Cryogenics 68 (2015) 44–58

Contents lists available at ScienceDirect

Cryogenics
journal homepage: www.elsevier.com/locate/cryogenics

Cryogenic mechanical behavior of 5000- and 6000-series aluminum

alloys: Issues on application to offshore plants
Doo-Hwan Park a, Sung-Woong Choi b, Jeong-Hyeon Kim a, Jae-Myung Lee a,⇑
a
Department of Naval Architecture and Ocean Engineering, Pusan National University, 30, Jangjeon-Dong, Geumjeong-Gu, Busan 609-735, Republic of Korea
b
Department of Extreme Energy Systems, Korea Institute of Machinery and Materials, Daejeon 305-343, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: The mechanical behavior of aluminum alloys was investigated in terms of four aspects: temperature,
Received 1 September 2014 strain rate, material type, and fracture shape. The candidate materials were 5000- and 6000-series alloys.
Received in revised form 1 February 2015 The material characteristics were investigated and summarized as a function of low temperature (110–
Accepted 9 February 2015
293 K) and quasi-static strain rate (104 and 102 s1). The results confirmed that the strength and duc-
Available online 18 February 2015
tility of aluminum alloys improved with a decrease in the temperature. The aluminum alloys showed a
strain rate effect only in terms of the ductility of the 5000-series alloys. In addition, fractography analyses
Keywords:
were performed on the fracture specimens to explain the material behavior at cryogenic temperatures.
Aluminum alloy
Cryogenic mechanical test
Ó 2015 Elsevier Ltd. All rights reserved.
Fracture
Strain rate sensitivity
Scanning electron microscopy (SEM)

1. Introduction at low temperatures and superior corrosion resistance. However,

The focus of natural resource development is now shifting from
continental to the polar and subsea regions. In keeping with this
trend, it has become increasingly important to build vessels that
can safely travel to and carry people and goods into these areas.
Many vessels, such as icebreakers, liquid natural gas (LNG) carriers,
drillships, and offshore plants such as FPSO (floating, production,
storage, and offloading) and jacket structures, are already operat-
ing in this field. Most of these vessels and structures are pioneering
new routes into the Arctic to contribute to subsea resource devel-
opment projects. Therefore, it is essential to consider low tem-
peratures when designing a structure for operating in these
harsh conditions. In particular, an understanding of structural
behavior at cryogenic temperatures is vital for designing LNG car-
riers, pipelines, etc., because the inner walls of these structures are
in direct contact with LNG at 110 K (163 °C). To predict the struc-
tural behavior at low temperatures reliably, a precise understand-
ing of the mechanical behavior and fracture characteristics of the
main materials used in the structure is needed first.
Recently, stainless steel, nickel steel, and invar alloys have
mainly been used as cryogenic materials in ships and offshore
structures. These materials show excellent mechanical properties
⇑ Corresponding author. Tel.: +82 51 5102342; fax: +82 51 5128836.
E-mail address: jaemlee@pusan.ac.kr (J.-M. Lee).
cryogenic materials are expensive because nickel, which is com-
monly used in such materials, has become increasingly expensive
over the years. Therefore, various studies have attempted to reduce
the cost of cryogenic materials. One representative approach used
toward this end is overlay welding, in which a stainless steel pipe
for the transportation of crude oil is manufactured by substituting
the inside surface of an existing general steel pipe, which is in
direct contact with the crude oil, with a stainless steel weld
overlay.
In keeping with this trend, aluminum alloy, which is widely
used in ships and offshore structures, has been focused on as an
alternative material for cryogenic conditions. These alloys have
shown great promise in various industries on account of their
low weight, high strength-to-weight ratio, superior mechanical
properties, and good formability. In addition, aluminum alloys
show significantly higher strength and ductility at cryogenic tem-
peratures than at room temperature. Thus, there are many appli-
cation examples of aluminum alloys, such as LNG carrier
insulation systems (Moss- and SPB-type tanks), Arctic equipment
(Arctic chemical processing equipment, pressure vessels, and heat
exchangers), and offshore structures (helidecks, subsea pipelines,
and drill pipes). Until 2000, the 5083 alloy was the main material
used in the Moss-type LNG tank. Therefore, aluminum alloys have
already been used as cryogenic materials for a long time, and their
superior properties in such applications have already been
verified [1].

http://dx.doi.org/10.1016/j.cryogenics.2015.02.001
0011-2275/Ó 2015 Elsevier Ltd. All rights reserved.
 D.-H. Park et al. / Cryogenics 68 (2015) 44–58 45

Several experimental studies have investigated the mechanical Therefore, to confirm the applicability of aluminum alloys for
behavior and temperature dependency of aluminum alloys under offshore plants and provide quantitative data about these materi-
quasi-static or dynamic strain rates. Picu et al. [2] investigated als under cryogenic condition, the material behavior changes of
the mechanical behavior of the commercial aluminum alloy aluminum alloys were specifically investigated in this study in
AA5182-O at temperatures of 153–423 K and strain rates of terms of four aspects: temperature, strain rate, material type, and
106–101 s1 through a uniaxial tension test. As a result, the fracture shape. In addition, two aluminum alloys (5000- and
Portevin–Le Chartelier (PLC) effect was verified at temperatures 6000-series), which are widely used in the shipbuilding and off-
of 193–430 K and strain rates below 101 s1. The effects of shore plant industries, were selected as candidate materials. The
dynamic strain aging on ductility and strain hardening were also material characteristics were investigated and summarized as a
studied. Hadianfard et al. [3] showed that AA5754 and AA5182 function of low temperature (110–293 K) and quasi-static strain
sheets had negative strain rate sensitivity at quasi-static rates rate (104 and 102 s1) [11,12].
but slightly positive sensitivity at dynamic rates. This result was
proved by tensile tests under quasi-static (less than 101 s1) and
dynamic (600–1500 s1) strain rates at elevated temperatures 2. Experiments
(296–573 K). Fan et al. [4] studied the dynamic mechanical behav-
2.1. Materials and preparation
ior of the 6061 aluminum alloy at elevated temperatures and vary-
ing strain rates. They showed that the mechanical properties of this
The experimental materials used in this study were 5000- and
alloy were more sensitive to the temperature than the strain rate.
The experimental results proved that the effect of the strain rate 6000-series aluminum alloys. Various aspects of these materials
have been investigated extensively [13–20].
was more critical at comparatively high temperatures (573–
673 K), whereas the effect of strain hardening was more important 5000-series aluminum alloys are mainly composed of Al–Mg
and non-heat-treated alloys. As the proportion of Mg increases,
at comparatively low temperatures (293–423 K). Wang and Jiang
[5] performed dynamic compressive tests on 2024-T6 and 7075- the tensile strength and stiffness of these alloys increase, but the
formability decreases. These alloys show high strength, good weld-
T6 aluminum alloys by using a Split-Hopkinson pressure bar
(SHPB) at low temperatures. The test results showed that the effect ability, and excellent corrosion resistance in a seawater environ-
ment [21]. Therefore, they are mainly used in shipyard sheets
of flow stress decreased linearly with increasing temperature. In
addition, the relatively low-temperature environment adopted in and plates, bulkheads and superstructures, materials for welded
structures, plants for low-temperature liquefied gas, etc. Represen-
the experiments did not affect the strain-hardening behavior.
In recent years, the cryogenic behavior of aluminum alloys has tative 5000-series alloys include 5052, 5083, 5086, and 5454.
6000-series aluminum alloys are mainly composed of Al–Mg–Si
been investigated under various loads. Park et al. [1] performed
tensile tests under different temperatures (110–293 K) and strain and heat-treated alloys. They show good formability and corrosion
resistance with high strength, and they are widely used as struc-
rate ranges (0.00016–0.01 s1) to understand the material behav-
ior of austenitic stainless steels (ASSs), aluminum alloys, and nickel tural materials. These alloys are used in general structural materi-
als for vehicles and architecture, pressure vessels, aluminum
alloys. The increase in ductility, yield, and ultimate tensile stress
for six types of cryogenic materials (AISI 304L, 316L, 321, 347, helidecks, etc. Representative 6000-series alloys include 6061,
6063, 6N01, and 6082.
AA5083, and Invar steel) used in LNG carriers was investigated at
low temperatures. In addition, to demonstrate these nonlinearities, A hyphenated suffix to the basic alloy number indicates the
temper designation; specifically, -H and -T temper designations
the strain rate sensitivity and strain-hardening rate were investi-
gated in terms of the flow strain, temperature, and strain rate. indicate strain hardened and thermally treated, respectively
(Table 1). In this study, the 5052 alloy was manufactured by per-
Glazer et al. [6] investigated the cryogenic tensile behavior of alu-
minum–lithium alloys. Tensile tests were conducted along the lon- forming strain hardening with 37.5% reduction in area, followed
by stabilization at room temperature. In addition, 6061 and 6082
gitudinal and long transverse directions at cryogenic temperatures
(298, 77, and 4 K), with the tested materials being AA2090-T81, alloys were manufactured by performing solution heat treatment
for 3 h at 803 K and artificial aging for 8 h at 453 K. The compo-
AA2090-T4, and a binary aluminum–lithium alloy. This study
demonstrated that the tensile strength and elongation of all tested nents of various alloys are listed in Table 1.
materials increased as the temperature decreased. In addition, the The experiments were mainly conducted using a universal test
strain-hardening rate in the longitudinal direction increased under machine (UH 1000KN, SHIMADZU) equipped with a cryogenic
the same conditions. Tarlupa [7] studied the deformation of an alu- chamber (Fig. 1) to measure the cryogenic mechanical behavior
of the aluminum alloys. In the cryogenic chamber, the temperature
minum alloy and its plasticity effect at low temperatures (77–
173 K). It was proved that the allowable level of strain increased of the test specimen was controlled using a temperature system;
the chamber was cooled using liquid nitrogen supplied through a
as the plasticity of the alloys increased at the experimental tem-
peratures. Furthermore, alloys that were cooled to low tem- mass flow control valve that was connected to the liquid nitrogen
supply. The test temperature was maintained for 30 min to achieve
peratures (77–123 K) after the hardening process (heat
treatment) exhibited an enhanced capacity for plastic deformation, thermal equilibrium between the environmental temperature and
the temperature of the test specimen after pre-cooling the
especially the aluminum–copper–lithium alloy.
Many other studies have also focused on aluminum alloys. In
addition, aluminum alloys have been widely studied experimental- Table 1
ly at room temperature [8–10]. However, the specific behavior of Components of selected aluminum alloys in wt (%).
aluminum alloys at low temperatures has yet to be clarified. In par-
Alloy Chemical composition (%)
ticular, few studies have focused on the low-temperature and
Si Fe Cu Mn Mg Cr Zn Ti Al
strain-rate behavior of these alloys for shipbuilding and offshore
plant industries. Therefore, aluminum alloy structures are 5052-H34 0.05 0.12 0.01 0.46 3.50 0.15 0.01 – Bal.
5086-H112 0.05 0.16 0.01 0.01 2.55 0.26 0.01 0.02 Bal.
designed based on empirical approaches, because quantitative ref-
6061-T6 0.66 0.36 0.17 0.08 0.84 0.07 0.12 0.02 Bal.
erence data that can be used immediately in the relevant field of 6082-T6 0.95 0.19 0.04 0.72 0.89 0.08 0.02 0.02 Bal.
industry are lacking.
46 D.-H. Park et al. / Cryogenics 68 (2015) 44–58

Fig. 1. (a) Photograph of the universal testing machine with cryogenic chamber and (b) schematic overview of the experimental apparatus.

Fig. 2. Tensile test under (a) ambient and (b) cryogenic conditions.

specimen. The tensile test environment is shown in Fig. 2 under
ambient and cryogenic conditions in the chamber. As shown in
Fig. 3, all material specimens were prepared according to the Kore-
an Industrial Standards (KS B0801 10). For more accurate data, a
cryogenic extensometer (K type, 3542-050M-100-LT, Epsilon Tech)
was mounted according to the gauge length of the specimen
(50 mm).
2.2. Experimental scenario
The test temperatures reached the temperature of liquid hydro-
gen (110 K), and the low-temperature range was divided into four
short ranges to investigate the cryogenic properties in detail. A
strain rate below 101 s1 is widely known as a quasi-static
condition, and the mechanical behavior of aluminum alloys is inde-
pendent of the quasi-static strain rate condition [4]. In addition,
when measuring the strain rate sensitivity using a strain rate jump
test, a strain rate difference of 1:100 is mostly used [2]. Therefore,
quasi-static strain rates of 104 and 102 s1 were adopted in the
present study to identify the rate dependence of aluminum alloys.
The experimental scenario is summarized in Table 2. To obtain
repeatable test results, three tests were performed for each scenar-
io (120 tests in total). Fig. 4 shows the variance of the test results in
the case of No. 10 for AA5052 and AA6061. As the stress and strain
values were significantly low, their average values were used to
prepare the representative stress–strain curves. Finally, fractogra-
phy analyses of the fracture specimens were conducted using an
S-4800 SEM (Hitachi Inc., Japan) at an accelerating potential of
20 kV to analyze the behavior of each material.
3. Comparative analysis
3.1. Temperature
Fig. 5 shows the temperature-dependent stress–strain relation-
ship of the aluminum alloys at a strain rate of 104 s1. The
 D.-H. Park et al. / Cryogenics 68 (2015) 44–58 47

Fig. 3. Test specimen shape and dimensions of aluminum alloy (KS B0801 10).

Table 2
and the increase in yield stress (5000-series: 12.49–16.05 MPa;
Conditions for cryogenic tensile tests.
6000-series: 52.23–107.26 MPa) of the 5000-series alloys was low-
No. Test temperature (K) Strain-rate (s1) Specimen (EA) Material er than that of the 6000-series alloys. Similarly, the yield, ultimate
1 104 3 tensile stress, and increase rate of the 6000-series aluminum alloys
2 293 102 3 were significantly superior to those of the 5000-series aluminum
3 104 3 AA5052
223 alloys. However, the ultimate tensile stress of the 5000-series
4 102 3 AA5086
5 104 3 AA6061 alloys showed a considerable increase compared to its yield stress.
153
6 102 3 AA6082 For AA5052, the increase rate of the ultimate tensile stress on yield
7 104 3 stress was 137% at room temperature (293 K) and 221% at cryo-
133
8 102 3 genic temperature (110 K); for AA5086, these rates were 174%
9 104 3
10
110
102 3
and 204%, respectively.
A distinct hardening characteristic of 5000-series aluminum
alloys was confirmed by performing a comparative investigation
with 6000-series aluminum alloys; a 9% increase was observed
under cryogenic conditions. In general, 5000- and 6000-series alu-
minum alloys have been reported as non-heat-treated and heat-
treated alloys, respectively. To increase the mechanical strength
of these aluminum alloys, a different method is normally adopted.
For 5000-series aluminum alloys, the material strength can be
determined by controlling the amount of strain because they are
strengthened by strain hardening. Therefore, when a tensile load
is applied to non-heat-treated 5000-series alloys, a clear hardening
phenomenon can be observed after the yield strength. However, in
the case of 6000-series aluminum alloys, the mechanical strength
can be improved by alloying additional elements such as Cu, Mn,
Zn, and Si for precipitation hardening. Precipitation hardening is
used to increase the strength and stiffness of the material by form-
ing minute particles in the original phase matrix. The material is
hardened by the precipitation and solution of alloying elements
during heat treatment [22]. Therefore, the distinct hardening of
6000-series alloys is difficult to confirm under simple tensile loads
alone.
On the other hand, the variation of the fracture strain with
decreasing temperature showed a different trend depending on
the type of aluminum alloy (Fig. 6). The fracture strain of 5000-ser-
Fig. 4. Stress–strain curves for AA5052 and AA6061 at 110 K. ies alloys increased considerably as the temperature decreased,
whereas that of 6000-series alloys did not exhibit any significant
temperature dependence. However, for 5000-series alloys, the
ultimate tensile stress of all adopted testing materials increased fracture strain increased steadily only until 133 K and subsequent-
steadily as the temperature decreased. All of the tested aluminum ly decreased at 110 K. For AA5052 and AA5086, the increase rates
alloys showed a rate of increase over 15.00% at 110 K compared to of fracture strain, based on the maximum value, were 43.54% and
room temperature conditions. The highest rate of increase rate was 59.04%, respectively. In order words, 5000-series alloys were
30.80% for the AA6082 aluminum alloy, and the lowest was 14.94% lengthened to one-and-a-half times their original length at cryo-
for the AA5052 aluminum alloy. The yield stress showed similar genic temperatures. For 6000-series alloys, the increase rates of
characteristics. The yield stress of AA6061 and AA6082 increased fracture strain were 11.93% and 1.23% for AA6061 and AA6082,
with decreasing temperature at a rate of 15.55% and 31.48% respectively. This confirmed that 5000-series alloys were superior
respectively. The increase rate (AA5052: 13.29%, AA5086: 16.18%) to 6000-series alloys in terms of fracture strain.
48 D.-H. Park et al. / Cryogenics 68 (2015) 44–58

Fig. 5. Temperature dependent stress–strain curves of the aluminum alloys: (a) AA5052, (b) AA5086, (c) AA6061 and (d) AA6082.

The microscopic behavior of the aluminum alloys is related to
the characteristics of their metallic materials. Aluminum alloys
have a face-centered cubic (FCC) crystal structure that shows a
large number of slip systems caused by the movement of disloca-
tions in the plastic deformation process. From a macroscopic view-
point, plastic deformation signifies a permanent deformation due
to slip. Thus, materials with many slip systems, such as copper,
nickel, and aluminum, have good ductility, resulting in greater per-
manent deformation. Furthermore, at low temperatures, the FCC
crystal structure activates the movement of dislocations without
changing the lattice structure, thereby causing the ductility of the-
se materials to increase or be maintained at low temperatures
[22,23].
Therefore, as reported in previous studies, aluminum alloys
show unique behavior like stress and ductility increase (or con-
stant behavior) even under low-temperature conditions, unlike
common metals [1]. Accordingly, the present study discussed the
temperature-dependent material characteristics, such as strength
and fracture strain, and the applicability to cryogenic temperature
conditions. Furthermore, it was identified that these alloys were
temperature-dependent materials and were suitable as cryogenic
materials in terms of strength and fracture strain.
3.2. Strain rate
Fig. 7 shows the strain-rate-dependent stress–strain relation-
ship of the aluminum alloys at room and 110 K. There was no
strain rate dependence in terms of the tensile stress and fracture
strain at quasi-static strain rates for the 6000-series alloys
(Fig. 7(c) and (d)). Similarly, for the 5000-series alloys, the tensile
stress was independent of strain rate (Fig. 7(a) and (b)).
Fig. 8 shows the strain-rate dependent yield and ultimate ten-
sile stress and fracture strain for all tested aluminum alloys. As
indicated in this figure, variation for stress and strain values
 D.-H. Park et al. / Cryogenics 68 (2015) 44–58 49

Fig. 6. Temperature dependent ultimate yield stress and fracture strain of the aluminum alloys: (a) AA5052, (b) AA5086, (c) AA6061 and (d) AA6082.

depending on strain-rates in all tested aluminum alloys can be
easily confirmed. For the 5000-series alloys, the effect of the strain
rate dependence on fracture strain was different for room (293 K)
and cryogenic (110 K) temperatures. At room temperature, the
fracture strain was longer when the strain rate was relatively fast
(102 s1) than when it was relatively slow (104 s1; Fig. 7). How-
ever, at cryogenic temperature, these trends were opposite to that
at room temperature; in other words, at 110 K, the fracture strain
was long when the strain rate was relatively slow. Furthermore,
the increased rate of fracture strain at the slow strain rate
(104 s1; AA5052: 43.54%, AA5086: 59.04%) was higher than that
at the fast strain rate (102 s1; AA5052: 18.58%, AA5086: 27.05%).
On the other hand, for the 6000-series alloys, no effect on strain
rate dependent fracture strain was obtained. This trend could be
confirmed by the imaginary extension line connecting the two
fracture strains of the 5000-series alloys intersecting (Fig. 8(a)
and (b)) and that of the 6000-series alloys (Fig. 8(c) and (d)) being
almost parallel. In addition, the fracture strain between two strain
rates could be generally predicted from previous research, as the
fracture strain behavior for 5000- and 6000-series aluminum alloys
has been linearly increased in the range of the quasi-static strain
rates [3,12,24,25].
Thus, following on from Section 3.2, it was identified that the
strength of all the tested materials and the fracture strain of the
6000-series alloys were independent of the strain rate. In other
words, the strain rate dependency of aluminum alloys could only
be identified in terms of the fracture strain of the 5000-series
alloys. In this regard, it has been reported that the effect of strain
rate on the strength and fracture strain of aluminum alloys was
insignificant at room and cryogenic temperatures, because
50 D.-H. Park et al. / Cryogenics 68 (2015) 44–58
Fig. 7. Stress–strain relationship with regard to the strain-rate of the aluminum alloys at 110 and 293 K, for (a) AA5052, (b) AA5086, (c) AA6061 and (d) AA6082.
aluminum alloys has a small strain rate sensitivity [9,26]. Also, the
strain rate sensitivity drops with an increase of the alloy content
and tempering. Therefore, the 6000-series alloys were less influ-
enced by the variation of strain rate than the 5000-sereis alloys,
as heat treatment and cold working reduces strain rate sensitivity
of 6000-series alloys.
3.3. Aluminum alloy type and its applicability
Offshore structures such as FPSO structures, drill ships, and
jack-up rigs consist of several structural components to extract
deepwater natural resources such as crude oil and natural gas. In
their structural design, harsh environmental conditions should be
considered in view of the safety aspect. In general, because the per-
formance of the structural members depends on the material capa-
city, various considerations must be made before structural design.
In the last few decades, stainless steel, nickel alloys, and aluminum
alloys have been widely used in such offshore structures. Among
these, the present study focuses on the mechanical characteristics
of four aluminum alloys and performs a comparative study with
AA5083. AA5083 is the most widely used material for low-
temperature conditions in marine structures, and its applicability
to ship and offshore structures has been proved in a previous
study [1].
Fig. 9 shows the comparative investigation between AA5083
and the tested aluminum alloys. All data were selected based on
the maximum properties under low-temperature conditions from
110 to 153 K. Fig. 9(a) shows that the yield, ultimate tensile
strength, and fracture strain of each material increased at cryo-
genic temperatures compared to room temperature conditions. In
other words, aluminum alloys with a value above 1 (reference val-
ue) showed improved mechanical properties at low temperature.
All tested alloys showed a value above 1. Therefore, the mechanical
properties of these alloys, including AA5083, improved at low tem-
perature. In comparison to AA5083, the rate of increase of the four
aluminum alloys showed a similar tendency, except for fracture
strain. In addition, AA6082 exhibited the highest increase rate in
terms of the yield (1.30) and ultimate tensile stress (1.31), and
AA5052 exhibited the highest increase rate in terms of fracture
strain (1.58).
 D.-H. Park et al. / Cryogenics 68 (2015) 44–58 51

Fig. 8. Strain-rate dependent yield and ultimate tensile stress and fracture strain, for (a) AA5052, (b) AA5086, (c) AA6061 and (d) AA6082.

In Fig. 9(b), the yield and ultimate tensile stress and fracture
strain of AA5083 were fixed at 1, and the properties of the other
aluminum alloys were expressed in terms of their ratio to
AA5083. If the property value of an aluminum alloy was above 1,
it indicated that this property was superior compared to that of
AA5083. It was confirmed that the ultimate tensile stress of the
5000-series alloys was markedly lower than that of AA5083
(AA5052: 0.43, AA5086: 0.28), whereas the fracture strain was
higher (AA5052: 1.64, AA5086: 2.01). On the contrary, the ultimate
tensile stress of 6000-series alloys was higher than that of AA5083
(AA6061: 1.23, AA6082: 1.42), whereas the fracture strain was
lower (AA6061: 0.74, AA6082: 0.55). In terms of the yield and ulti-
mate stress, except for AA5083, AA6082 exhibited the highest val-
ue (yield stress: 1.16 and ultimate stress: 1.42), followed by
AA6061, AA5052, and AA5086. In addition, the fracture strain of
the 5000-series alloys was higher than that of the 6000-series
alloys. It was noted that the order of the aluminum alloys for both
tests was exactly the opposite for strength compared to strain. In
other words, the aluminum alloy with high strength had lower
fracture strain both at room and low temperatures. The com-
parative study with AA5083 confirmed that the mechanical prop-
erties of all tested aluminum alloys increased similarly to those
of AA5083 at cryogenic temperatures. In addition, the four alu-
minum alloys had different strength and ductility in comparison
with AA5083. This means that these alloys could be applied in
cryogenic conditions in various fields.
The fracture strain of the 5000-series alloys increased consider-
ably at low temperatures. Therefore, these alloys are suitable for
52 D.-H. Park et al. / Cryogenics 68 (2015) 44–58

Fig. 9. (a) Temperature dependent increase rate of mechanical properties (value at cryogenic temperature/value at room temperature) and (b) material-type dependent ratio
of mechanical properties (value of each aluminum alloy/value of AA5083).

use in structures that require large displacements. However, neces-
sary precautions are required when 5000-series alloys are used at
temperatures below 110 K, because the fracture strain of the mate-
rial decreases and certain portions of the material show brittle
fracture. On the other hand, the superior tensile strength of
6000-series alloys could be used for structures that are subjected
to significant loading conditions under cryogenic conditions,
because the strength of these alloys increased steadily with
decreasing temperature. However, precautions are again required
for the application of these materials in structures that experience
large displacements. Therefore, 5000-series alloys are generally
used for thin plates and 6000-series alloys, for extrusions and thick
plates. In addition, the strength and ductility of each aluminum
alloy was complementary; for example, the low strength of
5000-series alloys could be sufficiently reinforced through proper
hardening of the material, instead of decreasing fracture strain.
In addition, because the 5000- and 6000-series alloys are of seawa-
ter-resistant grade, these materials are highly advantageous for use
in marine structures that are submerged for a long time in
seawater.
Similarly, aluminum alloys offer strong advantages for cryo-
genic applications in a marine environment. As frequently pointed
out, however, proper aluminum alloys should be selected by first
predicting the stress and displacement of the structures. To pro-
vide a quantitative indicator for the use of an aluminum alloy as
a cryogenic material, the experimental results for all the aluminum
alloys are summarized in Tables 3–6. The temperature dependen-
cies were calculated using the reference values at room tem-
perature and two strain rates. In Tables 3–6, ‘‘ref⁄’’ signifies the
reference values for the increment calculations. In addition, to
understand the test results clearly for different aluminum alloys,
the mechanical properties were plotted together in Fig. 10.

Table 3
Experimental results for AA5052.

e_ (s1) Temp. (K) rY (MPa) rY Increment (%) rT (MPa) rT Increment (%) eF eF Increment
4
10 293 120.76 ref⁄ 305.22 ref⁄ 0.303 ref⁄
223 126.49 4.74 283.23 7.20 0.338 11.35
153 131.21 8.64 314.64 3.08 0.419 38.09
133 136.30 12.86 333.50 9.26 0.436 43.54
110 136.81 13.29 350.82 14.94 0.372 22.61
102 293 126.62 ref⁄ 281.52 ref⁄ 0.320 ref⁄
223 130.80 3.29 279.18 0.83 0.334 4.44
153 131.97 4.22 310.82 10.40 0.380 18.58
133 122.24 3.46 330.14 17.26 0.375 17.01
110 89.42 29.37 346.36 23.03 0.303 5.28

Table 4
Experimental results for AA5086.

e_ (s1) Temp. (K) rY (MPa) rY Increment (%) rT (MPa) rT Increment (%) eF eF Increment
104 293 77.19 ref⁄ 211.97 ref⁄ 0.336 ref⁄
223 79.23 2.64 208.15 1.80 0.380 13.18
153 79.49 2.97 230.57 8.77 0.466 38.51
133 83.05 7.59 245.60 15.86 0.535 59.04
110 89.68 16.17 268.28 26.56 0.533 58.53
102 293 73.37 ref⁄ 201.01 ref⁄ 0.359 ref⁄
223 76.94 4.86 211.46 5.19 0.408 13.51
153 82.80 12.84 229.29 14.06 0.440 22.42
133 89.17 21.52 246.87 22.81 0.457 27.05
110 86.87 18.40 268.53 33.58 0.455 26.56
 D.-H. Park et al. / Cryogenics 68 (2015) 44–58 53

Table 5
Experimental results for AA6061.

e_ (s1) Temp. (K) rY (MPa) rY Increment (%) rT (MPa) rY Increment (%) eF eF Increment
4
10 293 335.79 ref⁄ 353.63 ref⁄ 0.177 ref⁄
223 337.07 0.37 356.94 0.93 0.192 8.86
153 367.64 9.48 398.21 12.60 0.191 8.25
133 372.99 11.07 408.40 15.48 0.198 11.93
110 388.02 15.55 425.47 20.31 0.185 4.90
102 293 337.57 ref⁄ 352.10 ref⁄ 0.175 ref⁄
223 342.42 1.43 362.03 2.82 0.175 0.19
153 379.61 12.45 405.35 15.12 0.185 5.59
133 385.22 14.11 417.57 18.59 0.192 9.59
110 395.41 17.13 433.88 23.22 0.198 12.89

Table 6
Experimental results for AA6082.

e_ (s1) Temp. (K) rY (MPa) rY Increment (%) rT (MPa) rT Increment (%) eF eF Increment
104 293 340.63 ref⁄ 375.54 ref⁄ 0.146 ref⁄
223 379.36 11.36 412.99 9.97 0.138 5.50
153 405.09 18.92 443.05 17.97 0.148 1.23
133 412.22 21.01 450.70 20.01 0.138 5.50
110 447.89 31.48 491.21 30.80 0.136 6.76
102 293 343.69 ref⁄ 368.40 ref⁄ 0.137 ref⁄
223 356.17 3.63 381.40 3.52 0.138 0.24
153 423.43 23.20 455.03 23.51 0.140 1.51
133 430.31 25.20 462.42 25.51 0.126 8.65
110 463.69 34.91 501.14 36.03 0.127 7.90

4. Fracture analysis be explained and predicted by examining the simple macroscopic

4.1. Macroscopic analysis
Aluminum alloys have an FCC crystal structure. In general,
metallic materials with this structure have numerous slip systems
and show plastic deformation owing to the movement of disloca-
tions along the slip system. Therefore, the ductility of aluminum
alloys is very high, and their tensile fracture surfaces show ductile
fractures such as cup-and-cone or shear fractures [22].
The fracture surfaces were characterized by ductility (fracture
strain  100%), and fractured materials with higher ductility, as
seen in the fractured cup-and-cone shape, were arranged from
the left in Fig. 11(a). This figure shows the typical macroscopic
fracture surfaces of the specimens, which exhibit representative
fracture shapes and fracture surfaces as a result of the tensile tests.
The most outstanding materials from each aluminum alloy series,
AA5086 and AA6061, were fractured in a cup-and-cone shape.
The total length of the AA5086 specimen after the tensile test
was longer, and the diameter of the fracture surface (circle) was
smaller than that of AA6061. In addition, for AA5086, which had
the highest ductility, a snail-shaped spiral shape was observed.
The holding trace of the particles in the internal structure was also
partly seen for AA5052. However, for AA6061, the shape was char-
acterized by a flat plane surface in the fracture surface. In
Fig. 11(b), the ductile fractures exhibited a characteristic surface
with a small shear lip, where the shear fracture had a 45° fracture
surface relative to the applied stress. The shear lip, which indicated
that slip occurred, was clearly observed in AA5052 and AA6082
specimens. Similar to the results of cup-and-cone-shaped frac-
tures, a holding trace of the particles was seen in the internal struc-
ture of AA5052, whereas the surfaces of AA6082 consisted mainly
of a flat shape. Therefore, even if the aluminum alloy was classified
by the same type of ductile materials, cohesion between particles
depended on the internal structure, which determined the ductility
of the material. This means that the ductility of the material could
observations of the fracture surface.
On the other hand, the fracture geometry for each material was
clearly classified by the environmental temperature. For AA5052, a
shear fracture was observed under room temperature and a cup-
and-cone-shaped fracture, at the other experimental temperatures
(Fig. 12). For AA6082, a shear fracture was observed under room
temperature. However, shear and cup-and-cone-shaped fractures
were observed in coexistence, as the temperature decreased to
cryogenic levels. In this regard, a recent study proved that surface
roughness and excessive hardness could result in shear fractures,
because the notch sensitivity was severe for very hard surfaces
because of the heat treatment schedule (oxidation and hot rolling)
leading to increasing stress [27]. In addition, through this study, it
was confirmed that the temperature is an important parameter
influencing the fracture surface, and low temperatures lead to
more cup-and-cone-shaped fractures.
4.2. Microscopic analysis
The fracture features of the specimens were examined using
SEM and by analyzing the microscopic fracture surfaces. Figs. 13–
16 show the fracture surfaces for the AA6061, AA6082, AA5052,
and AA5086 specimens at different temperatures and strain rates.
In each figure, the fracture surfaces are shown for different tem-
peratures, ranging from 110 to 293 K, under constant strain rates
of 104 s1 and 102 s1, respectively.
For AA6061 and AA6082, the fractured surface at low tem-
peratures was characterized by a dimple-like structure. The densi-
ty of the dimples increased with decreasing temperature, in a
manner typical of ductile fractures, regardless of the different
strain rates (Figs. 13 and 14). However, the temperature did not
have effect on the difference in ductility. This indicates that the
AA6061 and AA6082 specimens showed some ductile dependency
behavior as the temperature decreased, and the difference was not
much. Furthermore, the strain rate had little influence on the
54 D.-H. Park et al. / Cryogenics 68 (2015) 44–58

Fig. 10. Mechanical properties of different aluminum alloys.

ductile behavior. The effect of the strain rate on the ductility of the
AA6061 and AA6082 specimens was found to be insignificant
judging from the small strain rate sensitivity from the SEM
observations.
The SEM observations of the fracture surfaces of AA5052 and
AA5086 at different temperatures and strain rates are shown in
Figs. 15 and 16. For AA5052, it was observed that at constant strain
rates of 104 s1, the density of the dimples increased with
decreasing temperature until 153 K. The fracture surfaces at
110 K (Fig. 15(d)) showed a fragile aspect with a small portion of
brittle fraction surfaces. In some parts of the fracture surfaces,
the fractured grain was flat, and there were fragile aspects with
large brittle fracture surfaces and a small fraction of voids
compared to the fracture surfaces at 153 K. This indicates that
AA5052 had a brittle fracture that was shown along with a ductile
fracture under a temperature of 110 K. Fig. 16(a)–(d) show the
fracture surface of an AA5086 specimen that was fractured under
a strain rate of 104 s1. The low-temperature fracture surface
was characterized by dimple-like structures, with finer and smaller
dimples, and the depth and density of the dimples in the ductile
mode increased at cryogenic temperatures. Therefore, it can be
inferred that at a constant strain rate, the ductility increased as
the temperature decreased.
For AA5052 and AA5086, different fracture features were
observed in the specimens fractured at different strain rates and
temperatures. Fig. 15(a) and (e) show the typical fracture surfaces
 D.-H. Park et al. / Cryogenics 68 (2015) 44–58
Fig. 11. Photograph of (a) cup-and-cone and (b) 45° shear fractures.
Fig. 12. Fracture mode of (a) AA5052 and (b) AA6082 at each temperature.
of the AA5052 specimen and Fig. 16(a) and (e), those of the AA5086
specimen, at a temperature of 293 K under constant strain rates of
104 and 102 s1, respectively. Large void fractions and a higher
density of dimples were observed at the higher strain rate. The
dimples formed during the coalescence of voids suggested that
the fractures were in a more ductile manner at the higher strain
rate. It was noted that the ductility of AA5052 and AA5086
increased under a higher strain rate at a temperature of 293 K
(Fig. 8(a) and (b)). However, under a temperature of 110 K, differ-
ent fracture features were observed in the two cases. Fig. 15(d) and
55
(f) show the typical fracture surfaces of the AA5052 specimen and
Fig. 16(d) and (f), those of the AA5086 specimen, at a temperature
of 110 K under constants strain rates of 104 and 102 s1, respec-
tively. The voids and dimples that were formed at the low strain
rate were observed to be relatively fine and clear, although their
depth and density were reduced at the higher strain rate. Pre-
cipitate particle fractures were observed along with dimple-like
structures. In contrast to the strain rate dependency of the ductility
at 293 K, low-temperature fractures resulted in a combination of
voids and dimples as well as flatter fragile fracture surfaces, which
showed gradually decreasing ductility at the higher strain rate
(Fig. 8(a) and (b)).
4.3. Discussion
In general, ductile–brittle transition issues were more critical
for a metal’s behavior in environments with high loading rates
and large changes in temperature. It is well known that brittle frac-
tures are likely to occur for most metals when the loading rate is
increased and the temperature decreases [28]. The present study
demonstrated that shear fractures occurred at room temperature
(293 K) and cup-and-cone-shaped fractures, at a cryogenic tem-
perature (110 K), for AA5052 and AA6082. This indicates that the
ductile failure mode was the dominant character trait with
decreasing temperature. This result is opposite to what is com-
monly known.
To quantitatively describe this type of phenomenon, the frac-
ture behavior and its relationship with toughness had to be
explained. Toughness generally means the ability of a material to
plastically deform and absorb energy without rupturing. This
56 D.-H. Park et al. / Cryogenics 68 (2015) 44–58

Fig. 13. Fracture surfaces of the AA6061 specimen at different temperatures and strain-rates: (a) 293, (b) 222, (c) 153 and (d) 110 K at 104 s1, (e) 293 and (f) 110 K at
102 s1.

Fig. 14. Fracture surfaces of the AA6082 specimen at different temperatures and strain-rates: (a) 293, (b) 222, (c) 153, (d) 110 K at 104 s1, (e) 293 and (f) 110 K at 102 s1.

material property can be expressed by the modulus of toughness,
which is the energy of mechanical deformation per unit volume
prior to fracture [29]. The modulus of toughness in the static state
loading condition can be obtained by tensile tests and can be cal-
culated by using the area underneath the stress–strain (r–e) curve,
which gives the tensile toughness value. The quantitative numeri-
cal values for each aluminum alloy are listed in Table 7.
In the present study, the strength of the aluminum alloy
increased under cryogenic conditions, and ductility either
remained constant or increased (Figs. 5 and 6). Therefore, the mod-
ulus of toughness of all tested alloys, including AA5052 and
AA6082, which presented a marked change in the macro analysis
of the fracture surface for temperature, had a higher value at tem-
peratures below room temperature (Fig. 8). The value was
increased by 22.78% for AA6082 to 102.83% for AA5086. When
comparing the modulus of toughness values and fracture analyses
of the four aluminum alloys, it was not evident whether the tough-
ness modulus value itself was directly related to the density of the
dimples and voids in the SEM observations. It therefore has to be
applied differently in each material. In addition, the modulus of
toughness was not an efficient measure to obtain the degree of
ductility and fracture shape of materials, although a high tough-
ness modulus absorbed lots of energy, which delayed fracture.
As the temperature decreased, the modulus of toughness
increased accordingly, and the aluminum alloys could absorb more
energy. As a result, the fracture of the material was delayed, chang-
ing, for example, the fracture surface of the specimen from a shear
fracture at the ambient condition to a cup-and-cone fracture at
cryogenic temperature (Fig. 12). In addition, the SEM analysis con-
firmed that the density and depth of the dimples increased with
decreasing temperature (Figs. 13–16). From a macroscopic, micro-
scopic, and quantitative viewpoint, it was observed that aluminum
 D.-H. Park et al. / Cryogenics 68 (2015) 44–58 57

Fig. 15. Fracture surfaces of the AA5052 specimen at different temperatures and strain-rates: (a) 293, (b) 222, (c) 153, (d) 110 K at 104 s1, (e) 293 and (f) 110 K at 102 s1.

Fig. 16. Fracture surfaces of an AA5086 specimen at different temperatures and strain-rates: (a) 293, (b) 222, (c) 153, (d) 110 K at 104 s1, (e) 293 and (f) 110 K at 102 s1.

Table 7
Modulus of toughness of aluminum alloys at 293 and 110 K.

Aluminum alloy AA5052 AA5086 AA6061 AA6082

Temp. (K) Property
293 ut ðMPa) 76.9 59.9 54.7 50.9
Fracture type Shear Cup-cone Cup-cone Shear
110 ut ðMPa) 112.3 121.5 69.6 62.5
Fracture type Cup-cone Cup-cone Cup-cone Cup-cone
Increment (increase rate (%)) 35.4 (46.03) 61.6 (102.83) 14.9 (27.23) 11.6 (22.78)

alloys were more ductile at low than at room temperature. 5. Conclusions

Through the stress–strain relationship and fractography analysis
of AA5052, however, it was predicted that this tendency did not
last continually and that brittle fracture characteristics would be
observed at certain temperatures, such as at 110 K for AA5052
(Fig. 15(d)).
In this study, the cryogenic mechanical behavior of four types of
aluminum alloys, which have been used in ships and offshore
industries, was investigated using tensile tests with different tem-
peratures and quasi-static strain rate conditions. In addition, the
58 D.-H. Park et al. / Cryogenics 68 (2015) 44–58
fracture surfaces were investigated to obtain microstructural
results. The main findings are summarized as follows:
 Aluminum alloys are temperature-dependent materials, except
for the fracture strain of 6000-series alloys. As the temperature
decreased, the yield and ultimate tensile stress and fracture
strain of aluminum alloys increased. However, the fracture
strain of 6000-series alloys did not exhibit any significant tem-
perature dependence.
 Aluminum alloys are not strain-rate-dependent materials,
except for the fracture strain of 5000-series alloys at quasi-stat-
ic strain rates. The fast strain rate leads to an increase and
decease in the fracture strain of 5000-series alloys at low and
room temperatures, respectively. In other words, the fracture
strain of 5000-series alloys showed the opposite tendency for
different temperatures.
 Aluminum alloys were more ductile and absorbed more energy
during fracture at low temperature. The fracture surface of the
specimen was confirmed to change from a shear fracture to a
cup-and-cone fracture (AA5052 and AA6082). In addition, other
results such as the increase in density, depth of dimples, and
modulus of toughness with decreasing temperature support
this finding.
 5000-series alloys with superior ductility are suitable for use in
structures that require large displacement. In addition, the high
tensile strength of 6000-series alloys could be used for struc-
tures subjected to large loads and operated under cryogenic
conditions. However, necessary precautions are required when
5000-series alloys are used at temperatures below 110 K or
when 6000-series alloys are used in structures that experience
large displacement.
This study verified that aluminum alloys had appropriate prop-
erties for use in cryogenic conditions and that they could be
applied in cryogenic conditions in various fields. In addition, quan-
titative data were provided for the use of aluminum alloys as cryo-
genic materials.
Acknowledgments
This work was supported by a National Research Foundation of
Korea (NRF) grant funded by the Korean government (MSIP)
through GCRC-SOP (No. 2011-0030013). This research was also sup-
ported by the Basic Science Research Program through the NRF,
funded by the Ministry of Education (No. 2013R1A1A2A10011206).
References
[1] Park WS, Chun MS, Han MS, Kim MH, Lee JM. Comparative study on
mechanical behavior of low temperature application materials for ships and
offshore structures: Part I-Experimental investigations. Mater Sci Eng, A
2011;528(18):5790–803.
[2] Picu RC, Vincze G, Ozturk F, Gracio JJ, Barlat F, Maniatty AM. Strain rate
sensitivity of the commercial aluminum alloy AA5182-O. Mater Sci Eng, A
2005;390(1–2):334–43.
[3] Hadianfard MJ, Smerd R, Winkler S, Worswick M. Effects of strain rate on
mechanical properties and failure mechanism of structural Al–Mg alloys.
Mater Sci Eng, A 2008;492(1–2):283–92.
[4] Fan X, Suo T, Sun Q, Wang T. Dynamic mechanical behavior of 6061 Al alloy at
elevated temperatures and different strain rates. Acta Mater
2013;26(2):111–20.
[5] Wang Y, Jiang Z. Dynamic compressive behavior of selected aluminum alloy at
low temperature. Mater Sci Eng, A 2012;553:176–80.
[6] Glazer J, Verzasconi SL, Sawtell RR, Morris JW. Mechanical behavior of
aluminum–lithium alloys at cryogenic temperatures. Metall Mater Trans A
1987;18(10):1695–701.
[7] Tarlupa TS. Effect of low-temperature deformation on the plasticity of
aluminum alloys. Mater Sci 1996;32(5):579–82.
[8] Claunsen AH, Borvik T, Hopperstad OS, Benallal A. Flow and fracture
characteristics of aluminum alloy AA5083-H116 as function of strain rate,
temperature and triaxiality. Mater Sci Eng, A 2004;364(1–2):260–72.
[9] Oosterkamp LD, Ivankovic A, Venizelos G. High strain rate properties of
selected aluminum alloys. Mater Sci Eng, A 2000;278(1–2):225–35.
[10] Chen Y, Clausen AH, Hopperstad OS, Langseth M. Stress–strain behavior of
aluminum alloys at a wide range of strain rates. Int J Solids Struct
2009;46(21):3825–35.
[11] Wowk DL. Effects of prestrain on the strain rate sensitivity of AA5754 sheet:
Department of mechanical and materials engineering [dissertation]. Kingston
Ontario: Queen’s University; 2008.
[12] Mukai T, Kawazoe M, Higashi K. Strain rate dependence of mechanical
properties in AA5056 Al–Mg alloy processed by equal-channel-angular-
extrusion. Mater Sci Eng, A 1998;247(1–2):270–4.
[13] Kim SJ, Jang SK, Han MS, Park JC, Jeong JY, Chong SO. Mechanical and
electrochemical characteristics in sea water of 5052-O aluminum alloy for
ship. Trans Nonferrous Met Soc China 2013;23(3):636–41.
[14] Mahabunphachai S, Koc M. Investigations on forming of aluminum 5052 and
6061 sheet alloys at warm temperatures. Mater Des 2010;31(5):2422–34.
[15] Aval HJ, Serajzadeh S, Kokabi AH. Evolution of microstructures and mechanical
properties in similar and dissimilar friction stir welding of AA5086 and
AA6061. Mater Sci Eng, A 2011;528(28):8071–83.
[16] Moreira PMGP, Santos T, Tavares SMO, Richter-Trummer V, Vilaca P, de Castro
PMST. Mechanical and metallurgical characterization of friction stir welding
joints of AA6061-T6 with AA6082-T6. Mater Des 2009;30(1):180–7.
[17] Abdulhamid H, Kolopp A, Bouvet C, Rivallant S. Experimental and numerical
study of AA5086-H111 aluminum plates subjected to impact. Int J Impact Eng
2013;51:1–12.
[18] Sritharan T, Chandel RS. Phenomena in interrupted tensile tests of heat treated
aluminum alloy 6061. Acta Mater 1997;45(8):3155–61.
[19] Tucker MT, Horstemeyer MF, Whittington WR, Solanki KN, Gullett PM. The
effect of varying strain rates and stress states on the plasticity, damage, and
fracture of aluminum alloys. Mech Mater 2010;42(10):895–907.
[20] Pereira AM, Ferreira FM, Antunes FV, Bartolo PJ. Study on the fatigue strength
of AA6082-T6 adhesive lap joints. Int J Adhes Adhes 2009;29:633–8.
[21] Mazzolani FM. Aluminum alloy structures. Boston: Pitman; 1985.
[22] Callister Jr WD. Materials science and engineering and introduction. 7th
ed. New York: John Wiley & Sons; 2007.
[23] Hibbeler RC. Mechanics of materials. 7th ed. Upper Saddle River & New
Jersey: Pearson & Prentice Hall; 2008.
[24] Srivatsan TS, Hoff F, Sriram S. The effect of strain rate on flow stress, strength
and ductility of an Al–Li–Mg alloy. J Mater Sci Lett 1990;9:297–300.
[25] Manes A, Peroni L, Scapin M, Giglio M. Analysis of strain rate behavior of an Al
6061 T6 alloy. Proc Eng 2011;10:3477–82.
[26] Darras BM, Abed FH, Pervaiz S, Abdu-Latif A. Analysis of damage in 5083
aluminum alloy deformed at different strain rates. Mater Sci Eng, A
2013;568:143–9.
[27] Kabir IR, Islam MA. Hardened case properties and tensile behaviours of TMT
steel bars. Am J Mech Eng 2014;2(1):8–14.
[28] Ugural AC, Fenster SK. Advanced strength and applied elasticity. 4th ed. New
Jersey: Prentice Hall; 2003.
[29] Gere JM, Goodno BJ. Mechanics of materials. 7th ed. Toronto: Cengage
Learning; 2009.