,
and fires making pressure vessel safety a priority as seen in Figure 4. In the 18th and early 19th centuries,
low material quality and insufficient manufacturing processes coupled with ignorance in design,
operation, and maintenance contributed to numerous and often fatal explosions of boilers and pressure
vessels in the United States (Pressure Vessel, 2019). One such instance was documented on March 20,
1905 in which a boiler exploded in a shoe factory killing 58 people and injuring another 177 people in
Brockton, Massachusetts. Prior to this incident, these explosions were often viewed as, “Acts of God.”
However, this tragedy lead to the creation of a standard, the American Society of Mechanical Engineers
(ASME) Boiler and Pressure Vessel Code (BPVC), to ensure pressure vessel safety (Ellenberger, 2004) .
The ASME BPVC is the standard in the United States and several other nations that governs the design,
manufacturing, maintenance, and inspection of pressure vessels. Since pressure vessel requirements are
often application driven, every vessel has its own unique operation limits that are designed for this
includes the design pressure and design temperature. Additional requirements can be imposed for
special cases (corrosion, operating in low temperature environments, external loads, etc.) (ASME,
2019).

Figure 4: Falcon 9 Rocket Explosion due to Pressure Vessel Failure

3
 Traditionally, pressure vessels are constructed from metals such as steels (low alloy or chrome-
moly), stainless steel (austenitic, ferritic, martensitic, or duplex), or nickel alloys (Inconel 600 or 800)
(Ellenberger, 2004). However, with special considerations composite materials can be used to construct
pressure vessels with some distinct advantages and disadvantages. Composite materials provide a high
specific strength and modulus of elasticity when compared to traditional engineering materials such as
metals as seen in Figure 5. This allows composite pressure vessels to be more efficient when compared
to their metal counterparts in terms of strength to weight ratios (Ashby, 2010). This allows lighter
weight pressure vessels to be fabricated that are able to withstand the same pressures allowing for
significant weight savings which is key in aerospace, automotive, and other industries where weight
savings are of significant concern. This advantage is provided by the reinforcement fibers used to make
composite materials as the fiber reinforcement is thin and will ultimately have fewer defects per unit of
cross-sectional area when compared to bulk materials (Vasiliev, 2009). Moreover, specific modulus of
bulk materials is not easily altered via processing. While strength can be increased by alloying, heat
treatment, or mechanical deformation the modulus can only be changed predictably in a composite
material (Ashby, 2010). Additionally, the strength and other properties such as ductility can be tailored
in composite materials by the selection of the matrix, reinforcement, fiber volume, or fiber orientation
for specific pressure vessel applications. This provides the designer with a variety of options; however,
these properties have other design implications. The fibers in composite materials provides natural
crack arrests in the material that help resistance to fatigue failure which is imperative for pressure
vessels that undergo cyclic pressure fluctuations (Campbell, 2010).

Figure 5: Left: Ashby diagram showing the specific modulus advantage of composite materials to that of steel; Right: Ashby
diagram showing the specific strength advantage of composite materials to that of steel

Traditional engineering materials such as steel are typically isotropic (perhaps not thin-wall
sheet but most other available forms are) meaning that the material properties are the same in all
directions so that normal loads only generate normal strains. However, composite materials are
anisotropic materials. Thus, the material will have different material properties in all directions at a
position in a given composite material. As a result, the reinforcement in a composite material must be
oriented in relation to the expected applied loads which requires additional structural analysis in order
to ensure the properties are optimal for an application (Campbell, 2010). Composite materials are also
more resistant to corrosion. As a result, stress corrosion cracking will not be a design concern in a
pressure vessel made of composite material (Vasiliev, 2009). Composite pressure vessels are often
fabricated as a “one-piece” vessel which saves on assembly costs and eliminates expensive and time-

4
consuming welding processes traditionally seen with metal pressure vessels. Also, the elimination of
welding removes many internal stresses and weld distortion from being present in the finished product.
These factors help to increase the life span of a composite pressure vessel and reduce the necessity for
costly repairs.

However, there are definitely some concerns when using composites for pressure vessel
fabrication. The mechanical properties of a composite have a greater degree of variation which forces
real-world testing and determination of design allowables as opposed to traditional materials such as
steel which are procured to a set of requirements (Campbell, 2010). Moreover, a key principle of safe
pressure vessel design is to have a ductile failure in which a pressure vessel leaks before it bursts (ASME,
2019; Ashby, 2010). Metals are ductile allowing them to yield and locally relieve stresses to adjacent
material and by absorbing impact by deforming rather than breaking. Composite materials are not as
ductile as most metals and their damage tolerance is significantly less in terms of static performance
since losing fibers becomes directly proportional to loosing strength. Composites also lack notch ductility
which limits their use in pressure vessels that are subject to external loading or shock loading.
Furthermore, composite materials are subject to different types of defects such as delamination,
handling damage that is not as obliviously discovered, moisture absorption, and other factors that
metals are not subject to (Campbell, 2010). Additionally, composite pressure vessels often require metal
fittings due to the difficulty of fabrication such connections out of composite materials which act as
stress risers in the finished product (Ellenberger, 2004; Nabors, 2005). All these factors prevent the use
of composites in certain pressure vessels without serious design consideration and proof tests to verify
satisfactory factor of safety (ASME, 2019). Additionally, most composites cannot withstand high
temperature applications which limits most applications of composite pressure vessels to temperatures
less than 250°F while certain alloys can perform satisfactorily in excess of 1200°F (Ellenberger, 2004).
Composites are much more costly to produce than traditional pressure vessel materials such as steel.
This is largely due to raw material, preparation steps, specialized tooling, fabrication/difficulty to
produce certain complex parts, extra production controls, and specialized inspection costs (Campbell,
2010).

Overall, when weight is a key consideration in designing a pressure vessel, composite materials
have the potential to provide some serious advantages over metals.

5
 II. Material Description
Materials selection is a crucial consideration when designing pressure vessels. Pressure vessels
are structures that are governed by strict specifications due to the criticality of their function and
disastrous consequences upon failure. As a result, pressure vessels are categorized into several groups
depending on their construction. Type I pressure vessels are of all-metal construction, these are the
least costly of the pressure vessel classes costing roughly $5 per liter of volume. Type II pressure vessels
are mostly metal (steel or aluminum) which has glass-fiber composite overlap in the direction of hoop
stresses. Type III pressure vessels consist of a metal liner (typically aluminum) with a full composite
overlap (carbon fiber) in which the composite material carries the structural load as seen in Figure 6
(Legault, Pressure Vessel Tank Types, 2012). Type VI pressure vessels are relatively new being available
since 2010 and consist of a polymer liner (thermoplastic) which is overlapped with composite material
(glass or carbon fibers) (Legault, Next-Generation Pressure Vessels, 2012). Type V pressure vessels are
very uncommon and feature a full composite fabrication as seen in Figure 7 (glass or carbon fibers)
(Pressure Vessel, 2019). The cost of the pressure vessel increases with type, with Type V pressure
vessels being the costliest (Legault, Pressure Vessel Tank Types, 2012). Nevertheless, there are many
weight critical applications such as rocket motor cases and fuel storage tanks that require pressure
vessels to be lightweight while holding severe pressures.

Figure 6: Construction of a pressure vessel with metal liners

Figure 7: All composite pressure vessel

While Type III and IV pressure vessels cost 3.5 times as much as Type I pressure vessels there are
clear advantages. For commercial transportation applications, Type III and IV pressure vessels can
reduce the weight of gas containment systems by approximately 1,000 lbs (Legault, Pressure Vessel
Tank Types, 2012). Weight reduction of this magnitude does more than providing better fuel economy,
it would augment the capacity of the equipment and allow for other operational benefits such as better
handling. Moreover, Type III and IV pressure vessels enhance the practical limit of gas containment
pressures by providing more efficient potential energy storage. As a result, for storing pressures above
5,000 psi composite pressure vessels often provide a more practical solution (Legault, Pressure Vessel
Tank Types, 2012).

6
 Typically, composite pressure vessel design begins with the fiber selection. The fiber selection
process is based on the unique performance requirements including burst strength, weight, or stiffness
requirements. Typical features of different fiber types are shown in Figure 8. In the case of
manufacturing composite rocket motor cases, the key design concern is the strength to weight ratio.
Carbon fiber is employed when weight and stiffness considerations are paramount. Moreover, carbon
fiber has very high tensile strength and modulus. Hence, the ideal fiber choices would be either
intermediate-modulus polyacrylonitrile (PAN) based carbon fibers or intermediate-modulus high strain
PAN-based carbon fibers. The best fibers available for rocket motor cases are T1000GB and M30SC.
T1000GB is a quality intermediate-modulus high strain polyacrylonitrile (PAN) based carbon fiber.
T1000GB provides one of the greatest tensile strengths available at 924 × 106 psi (6371 GPa) while
maintaining a modulus of 42.6 × 106 psi (294 GPa). However, this comes with a significant price of $85
per lb. A cheaper alternative is M30SC which is an intermediate-modulus PAN based carbon fiber that
only costs $29 per lb. M30SC provides a tensile strength of 796 × 106 psi (5490 GPa) while maintaining
a modulus of 42.6 × 106 psi (294 GPa) (Peters, 2011). However, in the case of composite pressure
vessels used for compressed gasses and liquids, the cost is the key driver. As a result, other fibers are
selected such as fiberglass. Fiberglass has exceptional tensile properties and is the least expensive
filament choice, however, it is significantly denser and less resistant to fatigue loading than carbon fiber
(Campbell, 2010). Characteristic fiberglass fibers chosen for composite pressure vessel fabrication
include E-glass and S-glass. E-glass provides a tensile strength of 500 × 106 psi (3447 GPa) while
maintaining a modulus of 10.5 × 106 psi (72.5 GPa). S-glass provides better properties for a greater
cost with a tensile strength of 650 × 106 psi (4482 GPa) while maintaining a modulus of
12.6 × 106 psi (86.9 GPa) (Peters, 2011). After fibers are selected, it is also important to consider the
orientations of the fibers and the specific layup of the fibers. There the fiber orientations will be at
different angles to facilitate proper filament winding of the vessel and ensure that the stress is uniformly
distributed without creating stress risers. In a typical cylindrical pressure vessel, the hoop stress
dominates the longitudinal stress requiring more reinforcement in the circumferential direction.

Figure 8: Chart showing the advantages and disadvantages of different fiber types

Following fiber selection, the matrix must be selected. Typically, most pressure vessels are
manufactured using epoxy resins. Epoxy resins are often selected as they deliver a favorable
combination of strength, adhesion, low shrinkage, and processing versatility (Campbell, 2010). Pressure
vessels are required to be reliable (water/airtight), see immense operating pressures and cyclic loading,
and consist of complex geometry making strength, adhesion, and versatility key matrix selection
criterion (Mohan , Yarrapragada, & Kiran, 2012; Vasiliev, 2009). Generally, heat-cured epoxy resins that

7
can withstand 250°F to 350°F are employed (Peters, 2011). Composite pressure vessels are
manufactured via either filament winding or autoclave molding (Nabors, 2005). The vast majority of
composite pressure vessels are made via filament winding in which the fibers are spun around a
mandrel to form the geometry of the pressure vessel which may be removed. However, an autoclave
molding process can also be used by molding around a core which is later removed (Vasiliev, 2009). As a
result, the fibers would either be purchased as wound spools which can come dry or pre-impregnated
with the epoxy resin for filament winding. Resins used for filament winding should permit wet winding
in which the fibers are impregnated just before being wound onto the mandrel (Vasiliev, 2009). Pre-
impregnated fabric may be used for autoclave molding and makes the layup easier and prevents defects
such as lack of resin, voids, which can occur in wet layups as seen in Figure 9. A common epoxy that is
widely employed commercially and available in both winding resin systems and prepreg systems is
diglycidyl ether of Bisphenol A (DGEBA) (Campbell, 2010; Vasiliev, 2009).

Figure 9: Defects seen in composite fabrication

Material selection in composite pressure vessels must also consider application specific use
cases such as fatigue life, environmental effects, and shock loading. Properties such as fatigue life are
improved by proper fiber selection as carbon fibers are much more resistant to fatigue. Moreover,
environmental effects are usually of less concern since composite pressure vessels do not corrode as
readily as their metal counterparts. However, using a less moisture impacted matrix such as a
thermoplastic and use of special coatings on the surface can prevent degradation by ultraviolet light.
Impact resistance can be provided by using specially formulated tougher epoxy matrixes and the use of
aramid fibers.
Overall, the material selection in pressure vessel fabrication must consider the design
requirements of the specific application.

8
III. Mechanical Properties
A pressure vessel’s basic function is to store liquids and gases at high pressure. As a result, the
strength of a vessel is critical to prevent explosions due to rupture. However, weight constraints by
many industries such as aerospace have begun to drive for increased structural efficiency. In these types
of weight and safety-critical applications, fiber-reinforced composites are often selected due to their
high specific strength and low density.

In the case of a pressure vessel with no external loads, the internal pressure is contained by the
internal surfaces of the pressure vessel. In the case of a thin-walled pressure vessel with a circular
cylindrical shell of radius “r” loaded with internal pressure “p,” there are several stresses that need to be
considered as seen in Figure 10. First, there are axial forces “q” that are uniformly distributed over the
caps of the pressure vessel (end cross-section contours). This can be quantified by balancing the forces
in the axial direction as seen in Equation 1.

Figure 10: Cylindrical Pressure Vessel Showing Applied Loads and Stresses
𝑝𝑟
Equation 1: 𝜋𝑟 2 = 2𝜋𝑟𝑞 → 𝐴𝑥𝑖𝑎𝑙 𝑆𝑡𝑟𝑒𝑠𝑠 (𝑒𝑛𝑑𝑐𝑎𝑝) = 𝑞 = 2

The cylindrical shell of the pressure vessel experiences two key stresses as well. These key
stresses are the circumferential (Hoop) stress and longitudinal (Axial) stress as seen in Figure 10. The
circumferential (Hoop) stress acts about the diameter of the vessel (the circumferential direction) and is
quantified by Equation 2 for a cylindrical pressure vessel. The longitudinal (Axial) stress acts about the
longitudinal direction of the pressure vessel and is quantified by Equation 3 for a cylindrical pressure
vessel (Vasiliev, 2009). By comparing Equation 1 and 2 it is clear that the axial stress in a cylindrical
pressure vessel is one-half of the circumferential stress. However, since a pressure vessel typically
undergoes cyclic loading, it is often undesirable to have a significant variation in the material to prevent
localized stress concentrations (Mohan , Yarrapragada, & Kiran, 2012).
𝑝𝑅
Equation 2: Circumferential Stress = 𝜎𝑐 = 𝑡
𝑝𝑅
Equation 3: Longitudinal Stress = 𝜎𝑎 = 2𝑡

Where:

𝑝 = 𝑖𝑛𝑡𝑒𝑟𝑛𝑎𝑙 𝑔𝑎𝑢𝑔𝑒 𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒

𝑟 = 𝑟𝑎𝑑𝑖𝑢𝑠 𝑜𝑓 𝑐𝑦𝑐𝑙𝑖𝑛𝑑𝑒𝑟

9
 𝑡 = 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 𝑜𝑓 𝑡ℎ𝑒 𝑐𝑦𝑐𝑙𝑖𝑛𝑑𝑒𝑟 𝑤𝑎𝑙𝑙
To determine the appropriate materials are selected for the design several factors need to be
considered. First comes the material selection which includes the selection of the fiber, resin, and fiber
volume fraction. Following this, the optimal fiber orientation for the laminate and the lamina stacking
sequence can be determined. Following this, the number of lamina in the laminate and their orientation
can be determined which also determines the thickness of the pressure vessel shell (Mohan ,
Yarrapragada, & Kiran, 2012). Lamina are a flat armament of unidirectional (or woven) fibers that are
held together by a matrix material as seen in Figure 11. The lamina can be assumed to be orthotropic
meaning that they have different mechanical properties in all three axes. The thickness of the lamina
depends on the fibers and matrix from which it is made and can vary in thickness. A laminate is a stack
of lamina in a specific orientation that forms a composite material. Thus, to form a laminate, several
lamina are stacked with specific orientations as seen in Figure 12. Individual laminates have properties
that depend on the lamina that form it (Campbell, 2010).

Figure 11: Unidirectional and Woven Lamina

Figure 12: Laminate made up of multiple lamina

In the previous section, Section II, it was determined that for a high-end application, the ideal
constituents for a lamina would be intermediate-modulus high strain polyacrylonitrile (PAN) based
carbon fiber (T1000GB) for the reinforcement, diglycidyl ether of Bisphenol A (DGEBA) epoxy as the
matrix, with a fiber volume fraction of 65 percent (Peters, 2011). Since multiple layers can be used, the
lamina will be a unidirectional form as seen in Figure 13 (Hyer, 1998).

10
 Figure 13: Unidirectional Lamina with Directions Shown

Given the tensile strength, modulus, density, and Poisson’s ratio of the fiber and the matrix
many of the lamina’s properties can be determined as seen below (Mohan , Yarrapragada, & Kiran,
2012). The properties in many of these cases are a function of the rule of mixtures which only provides
an approximate theoretical value. To determine real-world properties, actual testing has to be
performed (Campbell, 2010). Some common properties of various fibers, resins, and unidirectional
composite materials are provided in Tables 1, 2, and 3 below (Campbell, 2010; Peters, 2011).

Tensile Strength of Lamina:

Longitudinal Strength:

𝜎11 = 𝜎𝑓 𝑉𝑓 + 𝜎𝑚 𝑉𝑚 = (6371 𝑀𝑃𝑎)(0.65) + (97.2 𝑀𝑃𝑎)(0.35) = 𝟒𝟏𝟕𝟓. 𝟏𝟕 𝑴𝑷𝒂

Transverse Strength:
𝜎𝑓 𝜎𝑚 (6371 𝑀𝑃𝑎)(97.2 𝑀𝑃𝑎)
𝜎22 = = = 𝟐𝟕𝟎. 𝟎𝟔 𝑴𝑷𝒂
𝜎𝑚 𝑉𝑓 + 𝜎𝑓 𝑉𝑚 (97.2 𝑀𝑃𝑎)(0.65) + (6371 𝑀𝑃𝑎)(0.35)

Modulus of Lamina:

Longitudinal Modulus:

𝐸11 = 𝐸𝑓 𝑉𝑓 + 𝐸𝑚 𝑉𝑚 = (294 𝐺𝑃𝑎)(0.65) + (3.58 𝐺𝑃𝑎)(0.35) = 𝟏𝟗𝟐. 𝟑𝟓 𝑮𝑷𝒂

Transverse Modulus:
𝐸𝑓 𝐸𝑚 (294 𝐺𝑃𝑎)(3.58 𝐺𝑃𝑎)
𝐸22 = = = 𝟏𝟎 𝑮𝑷𝒂
𝐸𝑚 𝑉𝑓 + 𝐸𝑓 𝑉𝑚 (3.58 𝐺𝑃𝑎)(0.65) + (294 𝐺𝑃𝑎)(0.35)

Major Poisson’s Ratio:

𝑣12 = 𝑣𝑓 𝑉𝑓 + 𝑣𝑚 𝑉𝑚 = (0.25)(0.65) + (0.35)(0.35) = 𝟎. 𝟐𝟖𝟓

Minor Poisson’s Ratio:

𝐸22 (10 𝐺𝑃𝑎)
𝑣21 = 𝑣12 = (0.285) = 𝟎. 𝟎𝟏𝟓
𝐸11 (192.35 𝐺𝑃𝑎)

11
Shear Modulus:
𝐸𝑓 (294 𝐺𝑃𝑎)
𝐺𝑓 = = = 𝟏𝟎𝟖. 𝟖𝟗 𝑮𝑷𝒂
2(1 + 𝑣𝑓 ) 2(1 + (0.35))
𝐸𝑚 (3.58 𝐺𝑃𝑎)
𝐺𝑚 = = = 𝟏. 𝟒𝟑𝟐 𝑮𝑷𝒂
2(1 + 𝑣𝑚 ) 2(1 + (0.25))
𝐺𝑓 𝐺𝑚 (108.89 𝐺𝑃𝑎)(1.432 𝐺𝑃𝑎)
𝐺12 = = = 𝟓. 𝟔𝟐𝟒 𝑮𝑷𝒂
𝐺𝑓 𝑣𝑚 + 𝐺𝑚 𝑣𝑓 (108.89 𝐺𝑃𝑎)(0.25) + (1.432 𝐺𝑃𝑎)(0.35)

Density:
𝑔 𝑔 𝒈
𝜌𝑐𝑜𝑚𝑝 = 𝜌𝑓 𝑉𝑓 + 𝜌𝑚 𝑉𝑚 = (1.8 3
) (0.65) + (1.202 3
) (0.35) = 𝟏. 𝟓𝟗𝟏
𝑐𝑚 𝑐𝑚 𝒄𝒎𝟑
From this comparison, it is clear that the properties are best in the direction of fiber orientation (11
direction in this case). However, using proper layering of lamina can provide composites with properties
that are far superior to traditional materials per unit mass. However, effective use of these in-plane
strengths is critical to achieve the best use of specific strength possible. Pressure vessels are subject to a
hoop stress that is twice that of the longitudinal (axial) stress. As a result, there needs to be twice the
reinforcement around the circumference of the pressure vessel as opposed to the longitudinal direction
to ensure that the strength provided by the reinforcement is optimally employed. However, layups on
pressure vessels are a complicated matter since the vast majority of vessels are made using filament
winding as opposed to unidirectional lamina (Peters, 2011). As a result, individual lamina are usually
oriented at a +45/-45-degree angle to build the majority of the wall thickness, afterwards extra
reinforcement is applied in the hoop direction to achieve the required operating pressure with a large
safety margin built-in.

Pressure vessels undergo extreme service conditions and are expected to perform. Failure of a
pressure vessel can be disastrous and even cost lives. Hence, the failure mode of pressure vessels is of
principal concern. Typically, pressure vessels are designed to either yield before burst or to leak before
burst depending on the contents and application (Ashby, 2010). Regardless there are several short-term
and long-term failure modes for pressure vessels.

Pressure vessels can fail in the short-term due to brittle fracture, ductile rupture, or crack
formation and propagation. In brittle fracture, there is no plastic deformation or stress that exceeds the
limit of the material. Instead, the material embrittles due to many factors including improper material
processing, strain aging, or poor operating environment which causes this embrittlement and
subsequent failure. In ductile rupture, the strength of the material is exceeded until a fracture occurs.
The crack formation would occur due to existing defects combined with excessive localized stress
concentrations (DFC Pressure Vessel Manufacturer Co, Ltd., 2019).

Long term failure of pressure vessels can be attributed to creep rupture, erosion/corrosion, and
cyclic loading. In creep rupture, the pressure vessel is loaded with given stress at a high temperature for
an extended duration. The material then slowly begins to plastically deform with time leading to
localized thickness reduction or bulging. This deformation eventually leads to the pressure vessel
fracturing. Erosion/corrosion can occur due to exposure to corrosive chemicals or exposure to steam
which can corrode or erode material over time. This creates initiation points for later failure to occur

12
due to stress concentrations. Cyclic failure is common in pressure vessels and can be due to progressive
plastic deformation, alternating plasticity, or fatigue under elastic or elastic-plastic strains (DFC Pressure
Vessel Manufacturer Co, Ltd., 2019).

Environmental effects must, therefore, be considered for pressure vessels. Common

environmental effects that impact composite pressure vessels and must be considered include moisture
absorption, damage due to chemical exposure, embrittlement due to ultraviolet radiation, erosion due
to fluid motion, and heat damage. Composites inherently absorb moisture, as a result, the matrix is
degraded over time. This can cause delamination due to the expansion of water during freezing
temperatures. Exposure to elevated temperatures can cause steam formation which can cause matrix
cracking (Campbell, 2010). However, if the pressure vessel is of sufficient thickness and proper
precautions are taken during fabrication, this risk can be greatly mitigated. Fluids such as gasoline or
other solvents can adversely affect and even dissolve many polymers and cause matrix failure
(Campbell, 2010). However, liners of more chemically inert materials such as titanium or Monel can be
used to prevent this from occurring. Ultraviolet radiation can cause degradation of the matrix due to
molecular weight change and cross-linking damage. Some fibers such as aramid fiber are also
susceptible to damage (Campbell, 2010). However, coatings or paint can be used to prevent damage
from ultraviolet radiation exposure. Erosion may occur due to fluid motion such as cavitation erosion
from steam expansion. This type of degradation can be avoided by using liners made of a more wear-
resistant material such as stainless steel. Lastly, heat damage can occur due to high-temperature service
(Campbell, 2010). This type of damage is difficult to prevent and would have to be accounted for in the
design phase by selecting appropriate materials for the design temperature of the pressure vessel.
Table 1: Fiber Types with their Respective Properties

13
Table 2:Typical Filament Winding Matrixes with their Respective Properties

Table 3:Typical Properties of Unidirectional Composite Materials

14
IV. Fabrication process
Composite vessels have grown in popularity for certain applications due to their unique
advantages such as high specific strength and modulus. Nevertheless, their manufacture can be much
more complex than their metal counterparts. Metal pressure vessels are typically made of formed metal
which is prepped and welded together. Composite pressure vessels typically come in two forms one
which includes a metallic or nonmetallic liner and one that is purely composite. Regardless, the
manufacturing process is similar.

Composite pressure vessels are usually manufactured by a fiber placement process. The most
common process is filament winding. Filament winding is used primarily due to the vast majority of
pressure vessels being shells that are revolved around a central axis. The process of filament winding is
relatively straight forward. Richard E. Young initiated what we could consider the modern filament
winding process in the early 1940s. He started with a simple mechanical winder that wound fiberglass
onto a wooden mandrel. After the winding process was complete, a paintbrush was used to apply epoxy
resin to the outer diameter of the wound part. The wound part was subsequently placed in an oven for
curing. Once cured, the wooden mandrel would be removed completing the fabrication (Peters, 2011).
While today’s filament winding operations are identical at a fundamental level, they can be much more
complex and automation has made the process much more consistent and controlled.

First, filament is procured in the form of fiber spools which are placed on creels. The fiber spools
can consist of un-impregnated tow or pre-impregnated (pre-preg) tows. The fiber is typically either
fiberglass or carbon fiber for most pressure vessels. Unimpregnated tow is typically used for fiberglass
while prepreg tow is typically used for carbon fiber vessels. The fiber spool is fed into the filament feed.
In the case of unimpregnated tow, the filament is feed into a resin bath to get saturated with resin while
pre-preg tow can bypass this step (Filament winding, 2019).

There is a mandrel which essentially acts as a mold for the finished internal shape of the
pressure vessel. The mandrel is rotated about its central axis by a machine drive which precisely controls
the speed of the mandrel. The filament is then wound around the mandrel in a prescribed way to meet
certain stress conditions. This the filaments are wound around the rotating mandrel under tension to
hold the fibers taut to prevent fiber buckling. There is a minimum of two axes in filament winding but
there can be many more depending on the complexity of the part and desired winding pattern. The two
basic axes function as a typical lather with the mandrel rotating around the spindle in the X-axis. The
delivery eyelet sits on a carriage that can traverse horizontally in the Y-axis parallel to the central axis of
the rotating mandrel laying down fibers in the desired pattern or angle (Filament winding, 2019). For
pressure vessel manufacturing, a minimum of 4 axes are required in order to achieve the desired
geometry and filament orientation (Campbell, 2010).

After the mandrel is covered with filament to the desired thickness using the prescribed
filament orientation, the resin is cured. This can be accomplished in many ways but the most common
method is to place the mandrel in an oven or under radiant heaters until the resin cures and the part
hardens. Once the resin is cured, the mandrel can be removed leaving the final hollow part (Filament
winding, 2019). However, for pressure vessels, the openings are rather small making mandrel extraction
difficult. For this reason, two types of mandrels are typically used. One type is an integral mandrel in
which the mandrel (hollow metal or plastic) becomes a permanent part of the final product acting as a

15
liner to prevent leakage and also working to affix fittings to the vessel. The second type is to use an
investment style mandrel which can be dissolved or removed after the resin is cured such as salt, sand,
or wax leaving only the shell behind (Campbell, 2010) (Nabors, 2005). Figure 14 below shows the basic
process of filament winding while Figure 15 shows the typical axes used in a filament winding machine
(ASM International, 2001).

Figure 14: Typical filament-winding machine

Figure 15: 6 axes of filament winding motion

There are three types of filament which allow three major variations in the filament winding
process. There is wet winding in which the dry reinforcement is impregnated by a liquid resin such as
diglycidyl ether of Bisphenol A, a typical epoxy resin used in both wet and pre-impregnated winding
operations, just before winding. Wet-rolled pre-preg winding is an alternative. In this method, the dry
reinforcement is impregnated with the liquid resin and then rewound prior to filament winding. The
third and most common process for high-end applications is tow-preg winding in which the tow is
purchased in pre-preg form. Having a wet winding process is common in many filament winding
operations and it is essential to have resin with the appropriate viscosity and pot life in this process. Pre-
preg tows, while more expensive than wet winding resin systems, are the preferred method for filament
winding due to several reasons. By using pre-preg tow the winding speed can be greatly increased, the
best control of resin content is achieved, and the tack can be modified to allow less slippage when
winding shallow angles (Campbell, 2010).

16
 Another key consideration in filament winding is the winding pattern. For pressure vessels, the
winding pattern is a critical consideration as it determines the internal pressure limit of the final product
(Filament winding, 2019). Typically, two winding methodologies are combined and are often controlled
using advanced numerical control (NC) programs. First, helical winding is used due to its versatility and
ability to form the pressure vessel heads. Helical winding is ideal since it can generate nearly any
combination of length and diameter. In helical winding, the mandrel rotates while the fiber carriage
traverses back and forth at a speed necessary to generate the desired helical angle. This is first done to
provide the basic body of the pressure vessel and generate the heads of the vessel. The other winding
methodology is hoop winding in which the mandrel rotates while the carriage feeds fibers in the
circumferential direction. Once the helical winding concludes, the hoop winding is performed on the
cylindrical section of the pressure vessel. This strategy leverages the fiber’s strength both primary
loading directions. First, the helical winding acts to withstand the axial stress while the hoop winding
acts to withstand the hoop stress. Twice as much hoop winding is needed for a cylindrical pressure
vessel due to the hoop stress being greater than the axial stress. This methodology advantageous since
hoop winding is much faster than helical winding. An illustration of the winding patterns and a graphic
from an actual pressure vessel winding operation is provided in Figure 16 (Campbell, 2010).

Figure 16: Top: Helical Winding, Middle: Hoop Winding, Bottom: Actual helical winding operation

Composite pressure vessels may also be manufactured by two other methods but their
use is much less likely. Automated fiber placement (AFP) is one of the most advanced methods for
fabrication of composite components. AFP machines place fiber reinforcements on molds or mandrels in
an automatic fashion as seen in Figure 17. A number of individual narrow width tows (~8mm) of pre-
impregnated material are placed to form composite layups. The process relies on a robot to place the
tows and build a part layer by layer. This allows fabrication of highly complex parts as each ply can be
placed in different configurations/angles in order to efficiently carry the intended loads in the

17
completed structure. This process relies on a head to place and cut the individual tows. The tows are
held taut by head will apply a tensile low to prevent fiber buckling as they are placed. Since the diameter
of the roller is the only limit of placement, complex contours can be fabricated with this process. The
head also applies pressure via the roller which can effectively debulk the laminate during lay-up.
Moreover, the heads contain heating and cooling capability which can be used to either increase the
tack and compaction (during lay-down) or decrease the tack (for cutting, clamping, restarting). While
AFP processes have their advantages, there are some key disadvantages that must be considered. The
software programming is quite complex with multiple axes moving especially on curved geometries
while trying to keep the fiber under tension to prevent bucking. AFP machines are complex and can be
very large making them useful in only some applications. Also, the current machines are very expensive
and complex and deposition rates are slow compared to conventional filament winding operations
(Campbell, 2010). Typically, pressure vessels can be made much faster using traditional filament winding
processes making AFP a less desirable choice. However, if a pressure vessel has numerous penetrations
or other complex features, AFP can prove to be an effective technique.

Figure 17: A Large Composite Rocket Fuel Tank made using an AFP process

Another method of manufacturing a composite pressure vessel is to perform a manual layup

over a mold using fabric (ASME, 2019). The fabric can be either pre-preg or a dry fabric that can be used
for a wet-layup. The fabric would be cut by hand or machine. Following this, the layup can be performed
forming the laminate. Then additional breather material can be added along with flow enhancing media.
Afterward, a vacuum bag can be used to debulk the laminate and remove voids while aiding resin flow
during cure. Typically, an oven can is used to cure the resin. An autoclave can be used to provide
additional pressure to produce a better laminate during the curing process (Campbell, 2010). Typically,
this type of process would involve much more manual labor and would not be efficient for mass
production. As a result, this would be reserved for custom or prototype pressure vessel applications.

Typically, the curing cycle for a composite pressure vessel made using epoxy resins by the above
methods involves curing in an autoclave. The completed pressure vessel is placed in the autoclave and
the temperature is ramped to the range of 240-280°F and held for 1 hour. The purpose of this ramp is to
allow the resin to flow and volatiles to escape. High pressures (100 psig) are commonly used during
autoclave processing to provide ply compaction and suppress void formation. The second part of the
cycle is the polymerization portion of the cure cycle. The temperate is ramped to 340–370°F and held at
this cure temperature for 4–6 hours. During this portion of the cure cycle, the resin viscosity initially
drops slightly due to the application of additional heat. However, it quickly rises as the chemical reaction
takes place initiating the cross-linking process. The resin gels into a solid and the cross-linking process

18
continues the pressure vessel is cured. After cure, a post cure can be performed at 175°F for 8-24 hours
to maximize some of the epoxy’s physical properties and allow a higher crosslinking density.

Like all manufacturing processes, filament winding, AFP, and wet lay-ups are subject to
manufacturing defects. Typically, with filament winding the key concerns are with resin application since
improper saturation can lead to voids or porosity in the final product. Also, foreign object inclusion may
be of concern since the wet fibers may allow small particulates into the pressure vessels which can
cause stress risers in the structure. Also, the fiber orientation may be off if the axes are not properly set-
up in either a filament winding or AFP process. In a wet layup, there are likely to be issues in fiber
orientation, wrinkling between layers, resin rich and resin starved regions, along with the possibility of
foreign object insertion. However, all of these methods should produce an acceptable product if proper
procedures and good manufacturing process are used.

Figure 18: Typical Defects in Composites

All in all, composite pressure vessels are typically manufactured using filament winding, but
there are several alternative means of manufacture.

19
 V. Composite analysis
Pressure vessels are critical structures that are subject to significant hoop and longitudinal
stresses along with numerous other application specific constraints such as service temperature,
corrosion, and cyclic loading. As a result, when designing composite pressure vessels, a key
consideration is the stacking sequence and interlaminar free edges.

In the previous section, Section II, it was determined that for a high-end application, the ideal
constituents for a lamina would be intermediate-modulus high strain polyacrylonitrile (PAN) based
carbon fiber (T1000GB) for the reinforcement, diglycidyl ether of Bisphenol A (DGEBA) epoxy as the
matrix, with a fiber volume fraction of 65 percent (Peters, 2011). Since multiple layers can be used, the
lamina will be a unidirectional form as seen in Figure 13 (Hyer, 1998).

Figure 19: Unidirectional Lamina with Directions Shown

As discussed in section III, given the modulus, density, and Poisson’s ratio of the fiber and the
matrix many of the lamina’s properties can be determined as seen below (Mohan , Yarrapragada, &
Kiran, 2012). The properties in many of these cases are a function of the rule of mixtures which only
provides an approximate theoretical value. To determine real-world properties, actual testing has to be
performed (Campbell, 2010).

Modulus of Lamina:

Longitudinal Modulus:

𝐸11 = 𝐸𝑓 𝑉𝑓 + 𝐸𝑚 𝑉𝑚 = (294 𝐺𝑃𝑎)(0.65) + (3.58 𝐺𝑃𝑎)(0.35) = 𝟏𝟗𝟐. 𝟑𝟓 𝑮𝑷𝒂

Transverse Modulus:
𝐸𝑓 𝐸𝑚 (294 𝐺𝑃𝑎)(3.58 𝐺𝑃𝑎)
𝐸22 = = = 𝟏𝟎 𝑮𝑷𝒂
𝐸𝑚 𝑉𝑓 + 𝐸𝑓 𝑉𝑚 (3.58 𝐺𝑃𝑎)(0.65) + (294 𝐺𝑃𝑎)(0.35)

Major Poisson’s Ratio:

𝑣12 = 𝑣𝑓 𝑉𝑓 + 𝑣𝑚 𝑉𝑚 = (0.25)(0.65) + (0.35)(0.35) = 𝟎. 𝟐𝟖𝟓

Shear Modulus:

20
 𝐸𝑓 (294 𝐺𝑃𝑎)
𝐺𝑓 = = = 𝟏𝟎𝟖. 𝟖𝟗 𝑮𝑷𝒂
2(1 + 𝑣𝑓 ) 2(1 + (0.35))
𝐸𝑚 (3.58 𝐺𝑃𝑎)
𝐺𝑚 = = = 𝟏. 𝟒𝟑𝟐 𝑮𝑷𝒂
2(1 + 𝑣𝑚 ) 2(1 + (0.25))
𝐺𝑓 𝐺𝑚 (108.89 𝐺𝑃𝑎)(1.432 𝐺𝑃𝑎)
𝐺12 = = = 𝟓. 𝟔𝟐𝟒 𝑮𝑷𝒂
𝐺𝑓 𝑣𝑚 + 𝐺𝑚 𝑣𝑓 (108.89 𝐺𝑃𝑎)(0.25) + (1.432 𝐺𝑃𝑎)(0.35)

As a result of the above calculations the lamina properties are as follows:

Longitudinal Modulus: 𝐸11 = 𝟏𝟗𝟐. 𝟑𝟓 𝑮𝑷𝒂

Transverse Modulus: 𝐸22 = 𝟏𝟎 𝑮𝑷𝒂

Major Poisson’s Ratio: 𝑣12 = 𝟎. 𝟐𝟖𝟓

Shear Modulus: 𝐺12 = 𝟓. 𝟔𝟐𝟒 𝑮𝑷𝒂

A composite pressure vessel can use a number of different stacking sequences. However, for the
purposes of this discussion, two common stacking sequences will be considered. The two common
stacking sequences include a [0/90/0]s and [0/-45/45]s laminate.

[0/90/0]s:
Step 1: Define Laminate:

Laminate Constituents
Layer Angle
1. 0°
2. 90°
3. 0°
4. 0°
5. 90°
6. 0°

Assumed: 𝑡1 = 0.127 𝑚𝑚

ℎ = 6(0.127) = 0.762 𝑚𝑚

21
 𝑡
𝑧0 = −
2
𝑧1 = 𝑧0 + 𝑡1
𝑧2 = 𝑧1 + 𝑡1
𝑧3 = 𝑧2 + 𝑡2
Step 2: Calculate [ABD] matrix:

Step 3: Calculate [A*] matrix:

[𝐴∗ ] = [𝐴]−1

22
 100.7 2.181 0
[𝐴] = [ 2.181 54.15 0 ] 𝐺𝑃𝑎 − 𝑚𝑚
0 0 4.285
0.010009 −0.000403 0
[𝐴]−1 = [ −0.000403 0.018483 0 ] GPa-mm
0 0 0.233372
Step 4: Calculate Effective Laminate Properties:
1 1
Effective in-plane longitudinal modulus: 𝐸𝑥 = ℎ𝐴∗ = .762∗0.010009 = 𝟏𝟑𝟏. 𝟏𝟏𝟔 𝑮𝑷𝒂
11

1 1
Effective in-plane transverse modulus: 𝐸𝑦 = ℎ𝐴∗ = .762∗0.018483 = 𝟕𝟏. 𝟎𝟎𝟐 𝑮𝑷𝒂
22

1 1
Effective in-plane shear modulus: 𝐺𝑥𝑦 = ℎ𝐴∗ = 0.762∗0.233372 = 𝟓. 𝟔𝟐𝟑 𝑮𝑷𝒂
66

𝐴∗ −0.000403
Effective in-plane Poisson’s Ratio: 𝑣𝑥𝑦 = − 𝐴12
∗ = − 0.010009
= 𝟎. 𝟎𝟒𝟎
11

𝐴∗ −.000403
Effective in-plane Poisson’s Ratio: 𝑣𝑦𝑥 = − 𝐴12
∗ = − 0.018483 = 𝟎. 𝟎𝟐𝟐
22

[0/-45/45]s:
Step 1: Define Laminate:

Laminate Constituents
Layer Angle
1. 0°
2. -45°
3. 45°
4. 45°
5. -45°
6. 0°

Assumed: 𝑡1 = 0.127 𝑚𝑚

ℎ = 6(0.127) = 0.762 𝑚𝑚
𝑡
𝑧0 = −
2

23
 𝑧1 = 𝑧0 + 𝑡1
𝑧2 = 𝑧1 + 𝑡1
𝑧3 = 𝑧2 + 𝑡2
Step 2: Calculate [ABD] matrix:

Step 3: Calculate [A*] matrix:

[𝐴∗ ] = [𝐴]−1
78.44 24.40 0
[𝐴] = [ 24.40 31.94 0 ] 𝐺𝑃𝑎 − 𝑚𝑚
0 0 26.50

24
 0.016722 −0.012775 0
[𝐴]−1 = [ −0.012775 0.041068 0 ] GPa-mm
0 0 0.037736
Step 4: Calculate Effective Laminate Properties:
1 1
Effective in-plane longitudinal modulus: 𝐸𝑥 = = = 𝟕𝟖. 𝟒𝟕𝟗 𝑮𝑷𝒂
ℎ𝐴∗11 .726∗0.016722

1 1
Effective in-plane transverse modulus: 𝐸𝑦 = ℎ𝐴∗ = .726∗0.041068 = 𝟑𝟏. 𝟗𝟓𝟓 𝑮𝑷𝒂
22

1 1
Effective in-plane shear modulus: 𝐺𝑥𝑦 = ℎ𝐴∗ = 0.726∗0.037736 = 𝟑𝟒. 𝟕𝟕𝟔 𝑮𝑷𝒂
66

𝐴∗12 −0.012775
Effective in-plane Poisson’s Ratio: 𝑣𝑥𝑦 = − =− = 𝟎. 𝟕𝟔𝟑
𝐴∗11 0.016722

𝐴∗ −.012775
Effective in-plane Poisson’s Ratio: 𝑣𝑦𝑥 = − 𝐴12
∗ = − 0.041068 = 𝟎. 𝟑𝟏𝟏
22

From this analysis, it is clear that there are distinct advantages and disadvantages associated
with both layer stacking sequences. For a [0/90/0]s laminate the effective in-plane longitudinal and
transverse modulus are high while the Poisson’s Ratios are low. However, the effective in-plane shear
modulus is low. For a [0/-45/45]s laminate, the effective in-plane longitudinal and transverse modulus
are low while the Poisson’s Ratios are high. However, the effective in-plane shear modulus is high. As a
result, it may be desirable to create a [90/45/-45/0]s laminate to better balance the shear modulus with
the other properties. However, since a pressure vessel is primarily loaded in the hoop and longitudinal
directions, the 0- and 90-degree orientations require the most reinforcement. Also, the hoop stress is
twice that of the longitudinal stress so that direction requires the most reinforcement. At the same time,
having the 90-degree lamina on the outside and inside surfaces aids to inhibit bucking of the pressure
vessel during service.

Composite materials have been selected in numerous applications including pressure vessels
due to their high strength-to-weight ratio and stiffness-to-weight ratio. Nevertheless, they are they are
vulnerable to various types of damage such as delamination. Hence, a principle consideration of
composite layer stacking is the free edge interlaminar stresses since they can cause delamination of
composite materials. The free edge interlaminar stresses are formed due to the differences in elastic
properties between the layer lamina that form the composite material. Due to the differences in elastic
properties (Poisson’s Ratios and coefficients of mutual influence) of the lamina, stresses develop near
the free edges of the laminate. Any free edge, including holes can introduce interlaminar stresses. Since
pressure vessels act to store pressure, there is inherently at least one-hole present. Thus, one must use
good design practices to avoid issues arising from this. A suggested design practices includes making the
lamina stacking sequence symmetric to prevent the generation of an extension-bending coupling.
Moreover, one should separate 90-degree lamina by a 0-degree ply to minimize interlaminar shear and
normal stress. Sequencing lamina and avoiding lamina in the same orientation is advised to create a
more uniform laminate (Sarvestani & Naghashpour, 2014). In the laminates discussed above, the free
edge interlaminar stresses of a machined hole should be relatively low for the [0/90/0]s and [0/-45/45]s
laminates since these good practices are maintained. The only concern is that pressure vessels are
subject to cyclic loading and even small free edge interlaminar stresses can be troublesome. This can be

25
further mitigated by having a liner made of isotropic material such as stainless steel or titanium which
can aid in preventing these types of stress.

All in all, the lamina stacking sequence and free edge interlaminar stresses must be considered
with designing a composite pressure vessel.

26
VI. Fatigue and toughness
Fatigue and toughness are two critical design considerations for many applications including
composite pressure vessels. Fatigue failure occurs due to variable stresses that are much lower than the
stress required to cause failure during a single loading. Fatigue is the result of three factors: 1) a
maximum stress of sufficiently high value, usually a tensile stress, 2) a large enough variation or
fluctuation in the applied stress, and 3) a sufficiently large number of cycles of the applied stress.
Moreover, there are three types of fluctuating stresses that commonly occur. First, there is fully
reversed loading in which the maximum and minimum stresses are equal. Second, there is repeated
stress cycles in which an average stress is applied on top of the maximum and minimum stresses. Third,
there random stress cycles in which a component is subjected to random loads during service (Campbell,
2010).

Pressure vessels function to store liquid or gas at pressure. As a result, they are often loaded
and unloaded with varying types of fluctuating stresses depending on their application. For instance, gas
storage cylinders are periodically emptied ands refilled with varying degrees of pressure extracted and
refilled depending on their application. The same gas storage cylinders may even be used for multiple
applications. Occasionally they may be fully or partially emptied and refilled and every so often the
frequency of these loads can fluctuate depending on the end use. Thus, depending on the application of
the pressure vessel, the number of cycles can vary from 10-100 for spacecraft applications, up to 10,000
for automotive fuel tanks, and up to 100,000 for hydropneumatics accumulators. The dependence of
pressure under which a composite pressure vessel fails (𝑝̅𝑛 ) after n-cycles of loading from zero pressure
(𝑝 = 0) up to 𝑝 = 𝑝̅𝑛 can be approximated by the following linear relationship: 𝑝̅𝑛 = 𝑝̅ (1 − 𝑟𝑐 log(𝑛)).
In this relation, 𝑝̅ is the burst pressure under static tests, n is the number of cycles ensured, and 𝑟𝑐 is the
experiential coefficient. Theoretically, since the dependence is linear only one fatigue test is required to
determine 𝑟𝑐 (Vasiliev, 2009). However, fatigue data characteristically has more scatter than static tests
and composite test data always shows more scatter than metals in static testing. As a result, fatigue
data for composites has a substantial amount of scatter in the results and the only way to obtain reliable
data is to conduct a number of tests at each pressure level (Campbell, 2010).

Nevertheless, carbon fiber composites have been shown to have superior fatigue performance
compared to most metals as seen in Figure 20. It is clear from the figure that composites can exhibit
fatigue properties far superior to metals such as titanium and high strength steel. As a result, fatigue life
in composite structures is not as severe of a concern as that of metal structures. For a well-designed
carbon fiber composite structure, static load capacity is a greater concern compared to fatigue life which
is normally the critical design parameter in metal structures (Campbell, 2010). This is a result of the
differences in the fatigue failure mechanisms in metals and composites.

Fatigue failure in metals is usually the result of a single crack that slowly propagates through the
thickness of a section until the remaining cross section cannot support the load causing failure to occur.
In a composite, fatigue failure is very different. In lieu of one individual crack, numerous different types
of damage occur at different locations and eventually link up to cause failure. There are five major
damage mechanisms: matrix cracking, fiber fracture, crack coupling, delamination initiation, and
delamination growth and fracture. For a composite materiel subjected to fatigue, the fist step of failure
is matrix cracking due to tensile loads in the off-axis plies. Next, the matrix cracks eventually intersect

27
the main load-bearing plies which act as crack arrests blunting crack growth. However, eventually stress
concentrations cause fiber fracture eventually leading to longitudinal cracking. This allows delaminations
to grow causing fracture of the composite (Campbell, 2010). In order to have an improved fatigue
strength, carbon fibers should be selected rather than fiberglass or aramid fibers.

Figure 20: Fatigue Strength of Composites vs Metals

Most components are subject to impacts and loads during service. As a result, fracture
toughness and damage tolerance are key considerations in any design. Fracture toughness describes the
ability of a material to resist fracture while damage tolerance is the ability of a component to continue
functioning with defects. The way composite materials behave is very different than traditional
engineering materials such as steel. Most fiber reinforced polymer composites consist of brittle fibers
(glass or carbon fibers) in a weak, brittle polymer matrix (epoxy or polyester resin). Despite both the
reinforcement fibers and the matrix being brittle in both cases, the composites provide surprisingly good
fracture toughness as seen in Figure 21 (Ashby, 2010).

Figure 21: Fracture Toughness of Composites

28
These properties are due to composite’s unique construction which is different from traditional
materials. Matrix cracking, fiber-matrix interfacial debonding, fiber fracture, fiber pull-out and
separation of bonded plies (delamination) are unique failure mechanisms to composite materials
(Campbell, 2010). They permit energy from impacts to be dissipated by individual failures while still
having integrity in the bulk material. Composites have a network of dispersed reinforcement and
numerous interfaces between the fibers and the matrix. During deformation, the damage is dispersed
throughout the composite. However, the construction of composites allows a significant amount of
damage to be present before the ability to bear loads is significantly degraded since fibers fail
individually and the loads are transferred to adjacent fibers. After a significant amount of damage,
cracks can be formed in composites but their behavior is very different than that of cracks in steel.
Composites have numerous interfaces (between the fibers and matrix and between plies), these
interfaces aid in creating roadblocks that prevent cracks from propagating. However, foreign object
damage or low velocity impacts are can cause delamination along interfaces forming a complex network
of delaminations and matrix cracks. The sources of delamination can have a serious impact on damage
tolerance where the delamination can lead to the formation of thinner sub-laminates that are not as
resistant to buckling. However, it is possible to increase the toughness of a composite pressure vessel by
selecting appropriate reinforcement and matrix materials. Aramid fibers such as Kevlar 49 or S2 fiber
glass fibers offer improved toughness. Additionally, toughened epoxy resins may be used to improve
both the toughness and damage tolerance of a pressure vessel.

Another key consideration with composite pressure vessels is proof testing. Composite pressure
vessels are almost universally subjected to proof tests that often exceed the design pressure in order to
ensure that the pressure vessel is appropriately designed and manufactured for service (ASME, 2019).
However, this type of loading can cause matrix cracks and lead to internal damage that can severely
impact the life of the vessel.

All things considered, composite pressure vessels can offer superior fatigue properties and have
some key advantages and disadvantages when it comes to damage tolerance.

29
VII. Recommendations
Composite pressure vessels offer some key advantages over their metal counterparts due to the
high specific strength and modulus of composite materials. As a result, composite pressure vessels can
offer significant weight savings which make their use desirable in many industries such as aerospace and
transportation. However, pressure vessels and their fabrication are highly dependent on the end use
and the specific materials are largely a function of use case.

In the case of pressure vessels used in weight critical applications such as rocket motor cases; the
key design consideration is the strength to weight ratio. In these applications, the pressure vessel is not
undergoing as many cycles and the lifespan is rather short with 10-100 cycles. As a result, the key design
limitation is strength to preform the job. In this type of application, the weight is minimized by using a
carbon fiber reinforcement with an epoxy matrix with a 65 percent fiber volume (Peters, 2011).

In the case of pressure vessels for transportation, which may be used to store gases such as
hydrogen for fuel cell vehicles the design considerations are different. First, the pressure vessel must be
capable of achieving the required strength but cost and mass must also be considered. Additionally,
these applications have many more cycles from 100-1000 cycles making fatigue a more serious
consideration. Moreover, in these types of applications toughness and damage tolerance must be
considered to avoid disastrous results if accidents were to occur. As a result, the ideal composite would
employ carbon and aramid fiber reinforcement with a 65 percent fiber volume fraction with a
toughened epoxy resin matrix (Campbell, 2010; Peters, 2011).

In all cases, the ideal manufacturing technique to manufacture the pressure vessel would be
filament winding. This method provides the fastest and most cost-effective manufacturing method while
allowing for good quality. Since most pressure vessels are bodies of revolution, they can be easily made
using this method (Campbell, 2010). The fiber orientation is carefully controlled and the resin is applied
to the fiber tow allowing for proper fiber saturation. Moreover, the tow is held taught as it is applied to
the rotating mandrel allowing the fiber to be uniformly placed without bucking. This method also allows
for the metal fittings to be integrated into the vessel producing a lower potential for defects. However,
the geometry of the fittings must be optimized to allow them to be secured properly. It is recommended
that the fittings be designed on same principle as Figure 22 to ensure that the pressure is uniformly
distributed in the hemispherical heads while also preventing separation as the composite would hold
the fitting in during service.

Figure 22: Recommended Fitting Design

30
It is advised that wet-up methods be avoided for any critical applications as the likelihood of defects will
increase and seams can act as stress risers in the final pressure vessel which may lead to premature
failure. Also, autoclave curing should be employed to generate a uniform cross section with minimal
voids. A post-cure should be performed to ensure that the maximum amount of cross-linking is
generated in the matrix to ensure optimal strength. The ideal layup for fabrication would be a [0/-
45/45/90]s laminate for several reasons. First, this laminate allows for proper manufacturability with the
filament winding process. Additionally, it allows strength to be concentrated about the circumference
with the 90° lamina and about the longitudinal axis. The +/- 45° lamina help ensure that stresses are
distributed evenly along the surfaces to prevent delamination between the layers.

Another recommendation would be to forgo the traditional proof test for an alternative quality
examination. Traditionally, proof tests are used to verify that the pressure vessel is manufactured
correctly by loading the pressure vessel beyond the design pressure and holding. However, in composite
structures, there may be irreversible damage that is caused by this methodology. Alternatively, strain
gauges can be applied to the pressure vessel and the max design pressure can be applied. If the stain
exceeds the expected, it shows that there is a defect with the manufacturing process. Moreover, non-
destructive techniques such as ultrasonic or x-ray techniques can be used. CT scanning is a relatively
new technique that can identify porosity location and size and monitor that the laminate plies are in
good condition as seen in Figure 23 (Good & Steiner, 2017).

Figure 23: Fig. 2: CT scan of a helicopter blade section which shows porosity (location and size) as well as the individual laminate
plies.

It is also recommended that a thin liner made of aluminum, stainless steel, HDPE is used for
most types of pressure vessels. Composites typically have some degree of porosity which can allow the
leakage of gas or liquid at high pressures. The liner will provide a solid barrier that transmits the
pressure load to the composite material while preventing leakage of any gas or liquid. It would also act
to prevent moisture from degrading the composite over time while also providing an extremely
chemical resistant liner if HDPE is used.

All things considered, if the application of the composite pressure vessel is fully analyzed and
the materials, manufacturing techniques, and testing techniques are appropriately selected composite
pressure vessels provide a number of key advantages over traditional metallic pressure vessels.

31
VIII. Summary
Traditionally, pressure vessels are constructed from metals such as steels. However, with special
considerations composite materials can be used to construct pressure vessels with some distinct
advantages and disadvantages.

Composite materials provide a high specific strength and modulus of elasticity when compared to
traditional engineering materials such as metals. This allows composite pressure vessels to be more
efficient when compared to their metal counterparts in terms of strength to weight ratios (Ashby, 2010).
This allows lighter weight pressure vessels to be fabricated that are able to withstand the same
pressures allowing for significant weight savings which is key in aerospace, automotive, and other
industries where weight savings are of significant concern.

For commercial transportation applications, composite pressure vessels can reduce the weight
of gas containment systems by approximately 1,000 lbs (Legault, Pressure Vessel Tank Types, 2012).
Weight reduction of this magnitude does more than providing better fuel economy, it would augment
the capacity of the equipment and allow for other operational benefits such as better handling.
Moreover, composite pressure vessels enhance the practical limit of gas containment pressures by
providing more efficient potential energy storage. As a result, for storing pressures above 5,000 psi
composite pressure vessels often provide a more practical solution (Legault, Pressure Vessel Tank Types,
2012). Moreover, composite materials are not degraded by corrosion as are metallic pressure vessels.
Advantages aside, there are a number of things that must be considered to design a composite
pressure vessel. First, the intended application must be analyzed to determine the key design
considerations such as weight savings, fatigue life, cost, etc. Next, the fiber reinforcement and matrix
must be selected. From there the layup of the lamina must be considered along with the optimal
manufacturing methodology. Process characteristics such as cure cycles, or individual process
parameters such as the use of an autoclave to apply pressure during cure must be considered.
Afterwards, quality evaluation must be considered in order to verify that the design’s intent has been
met.
While composite pressure vessels offer some key advantages over traditional metallic pressure
vessels. The key in implementing a material insertion is to analyze the intended application and
determine what is necessary for a composite pressure vessel in a specific application.

32
IX. References
Ashby, M. F. (2010). Materials Selection in Mechanical Design. Boston: Butterworth-Heinemann.

ASM International. (2001). ASM Handbook, Volume 21, Composites. ohiop: ASM International.

ASME. (2019). Boiler and Pressure Vessel Code (BPVC) Section X: Fiber-Reinforced Plastic Pressure
Vessels. New York: ASME.

Campbell, F. C. (2010). Structural Composite Materials. Materials Park: ASM International.

Christensen, R. (2012). Mechanics of Composite Materials. Newburyport : Dover Publications.

DFC Pressure Vessel Manufacturer Co, Ltd. (2019, April 3). Failure modes of pressure vessels. Retrieved
from DFC Tank: https://www.dfctank.com/news/three-kinds-of-failure-modes-of-pressure-
vessels.html

Ellenberger, P. (2004). Pressure Vessels: ASME Code Simplified. McGraw-Hill.

Filament winding. (2019, September 29). Retrieved from Wikipedia:

https://en.wikipedia.org/wiki/Filament_winding

Good, A., & Steiner, D. (2017, January 31). X-rays for NDT of composites. Retrieved from Composites
World: https://www.compositesworld.com/articles/x-rays-for-ndt-of-composites

Hyer, M. W. (1998). Stress Analysis of Fiber-Reinforced Composite Materials. Singapore: McGraw-Hill

Book Co.

Legault, M. (2012, July 2). Next-Generation Pressure Vessels. Retrieved from Composites World:
https://www.compositesworld.com/articles/next-generation-pressure-vessels

Legault, M. (2012, July 2). Pressure Vessel Tank Types. Retrieved from Composites World:
https://www.compositesworld.com/articles/pressure-vessel-tank-types

Mohan , K., Yarrapragada, R., & Kiran, V. (2012). Composite Pressure Vessels. International Journal of
Research in Engineering and Technology.

Nabors, S. (2005, March 1). Making a Metal-Lined Composite-Overwrapped Pressure Vessel. Retrieved
from Tech Briefs:
https://www.techbriefs.com/component/content/article/tb/techbriefs/manufacturing-
prototyping/747

Peters, S. T. (2011). Composite Filament Winding. Materials Park: ASM International.

Pressure Vessel. (2019, August 3). Retrieved from Wikipedia:

https://en.wikipedia.org/wiki/Pressure_vessel

Sarvestani , H., & Naghashpour, A. (2014). Analysis of Free Edge Stresses in Composite Laminates Using
Higher Order Theories. Indian Journal of Materials Science.

33
Vasiliev, V. (2009). Composite Pressure Vessels: Analysis, Design, and Manufacturing. Blacksburg: Bull
Ridge Publishing.

34