Idea Transcript
MECHANICAL PROPERTIES OF AN EXTRUDED ALUWHNttiM-LITHtUM-ALLOY by Ming-JenTAN B.Sc(Eng), A.R.S.M.
A thesis submitted for the degree of Doctor of Philosophy of the University of London and for the Diploma of Imperial College
John Percy Research Group Department of Materials Royal School of Mines Imperial College of Science Technology & Medicine
August 1989
ABSTRACT The addition of Lithium to Aluminium results in lower density, increased modulus of elasticity and a potential for high strength, attractive for the aerospace industries.
properties that make it very
However, the fracture toughness is poor
compared to the conventional aluminium alloys currently in use.
This research is
aimed at implementing thermomechanical treatments in an attempt to improve the extruded properties of an Aluminium-Lithium alloy AA2091. The effects of variations in homogenisation treatment, extrusion parameters, solution heat treatment and differing ageing and thermomechanical treatments were examined.
The effects of a 2% pre-ageing stretch on the microstructure and the
resultant mechanical properties for underage, peak-aged and overaged specimens were studied for extrusions at different reduction ratios and extrusion temperatures.
The
inclusion of a natural ageing cycle (at room temperature) before artificial ageing was also examined.
Artificial ageing was done at different temperatures from 170°C to
210°C. Both scanning and transmission electron microscopy were used together with the light microscope to study the microstrucutres obtained at various stages of processing, and tensile and fracture toughness testing used to evaluate the mechanical properties.
Additional tests include the Differential Scanning Calorimetry (DSC) and
Secondary Ion Mass Spectrometry (SIMS) to determine homogenisation practice and to identify compositions of unknown precipitates respectively. The mechanical properties after extrusion
and subsequent treatments was
found to be dependent on the processing parameters and its consequent structures obtained. It was found that the best ageing condition in terms of fracture toughness was the underaged temper at a low artificial ageing temperature, and the implementation of a 2% pre-ageing stretch did no improve properties.
A natural ageing cycle prior to
artificial ageing improved both strength and properties and represents the most attractive route. The presence of Copper-rich i- and c-phases were found to be present throughout processing and shown to be detrimental to properties in that they assume the role of crack initiation sites during fracture.
Ttf my P arents
3
CONTENTS PAGE
Chapter One
ABSTRACT
2
LIST OF TABLES
7
LIST OF FIGURES
7
INTRODUCTION
1.1
Aerospace Applications
13
1.2
Alloy
15
Chanter Two
Development
LITERATURE SURVEY
2.1
Aluminium-Lithium
Alloys
2 .2
Physical
ofAl-Li
2 .3
Metallurgy
18
Alloys
19
2.2.1
Al-Li Binary System
24
2.2.2
Al-Li-Mg Ternary System
24
2.2.3
Al-Li-Cu Ternary System
27
2.2.4
Al-Li-Cu-Mg Quaternary System
29
2.2.5
Dispersoid Forming ElementAdditions
30
2.2.6
Impurities
Mechanical
2.3.1
Properties
Density
of Al-Li
Alloys
31
31
2.3.2 Stiffness
33
2.3.3 Strength
36
2.3.4
38
Toughness
4
2 .4
Extrusion Processing of Aluminium Alloys
41
2.4.1 Homogenisation
41
2.4.2 Hot Working
41
2.4.3 Heat Treatment
47
Chanter Three EXPERIMENTAL DETAILS 3.1
Introduction
49
3.2
M aterial
49
3 .3
Thermal
3 .4
Extrusion
51
3 .5
Light Microscopy
51
3 .6
Transmission
3 .7
Scanning Electron Microscopy
55
3 .8
Scanning Ion Mass Spectroscopy
56
3 .9
Hardness
56
3 .1 0
Tensile Testing
57
3.11
Fracture Toughness Testing
57
Analysis
Electron
49
Microscopy
54
Characterisation
Chanter Four INGOT PRE-HEATING 4.1
As-Cast
Microstructures
4 .2
Homogenisation Treatment
62
4 .3
Air Cool/ Quench after Homogenisation
74
4. 4
Differential
Scanning
4. 5
Conclusions
Calorimetry
61
Analysis
79 84
Chapter Five EXTRUSION PROCESSING AND PRE-TREATMENTS 5.1
Extrusion
Peak Pressures
86
5.1.1 Pressure Displacement Plots
86
5.1.2 General Pressure Equation
88
5.1.3 Extrusion Pressure Requirements
91
5
5 .2
Extrusion
5 .3
As-Extruded
5 .4
Temperature
Structure
Rise
and
94
Properties
5.3.1
Recrystallisation
99
5.3.2
As-Extruded Structures
1 09
5.3.3 Sub-Grain Sizes
11 8
5.3.4
120
As-Extruded Properties
Solution Heat Treatments
5.4.1
124
Structures
125
5.4.2 Time Variations
5 .5
Ageing
5 .6
Conclusions
Chanter Six
6.1
6 .2
6 .3
99
126
Practice
130
138
THERMOMECHANICAL PROCESSING
Effects of Ageing Temperature
170°C
141
6.1.1 Reduction Ratio 20:1
142
6.1.2 Reduction Ratio 30:1
147
Effects of Ageing Temperature 190°C
147
6.2.1 Reduction Ratio 20:1
147
6.2.2 Reduction Ratio 30:1
150
6.2.3 Notched Tensile Strengths
157
Effects of Pre-Age Streching
159
6.3.1
2% Pre-Age Stretch, Artificial Aged 190°C
6.3.2 Variations in % Stretch
159 1 65
6 .4
Effects of Ageing Temperature 210°C
169
6 .5
Natural Ageing
171
6.5.1
Prior to Artificial Ageing
Natural Ageing Response
171
6
6 .6
6 .7
6.5.2 2 Weeks Natural Age, Artificial Aged 190°C
173
6.5.3 6 Weeks Natural Age, Artificial Aged 190°C
1 73
6.5.4 6 Weeks Natural Age, Artificial Aged 170°C
1 83
Fracture Study
184
6.6.1
Fracture Surfaces
184
6.6.2
Fracture Particles
189
Conclusions
201
Chanter Seven CONCLUSIONS & SUGGESTIONS FOR FURTHER WORK
7.1
Conclusions
203
7 .2
Suggestions for Further Work
206
ACKNOWLEDGEMENTS
207
REFERENCES
208
7
LIST OF TABLES 1
Nominal Compositions of Al-Li Alloys
19
2
Comparison of Cu-rich compositions in Homogenised,
83
As-Extruded and Fractured Specimens 3
Ageing Practice of AA2091
137
LIST OF FIGURES Chapter One
1.1
% Use of Materials in Airframes
14
1.2
Dominating Design Requirements
16
Chapter Two
2.1
Al-Li Binary Phase Diagram
20
2.2
as above with 6 and 6' solvus
21
2.3
Isothermal section of Al-Li-Mg Ternary at
2.4
A
Isothermal section of Al-Li-Cu Ternary at 350°C
B
Isothermal section of Al-Li-Cu Ternary at 165°C
200°C
23 25
2.5
Isothermal sections of Al-Cu-Mg Ternary
26
2.6
A
Structural wt. savings possible in range of properties
32
B
Variation of plane stress fracture toughness with 0.2% PS
for various Al-Li-X Alloys compared with target values 2.7
Contributions to Young's Modulus
34
2.8
Density vs Lithium content
35
2.9
Specific modulus vs Lithium content
35
2.10
Layout of an Extrusion Press
42
Chanter Three
3.1
D.T.A. Set-up
50
3.2
Extrusion Press
53
8
3.3
Short-Rod Fracture Toughness Specimen Dimensions
58
34
Design of Frackjack for use S.R. Fracture Toughness Testing
59
Chapter
Four
4.1
As-Cast Microstructure
64
4.2
Differential Thermal Analysis of As-Cast Material
65
4.3
A Homogenisation at 526°C
68
B Homogenisation at 536°C C Homogenisation at 546°C 4.4
A
Homogenised, with i-phase, Bright Field
70
B SADP of i-phase C
Dark Field
4.5
SIMS Analysis at grain boundary particle
71
4.6
SIMS grain boundary photographs
73
4.7
A Air Cool after Homogenisation
76
B 4.8
Ice-Water Quench after Homogenisation
Mech. Props, of Air Cool/Quench after Homogenisation
77
for 400°C Extrudes 4.9
Mech. Props, of Air Cool/Quench after Homogenisation
78
for 500°C Extrudes 4.10
DSC of As-Extruded Material
80
4.11
DSC of Solution Treated Material
81
Chapter
Five
5.1
Load vs Time plot (Temperature Variations)
87
5.2
Load vs Time plot (Reduction Ratios Variations)
87
5.3
Peak Extrusion Pressures vs Ln (Z/A)
92
5.4
Correlation between Predicted Peak Pressure
92
vs Actual Peak Pressure 5.5
Correlation for 10:1, 20:1 and 30:1 Ratios
93
5.6
Correlation for 40:1 and 50:1 Ratios
93
5.7
Schematic Diagram of Heat Flow during Extrusion
96
5.8
Heat Loss from Different Sources During Extrusion
97
5.9
Extrusion Temperature Rise
98
5.10
Recrystallisation Structures
102
A
As-Extruded 20:1/400°C
9
5.11 5.12
B
As-Extruded 30:1/400°C
C
As-Extruded 30:1/500°C
D
Solution Treated 20:1/400°C
E
Solution Treated 30:1/400°C
F
Solution Treated 30:1/500°C
A
As-Extruded 40:1/500°C
B
Solution Treated 40:1/500°C
A
As-Extruded 50:1/400°C
B
Solution Treated 50:1/400°C
1 04 1 06
5.13
% Volume Recrystallised vs Extrusion Temperatures
1 07
5.14
As-Extruded Structures
1 11
5.15
5.16
5.17
A
30:1/500°C
B
30:1/400°C
C
20:1/500°C
D
20:1/400°C
As-Extruded Structures A
20:1/350°C
B
30:1/525°C
C
50:1/500°C
A p' on grain boundary B
Recrystallised grain
C
Recrystallised grain
D
Loops and helices
A
1/Subgrain size vs Ln Z
B
Subgrain size vs Actual Extrusion Temperature
114
116 1 17 119
5.18
Hardness values for As-Extruded Material
1 21
5.19
As-Extruded Properties
122
5.20
As-Extruded Notched-tensile Strengths
123
5.21
A
1 28
B
Solution Heat Treated Structures - 24 hours Solution Heat Treated Structures - 40 minutes
5.22
Variations in Props, for Different Solution Treatment Times
129
5.23
Ageing curves for 20:1/400°C
132
A
170°C
B
190°C
C 210°C 5.24
Ageing curves for 30:1/450°C
133
A
170°C
B
190°C
C 210°C 5.25
210°C aged Structure
135
Chapter Six 6.1
Mechanical Properties of 20:1 Extrusions, Aged 170°C
143
6.2
20:1/400°C, Aged 170°C Structures
146
A
Bright Field showing d' (UA)
B
Dark Field showing d' (UA)
C
Peak-aged
D SADP of above E Overaged F
as above
6.3
Mechanical Properties of 30:1 Extrusions, Aged 170°C
148
6.4
Mechanical Properties of 20:1 Extrusions, Aged 190°C
1 49
6.5
Mechanical Properties of 30:1 Extrusions, Aged 190°C
151
6.6
A
153
Structure of 30:1/525°C Extrusion, 190°C Underaged
B as above C
Structure of 30:1/525°C Extrusion, 190°C Peak-aged
154
D as above 6.7
A
Structure of 30:1/400°C Extrusion, 190°C Overaged
B
Bright Field of 30:1/525°C, 190°C Overaged
156
C Dark Field of above D
Bright Field of 30:1/525°C, 190°C Overaged
E Dark Field as above 6.8
Notched-Tensile Strengths A
20:1 Extrusions
B
30:1 Extrusions
C
40:1 Extrusions
158
6.9
Props. 2% Pre-Age Stretching on 20:1/400°C, Aged 190°C
1 60
6.10
Aged /Pre-Age Stretched Structures
1 62
A
20:1/400°C, 190°C Peak-aged
B
as above but with 2% Pre-Age Stretching
C
30:1/500°C, 190°C Peak-aged
D
as above but with 2%Pre-Age Stretching
6.11
Tensile Fracture Surfaces A
20:1/400°C, 190°C Peak-aged
B
as above but with 2% Pre-Age Stretching
C
30:1/400°C, 190°C Peak-aged
D
as above but with 2%Pre-Age Stretching
E
30:1/500°C, 190°C Peak-aged
F
as above but with 2% Pre-Age Stretching
164
6.12
Mech. Props, for Varying Amounts of Pre-Age Stretch
1 66
6.13
Structures showing Varying Degrees of Pre-Age Stretch on
168
30:1/500° Extrusion , 190°C Peak-Aged A
1% Stretch
B
2% Stretch
C
4% Stretch
6.14
Mechanical Properties of 20:1/400°C Aged at 210°C
170
6.15
Hardness Variations with Natural Ageing
172
6.16
Ageing at 190°C with and w/o 6 Weeks Prior Natural Ageing
172
6.17
Mech. Props, of 2 Weeks Natural Ageing Prior to 190°C Age
1 74
6.18
Natural Aged Structures for 30:1/500°C Extrusions
176
A 2 Weeks Natural Ageing, 190°C Peak-Aged B 6 Weeks Natural Ageing, 190°C Peak-Aged 6.19
Mech. Props, of 6 Weeks Natural Ageing Prior to 190°C Age
1 77
6.20
Mech. Props, of 6 Weeks Natural Ageing Prior to 170°C Age
1 78
6.21
Natural Aged Structures for 30:1/400°C Extrusions
180
A 6 Weeks Natural Ageing, 170°C, Underaged B as above C SADP of above 6.22
Natural Aged Structures for 30:1/400°C Extrusions A
182
6 Weeks Natural Ageing, 170°C, Peak-aged
B Dark Field of above 6.23
Tensile Fracture Surfacesof As-Extruded Specimen A
1 86
20:1/400°C Extrusion
B as above C
20:1/500°C Extrusion
D as above 6.24
Tensile Fracture Surface of 20:1/400°C , 170°C Aged A Underaged
1 88
B Overaged 6.25
Short Rod Fracture Surface of 20:1/400°C, 170°C Aged
1 91
A Underaged B Overaged 6.26
as for Figure 6.25
193
6.27
A
1 96
Tensile Cross Section of 30:1/500°C ,190°C Overaged
B as above C as above 6.28
A
Fractured Particle of Figure 6.27
198
B as above C 6.29
Ion Micrographs of above
As-Extruded Structure of 50:1/500°C A
Microstructure
B
c-phase present
C SADP of above
200
0 9 &
lW ¥ K P‘D ‘UC(TlOftC
INTRODUCTION
1.1 AEROSPACE APPLICATIONS
Aluminium’s light weight, strength, ductility, corrosion resistance, ease of assembly and cost has made it the dominant material in the aircraft industry for over half a century.
However, with the advent of titanium and
plastic composites,
aluminium's dominance is threatened. Figure 1.1 gives the % usage of materials up to the year 2000; the figures of the latter years are projected figures [112].
The
aluminium industry is countering this threat by developing advanced technology materials of its own. These include rapidly solidified alloys, metal matrix composites and light weight Aluminium-Lithium alloys.
Performance and cost considerations
favour the latter, and commercialisation of Aluminium-Lithium products are now underway for widespread use in the aerospace industry. It should be pointed out that titanium and aluminum are rarely in direct competition as extra costs and fabrication difficulties have limited titanium's use to those parts where superior elevated temperature properties and corrosion resistance are required.
Figure 1.2 gives the dominating design requirements for different
sections of an airframe in which the use of Aluminium alloys are most widespread because of their attractive properties [112]. Carbon fibre composites are expensive and difficult to fabricate and join to metallic structures. Amongst the other problems encountered by the composites are their inferior properties in terms of shear, adhesion and confidence in usage, and absence of any established testing procedures compared to conventional metal alloys. Improvements continue to be made by impurity control, minor elemental additions, better heat treatment and thermomechanical working practices in the aluminium industry.
New manufacturing techniques such as superplastic forming of
complex shapes, precision die forging and one-piece non-critical casting are also being used to reduce the weight and cost of some components.
Figure 1.1
Percentage use of Materials in airframes
1.2
ALLOY DEVELOPMENT It was not the potential weight saving which prompted the Aluminium Company
of America (ALCOA) in the 1950s to develop AA2020, the first commercially available Aluminium-Lithium (Al-Li) alloy, but the advent of supersonic flight and associated problem of kinetic heating. The introduction of a small amount of lithium (1.1%) into a copper-based aluminium alloy enabled the material to be used at modestly elevated temperatures.
Almost incidentally, the alloy was 3% lighter and 8% stiffer than
Duralumin. Introduced in the late 1950s, AA2020 exhibited poor fracture toughness, rapid crack propagation and a tendency to shatter during manipulation. The alloy was soon taken off the market, but not before it was used fairly successfully for the upper wing skins of the US Navy's North American RA-5C Vigilante. This aircraft has just recently been phased out of service, experiencing no fatigue or cracking problems and exhibiting particularly good corrosion resistance. Interest in Al-Li alloys renewed in 1967 when the Soviet Union unveiled an Al-Mg-Li alloy called 01420 (used in the MiG-25 Foxbat).
The combination of 4%
Magnesium and 2% Lithium produced an exceptionally light and fairly stiff, but not particularly strong alloy.
Relative to high strength zinc-based AA7075, 01420 was
10.5% lighter and 7% stiffer. The threat of superior Soviet metals spurred Aluminium-Lithium development in the UK and the USA. The initial objective was to match the Soviet alloy, but it proved difficult to cast and fabricate. Such were the difficulties that in the mid 1970s ALCOA concluded that Aluminium-Lithium could not be produced by conventional ingot metallurgy and elected to follow the powder metallurgy route. The company hit cost problems, however, and development lapsed, only to resume in 1978 along the ingot metallurgy route. Britain acheived only limited success in duplicating the Soviet alloy, but in 1977 Aluminium-Lithium development in the UK accelerated rapidly, driven by two forces : the fuel crisis and the "black spectre" of carbon fibre. In 1981, the Royal Aircraft Establishment (RAE), Farnborough, made something of a breakthrough when instead of using copper (as AA2020) or magnesium (as 01420) as the dominant component, theyused lithium itself. The resulting medium strength alloy, code-named DTD XXXA (AI-2.5% Li-1.4% Cu-0.5% Mg) has not only a 10% density reduction , but more importantly, the alloy has better fracture behaviour
Faiiguc/iiamago
Figure 1.2
than its predecessors. To date, the major aluminum companies
viz.
British ALCAN, ALCOA and
Cegedur-Pechiney have introduced Aluminium-Lithium products onto the market. The main concern of the aerospace manufacturers is the cost factor. The 3% lithium content in a typical alloy costs as much as the aluminium itself, doubling the cost of the alloy. Added to that, is the cost of development of the alloy and the depressed oil prices in recent years, which may not put such a premium on weight savings.
Nevertheless,
Aluminium-Lithium alloys are already incorporated in the design of new aircrafts such as the Boeing 757, and to a wider degree the 7J7.
CrtWPltE 2l(l W 0
svn&Ey
LITERATURE SURVEY
2.1
ALUMINIUM-LITHIUM ALLOYS
Commercial production of several Aluminium-Lithium alloys has been initiated by the aluminium companies. Samples of extrusions, forgings, sheet and plate have recently been provided to aircraft companies for evaluation and qualification testing. These individual and joint development programmes will accelerate alloy optimization, registration and subsequent use of these new products.
Table 1 lists the nominal
compositions of the Aluminium-Lithium alloys and their targeted applications.
2.2
PHYSICAL METALLURGY OF ALUMINIUM-LITHIUM ALLOYS
2.2.1____ Aluminium-Lithium
Binary
When Aluminium-Lithium binary alloys (containing more than 1% lithium) are quenched from the single-phase field and aged at a temperature below the metastable solvus line [1], homogeneous precipitation of the metastable phase 3' (AI3 U) occurs. The atomic arrangement of aluminium and lithium is ordered 3' (LI2 structure) and it is geometrically similar to the F.C.C. lattice of the solid solution thus facilitating the cube/cube orientation dependence observed [2j. The close match between the lattice parameters of the precipitate and matrix results in a small misfit strain [3j and leads to a homogeneous distribution of coherent, spherical 3' precipitates. Silcock [4] first showed the precipitation sequence during the ageing of binary Aluminium-Lithium alloys to be : supersaturated solid solution
-> 3’ (AI3 U) -> 3 (AILi)
and determined the following orientation relationship between the cubic 3 phase and the matrix:
(1OO)0//(11O)AI ;
(0ii)a//(in)Ai;
( o n ) a / / ( i 12) A |
with a { 1 1 1 Jai habit plane. The equilibrium 3 phase has a plate like morphology and a
TABLE 1
NOMINAL COMPOSTIONS OF AL-LI ALLOYS (BY WEIGHT %)
ALLOY
LI
CU
MG
ZR
TARGET
8090 (LITAL A/CP271)
2.2-2.7
1.0-1.6 0.6-1.3
0 .0 4 -0 .0 6
8090A (ALITHALITE A)
2.1-2.7
1.1-1.6
0.8-1.4
0.08-0.15
2024-T3X
8091 (LITAL B)
2.4-2.8
1.6-2.2
0.5-1.2
0.08-0.16
7075-high
8192 (ALITHALITE C)
2.3-2.9
0.4-0.7
0.9-1.4
0.08-0.15
medium strength
2014-medium
8092 (ALITHALITE D)
2.1-2.7
0.5-0.8
0.9-1.4
0.08-0.15
7075-T73X
2090 (ALITHALITE B)
1.9-2.6
2.4-3.0
0-0.25
0.08-0.15
7075-T6X
1.7-2.3
1.8-2.5
1.1-1.9
0.04-0.16
2024-T351
2091 (CP274)
strength
strength
2 0
WEIGHT PERCENT LITHIUM
ATOMIC
CONCENTRATION OF Li
FIGURE 2.1 Phase Diagram of Al-Li with metastable 3’ solvus (dashed line) from McAlister [9] and the predicted metastable miscibility gap (dotted line) from Sigli and Sanchez [8]
FIGURE 2.2 Phase Diagram of Al-Li showing the solvus, and the apparent and coherent 3' solvus lines.
cubic NaT1 structure with a lattice parameter of 6.37
A [5].
The work was confirmed by Noble and Thompson [2 ] and Williams and Edington [3] using the Transmission microscope; and Nazato and Nakai [6 ] and Cocco, Fagherazzi and Schiffini [7] employing thermal analysis and small angle x-ray scattering respectively. More recent work by Sigli and Sanchez [8 ] introduces the presence
of a
miscibility gap which is metastable with respect to the a-9' solvus and lies within the a + 9' phase field.
Figure 2 . 1
incoporates both the 9 solvus and the apparent 9'
solvus lines with the presence of the miscibility gap [9]. The section of the phase diagram showing the 9 solvus and the apparent and coherent 9' solvus lines as determined by TEM [2,11] and SAXS [7,11] is given in Figure 2.2. The metastable a + 9' solvus plays a significant role ensuring that the heat treatment of alloys of commercial interest involves compositions and temperatures lying within it. Coarsening during ageing has been shown to follow Lifshitz-Wagner kinetics, i.e. the average particle size is proportional to the cube root of time [2,3]. In addition to the diffusion controlled growth of 9' in the matrix, local coarsening of 9' appears to be accelerated by the proximity of grain boundaries
[10] and dislocations
[3].
Preferential coarsening near these features results in the development of adjacent precipitate-free zones (PFZs), the rate of growth of which also follows the cube root of time [1 1 ,1 2 ]. There is a disagreement about the mechanism of 9 (AILi) formation.
Some
workers [10,13,14] have reported that 9 exists even after very short ageing times. It is unclear whether 9 is formed from 9' or that it nucleates heterogeneously at the grain boundaries. Nozato and Nakai [6 ] proposed that 9 forms from the metastable 9' on the grain boundaries as a result of the preferential coarsening of
9' such that the
boundaries provide the energy required to overcome the interfacial energy barrier of the 9' to 9 transformation. On the other hand, Williams [14], Sanders et al [16] and Jha [17] showed
9 nucleation independent of
9' and growth of
accompanied by dissolution of surrounding metastable 9'.
9-phase particles
2 3
J3 : AI3M g2
,
: Al.l2M g 17 ,
6 : AILi
FIGURE 2.3 Isothermal section of the Al-Li-Mg system at 200°C [115].
2.2.2
Aluminium-Lithium-Maanesium
Ternary
Additions of copper and magnesium to the Aluminium-Lithium system have been shown to be solid solution strengthening [18].
They also can alter the
precipitation sequence, the solubility of other alloying elements in aluminium, and form binary phases and co-precipitate with 3'. They may also combine with lithium and precipitate as phases peculiar to the ternary and quaternary systems. Lewinson and McPherson [19] studied and constructed Al-Mg-Li isothermal sections. Thompson and Noble [20] investigated the precipitation characteristics and defined them a s : e (AI2 Cu)
T 1 (AI2 CuLi) 3' (AI3 Li) -> 3 (AILi) depending on the Cu:Li ratio. Work by Sanders and Balmuth [16] showed that the metastable precursors of e', which occur in binary Al-Cu alloys, also occur in the ternary system.
The
precipitation of e' has been shown to occur simultaneously with 3', but only in alloys of high Cu:Li ratios such
as 2020 (Al -4.5% Cu -1% Li -0.5% Mn -0.2% Cd) [21].
Copper ( w t °/0)
2 5
L ith iu m ( w t % )
Copper ( w t °lo)
» '
e"
» t
e" -8'
b
‘
e'c +T,) e '* T ,
/G P (1 ) + 0'+ T-; + 6 /
0'
0'
1 _______i -
0'
0' <
V 6 '
6'
_______________ L
/
/ /
/
/ ,
/ ^ /
GP zones -> S" -> S'
S (AI2 CuMg)
The intermediate phases S" and S' are similar to S in structure and orientation but have different lattice parameters [31].
The S' precipitate is semi-coherent and has an
orthorhombic crystal structure : a = 4.04
A, b = 9.25 A, c = 7.18A
S phase grows as rods in the , direction with an orientation relationship : (1 0 0 ) S '//(1 00) ai ;
( 010)S '//(021 )Ai ;
( 001)S '//(01 2 )A ,
and cell dimensions a = 4.04 A , b = 9.23 A , c = 7.14
A
The rods widen and forms laths on {210} habit planes. Lorimer did not find any metastable precursor to the S' phase, which later Cuisat
et a l
viz.
S" phase
[32] proposed. There is no conclusive evidence linking GP
zones and S phase [33], and no microstructural evidence showing the transition from S' to S phase, thus in these alloys, AI2CuMg based precipitates are generally termed S phase. 3' (AI3 Li) d
0 0
V O
-b. o o
on o o
oo
CT\
0.2% PS (MPa)
oo
ro cno ooo
oo-o
UTS (MPa)
oo
cn -k ro cno co oo
CO
gure 6.1 170°C AGEING FOR 20:1 EXTRUSIONS
TREATM ENT
1 44
matrix on bowing round these particles [8 6 ]. Fracture toughness, however, decreases on ageing and this can be attributed to the grain boundary weakening Hautefeuille
et al
[103]
caused by the coarsening of particles there.
attributed this toughness decrease on ageing to the
intergranular precipitation of quasi-crystalline T 2 (AlgCuLie) phase which leads to loss of grain boundary strength.
Another factor that accounts for the low fracture
toughness in Aluminium-Lithium alloys is associated with intergranular failure both in underaged and overaged condition, although different mechanisms work in each case. In the underaged condition, deformation produces strain localization
across the grains
leading to wedge-opening grain boundary failure, whilst in the overaged condition the presence of soft precipitate-free zones (P.F.Z.) imposes localized strain and precipitates adjacent to the grain boundary act as sites for microvoid formation. In the peak aged condition there is a combination of both mechanisms. The results clearly indicate that the PFZs have a more detrimental effect and the fracture toughness values observed of about 12 MPa m1 / 2 are very low.
It was observed that in the overaged
condition, there is an arrest in the decrease in toughness; here, looping and bowing of the dislocations occur around the increasingly incoherent precipitates and this increases the work hardening rate and thus holds up the toughness. The toughness values in the underaged condition are acceptable with minor but important variations being produced by varying the extrusion conditions.
The best fracture toughness is
obtained by extrusion at high temperature and high extrusion ratio (See next Section), and in the underaged condition. The microstructures of the 170°C aged extrude are shown in Figure 6.2 . Figure 6.2 A & B shows the widespread 3' precipitation in both the bright and dark field and there is no 3' PFZ in evidence. The 3' is thought to have existed from the as-extruded condition due to the rapidity of its formation. The presence of S and T-j phase has been detected (Figure 6.2 C & D) with alongside the precipitation of 3' phase. The SADP shows the (111) orientation; the superlattice spots come from the 3' phase, the spots occuring in pairs come from the S phase and streaking is due to the T-| phase. The overaged structures shown in Figure 6.2 E & F shows copious precipitation of the heterogeneous S-phase, which has coarsened compared to that of the peak-aged condition. This
contributes to the deleterious effect on fracture properties, rapid
deterioration of strength and fall in fracture toughness .
Figure 6.2 A B C D E F
20:1/400°C Extrusion, Artificially Aged 170°C
Underaged, showing d ' precipitation As above, Dark Field Peak-aged SADP of above, orientation Overaged Overaged
■■
Z '9
ejnB y
1 47
6.1.2
Reduction Ratio 30:1 The mechanical properties of the extruded material after solution soak at
530°C for 40 minutes for extrusion ratio 30:1 is given
for extrusion temperatures
400°C, 500°C and 525°C in Figure 6.3 . In terms of strength and compared to the 20:1 extrusions, there is very little difference. Strength decreases from the 400°C extrusion temperature to the 500°C and 525°C extrusion temperatures.
Similarly, the ductility also rises from 4% in the
underaged temper to almost 8 % in the overaged temper, being higher at 525°C than 500°C and 400°C.
Toughness is superior on ageing for the 30:1 reduction ratios
compared to the 2 0 : 1 reduction ratios for the 525°C extrusion temperature, but there is hardly any difference in the 400°C extrusion temperature. The values all lie within the range of 15 to 18 MPa m1 / 2 .
6.2
EFFECTS OF AGEING TEMPERATURE 190°C In most precipitation hardenable systems, a complex sequence of time and
temperature-dependent changes is involved.
Adequate control favours avoiding
short-time, high temperature combinations and a preference for temperatures that provide a broad maximum, although economic factors must also be taken into account. Thus although ageing at 190°C produces lower peak strength (from hardness ageing curves), it is commercially more viable. (The time to peak age at 170°C is about 200 hours, compared to 40 hours at 190°C. )
6.2.1
Reduction Ratio 20:1 The properties of the alloy in the aged conditions are shown in Figure 6.4 for
the 20:1 reduction ratio, and the 30:1 reduction ratio is given in the next section. The Ultimate Tensile Strength (U.T.S.) obtained was almost 50 MPa lower at peak age for the 400°C extrusion temperature compared to the 170°C ageing, although this is not as significant for the 525°C extrusion temperature.
Although as-extruded strength was
lower for the lower temperature extrusion , after solution soak at 530°C for 40 minutes, it has superior strength. This observation also applies to the 170°C ageing temperatures, and similarly, it is due to the increased potential for greater
K(ICSR)
TREATM ENT
U Ui A to O Q OO o o o n n o
II < >□
% EL
Ul Ul ^ to o 0 o 0)0 o o o non
n < >□
0.2% PS (M P a)
o
o
o n n
o
t_/l Lft -U to o o t-n o o
UTS (MPa)
-Cv
to o o W O O o o o n o n
t-/\
■£» co
;«re 6.3 170°C AGEING FOE 30:1 EXTRUSIONS
TREATM ENT
K(ICSR)
O O
cn l\D o cn oo
% EL
o o
cn 4^ ro cno ooo
0.2% PS (MPa)
UTS (MPa)
Figure 6.4 190°C AGEING FOR 20:1 EXTRUSIONS
TREATM ENT
TREATM ENT
150
strength after solution soak combined with the greater substructural hardening for the lower temperature extrude.
This also illustrates that the Solution Heat Treatment
chosen had an effective solutionising effect. There was very little difference in ductility in the underaged and peak aged condition but the higher temperature extrusion has slightly better ductility in the overaged condition. condition.
Ductility rises from 4% to between 8 and 10% in the overaged
In terms of fracture toughness, the underage condition again provided the
best values, but for both extrusion temperatures, they were about 1 MPam1 / 2 lower than that of their counterparts aged at 170°C .
6.2.2
Reduction
Ratio 30:1
Results of extrusions aged at 190°C for the 30:1 reduction ratios are given in Figure 6.5 .
In terms of strength there was very little difference with the 20:1
extrusions, and peak U.T.S. occurs at just over 600 MPa, and peak proof strength just under 600 MPa . The proof strength in the underage and overage tempers is highest for the highest extrusion temperature 525°C, followed by the 500°C extrusion temperature and the 400°C extrusion temperature, and the difference ranges from around 420 to around 520 MPa . The ductility obtained was much higher for the underaged temper compared to that of the 170°C ageing, and to that of the 20:1 extrusion ratio and aged at 190°C. It stabilises at about 8 % for the overaged temper, being highest for the lowest extrusion temperature, in agreement with the inverse relationship to strength differences. This higher ductility at the underage temper also corresponds to the lower underage strength as compared to the previous conditions. The toughness values are better than those of the 2 0 : 1 extrusions aged at this temperature, again confirming the better toughness properties at higher reduction ratios.
At the higher reduction ratios, there is greater
work hardening, which gives smaller subgrain and grain sizes. The microstructures of material aged at 190°C are given in Figure 6.6 and Figure 6.7 showing the underaged and overaged tempers respectively. In is apparent that there is no "c-phase PFZ" in the underaged temper but it develops during the course of ageing to the overage condition. In Figure 6.6 A, B & C, c-phases can be found amongst the numerous heterogeneous S phases. In Figure 6.6 B, fine spots, which
oo
O
to
4^
K(ICSR)
li t
TREATM ENT
ooo
ui in £ ro tnOooOooo
II <
o\
oo
%EL
u> U) Ko 8o 8 o n n n
i1 * *
w
S *
niffs-'* ^ ift
" *
’*
*\ • * r *
*
£v
"i 7 8
Figure 6.20 6 WEEKS NATURAL AGEING PRIOR TO 170°C AGEING □ ----o----
20:1/400°C 30:1/400°C
■ E----
30:1/500°C
20:1/400°C 30:1/400°C 30:1/500°C
20:1/400°C 30:1/400°C 30:1/500°C
20:1/400°C 30:1/400°C 30:1/500°C
Figure 6.21 A B C
Natural Aged Structures for 30:1/400°C Extrusions
6 Weeks Natural Ageing, 170°C Underaged (S phase) As above, showing Ti phase SADP of above orientation (presence of 3', S, T* phases)
Figure 6.21
Figure 6.22 A B
Natural Aged Structures for 30:1/400°C Extrusions
6 Weeks Natural Ageing, 170° Peak-Aged Dark Field of above, showing d' precipitates
Figure 6.22
0 . 05 | ± m
1 83 artificially aged at 190°C, strengths were higher than those that were naturally aged for only
2 weeks, and again there was not a lot of difference between the extrusion
conditions examined except that the
20:1 extrusion reaches a slightly lower peak
strength ( which was also the case for the
2 weeks Natural Aged speciemens ). Ductility
reached higher values for the 30:1/400°C values throughout ageing, and was lower for the 30:1/500°C and 20:1/400°C extrusions. over those that were naturally aged only for
Toughness was
significantly improved
2 weeks and values close to 18 MPa m1/2
were reached for the underaged treatments.
Figure 6.18 B shows that homogeneous S phase precipitation can be found within the matrix and this is widespread.
They are observed as "black spots", which
are end-on variants of homogeneous S phase. Coarse precipitates of heterogeneous S phase are found alongside its homogeneous counterparts, they occur as laths.
6.5.4
6 Weeks Natural Ageing. A rtificial Aged 170°C Using a lower temperature artificial age (170°C) after
6 weeks natural
ageing, we have far superior strength and toughness to those obtained by any other previous treatments (Figure 6.20), however, ductility was lower.
Toughness values
reached were around 18 to 20 MPa m1/2 in the underaged condition for all the different extrusion conditions, this approaches the value obtained by alloy AA2014, which is 24 MPa m1/2.
The microstructures obtained were similar to to those that
also had
6 weeks
Natural Ageing, but artificially aged at 190°C, with no perceptable difference. Figure
6.21 shows precipitation of S, T-| and
3' phase in the underaged specimens, as
confirmed by the Diffraction Patterns in Figure 6.21 C. bright and dark fields of
Figure
6.22 A and B are the
6 weeks Natural Aged structures for 30:1/400°C Extrusions
that were subsequently artificially peak-aged at 170°C showing a dense population of 3' phase, some with a "bull's eye" profile, as discussed for Figure 6.7 .
1 84
6.6
FR A C TO G R A PH Y
Microscopic examination, commonly done using Scanning Electron Microscopy can give important information about the nature of fracture from the fracture surface. In this section, fracture path and particle analysis is also discussed.
6.6.1___ Fracture Surfaces In Figure 6.23, the tensile fracture surface of the as-extruded material presented.
In
the
as-extruded
condition,
the
ductility
is
very
is
extrusion
parameter-dependent, and in this case for the 20:1/400°C and 20:1/500°C, the measured values for elongations were 9% and 5% respectively.
The fall in elongation
values can be attributed to the large increase in strength : almost 100MPa in UTS and more than 100 MPa in Proof Stress (Figure 5.19).
On the micrographs, they both
show a preponderance of ductile microvoid coalescence mode of deformation, with ligaments elongated along the axis of stress.
Figure 6.24 shows the micrograph of fracture surfaces of tensile specimens of the underaged and overaged temper. Features of ductile fracture are evident here with the formation of a continuous pattern of dimples or shallow depressions on each surface of the fracture.
They vary in size but in the overaged sample, the dimples were more
numerous and deeper indicating a higher strain to elastic instability. This also follows the load/displacement traces where the overaged specimens had a higher strain prior to elastic instability, and a corresponding higher % Elongation value. nearly all be associated with individual inclusions.
These dimples can
On the microscopic level, a crack is
formed by coalescing of microvoids that form as a result of particle-matrix decohesion.
Measurements of ductility in the tensile test such as % Elongation (and Reduction in Area), whilst providing convenient figures for specifying material quality in general terms, are not really suitable for predicting the material's resistance to fast crack propagation in service.
Fracture toughness testing is necessary if calculation of
the magnitude of applied stress to produce rapid crack propagation is to be known. Here pre-cracked specimens of standard geometry are loaded until they break, and then, if the fractures are macroscopically brittle, the fracture loads can be used to calculate the toughness directly.
Figure 6.23 A B C D
Tensile Fracture surfaces of As-Extruded Specimen
20:1/400°C Extrusion As above 20:1/500°C Extrusion As above
Figure 6.23
Figure 6.24 Tensile Fracture Surface of 20:1/400°C Extrusion,170°C Aged A Underaged B Overaged
Figure 6,24
1 89
Relationships between the strain hardening exponent (n) obtained in uniaxial tension test
(