Materials perspective

Creep resistant aluminium alloys and their

applications
J. S. Robinson, R. L. Cudd and J. T. Evans
The CEC funded thematic network CREEPAL (BRRT CT98 5101) was established in 1998 to collect and
disseminate information on the long term creep and thermal mechanical cycling behaviour of aluminium alloys. This
paper presents some of the information gathered by consortium members and some additional industrial information
related mainly to forging alloys. A short history of the use of aluminium alloys at elevated temperature is presented.
The principal considerations required when evaluating aluminium alloys for use at temperatures in the 100 350C
range are detailed. The important microstructural mechanisms of creep in aluminium alloys are reviewed and the
physical metallurgy of Hiduminium RR.58 (2618A) is examined. The alloys that have found commercial
application at elevated temperature are listed. Recent developments are described and opportunities for further
development considered.
MST/5593
At the time this work was carried out Mr Robinson was in the Department of Materials Science and Technology, University
of Limerick, Ireland (jeremy.robinson@ul.ie), Dr Cudd was with HDA Forgings Ltd (now Mettis Aerospace Ltd), Redditch,
Worcs B97 6EF, UK, and Professor Evans was in the Department of Mechanical, Materials and Manufacturing Engineering,
University of Newcastle upon Tyne, Newcastle upon Tyne NE1 7RU, UK. Dr Cudd is now retired. Manuscript received 30 July
2002; accepted 21 October 2002.
# 2003 IoM Communications Ltd. Published by Maney for the Institute of Materials, Minerals and Mining.

Introduction
It is just over 115 years since the discovery by C. M. Hall of
a commercial relatively low cost method of extraction of
aluminium from its oxide.1 This extraction of the earth s
most abundant metallic element led to gradual widespread
commercial use of this metal, primarily because of its low
density and useful strength. One of its rst commercial
uses, cast cooking utensils, also recognised another of its
important characteristics: high thermal conductivity. This
may be considered the rst elevated temperature application. In the intervening years the applications of aluminium
alloys has been diverse, however the contribution the metal
has made to the aerospace industry is dif cult to overestimate. The Wright brothers recognised the value of
aluminium in reaching the goal of powered ight. They used
an aluminium alloy to achieve critical weight reduction in
the four cylinder 9 kW internal combustion petrol engine
that powered their rst ight in 1903. 2
Aluminium alloys were developed for structural and skin
materials operating at ambient temperatures and for engine
applications where high temperature performance was an
important consideration. In the UK, aeroengine manufacturers such as Rolls Royce understandably took the lead in
the development of alloys for high temperature applications
in conjunction with companies such as High Duty Alloys,
which at that time produced aluminium alloy castings,
rolled products, extrusion and forgings. These high performance alloys for use at elevated temperatures were always
based on the precipitation hardenable Al Cu system.
The design of the Concorde supersonic transport (SST)
aircraft in the 1950s pinpointed the need for a new creep
resistant alloy for the skin of the aircraft. For example, the
in ight temperature pro le of the skin and structure of
Concorde is shown in Fig. 1. The starting point was the
alloy Hiduminium RR.58, which had originally been
developed as an engine alloy for pistons (Hiduminium
RR.59) and later for impellers in gas turbines that could
operate at 175 250 C for relatively short durations.
Production of Hiduminium RR.58 in sheet form 4 met
DOI 10.1179/026708303225009373

the requirements for the relatively high skin temperatures of

127 C at the nose and 100 C in the main structure generated
at the cruising speed of Mach 2 .04. 5 ,6
It is signi cant that the chemistry of Hiduminium
RR.58, which was very close to one for which Rolls Royce
led a patent in 1928, 7 is still the alloy most commonly
speci ed for applications at elevated temperature. (The
Aluminium Association, International Aluminium Alloy
Designation System (IADS) composition closest to
Hiduminium RR.58 is 2618A in the UK/Europe and
2618 in North America). Whilst other alloys have been
developed with improved high temperature characteristics
as will be detailed later, their use for high temperature
applications has not been widespread to date.

Concorde skin and internal structure temperature:

from Ref. 3

Materials Science and Technology

February 2003 Vol. 19

143

144 Robinson et al.

Creep resistant aluminium alloys and their applications

2 Variation in proof stress Rp0 .2 for Hiduminium RR.58

tested at soaking temperature: from Ref. 9

Designs for a SST aircraft to succeed Concorde continue

to be considered by a number of aerospace companies. The
question of the alloy to be used for the skin material is one
of a number of details that arises and this was the main
reason for the initiation of the CEC funded thematic network Creepal (BRRT CT98 5101). Creepal was a 3 year
programme established in 1998. 8 The objectives were to:
(i) increase the amount of creep data in the public
domain
(ii) de ne rules for best practice testing and data
acquisition
(iii) assess the microstructural basis of the resistance to
creep and stress relaxation
(iv) quantify characterisation and microstructural modelling of creep.
This paper presents some of the information gathered by
consortium members related to skin alloys and additional
industrial information related mainly to forging alloys.

Variation in proof stress Rp0 .2 for Hiduminium RR.58

at 200 and 250C: recovery properties measured at
room temperature; elevated temperature properties
tested at soaking temperature; from Ref. 9

REGRESSION (OR REVERSION)

Regression or reversion is often observed in heat treatable
alloys when the application temperature is higher than the
aging temperature. Exposure initially results in resolution of
precipitates characteristic of the aging temperature, leading
to a signi cant decrease in both recovery and elevated temperature strength. This is followed by a second precipitation, characteristic of the exposure temperature and leading
to partial recovery in strength. Figure 4 indicates the change
in Rp 0 . 2 , for the alloys, Hiduminium 66 L73 (similar to
2014T6), Hiduminium RR.58(DTD5070) and Hiduminium
72(W) (similar to 2024T4) in sheet form subject to rapid
heating and loading.1 0 Of these three alloys the 2024 type
alone exhibits a signi cant regression response, a direct
consequence of the T4 temper.

Elevated temperature use

ELEVATED TEMPERATURE PROPERTIES
The position of Hiduminium RR.58 as a tried and tested
skin alloy in a supersonic application implies that its
elevated properties can be taken as a benchmark for any
new developments. For instance, strength at elevated temperatures is indicated by the variation in 0.2% proof stress
Rp 0 . 2 as is shown in Fig. 2 for Hiduminium RR.58 in a T61
type temper. These data were obtained by testing specimens
manufactured from 25 mm diameter forged bar, and were
conducted at the elevated temperature shown after prior
soaking at that temperature for times of up to 1000 h. The
data in Fig. 2 thus indicate the result of two in uences, i.e.
the effect of prolonged aging treatment combined with the
subsequent thermal activation of plastic deformation.

RECOVERY PROPERTIES
Recovery properties relate to the effect of exposure at any
temperature on the property characteristics, compared with
those achieved before the exposure. Testing is normally
performed at room temperature. An example is shown in
Fig. 3 where the recovery properties of Hiduminium
RR.58 are compared with the properties at 200 and 250 C.
Materials Science and Technology

February 2003 Vol. 19

Variation of proof stress Rp0 .2 under conditions of

rapid heating and loading: sheet materials 10 min
soak; tested at soaking temperature; from Ref. 10

Robinson et al.

5 Creep curves of French equivalents of 2618T6 and

2024T8 tested at 130C: from Ref. 18

EXPANSION AND VOLUME CHANGES

Aluminium alloys have a comparatively high coef cient of
linear expansion in the range 16 24 610 6 K 1 .1 1 Depending on the alloy system, further precipitation at the exposure
temperature can lead to either small increases or decreases
in volume.

CREEP
Creep is the direct manifestation of the accumulation of
many microplastic events occurring over a (nominally long)
period of time. The application of stress at the exposure
temperature leads to a permanent strain in the direction
of the applied stress. Under severe conditions, this leads
to signi cant damage accumulation and nally rupture.
Recent reviews of creep are provided in Refs. 12 and 13.
The rate of creep is a strong function of the applied stress
and temperature and the resistance to this form of deformation is obviously of great importance when materials are
used at elevated temperatures.
Although creep is a phenomenon of major importance
in aerospace applications, there have been relatively few
systematic studies to elucidate detailed mechanisms of creep
in aluminium alloys. The intrinsic thermodynamic instability of the precipitation hardened microstructures in these
alloys means that creep is more complex than it is in simple
metals.
The mechanism of creep in aluminium and its alloys
remains a controversial subject. For pure aluminium the
widely accepted view is that different creep mechanisms
can occur depending on the combination of stress and
temperature. Power law creep expressions adequately
describe the behaviour over a wide range of stress and temperature. At high temperatures and low stresses Harper
Dorn creep involving dislocation processes is reported to
occur. Diffusional ow of the Nabarro Herring type is also
possible in ne grained material at high temperatures and
low stresses, but has rarely been reported in the literature.
The exact microstructural events giving rise to these types of
creep remain obscure and controversy still surrounds the
evidence of these phenomena.1 4 1 6
For precipitation hardened alloys in the regimes of
stress and temperature of technological interest, the mechanism of creep is believed to consist of glide and diffusion
assisted climb with the associated development of a
dislocation substructure.1 7 Unlike pure aluminium where

Creep resistant aluminium alloys and their applications 145

Variation of elastic modulus with temperature of

some wrought Al alloys: bar materials; from Ref. 10

dislocation dislocation interactions and the development

of a dislocation substructure are signi cant, in precipitation
hardened alloys the dislocation precipitate interactions are
of primary importance. The creep rate will be determined by
the rate at which dislocations interact with and ultimately
overcome obstacles provided by the precipitates. As has
already been indicated, commercial creep resistant aluminium alloys are always based on the Al Cu system. While
the solid state diffusivity of copper in aluminium is similar
to the other standard alloying elements, it is the stable
nature and ne distribution of the semicoherent precipitates
that can be formed that give rise to its selection. Microstructural stability is of paramount importance for long
term application. Figure 5 displays tensile creep curves
for the French equivalents of 2618T6(A U2GN) and
2024T8(A U4G1) tested at 130 C.1 8 These data are
relatively old and the long duration of these tests contrasts
with the shorter time scales seen in recent data. 1 9 Further,
the data in Fig. 5 highlight the danger of assessing the
performance of different alloys for long term use based on
accelerated tests. In Fig. 5 the short term performance of
2024T8 is signi cantly superior to 2618T6, however the
rapid acceleration of damage accumulation in 2024 owing
to strain accelerated phase transformation renders this alloy
inferior to 2618 in practical applications. Unfortunately,
the cost associated with long term creep tests has resulted
in a dearth of data useful to designers.
The authors have recently reviewed short term
Hiduminium RR.58 creep data from the 1960s released
by HDA Forgings Ltd (now Mettis Aerospace Ltd) to the
CREEPAL thematic network,2 0 ,2 1 and a large aluminium alloy creep data set of US origin has recently been
published.2 2

FATIGUE
The number of cycles to initiate fatigue is generally considered to be inversely related to the total elastic and plastic
strain cycle. The complete strain cycle to achieve a required
stress will be greater the lower the elastic modulus and yield
stress. For even moderate increases in temperature the
elastic modulus, Fig. 6 and yield stress (Fig. 2) decrease and
the fatigue life is reduced (Fig. 7) although this is mainly a
consequence of the lowered yield stress. In addition, if creep
is a signi cant factor then the creep damage is likely to be
deleterious to the fatigue life. The presence of pre-existing
Materials Science and Technology

February 2003 Vol. 19

146 Robinson et al.

Creep resistant aluminium alloys and their applications

7 High cycle fatigue resistance of Hiduminium RR.58

forged bar: RT room temperature; from Ref. 9

aws in materials operating in fatigue and creep environments raises the question of interactions of different damage
mechanisms occurring at the crack tip. The application of
the concepts of fracture mechanics has been shown to be
useful in predicting creep crack growth.2 3 ,2 4

Sheet tear toughness of aluminium alloys at room

temperature (solid symbols) and 150C (open symbols):
time at temperature 0.5 h; from Ref. 25

effect is the precipitation of the magnesium rich b phase

(Mg 5 Al8 ) from the solid solution. (All compositions in this
paper are given in wt-% unless shown otherwise.) This can
lead to a marked reduction in ductility, corrosion resistance
and stress corrosion cracking resistance.

TOUGHNESS
Aluminium alloys normally exhibit higher toughness at
elevated temperature when compared with room temperature. There is very little elevated temperature plane strain
type data in the public domain. The results of sheet tear
resistance tests have been published recently,2 5 and the
results of these tests at room temperature and 150 C for
alloys with potential elevated temperature applications
are presented in Fig. 8. The untypical behaviour of 2024
in the naturally aged condition (T3) is a consequence of
strengthening of the matrix during exposure to the elevated
temperature during the test. While dependent upon orientation, Hiduminium RR.58 has a plane strain fracture
toughness of between 18 22 MPa m 1 /2 , which is at the
lower end of the range when compared with other high
strength aluminium alloys (20 40 MPa m 1 /2 ).5

Classi cation of wrought alloys

Aluminium alloys can be categorised into two major
groups: non-heat treatable, and heat treatable.

NON-HEAT TREATABLE ALLOYS

These range from commercially pure aluminium containing
only small amounts of other elements, notably iron and
silicon, to alloys containing signi cant amounts of alloying elements such as manganese (3xxx series alloys) and
magnesium (5xxx series alloys). In all, strengthening can
be achieved by cold working and in the manganese and
magnesium containing alloys, solid solution strengthening
can be signi cant. Such alloys may be used at elevated
temperatures where stressing of any signi cance is not a
factor. Recovery properties will generally be the same as the
original properties unless the temperature induces softening
by static recovery or recrystallisation. With the higher
magnesium containing alloys ( 43 4 wt-%Mg), another
Materials Science and Technology

February 2003 Vol. 19

HEAT TREATABLE ALLOYS

The strength of such alloys is dependent on the ability
to solutionise the main elements by a high temperature
treatment, retain the solid solution by cooling at a suitable
rate, and then producing a ne submicrometre strengthening precipitate either at room temperature or elevated
temperatures generally in the range 100 200 C. Because of
this method of strengthening, the alloys can be considered
thermodynamically metastable. In the majority of cases
the instability is not perceptible at room temperature. This
instability can be considered in three ways depending on the
exposure temperature:
(i) as the slope of the solvus for precipitation on a
temperature y/concentration x basis is positive, after
aging at an elevated temperature and cooling to
room temperature, there is still a small proportion of
the age hardening constituents in solid solution.
Depending on the kinetics of precipitation in the
particular alloy system and the subsequent exposure
temperature, noticeable further aging may occur,
characteristic of the concentration remaining in
solid solution and the exposure temperature
(ii) if the temperature of exposure is close to that of
aging, then the material continues to age. With time,
this will generally lead to a decrease in strength as
the precipitates coarsen (overaging)
(iii) as mentioned above, if the exposure temperature
exceeds the aging temperature, then reversion can
occur.

Elevated temperature instability at

relatively low temperatures
Non-heat treatable alloys containing in excess of 4% magnesium retain some magnesium in solution after processing

Robinson et al.

at elevated temperature. Subsequent slightly elevated

temperature exposure ( 466 C) can lead to preferential
precipitation of Mg5 Al8 at the grain boundaries and intergranular stress corrosion failure in service.2 6 A similar
phenomenon occurs with the medium strength weldable
Al Zn Mg alloys although the effect can be mitigated by
the addition of small amounts of copper and/or silver.2 7
The high residual saturation of solute which is the driving
force for this instability has also been reported for certain
aged lithium containing aluminium alloys exposed for long
times at temperatures of 60 135 C, the exposure resulting
in an increase in strength and reduction in ductility and
toughness.2 8 3 0

Alloys used at elevated temperatures

CASTING ALLOYS
It is intended to give only a brief summary of casting alloys
as the main concern is for the wrought form. Wrought
alloys produced from conventional semicontinuous casting
techniques are limited to those that are amenable to this
casting technique, and to those that are subsequently
deformable to produce the nished part. This places a
not inconsiderable limit on the alloy compositions that
can be produced as wrought products. Whilst castability is
an obvious concern for cast alloys, there is generally less
limitation on the composition than for wrought alloys. One
example of this compositional complexity is an early alloy
developed for use up to 350 C, which was described as an
Al Cu Ni Co Sb Zr alloy.9
A much simpler casting alloy, Al 3Mn was developed
for use as a gas burner for domestic cookers.9 This latter
use probably represents the highest application temperature
for an aluminium alloy, being of the order of 500 C. The
relatively high manganese content compared with other
casting alloys, whilst not giving a high strength capability,
does increase the stiffness at temperature. In addition, manganese has a much higher capacity for retaining hydrogen
than other alloying elements such as chromium, iron, cobalt
or nickel.3 1 This characteristic is instrumental in preventing
in service blistering owing to hydrogen absorption.
Most aluminium casting alloys are based on the Al Si
system because the addition of silicon leads to high uidity,
good feedability and high resistance to hot cracking.
Additionally, silicon is unique in signi cantly reducing the
coef cient of linear expansion so that with hypereutectic
compositions, values as low as 16 610 6 K 1 can be
achieved.1 1 With such levels of silicon, wear resistance is
very good. Alloys containing 12 23Si in which the primary
silicon can be signi cantly re ned by the addition of small
amounts of phosphorus have been used for pistons in petrol
and diesel engines.
In the early stages of the development of motor vehicles,
extensive use was made of cast aluminium alloys in internal
combustion engines, but improvements in ferrous metallurgy with reduced costs compared with aluminium led to
aluminium alloys taking a minor role for many years. With
the current emphasis on increasing fuel economy, cast
aluminium alloys are again nding growing application as
engine components and many modern vehicles have both
aluminium alloy cylinder heads and engine blocks.
Heat treatable Al Si Cu Mg hypoeutectic alloys like
C355T61 (Al 5Si 1 .3Cu 0 .5Mg) can nd application in
the cold section of automotive turbo chargers as the centrifugal compressor wheels. These components are normally
manufactured using vacuum assisted gravity casting into
gypsum based moulds. This alloy and others, for example
A357T6 (Al 7Si 0 .5Mg), also have potential structural
aerospace applications with highly complex near net shape
components being manufactured using the investment

Creep resistant aluminium alloys and their applications 147

casting process known as Sophia* developed by Thyssen

Umformtechnik + Guss Gmbh.3 2

WROUGHT ALLOYS
The chemical compositions of some of the alloys referred
to in this section are given in Table 1. Where the highest
performance is required and the con guration allows,
engineers still favour wrought product forms. The precipitation hardening alloys offer the best outright creep
performance but non-heat treatable wrought alloys do nd
certain applications, an example being the use of 99 .5%
aluminium, and alloys containing either manganese or
magnesium as the main alloying elements for pressure
vessels at temperatures up to 200 C.2 6 The use of the higher
magnesium compositions is not permitted in this application for the microstructural instability reasons described
earlier.
For most elevated temperature applications the alloys
employed are developments of the rst age hardening
system, Al Cu Mg, discovered by Alfred Wilm just before
the First World War.3 3 Before the development of
Concorde, aluminium alloys had been used successfully
in aeroengines in cast and wrought forms at temperatures
rarely below 200 C. Short design lives up to 1000 h had
allowed this exploitation even up to temperatures approaching 350 C. In the UK circa 1950 there were a number of
wrought alloys for use in the range 150 350 C.3 4 It is
likely that some had characteristics suitable for speci c
applications, but rationalisation led to two alloys being the
most widely employed. These were initially known as
Hiduminium RR.58, originally referred to as Hiduminium
59 low silicon, and Hiduminium RR.57. The initial
development of Hiduminium RR.58 arose from the
requirement in the late 1930s and early 1940s for an
aluminium base alloy capable of operating in the 200
250 C range. The target application at this time was cold
section components of the Whittle gas turbine and the rst
forged impellers made in the alloy were tted to the Gloster
Meteor Mk1, which rst ew in 1943. The actual
composition of Hiduminium RR.58 was developed as a
low silicon variant of Hiduminium RR.59, an aeroengine
piston alloy used extensively during the Second World War.
This low silicon alloy was dif cult to cast consistently using
traditional chill casting techniques, however the introduction of semicontinuous casting processes enabled large
diameter ingots to be made. In the USA similar alloys
with slightly different compositional ranges were adopted
and designated 2618 and 2219. Notably 2219 intentionally
contains small additions of zirconium and vanadium.
Mention should also be made of the 2014 alloy that in the
UK was employed as a structural alloy for aerospace, whilst
in the USA, 7075 (Al Zn Mg Cu) was preferred. 2014
also found favour for some elevated temperature uses
such as aircraft wheels where high strength, good fatigue
and resistance to the heating generated by braking were
important. In this latter application when heating is only
for short periods, it is the recovery properties that are
important rather than those at elevated temperature and
long term creep is unlikely to be a consideration. 2014 has
also been used for impellers in diesel turbo chargers. Alloys
of the 4032 type (Al 12Si 1Cu 1Mg 1Ni) are used for
forged pistons where engineers prefer the integrity of the
wrought form to the equivalent cast alloys.
In the late 1950s Alcoa introduced an Al Li alloy,
2020, containing copper, magnesium and manganese having
good elevated temperature properties allied to low density
imparted by the lithium addition. Concerns for the fracture
behaviour of this alloy resulted in limited use. Although
considered for the supersonic Concorde and British Aircraft
*Sophia is a registered trademark.

Materials Science and Technology

February 2003 Vol. 19

148 Robinson et al.

Creep resistant aluminium alloys and their applications

Corp. TSR.2, it was only used for skinning the wings on one
large programme, the US Navy s North American Rockwell A3J/RA 5C Vigilante.3 5 Production of the alloy
ceased in 1969 although the aircraft were not retired from
service until 1979. 3 6
Of the 2618 and 2219 type alloys, the former has found by
far the most widespread use at elevated temperatures mainly
in aeroengines and turbo chargers. The alloy 2618 was
extensively used for compressor blades in earlier generation
aeroengines, a use that has declined but not completely
disappeared with the increased compressor temperatures
now employed. It nds extensive current use for structural
engine parts (rings), bleed valves, guide vanes, engine
casings, annulus llers and impellers. The alloy 2219 is not
similarly used, however its weldability and good toughness
at cryogenic temperatures are the major reasons for its
selection in many of its applications.
As mentioned earlier, the alloy used for skinning the
Concorde was developed from that of a Rolls Royce patent
of 1928. 7 This development was carried out at High Duty
Alloys on the basic 2618 alloy to optimise its performance
in sheet form at elevated temperatures.4 In the early 1960s
most of this work was complete, and investigations were

being carried out on an alloy showing improved elevated

temperature potential This alloy was based on the UK
version of 2219 but with a small addition of magnesium and
silicon.
Both the UK and USA versions of 2219 did not exhibit
particularly useful strength unless subjected to cold work
of 1 .5 3% between solution treatment and aging. The
addition of 0 .15 0. 40%Mg and 0.10 0 .20%Si gave similar
strength values to the cold worked alloys without the application of any cold work.3 7 Whilst the cold work required
by the original variants can be applied quite easily with
the required precision to sheet and plate by stretching, and
to rectilinear forgings by cold compression or in some
circumstances stretching, it is not so easily applied to die
forgings. The application of plastic deformation does have
the added bene t of reducing residual stresses induced by
the quenching stage of the heat treatment.
Two versions were initially developed from the UK
alloy, one with a nominal magnesium content of 0 .35% for
sheet and rivet stock, the other with a nominal magnesium
content of 0 .20% being preferred for other wrought forms.
It was postulated that the small additions of magnesium led
to a substantial increase in solute/vacancy binding energy

Table 1 Chemical composition of wrought elevated temperature aluminium alloys

Period

Alloy

Value

Si

Fe

Cu

Mn

Mg

Cr

Ni

Zn

Ti

Ag

Zr

Li

Cd

1930 1950

2031

Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum
Minimum
Nominal
Maximum

0.50
0.9
1.3

0.6
0.9
1.2

0.6
0.9
1.2

0.6
1.0
1.4

0.20

0.20

0.30

0.02

0.10

0.20
0.10
0.15
0.20

0.35
0.9
1.1
1.3
0.9
1.2
1.4

1.8
2.3
2.8
5.8
6.3
6.8
5.7
6.1
6.5
1.9
2.3
2.7
1.8
2.2
2.7

0.25

0.15
1.3
1.5
1.8
1.2
1 .5
1 .8

4.5

0.5

2219
2219
(UK variant)
2618 (USA)
2618A (UK)
2020
1950 1970,
High Duty
Alloys variants

HDA54
(forgings)
HDA543
(sheet)
HDA254
HDA543Ag

1970 1990

2519
Polmear
(?~small
amounts)
2650
Polmear
(dev)

1990 2000
McCook
(Reynolds)
Al Li alloys

2094
2095
2195
X2096

Materials Science and Technology

0.25
0.10
0.18
0.25
0.15
0.20
0.25

0.5
0.20
0.3
0.40
0.20
0.28
0.35

0.10
0.15
0.20

0.30

5.8
6 .0
6 .25

6 .0

0.15

6 .0

0.25

0 .22

0.15

0.20
0.25
0.30

0.25

0.10
0.9
1.0
1.2
0.8
1.1
1.4

0.15

0.10

0.10

0.12

0.30

5 .3
5 .9
6 .4

6 .3

0.25

2 .5

0.20

0.15

0.1

0.1

0.40

1 .6

0.12

0.15

0.12

0.15

0.12

0.15

0.12

0.15

0.3

February 2003 Vol. 19

0.25

0.25

0.25

0.25

0.10

0.11

0.15

0 .45

0.25

0.10

0.25

0.1

0.25

0.1

0.25

0.1

0.2

0. 40
0. 45
0. 50

0.1
0.12
0.15

0.10
0.18
0.25

0.18

0.05
0.10
0.15

0.1

0. 45

1. 1

0 .25
0 .5
0. 8
0. 25
0. 5
0. 8
0. 25
0. 5
0. 8
0. 25
0. 55
0. 89

3 .5 6.5

0.02
0.06
0.10

0.12

4 .4
4 .8
5 .2
3 .9
4 .3
4 .6
3. 7
4. 0
4. 3
2. 3
2. 7
3. 0

0.30

0 .05
0 .22
0 .40

0.10
0.3
0.50

0 .35

Ti + Zr
0.20

0.12

0.25

6 .0

0.10
0.12
0.15

0.05
0.1
0.15

0.20

0.15

0.10
0.18
0.25

0.10

0.20
0.04
0.07
0.10

0.10

0 .15
0 .22
0 .28
0 .30
0 .35
0 .40

0. 25
0. 4
0. 6
0. 25
0. 4
0. 6
0. 25
0. 4
0. 6
0. 25
0. 4
0. 6

0.16

0.04
0.11
0.18
0.04
0.11
0.18
0.08
0.12
0.16
0.04
0.11
0.18

0. 7
1. 0
1. 4
0. 7
1. 1
1. 5
0. 8
1. 0
1. 2
1. 3
1. 6
1. 9

Robinson et al.

Creep resistant aluminium alloys and their applications 149

and retarded diffusion rates. The density of precipitate

nucleation was thus increased and precipitate growth
limited.3 7
Two aging treatments were developed, the rst higher
strength version being 12 h at 190 C, the other 16 h at
215 C. The latter was based on a requirement for no further
non-reversible volume changes to take place during use at
temperature. Besides there being no requirement to cold
work the alloy to achieve useful properties, other advantages were that the alloy was not particularly cooling rate
(quench) sensitive and could be boiling water quenched, and
in thin sections salt quenched (200 C) to reduce residual
stresses and give stability during machining.
Further developments on this alloy system involved the
addition of up to 0 .5%Ag (Ref. 38) and separately the addition of 0 .12%Zr to improve weldability. The latter increased
the sensitivity to cooling rate when compared with the basic
alloy and led to an increase in quench induced stresses
in product forms that were not easily mechanically stress
relieved e.g. forgings. A further development by the Fulmer
Research Institute in the UK was the addition of small
amounts of germanium, which replaced the silicon in the
magnesium silicide compound.3 9 This led to increased
response to aging as evidenced by higher room temperature
strength but it was less ef cacious in retarding the coarsening and transformation of the h precipitate, and was
comparatively deleterious to the elevated temperature
properties.
Despite some of these developments showing signi cant
advantages for use at elevated temperature and numerous
proprietary speci cations being raised for all but the silver

bearing alloy, they found few applications and were never

submitted for either national or international standardisation. It is interesting that although this type of alloy has not
found widespread use to date, a similar alloy 2519 was
registered by the USA in 1985, and that alloys containing
silver have been proposed for extending the elevated
temperature range of aluminium alloys.4 0 4 3 A summary
of the development of alloys for use at elevated temperature
is given in Table 2. A list of some of the applications of
these alloys is presented in Table 3.

Physical metallurgy of Hiduminium RR.58

(2618A)
During the 1960s after selection of Hiduminium RR.58
for Concorde, the physical metallurgy of this alloy was
examined extensively and the reasons determined for its
creep resistance. When aged at temperatures below 200 C,
strengthening in an Al 2 .5Cu 1.2Mg alloy occurs from
formation of transition precipitates in the pseudo-binary
Al Cu Mg system.4 4 ,4 5
a

Table 2 Summary of development of wrought aluminium

alloys for use at elevated temperature
Period

Progress

1910 1920

Duralumin (Al Cu) age hardening alloys

discovered and development accelerated
by WWI
Development of Al Cu system continues with
numerous alloying additions notably Mg, Si,
Ni, and Fe for elevated temperature use
Widespread use of wrought elevated
temperature alloys particularly in
aeroengines during WWII
Majority of applications turn to use
of 2618/2618A/2219.
Development of clad 2618A for Concorde.
Development of Al 6 wt-%Cu base
to meet possible requirement for second
generation supersonic transport aircraft
by additions of Mg, Si, Ge, and Ag.
First commercial Al Li alloy 2020, having
high temperature capability, but does
not nd widespread use due to notch
sensitivity and falls into disuse.
Experimental alloys based on
Al 6 wt-%Cu with Ag proposed by
Polmear for turbo charger impellers.
Further development of Al Li alloys.
Rapid solidi ed processing offers potential
but no alloy makes full commercialisation.
Renewed interest in 2nd generation
supersonic transport aircraft. Europe
proposes cruising speed in line with use
of aluminium. USA propose higher speed
requiring use of titanium.
NEWAL150 European collaborative
development programme improves on
high temperature properties and damage
tolerance of 2618A.
Australian/French programme results in
Al Cu Mg Ag becoming candidate for
supersonic transport aircraft.
Al Li based alloys registered by USA with
various Cu/Mg ratios plus Ag which do not
require cold work to attain high strength.
Al Li high Cu ,high Cu/Mg, + Ag alloy
designed for elevated temperature use?

1920 1930
1930 1950
1950 1970

1970 1990

1990 2000

2000 +

-Al GPB zones S S (Al2 CuMg)

The GPB zones nucleate homogeneously and grow cylindrically, parallel to 100 A l matrix directions while S
particles are lath like platelets with a long axis also parallel
to the 100 A l matrix. S precipitates are semicoherent
and are structurally very similar to the equilibrium phase
Al2 CuMg which is orthorhombic and retains coherency in
100 A l directions over a large range of particle sizes. In the
as quenched condition dislocation loops and helices are
observed in the matrix. During arti cial aging these helical
dislocations grow by a climb process but also act as sites
for nucleation of S on {210} planes. Elevated temperature aging treatments will produce peak hardness when
~ 25% of the S phase has formed and a large proportion of GPB zones remain.
The controlled addition of 0 .18 0.25%Si in Hiduminium RR.58 retards the onset of hardening during natural
aging but increases the peak hardness and reduces the
time to peak properties during arti cial aging. 4 6 Silicon also
promotes a higher density of smaller S precipitates and
makes the alloy more resistant to overaging. Silicon additions reduce the size and number of dislocation loops and
helices, a result of reduced vacancy mobility. During arti cial aging the length of the S phase laths is noted to increase
more slowly compared with Si free alloys. In Si bearing
Table 3 Applications of elevated temperature alloys
Wrought

Cast

Cooking utensils
High duty pistons

Cooking utensils
Pistons
Brake drums
Wheels

Aircraft wheels and

torque plates

Cylinder liners
Impellers
Cylinder heads
Cylinder blocks
Bearings
Domestic gas burners
Pressure vessels

High duty impellers

Connecting rods
Pressure vessels
Compressor blades
Compressor discs
Guide vanes
Annulus llers
Bleed valves
Annular compressor
components(rings)
Aeroengine casings
Vacuum turbo devices
Supersonic transport
aircraft skin and structure
Supersonic missile skins

Aeroengine casings

Materials Science and Technology

February 2003 Vol. 19

150 Robinson et al.

Creep resistant aluminium alloys and their applications

10

a forged Hiduminium RR.58; b rolled sheet 2650T8

9 Typical recrystallised microstructure of forged Hiduminium RR.58 and rolled sheet 2650T8

alloys the distribution of S is more uniform owing to the

lower dislocation density. There is some evidence that
0 .25%Si is the optimum addition level for this alloy.
Hiduminium RR.58 also has 1%Ni, and 1%Fe controlled additions. The in uence of Fe and Ni on the structure and properties of the basic Al 2. 5Cu 1.2Mg alloy
has been investigated and the in uence on the kinetics of
precipitation hardening was found to be small. However,
when Fe and Ni are added separately the aging response
is diminished owing to the formation of insoluble intermetallics AlCuNi or (CuFe)Al 3 . Combined Fe and Ni form
FeNiAl9 and the age hardening response reverts to the basic
alloy.4 7 The FeNiAl9 intermetallic is usually dispersed into
clusters of particles associated with prior interdendritic
areas in the cast material. Individual intermetallic particles are ~ 1 3 mm in diameter. Areas containing clusters
are signi cantly harder than the surrounding matrix.
The other phase of note that can form in Hiduminium
RR.58 is Mg 2 Si and this phase is usually observed as
rounded isolated particles with size similar to the FeNiAl9
particles. Figure 9a is a typical microstructure from forged
Hiduminium RR.58. This example demonstrates a fully
recrystallised equiaxed microstructure with obvious clustering of the intermetallic particles.
In Fig. 10 a partially recrystallised structure is illustrated
with corresponding coarse recrystallised grains and areas
that are unrecrystallised containing a well de ned substructure. This specimen was prepared from a stress rupture
test conducted at 120 C and 321 MPa with rupture
occurring in 1200 h after a plastic strain of 7 .7%. Secondary
Materials Science and Technology

February 2003 Vol. 19

Hiduminium RR.58 microstructure from stress rupture specimen tested at 321 MPa and 120C

cracking along the gauge length was noted and some

intergranular wedge type cracks can be observed. Wedge
cracks are characteristic of grain boundary sliding at triple
points under the in uence of high shear stresses in contrast
to cavitation which is more typical of lower stresses. These
specimens were prepared using a concentrated nitric acid
(20 ml), hydro uoric acid (20 ml) and glycerol (60 ml)
etchant.
Surprisingly, the exact role of the FeNiAl9 intermetallic
particles is still unclear but there is some evidence that the
particles nucleate screw dislocations during the quenching
stage owing to differing thermal contraction with the
matrix. These dislocations interact with vacancies forming
sessile helical dislocations, which in turn act as potent
nucleation sites for the S phase during age hardening. The
particles also modify the recrystallisation and grain growth
behaviour and act to re ne the grain size and prevent grains
becoming very coarse during heat treatment. The relatively
high volume fraction of particles does result in low fracture
toughness compared with other Al Cu Mg alloys.

Creep performance of Hiduminium RR.58

Tensile creep curves generated by testing specimens
machined from Hiduminium RR.58 forged bar at 150 C
are shown in Fig. 11. 9 The alloy displays behaviour similar
to most engineering materials in that the time to failure and
creep rates are a function of stress (and temperature). For
exposure at a particular temperature, creep ductility tends
to decrease as the stress is decreased.
Hiduminium RR.58 was originally designed to operate
at temperatures up to 250 C. This temperature was above
the normal aging temperature of the alloy and component
lives were correspondingly short. At lower temperatures
the performance of the alloy is such that structural components can have lifetimes measured in tens of thousands
of hours. (The ight time of Concorde at the time of writing
is ~ 12 000 h and while this aircraft has had in service
problems, issues with the structural integrity of the
Hiduminium RR.58 airframe have not been reported).
Hiduminium RR.58 forgings and forged bar typically
have mean grain diameters of the order of 200 mm. Plate,
extrusions and sheet are normally much ner grained than
this and at higher temperatures than 200 C these product
forms can display lower creep resistance. This effect is of
less signi cance at lower temperatures (100 150 C).
The in uence of plastic deformation on the creep performance of Hiduminium RR.58 was investigated thoroughly
during its development for Concorde. These observations
and the application of thermomechanical type treatments

Robinson et al.

Creep resistant aluminium alloys and their applications 151

12
11

Tensile creep curves from Hiduminium RR.58 forged

bar tested at 150C: from Ref. 9

are reviewed in Ref. 48. Any treatment that increases the

dislocation density before creep loading appears to be
detrimental to the creep performance in terms of the
minimum creep rates observed. The origin of these dislocations could be statically recovered regions resulting
from forming at high temperatures or cold work introduced
after solution heat treatment but before aging (which is a
common procedure for the reduction of residual stress).
While the minimum creep rate is adversely affected by cold
work, there is evidence that the microstructural modi cation to the precipitate distribution is more stable and this
can enhance the long term performance of the alloy.4 8 ,4 9

Further developments
INGOT METALLURGY
Since the development of Hiduminium RR.58 for the
Concorde and the environmental, political and commercial
problems accompanying the introduction of this SST aircraft, the interest in a second generation aircraft has been
intermittent on both sides of the Atlantic. Many paper
projects have been proposed,5 0 with speeds ranging from
Mach 1 .6 up to Mach 10 .0, with Mach 2 .4 appearing to be
favoured by the recent High Speed Civil Transport
programme funded by NASA. 5 1 This speed is above that
at which conventional ingot metallurgy aluminium alloys
can be used. The speed requirement of the European Concorde successor that emerged in the early 1990s was de ned
as being Mach 2 .0 2. 2 and this decision was taken to
ensure that aluminium alloys could be candidate materials.
The proposed aircraft would however have a ying lifetime
of four times that of Concorde. From a creep perspective
the requirements for the fuselage material of the future
European supersonic aircraft were:
(i) stress 130 to 170 MPa
(ii) in service time 80 000 ight hours (20 000 cycles)
(iii) temperature 90 to 100 C with a maximum service
temperature of 130 C
(iv) deformation 50 .1% total plastic strain within
60 000 h at 130 C and 150 MPa.
In addition, improvements in damage tolerance and fatigue
with static properties 20% greater than subsonic aircraft
were expected.5 2 For this aircraft new materials were

Minimum creep rates at 150C for some creep resistant aluminium alloys: from Refs. 9, 41, 59, and 60

considered essential for the fuselage. Advanced aluminium

alloys and/or organic composites were both candidates.5 3 ,5 4
The CEC funded BRITE EURAM `NEWAL150 programme, which terminated in 1996, sought to develop
improved aluminium base alloys for this purpose with a
temperature capability of 150 C. The programme aimed
to develop alloys that could demonstrate superior creep
and damage tolerance performance compared with 2618A.
While the bulk of the data from NEWAL150 has yet to
emerge into the public domain, it has been reported that
a modi ed 2650 alloy could exhibit similar performance
in creep in conjunction with greatly superior damage
tolerance.5 5 ,5 6 The alloy 2650 is itself a minor modi cation
of 2618A with lower iron and nickel but increased manganese. The microstructure of 2650T8 sheet is shown in
Fig. 9 b. Recent investigations into the mechanism of creep
of 2650T8 at 150 C and high stresses have con rmed that
strain enhanced particle instability and coarsening of the S
phase is a major factor in accounting for the acceleration
of creep, with the transition precipitates growing at the
expense of ne GPB zone type particles.5 7 5 9 The high
manganese content of this alloy will render it quench
sensitive and while this will not be a concern for sheet and
plate, the alloy will be problematical for die forgings that
cannot be cold compressed.
At the same time as NEWAL150, a joint development
programme was being carried out by Australia and France,
evaluating a series of Al Cu 0 .45Mg 0 .45Ag 0. 30Mn
0 .15Zr alloys with Cu in the range 3 .6 6 .5%. The high
copper variant showed a marked improvement in creep
performance at 150 C (Fig. 12). It should be stated that
these tests were of relatively short duration. The advantage
of silver bearing alloys is the relative stability and coarsening resistance of plates of the V strengthening phase.6 1

OTHER MANUFACTURING ROUTES

The ingot metallurgy route for wrought products restricts
compositions to those that can be semicontinuously cast
and thereafter plastically deformed. Cast products can
suffer owing to comparatively gross second phase particles
whilst with wrought products these can be advantageously
fragmented. Even so, these second phase particles in
wrought alloys are coarse (50 100 mm) compared with the
aging precipitate. Other routes such as spray deposition
Osprey6 2 ,6 3 and rapidly solidi ed powder (RSP) metallurgy
Materials Science and Technology

February 2003 Vol. 19

152 Robinson et al.

Creep resistant aluminium alloys and their applications

(cooling rates up to 10 6 K s 1 ) give opportunities to

produce alloys that cannot be produced by conventional
ingot metallurgy, and control the size and distribution of
secondary phase particles. Large volume fractions of low
diffusivity transition metal based intermetallic dispersoids,
which resist coarsening at all but the highest elevated
temperatures, can be formed in the matrix during heat
treatment. Many alloys were investigated to try to raise the
service temperature above 300 C and most of the `better
alloys contained iron.6 4 These alloys displayed good short
term elevated temperature strength but there is less data
available detailing their long term creep resistance. Possibly
the best materials of this type were the planar ow cast
Al Fe V Si series of alloys.6 5 Unfortunately, these alloys
did not nd commercial application, and production ceased
in the early 1990s.
A range of silver bearing spray deposited alloys was
developed by Alusuisse Lonza in the late 1980s and early
1990s. 6 6 ,6 7 Initial development centred on an Al Cu Mg
Ag base with signi cant additions of transition metal
elements intended to form small dispersoids. The rates of
cooling during spray deposition were insuf cient to prevent
the formation of coarse constituent type particles and further alloys were developed with much lower concentrations
of dispersoid forming elements. The elevated temperature
properties of these alloys were better than 2618 and 2219
but the alloys did not nd major application.
These processes also facilitate the production of metal
matrix composites (MMCs), an early commercially exploited
example of which was SAP (sintered aluminium powder).
The very small oxide particles (10 6 10 8 m) evenly distributed and epitaxially bonded to the aluminium matrix
produced signi cant strengthening particularly at high
temperature.6 8
The Osprey process has much to commend it in that if a
market for large volumes of material could be established,
the price of stock may possibly become competitive with
ingot metallurgy routes. The current major use of Osprey
material is for cylinder liners produced in an Al Si alloy
for Mercedes Benz vee con guration internal combustion
engines.6 9 The cooling rate of this process is not high
enough to give the advantages offered with powder production by atomisation and other higher cooling rate
processes as described above.
Powder alloys and associated MMCs go through an
energy and labour intensive process to produce stock for
further conventional processing and in small quantities the
cost can be as much as 50 times that of ingot metallurgy
stock and is unlikely to fall below 25 times for the foreseeable future even for large quantities. Even at this price
the product can be cost effective. A current example is the
use of a 2124 + 25%SiC MMC for potential replacement
of titanium guide vanes in aeroengines.7 0 A further example
is the use of the same composition and an alternative
containing boron carbide for pistons in Formula One racing
engines, replacing the ubiquitous 2618. The SiC increases
the density very slightly whereas the boron carbide leads to a
reduction, whilst both give a signi cant increase in stiffness,
fatigue strength, and resistance to the elevated temperatures
experienced. These are specialised applications and one
could not anticipate the use of such pistons in other than
very high cost vehicles.

Discussion
The most widely used alloy for elevated temperature use is
2618/2618A and even though in the past 30 years alloys
have been developed with improved high temperature properties, its position is still unchallenged.
The reasons for this situation are numerous and include
the following:
Materials Science and Technology

February 2003 Vol. 19

(i) conservatism; its extensive use over the past 50 years

has resulted in a sizeable data base on which engineers can call which is unmatched by competitive
alloys
(ii) availability; because of its extensive use the alloy is
commercially viable and hence readily available at
a competitive price. Small call off for other alloys
leads to a disinclination for producers of the cast
stock to offer these in preference to 2618, and where
they do to increase price and lead times. This factor
becomes even more signi cant as rationalisation
proceeds in the industry and the small exible
casting facility is replaced by facilities having
minimum casts of tens of tonnes
(iii) exibility; the selection of one alloy to cover all parts
rather than a number of alloys, which may be the
optimum for different parts, allows the purchaser of
parts a greater exibility for order changes
(iv) low residual stress; the UK version of 2618, 2618A
allows up to 0. 25%Mn compared with 0 .05%Mn in
2618. Whilst this is not recorded as an intentional
alloying addition, it is believed that a level of
0 .15%Mn is bene cial to the properties of plate.
However, such `high manganese levels make the
alloy sensitive to the cooling rate subsequent to
solution treatment. Whilst there is not a problem
with plate that can be stress relieved by stretching, it
is not generally suitable for die forgings. Restricting
the manganese level to 50 .05%, makes the alloy
much less cooling rate sensitive and allows forgings to be boiling water quenched. This in turn
signi cantly reduces the residual stresses and subsequently the movement during machining. Thick
sections up to at least 200 mm can be boiling water
quenched without a signi cant reduction in properties when compared with the properties of 50 mm
section. The absence of the need to stretch or cold
compress rolled rings to reduce residual stress and
improve properties as would be required for the
cooling rate sensitive 2219, is a signi cant cost
reduction factor.
In the previous section mention has been made of recent
developments, although these do not appear to be much
advance on those of 30 years ago that were referred to
earlier. The major stimulus both then and now was a new
generation SST aircraft but this is not likely to proceed
either in the USA or Europe in the foreseeable future.7 1
In Europe, the attention of the aviation industry is
now focused on the very large subsonic aircraft (Airbus
A380), which has new aluminium alloy requirements
but not at elevated temperature. Having surrendered the
market for very large aircraft to Airbus, Boeing is taking
a more adventurous view of future aviation needs with
the announcement in March 2001 of the Sonic Cruiser.7 2
In the short term this 200 250 seat aircraft will y at
Mach 0 .95 0 .98, but it may herald the start of a family
that could include a supersonic commercial transport.7 3
This aircraft was due to y in 2006 but is now delayed to
at least 2008 because of the general downturn in civil
aviation.7 4 Conventional aluminium alloys are obvious
candidate materials for the fuselage and at Mach 1 .0 the
maximum skin temperature will only be of the order of
13 C.
The extended delay of a new SST aircraft raises the
question: is there another signi cant use to which creep
resistant alloys can be put, that can drive the development
or exploitation of new materials on a commercial scale large
enough to make economic sense?
In aeroengines, it appears unlikely that the moderately
higher temperature capability of these ingot metallurgy
alloys will be such to reverse the replacement of aluminium
base with titanium base alloys, although aluminium base

Robinson et al.

MMC and RSP may give an alternative in some selected

applications.
Impellers for power generation turbo chargers are an area
where substantial amounts of forgings in 2618 are used. The
higher performance versions use the titanium alloy Ti
6Al 4V because of the higher temperatures. There is a
possibility at the lower end of this range, where peripheral
temperatures reach up to 200 C, that an improved ingot
metallurgy aluminium alloy could replace titanium. ABB
in Switzerland have investigated a medium copper alloy
akin to those of Polmear, mentioned earlier.6 6 ,6 7 ,7 5 This was
produced by the spray deposition Osprey process and
forged. This Al Cu Mg Ag base alloy was strengthened
through the coprecipitation of h and V phase. This alloy
was reported to have superior creep resistance to 2618A up
to 180 C but at 200 C the improvement was minor leading
to its rejection as a candidate titanium replacement. With
rotating parts such as impellers where the stressing of the
part is related to the rotational speed, one avenue might be
to reduce the mass by using an alloy of lower density and
this might be brought about by the use of lithium as an
alloying element.
In general the developments in the past 40 years have
shown that high temperature performance is improved in
going from the S type precipitates in alloys such as 2618
(low Cu/Mg ratio) to h type precipitate in high Cu/Mg
alloys (2219 type) to V type precipitates in high Cu/Mg
alloys containing silver. This increased performance results
from the morphological differences in the aging precipitates
and the increasing resistance to coarsening. Other structural
features are also of importance such as degree of recrystallisation and grain size. With the latter, obtaining a
consistent grain structure can be dif cult if not impossible
owing to the wide variation in deformation that may be
experienced in different areas of a die forging. Other factors
such as temperature gradients owing to relatively long
contact times with the die and die temperature consistency
do not help in this respect. The relatively slow process of
cast stock breakdown exacerbates the situation when
extruded or rolled stock is not used. In addition the grain
ow can be less well de ned in die forgings than the other
wrought forms.
The alloy 2618 is particularly prone to variations in
thermal history and variable deformation during forging.
Within a forging temperature range of 330 460 C and with
the variable deformation characteristic of many con gurations, the structure of a part may vary from unrecrystallised,
coarse recrystallised, to ne recrystallised grains. Each of
these structures has different characteristics.
(i) the unrecrystallised structure is sensitive to the
cooling rate during heat treatment even when the
manganese level is below 0 .05%. These areas have a
signi cantly lower strength than the recrystallised
areas if the part is boiling water quenched to keep
the residual stresses low
(ii) the coarse recrystallised structure has little effect on
the room temperature properties, although it may be
detrimental to the fatigue life. For creep at elevated
temperature, the performance judged by the minimum strain rate is the best of the three structures but
the overall ductility is poor and the material can be
notch sensitive in creep rupture tests
(iii) the ne recrystallised structure generally has a higher
minimum creep rate than the coarser structure but
the overall creep ductility is signi cantly superior
and notch sensitivity in creep rupture tests is not
apparent
(iv) forging the 6%Cu alloy developed by High Duty
Alloys in the 1960s in the same temperature range
has not led to unrecrystallised structure. There is
however a signi cant difference in grain size between
material processed at 460 C (coarse) and that

Creep resistant aluminium alloys and their applications 153

processed at 300 C ( ne). The structural effect of

the addition of zirconium was to give a grain size
when forged at 460 C comparable to that of the
zirconium free alloy forged at 300 C. The zirconium
containing alloy exhibited improved tensile ductility
at room temperature and vastly improved tensile
ductility at elevated temperatures.

Recommendations for future alloy

development
Apart from the transitory period in the 1960s and 70s when
a high temperature creep resistant alloy was required for the
Concorde, and was used for the skin and structural elements
that were also subject to similar temperatures by conduction
(Fig. 1) the main market for such an alloy has been the
aeroengine and turbo charger. With the second generation
SST aircraft currently on hold, it seems probable that the
market for alloys with elevated temperature potential will
remain the same as it has for the past 50 years or so. The
proposals for a new SST aircraft in the 1980s and 90s did
generate a renewed interest in development of alloys for
elevated temperature use although they do not in general
appear to be signi cantly different to those that were being
investigated in the 1960s. Where commercially available,
the latter have not found widespread use at elevated
temperature.
It is considered that for any new alloy to be accepted for
widespread use, it would need to be seen as a replacement
for 2618 in areas where this alloy is currently acceptable. To
do this it must have some notable advantage other than its
improved elevated temperature properties, which in many
cases would not be of signi cance. A prime consideration
in all aircraft manufacture is weight, so that a reduced
density would be a signi cant characteristic. Another would
be good weldability, as this would allow rolled or extruded
pro les to be bent and joined to make rings. It was
suggested earlier that a density reduction for impeller
material would lead to the material being less highly stressed
in operation. This should mean that impeller temperatures
could be increased to at least the extent that the minimum
strain rate was not increased and creep rupture life
enhanced in service even if the intrinsic creep performance
was not improved.
From developments to date, it appears likely that the
optimum elevated temperature performance is to be found
in the Al Cu Mg Ag system particularly with a high
Cu/Mg ratio. Silver at an addition level of 0 .4 0.5% is
expensive and can increase the cast stock price by up to
50%. With regard to reducing the density, additions of
lithium appear to be the most promising option. Here the
cast stock price may be as much as 7 to 8 times that of a
current alloy such as 2618. In terms of semi nished and
nished components, the increased cost in percentage terms
will of course normally be signi cantly less than the gures
quoted above. The reduced thermal conductivity of lithium
containing alloys would be a negative characteristic with
regard to elevated temperature use.
Many of the aluminium lithium alloys currently available are of the Al Li Cu Mg Ag family, notably those
originating from the Weldalite* family of alloys originally
manufactured by Reynolds (formerly McCook Metals LLC,
now Pechiney). Other than the formation of lithium containing precipitates, d (Al3 Li), d (AlLi) and T 1 (Al 2 CuLi), the
precipitation systems are similar to those in the lithium
free aluminium base alloys, i.e., h (Al 2 Cu) and V (related to
Al2 Cu). This family of alloys is of particular interest for
forgings as the magnesium works in a similar manner to
*Weldalite is a registered trademark.

Materials Science and Technology

February 2003 Vol. 19

154 Robinson et al.

Creep resistant aluminium alloys and their applications

Table 4 Elevated temperature aluminium alloy objectives

Aim

Reason

At least equivalent properties to 2618

at room and elevated temperature
Density between 5 and 10% lower than 2618

Current parts in 2618 can be replaced without major redesign

Properties independent of cold working

between solution treatment and aging
Minimum cooling rate sensitivity for
quench after solution treatment
Higher elastic modulus than 2618
Improved damage tolerance compared with 2618
Good weldability
All of above

To encourage use in place of existing 2618 by weight

reduction and to reduce stress in rotating parts
To alleviate the requirement to cold work die forgings
in order to achieve high strength
To allow cold working of sheet, plate and hand forgings for straightness
and/or residual stress minimisation without signi cant effect on properties
To allow use of water quench temperatures above 80C to reduce residual
stress in die forgings without recourse to cold compression
To allow boiling water quenching for thick sections
To improve initiation fatigue resistance properties
For aerostructural advantage
To give alternative method of manufacture. In particular annular
components which would reduce input weight and help offset
probable increased basic cost of the alloy
To give one alloy suitable for widespread applications including
replacement of 2618, so that volume production and optimum use
of scrap would aid in cost reduction

when it is added to Al 6Cu alloys, so that cold working

is not a prerequisite to obtain high strength as with some
Al Li Cu alloys.7 6 The chemical compositions of this
alloy range are given in Table 1. The general requirements
of an alloy that it is considered would nd widespread
applications are presented in Table 4.

Acknowledgements
The authors wish to acknowledge the contribution of the
Creepal partners and the European Commission for the
support of this work through the thematic network
(BRRT CT98 5101), and HDA Forgings Ltd of Mettis
Aerospace Group, UK for supplying the mechanical
property data for Hiduminium RR.58.

Conclusion s
Heat treatable and non-heat treatable aluminium alloys
nd application at elevated temperatures up to a maximum
temperature of around 500 C. This temperature is representative of an extreme application and a more realistic
upper limit would be of the order of 350 C. Even this
temperature is a severe requirement if signi cant component loading or lifetime is required. If a component is to
be structurally ef cient and have a lifetime measured in
thousands of hours, 100 175 C is a more realistic operating
regime.
This relatively low temperature capability has resulted
in limited successful development of creep resistant aluminium base alloys. This is re ected in the fact that the alloy
that nds greatest application at elevated temperature
is Hiduminium RR.58/2618 and the chemistry of this
alloy was rst arrived at in the 1920s. Ingot and powder
metallurgy manufacturing routes have all demonstrated
alloys with superior short term creep performance compared with 2618 but none have gone on to successful
commercialisation.
The likelihood of the successful development and exploitation of an improved high temperature alloy appears to
be dependent on the commitment to a second generation
large supersonic transport aircraft, capable of ying at
Mach 2 .0 2.2. The alloy would be used for the fuselage of
this aircraft.
In the absence of a new SST aircraft the application of
a new alloy would be greatly improved if it could be used
as a direct replacement for 2618. Its replacement of 2618
in current applications would most probably depend on
advantages, such as weight reductions, improved fatigue
and damage tolerance, and weldability, rather than
enhanced resistance to elevated temperature degradation.
Once established its improved high temperature characteristics would extend its range of application beyond those
which 2618 satis es. It is considered that the most likely
candidate alloy will be one based on the Al Li Cu Mg
Ag system.
Materials Science and Technology

February 2003 Vol. 19

References
1. c. m. hall: `Process of reducing aluminium by electrolysis ,
Patent US400766, US Patent Of ce, Washington, DC, 1889.
2. In `Aluminium, Vol. 2 `In design and application, (ed. K. R. V.
Horn), 16; 1967, Metals Park, OH, American Society of
Metals.
3. k. owen: `Concorde: new shape in the sky ; 1982, London,
Jane s.
4. w. m. doyle: `Improvements in or relating to sheets of
aluminium alloy and methods of manufacturing same ,
Patent GB1069982, British Patent Of ce, London, 1967.
5. `The skin of concorde ; 1974, Milton Keynes, Open University.
6. w. m. doyle: Met. Form., August 1969, 216 220.
7. `An improved aluminium alloy , Patent GB300078, Britsh
Patent Of ce, London, 1928.
8. y. girard, j.-c. ehrstrO m and j. c. rantala: Proc. Symp. on
`Creep deformation: fundamentals and applications , Seattle,
WA, USA, 2002, (ed. R. S. Mishra et al.), 409 419; Warrendale,
PA, TMS.
9. `Hiduminium technical data ; 1967, Slough, High Duty Alloys.
10. w. m. doyle: J. R. Aeronautical Soc., 1960, 64, 535 548.
11. `The properties of aluminium and its alloys ; 1993, Birmingham,
Aluminium Federation.
12. b. wilshire: Metall. Mater. Trans. A, 2002, 33A, (2), 241 248.
13. w. blum, p. eisenlohr and f. breutinger: Metall. Mater.
Trans. A, 2002, 33A, (2), 291 303.
14. k. r. mcnee, h. jones and g. w. greenwood: Proc. 9th Int.
Conf. on `Creep and fracture of engineering materials and
structures , University of Wales, Swansea, UK, (ed. J. D.
Parker), 185 195; 2001, London, The Institute of Materials.
15. w. blum and w. maier: Phys. Status Solidi (a), 1999, 171, (2),
467 474.
16. b. wilshire and c. j. palmer: Proc. Symp. on `Creep deformation: fundamentals and applications , Seattle, WA, USA,
2002, (ed. R. S. Mishra et al.), 51 60; TMS.
17. w. blum: in `Hot deformation of aluminium alloys , (ed. T. G.
Langdon et al.), 181 209; 1991, Warrendale, PA, TMS.
18. h. martinod, c. renon and j. calvet: Mem. Sci. Rev. Metall.,
1969, 66, (4), 303 310.
19. b. wilshire and h. burt: Proc. 9th Int. Conf. on `Creep and
fracture of engineering materials and structures , University of

Robinson et al.

20.

21.

22.

23.
24.

25.
26.
27.
28.

29.
30.

31.
32.
33.

34.
35.

36.
37.
38.

39.
40.
41.

42.

43.
44.
45.
46.
47.

Wales, Swansea, UK, 2001, (ed. J. D. Parker), 261 268; 2001,

London, The Institute of Materials.
j. t. evans, j. s. robinson and r. l. cudd: Proc. Symp. on
`Creep deformation: fundamentals and applications , Seattle,
WA, USA, 2002, TMS, (ed. R. S. Mishra et al.), 289 298.
j. s. robinson, j. t. evans and j. l. lammas: Proc. 9th Int. Conf.
on `Creep and fracture of engineering materials and structures ,
University of Wales, Swansea, UK, 2001, (ed. J. D. Parker),
251 259; 2001, London, The Institute of Materials.
j. g. kaufman (Ed.): `Properties of aluminum alloys: tensile,
creep and fatigue data at high and low temperatures ; 1999,
Materials Park, OH, ASM International.
b. c. hamilton and a. saxena: Eng. Fract. Mech., 1999, 62, (1),
1 22.
g. a. webster: in `Engineering approaches to high temperature
design , Vol. 2, (ed. B. Wilshire and D. R. J. Owen), 1 55;
1983, Swansea, Pineridge.
j. g. kaufman (Ed.): `Fracture resistance of aluminium alloys ;
2001, Materials Park, OH, ASM International.
j. e. bowers and r. d. s. lushey: J. Inst. Met., 1972, 100, 257 267.
j. busby, j. f. cleave and r. l. cudd: J. Inst. Met., 1971, 99,
41 49.
m. j. kerr, e. d. sweet, c. g. bennett and b. c. muddle: in
`Aluminium alloys: their physical and mechanical properties ,
Vol. 217, 1079 1084; 1996, Zurich Uetikon, Trans Tech.
b. noble, s. j. harris and k. dinsdale: Acta Mater., 1997, 45,
(5), 2069 2078.
b. noble, s. j. harris and k. dinsdale: in `Aluminium alloys:
their physical and mechanical properties , Vol. 331 333,
1353 1358; 2002, Zurich Uetikon, Trans Tech.
m. v. lanker: `Metallurgy of aluminium alloys ; 1967, London,
Chapman and Hall.
http://www.thyssen-feinguss.com, Thyssen Umformtechnik +
Guss Gmbh, Soest, Germany.
a. wilm: `Process for the improvement of aluminium alloys
containing magnesium , Patent 244554 Klasse 48d Gruppe 5,
German Patent Of ce, Berlin, 1909.
p. l. thorpe, g. r. tremain and r. w. ridley: J. Inst. Met.,
1950, 77, 111 140.
e. s. balmuth and r. schmidt: Proc. 1st Int. Aluminium
Lithium Conf., Stone Mountain, GA, USA, 1980, Metallurgical
Society of AIME.
B. Gunston (Ed.), `Chronicle of aviation , 758; 1992, MO, JL
International.
w. k. armitage: J. Inst. Met., 1970, 98, 46 51.
w. m. doyle and s. j. ashton: `Improvements in or relating
to aluminium alloys and articles made therefrom , Patent
GB1089454, 1967.
Fulmer: Aluminium base alloy, Patent GB1211563, British
Patent Of ce, London, 1970.
i. j. polmear: J. Aust. Acad. Technol. Sci. Eng., July/August
1996, 93.
i. j. polmear, g. pons, y. barbaux, h. octor, c. sanchez,
a. j. morton, w. e. borbidge and s. rogers: Mater. Sci.
Technol., 1999, 15, (8), 861 868.
i. j. polmear, g. pons, h. octor, c. sanchez, a. morton,
w. borbidge and s. rogers: in `Aluminium alloys: their
physical and mechanical properties , Vol. 217, 1759 1764;
2002, Zurich Uetikon, Trans Tech.
s. m. kazanjian, n. wang and e. a. starke: Mater. Sci. Eng. A,
1997, A234, 571 574.
j. m. silcock: J. Inst. Met., 1960 61, 89, 203 210.
r. n. wilson and p. g. partridge: Acta Metall., 1965, 13,
1321 1326.
r. n. wilson, d. m. moore and p. j. e. forsyth: J. Inst. Met.,
1967, 95, 177 182.
r. n. wilson and p. j. e. forsyth: J. Inst. Met., 1966, 94, 8 13.

Creep resistant aluminium alloys and their applications 155

48. r. singer and w. blum: Z. Metallkd, 1977, 68, 328 336.
49. e. a. fell and w. m. doyle: J. Inst. Met., 1964 65, 93, 316
318.
50. k. owen: `Concorde, story of a supersonic pioneer ; 2001,
London, Science Museum.
51. g. endres: `Concorde; 2001, St Paul, MN, MBI Publishing Co.
52. Final report BRRT CT98 5101, Creepal Thematic Network,
1999.
53. y. barbaux, g. pons and g. lapasset: Proc. Eur. Conf. on
`Advanced materials processes and applications, Padova/Venice,
Italy, September 1995, Associazione Italiana di Metalurgia.
54. y. barbaux, d. guedra-degeorges, j. cinquin, p. fournier and
g. lapaset: Proc. ICAS Conference, Sorrento, Italy, September
1996, ICAS, 937 942.
55. y. barbaux: Proc. 5th Eur. Conf. on `Advanced materials
processes and applications , Maastricht, The Netherlands,
April 1997, Vol. 4, 559 566; 1997.
56. s. pantelakis, a. kyrsanidi, e. el-magd, j. dunnwald,
y. barbaux and g. pons: Theor. Appl. Fract. Mech., 1999,
31, (1), 31 39.
57. j. majimel, g. molE nat, m. j. casanove, d. schuster,
a. denquin and g. lapasset: Scr. Mater., 2002, 46, (2), 113
119.
58. j. majimel, g. molenat, f. danoix, d. blavette, g. lapasset
and m. j. casanove: in `Aluminum alloys 2002: their physical
and mechanical properties , Vol. 396 400, 1025 1030; 2002,
Zurich Uetikon, Trans Tech.
59. j. majimel, m. j. casanove, g. lapasset and g. molenat:
in `Aluminum alloys 2002: their physical and mechanical
properties , Vol. 396 400, 1383 1388; 2002, Zurich Uetikon,
Trans Tech.
60. b. wilshire and h. burt: Proc. Symp. on `Creep deformation:
fundamentals and applications , Seattle, WA, 2002, TMS,
(ed. R. S. Mishra et al.), 269 278.
61. c. r. hutchinson, x. fan, s. j. pennycook and g. j. shiflet:
Acta Mater., 2001, 49, (14), 2827 2841.
62. a. leatham: JOM-e, 1999, 54, (4).
63. j. white, k. mingard, i. r. hughes and i. g. palmer: Powder
Metall., 1994, 37, (2), 129 132.
64. m. a. zaidi, j. s. robinson and t. s. sheppard: Mater. Sci.
Technol., 1985, 1, 737 742.
65. J. R. Davies et al. (Eds.), `Properties and selection: non-ferrous
alloys and special-purpose materials , Vol. 2, 206 208; 1990,
Materials Park, OH, ASM International.
66. o. beffort, c. solenthaler and m. o. speidel: Mater. Sci. Eng.
A, 1995, A191, (1 2), 113 120.
67. o. beffort, c. solenthaler, p. j. uggowitzer and m. o.
speidel: Mater. Sci. Eng. A, 1995, A191, (1 2), 121 134.
68. l. f. mondolfo: `Aluminium alloys structure and properties ,
346; 1976, London, Butterworth.
69. p. stocker, f. rUckert and k. hummert: MTZ Motortechnische Z., 1997, 58, 9, (58), 1 7.
70. j. c. neal, j. a. hooker and a. d. tarrant: in `Advanced
materials for 21st century turbines and power plant , 813 821;
2000, London, Institute of Materials.
71. Economist, 1999, 350, 74.
72. Economist, 2001, 360, 56 57.
73. w. sweetman: Interavia Business Technol., 2001, 56, (657), 20
23.
74. Economist, 2002, 362, 58.
75. r. machler, l. eschbach, c. gerdes, c. solenthaler and
m. o. speidel: in `Aluminum alloys: their physical and mechanical properties , Vol. 217, 1771 1776; 1996, Zurich Uetikon,
Trans Tech.
76. b. p. huang and z. q. zheng: Acta Mater., 1998, 46, (12),
4381 4393.

Materials Science and Technology

February 2003 Vol. 19