Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Wallach
Lent Term 2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
This course deals with the design and use of metallic alloys with a focus on the development and
control of microstructure, the relationship between microstructure and properties, and applications.
1.
Introduction. Sustainability. Basis of lattice types. Hume Rothery rules for solid solutions.
Summaries of mechanical and physical properties.
Production of light alloys (Al, Ti, Mg).
2.
Aluminium alloys. Phase diagrams. Alloys and tempers. Alloy characteristics. Review of
preciptation hardening, oxidation, corrosion resistance and fatigue.
3.
Titanium alloys. Pure Ti. Alloying Ti. Specific alloys: , + , and . Superplasticity.
4.
5.
6..
Steels. Review of plain C, alloy steels and cast irons. Commercial steels: high-strength, lowalloy (HSLA), bainitic, dual-phase, transformation induced plasticity (TRIP). Stainless steels.
7.
8.
9.
Non-destructive testing
Reading list
1. Light Alloys: from traditional alloys to nanocrystals, Polmear I. J.,
th
pub. Elsevier Butterworth-Heinemann, 4 edition (2006)
www.msm.cam.ac.uk/phase-trans/newbainite.html
rob.wallach@msm.cam.ac.uk
Page (i)
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
huge range of alloys and tempers allowing optimisation of properties for diverse applications;
their generally high electric conductivity, thermal conductivity, strength, corrosion resistance;
ability to manufacture smart materials e.g. based on their superconducting, optical and
magnetic properties plus biocompatibility.
steels
low cost, high strength (over 90% by weight of all metal usage is steel);
titanium alloys
copper
nickel
Applications include:
Metal
Industries
Applications
Steels
Very wide
Aluminium alloys
Titanium alloys
Biomedical, aerospace
Copper alloys
Nickel superalloys
Aerospace
legislation Kyoto agreement (ended in 2012 and future very uncertain - Doha, Qata, 2012);
scarcity as key metals, especially rare earth metals are limited in their abundance.
greenhouse gas emissions when extracting and shaping (main emphasis normally on CO2).
See Figs. 1.1 and 1.2 on next page.
1 www.ceram.com/materials/metals/
rob.wallach@msm.cam.ac.uk
Page 1 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Figure 1.1 Reserves versus annual world production of various elements [CES software]
Figure 1.2 CO2 emissions versus embodied energy, both for primary production [CES software]
rob.wallach@msm.cam.ac.uk
Page 2 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Question arises of whether the periodic table can explain, in a first-order manner, the different lattice
type for metals since it is the electron distribution, and the resulting energy of a given arrangement
within the material, that determines the lattice type.
A simple approach, by Hume-Rothery, considered the correlation between electronic configuration
and crystal structure.
Crystal structures were shown to be the same for many non-transition metallic elements and alloys
having the same average number of valence electrons per atom, e/a.
This did not extend to the transition metals since, as explained by later researchers, their d electron
orbitals are relatively more localised than the outer-shell s and p electron orbitals; the latter have a
greater influence on the long-range order. It appeared that the ratios e/a are:
bcc structures form in alloys with < 1.5
outer-shell s, p electrons per atom,
hcp structures form in alloys with 1.7 2.1 outer-shell s, p electrons per atom,
fcc structures form in alloys with 2.5 3.0 outer-shell s, p electrons per atom,
diamond structures in alloys with
4.0
outer-shell s, p electrons per atom.
This fits for elements in the third row of the periodic table on adding an additional electron, namely
from Na, Mg, Al to Si. However, deviations are noted in the fourth row of the periodic table for both
Ca (fcc and not hcp) and Ga (orthorhombic and not fcc).
The majority of transition element metals near their melting temperatures have bcc structures. This
can be explained by the presence of vacant d orbitals which means that electrons from outer-shell
orbitals can be accommodated rather than going into p orbitals.
2 Smell-O-Mints, www.jschilling.net/sw_smellomints.php
rob.wallach@msm.cam.ac.uk
Page 3 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
More recent modelling is based on many-electron calculations but these tend to be complex and it is
often necessary to make approximations. This is problematic given that the energy differences of
possible crystal structures for a given element often are small, less than 1% of the bonding energy,
One approach is to consider the energy contributions from
a. the bonding or conduction electrons although, in metals, the potential and kinetic energies of
these electrons are substantially, though not totally, structure independent;
b. the Coulomb interaction energy between the positive metal ions and the sea of electrons .
b) Hume Rothery rules for solid solutions
The majority of metals used in commercial applications are alloys which are typically are solid
solutions or, alternatively, comprise two (or more) phases.
For substitutional alloys, the following Hume Rothery rules indicate the extent of solid-solubility.
Rule 1:
Atomic size factor or the 15% rule: extensive substitutional solid solubility may occur if the
relative difference between the atomic radii (r) of the two elements is less than 15%, i.e.
rsolute - rsolvent
for solid solubility,
15%
rsolvent
Conversely, if the difference > 15%, solubility generally is limited.
Rule 2:
Crystal structures of the two elements must be identical for appreciable solid solubility.
Rule 3:
Valency. The solute and solvent atoms should typically have the same valency in order to
achieve maximum solubility. For different valencies, a metal will dissolve a metal of higher
valency to a greater extent than one of lower valency.
Rule 4:
Electronegativities need to be similar for maximum solubility, i.e. a solute and solvent
should be close in the electrochemical series. When the difference in electronegativities
increases, intermetallic compounds tend to form rather than substitutional solid solutions.
Darken-Gurry maps can be used to show the effects on solid-solubility of electronegativity (vertical
axis) and atomic radius (horizontal axis). Ellipses are added to show the effects of the above rules.
Two DarkenGurry plots are shown in Fig. 1.4 for aluminium3 and Fig 1.5 for palladium4 as solvents.
Figure 1.4
High solubility expected if the solute atom
is within first ellipse (<0.2 difference in X and <7.5%
difference in r from the solvent, Al).
Solubility of at least 5% is expected if the solute atom
is within second ellipse (<0.4 difference in X and <15%
difference in r from the solvent, Al).
Note that Mg has high solid-solubility in Al, as does Cu
Yet neither would have been predicted from the above.
Figure 1.5
DarkenGurry map with ellipse around the
solvent Pd. The vertical line is tangent to
the ellipse at 15% of the diameter of Pd.
3 www.synl.ac.cn/org/non/zu1/knowledge/Hume-Rothery-rules.pdf
4 Fang S S, Lin G W, Zhang J L, Zhou Z Q, The maximum solid solubility of the transition metals in palladium,
International Journal of Hydrogen Energy, Volume 27, Issue 3, March 2002, pages 329332.
rob.wallach@msm.cam.ac.uk
Page 4 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
It has been suggested in the case of the Pd that the discrepancy between predicted and observed
extents of solid-solubility are due to the fact that the electrons in the d and f orbitals (especially) in
some elements do not behave as if completely delocalised. Such electrons retain some unpaired
electron characteristics, e.g. their spin, which accounts for the magnetic properties of these metals.
This is consistent with the Hume-Rothery explanation for transition metal structures.
For interstitial alloys, the following Hume Rothery rules indicate the extent of solid-solubility.
Interstitial solid solutions are more likely to be formed if
- a solute is smaller than the interstitial sites in the solvent lattice of a solvent;
- a solute has approximately the same electronegativity as the solvent.
In practice, there are very few elements that can form ions which are sufficiently small to fit in
interstitial sites, and so appreciable solubility is rare for interstitial solid solutions.
Possible metal ions that may form interstitial solid solutions are: Li, Na, B.
Plus non-ions H, C, N
Many interstitial solid solutions have a strong tendency to spontaneous ordering and examples of
ordered or partially interstitial solid solutions include Al-Li.
c) Intermetallics
Intermetallic compounds are metallic phases but, unlike the alloys described above, each generally
has a very limited composition range, i.e. intermetallics tend to have a narrow and fixed stoichiometry.
Ordering within the crystal lattice is thus common. One consequence is that typically they are quite
hard and strong.
d) Defects
All metallic crystal lattices have defects - perfect lattices, with all atoms of the same type in identical
positions and no missing atoms, do not exist. Defects are crucial to both mechanical and physical
properties of metals and their presence and distribution can be controlled in order to optimise desired
properties intentionally used to manipulate the mechanical properties of a material.
There are three major categories of crystal defect:5
point defects, where an atom is missing or irregularly placed in a lattice structure. Point defects
include:
lattice vacancies,
self-interstitial atoms,
substitutional alloying or impurity atoms,
interstitial impurity atoms;
line defects, which are groups of atoms in irregular positions, e.g. dislocations are line defects;
planar defects, which are interfaces between homogeneous regions of a material. Planar defects
include
grain boundaries,
stacking faults
external surfaces.
5 www.ndt-ed.org/EducationResources/CommunityCollege/Materials/Structure/crystal_defects.htm
rob.wallach@msm.cam.ac.uk
Page 5 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
grain refinement
work hardening
texture strengthening
solid solution hardening
solute pinning
precipitation hardening
ordered structures
two phase strengthening
intermetallic hardening
oxide dispersion strengthening
Toughness
grain refinement
dislocation motion
control of particles
stress distribution
composites
Creep
Fatigue
rob.wallach@msm.cam.ac.uk
Page 6 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Magnetic
soft magnets
hard (permanent) magnets
Density
high and three major crystal classes with long range atomic ordering
Melting temperature
The physical properties of transition metals strongly depend on the number of d electrons and the
strength of the d-orbital interactions, as is shown below in Fig. 1.6 for cohesive energy, bulk modulus
and melting temperature.
Cohesive energy
Bulk modulus
6 www.synl.ac.cn/org/non/zu1/knowledge/Hume-Rothery-rules.pdf
rob.wallach@msm.cam.ac.uk
Page 7 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Figure 1.6 Comparative weights of different alloys for the same stiffness.
th
[Polmear, Light Alloys, 4 edition, p 4, 2006]
Increasingly changing situation for usage of alloys from Al, Ti or Mg due to:
-
composite materials development, e.g. C fibre in epoxy (Boeing, British Aerospace aircraft)
metal matrix SiC in Al or Ti (e.g. car engines7)
ceramic matrix (e.g. Si-SiC nuclear fusion8)
recycling: all three metals are very energy intensive to produce from ores and so recycling is
increasingly promoted, especially for Al alloys.
Figure 1.7 Years of supply (based on known mineral reserves versus recycle rate for various metals.
[Norgate and Rankin, 2002]
7 http://en.wikipedia.org/wiki/Metal_matrix_composite
8 http://composite.about.com/od/aboutcompositesplastics/l/aa030205.htm
rob.wallach@msm.cam.ac.uk
Page 8 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
1.6 Production of light alloys Al, Ti and Mg [see next page for abundance of elements]
1.6.1 Aluminium production9
Aluminium is produced in a two-stage process:
(i) the ore bauxite - iron and aluminium hydroxides/oxides - is refined to produce alumina Al2O3
Bayer process:
Al2O3.xH2O + 2 NaOH
2 NaAlO2 + 2 H2O
2 NaOH + Al2O3.3H2O
(ii) the alumina (aluminum oxide trihydrate) is reduced electrolytically to give metallic aluminum
Hall-Hroult process:
2Al2O3 + 3 C 4 Al + 3 CO2
Four tons of bauxite produce 2 tons of alumina from which one ton of aluminium is formed.
1.6.2 Titanium production10
(a) Kroll process
(i) reduction of titanium ore rutile TiO2 or ilmenite FeO.TiO2 using coal and chlorine gas at 800C
TiO2 + 2 Cl2 + 2 C TiCl4 + 2 CO2
(ii) fractionally distillation of TiCl4 to separate from other chlorides, formed from other metals in
the ore, under an argon or nitrogen atmosphere, and is stored in totally dry tanks;
(iii) reduction of TiCl4 in a batch process using molten Mg (reaction at 1100C) or Na (550C)
under argon to yield Ti porous sponge (note that Na approach is used in the UK only).
TiCl4 + 2 Mg Ti + 2 MgCl2
The Ti is solid at the reaction temperature allowing easy separation from the MgCl2.
(iv) melting of sponge, or sponge plus a master alloy to form an ingot which then is primary
fabricated into billet, bar, sheet, strip or tube.
The Kroll process is both time-consuming and expensive, in part because of the batch processing
required for the TiCl4 reduction.
(b) FFC Cambridge process11
A completely different approach, the FFC Cambridge process, is based on the direct electrochemical
reduction of TiO2 to Ti in molten CaCl2. First developed here in the Department between 1996 and
1997, the process is being developed for commercial usage and also has potential for other oxides.
The process typically takes place between 900 - 1100C, with an anode (typically carbon) and a
cathode (the oxide being reduced) in a bath of molten CaCl2.
1.6.3 Magnesium production
(a) electrolytic reduction of MgCl2 at temperatures ~ 700C. The MgCl2 is obtained from sea
water. the two common ores dolomite MgCO3-CaCO3 and magnesite MgCO3.
(b) the Pidgeon thermic process in which Si or ferro-silicon is used as a reducing agent to extract
Mg from its two common ores dolomite MgCO3-CaCO3 and magnesite MgCO3. The reaction is
highly endothermic. Mg is produced as a vapour which allows its separation.
Si (s) + 2 MgO (s)
Mg (g)
From the Ellingham diagram, the reaction is thermodynamically unfavourable but occurs, utilising
Le Chatelier's principle, due to the Mg removal by condensation at temperatures < 1090C.
9 www.rocksandminerals.com/aluminum/process.htm
10 www.chemguide.co.uk/inorganic/extraction/titanium.html or www.madehow.com/Volume-7/Titanium.html
11 Chen G.Z., Fray D.J. & Farthing T.W., Nature, 407, 361-364 (21 September 2000)
rob.wallach@msm.cam.ac.uk
Page 9 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Page 10 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
rob.wallach@msm.cam.ac.uk
Page 11 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
2. Aluminium Alloys
2.1 Typical Phase Diagrams
Since aluminium exists solely with fcc crystal structure, there are no allotropic phase changes - unlike
Fe, C, Si, S all of which can exist individually in several different crystalline forms.
Control of microstructure and hence properties is by alloying, mainly through solid-solution hardening
and precipitation. Up to 70 wt% of zinc can dissolve in aluminium (at temperatures ~ 400C), followed
by magnesium (17 wt% at 450C), copper (5.7 wt%) and silicon (1.65 wt%).
Typical eutectic and peritectic phase diagrams are illustrated in Fig. 2.1; these two forms describe the
vast majority of phase diagrams for aluminium alloys.
Figure 2.1 Typical phase diagrams for aluminium alloys, illustrating a eutectic and peritectic.
The solubility of solute in the aluminium matrix phase (either denoted by or by ) depends on the
phase with which the is in equilibrium. In the AlCu system, the stable precipitate is CuAl2 but
metastable GP1 zones can form preferentially as they are easier to nucleate (they have lower surface
energy so the activation energy is less). Thus, the free energy curve for GP1 zones is located above
that for CuAl2, as shown in the figure below. The common tangent construction shows that this leads
to an increase in the solubility of copper in when it is in equilibrium with GP1 zones at a particular
temperature see points P and Q. In addition, note that for a given composition, a greater
undercooling is required before GP1 zones can precipitate (see points R and S in the figure below).
R
P
P Q
S
Figure 2.2 The solubility of solute in is larger when it is in equilibrium with GP1 zones compared
with when it is in equilibrium with CuAl2, as is also shown on free energy (G) diagram.
In the AlCu system, the enthalpy of mixing, HM is positive. Hence, at low temperatures, there will
be a tendency for like atoms to cluster, giving rise to a miscibility gap (Fig. 2.3). The enthapy of
mixing is given by:
HM = Na z (1 x) x
where = AA + BB 2 AB
x is concentration of B in binary A,B solution
z is the coordination number
Na is Avogadros number
By considering the entropy changes when mixing two types of atoms, the thermodynamics of solid
solutions can be described and the free energy of mixing GM is given by:
GM = HM - T SM
= Na z (1 x) x + Na kT {(1-x) ln (1- x) + x ln (x) }
(see www.doitpoms.ac.uk/tlplib/solid-solutions/thermodynamics.php)
rob.wallach@msm.cam.ac.uk
Page 12 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Figure 2.3 A eutectic phase diagram with a hidden miscibility gap (left) and free energy of
mixing plotted as a function of temperature and enthalpy HM of mixing.
- when HM < 0 the atoms prefer unlike neighbours and it is always the case that A/ xA > 0;
-
when HM > 0 the atoms prefer like neighbours so for low temperatures and for certain
composition ranges A/ xA < 0 giving rise to the possibility of uphill diffusion.
The miscibility gap at any temperature can be determined by the usual common tangent construction.
Noting that the regular solution model has symmetry about x = 0.5, the compositions corresponding
to the common tangent construction can in this special case be obtained by setting
GM
x
that is
Na z (1 2 x) + Na kT ln(
x
) = 0
1 x
kT
x = exp
The solubility therefore changes exponentially with the reciprocal of temperature, and increases as
HM tends to zero. This is illustrated for a variety of solutes in aluminium in Fig. 2.4 below.
Copper has the largest solubility, i.e. the smallest enthalpy of solution. Solid solution strengthening is
useful but it leads only to an increase of about 40 MPa in the strength of commercial alloys.
However, the copper solubility decreases exponentially as temperature falls and this is used to
facilitate precipitation hardening in aluminium alloys.
rob.wallach@msm.cam.ac.uk
Page 13 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Strengthening
Major applications
1xxx
pure aluminium
work harden
2xxx
Al-Cu
Ppte: CuAl2
3xxx
Al-Mn
work-harden
4xxx
Al-Si
Si needles
5xxx
Al-Mg
work-harden
6xxx
Al-Si-Mg
Ppte: Mg2Si
7xxx
Al-Zn-Mg
Ppte: MgZn2
8xxx
Ppte: Li3Al
work-hardening
solid-solution hardening
grain size control
two-phase alloys
precipitates of optimised spacing and coherency
oxide dispersion strengthening.
Temper designation
Aim: to describe the likely strength as a consequence of fabrication history.
Two families of suffixes after the alloy designation.
-Tx for precipitation hardened alloys where x is a number corresponding to a specific treatment,
e.g.
T4 solution treated and then naturally aged
T6 solution treated and artificially aged.
-F, O or Hx for wrought alloys
e.g.
F as fabricated
O annealed
Hx strain hardened and where the value of x denotes extent of hardening
H4 half hard, H8 fully hard.
T6 is a very common heat treatment for aircraft alloys (typically 2xxx, i.e. Al-Cu), which after solution
treatment might be aged for 68 h at 150170C to obtain the required tensile properties.
More complex heat treatments also used, e.g. better combination of tensile and fatigue properties if
alloys aged for a shorter time at 150170, followed by some natural ageing at ambient temperatures.
Natural ageing generally leads to secondary precipitation on a finer scale.
rob.wallach@msm.cam.ac.uk
Page 14 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
highest electrical conductivity at ambient temperature than any other element for a given weight;
high purity aluminium has a very low yield strength of ~7 MPa and so overhead power
distribution cables are reinforced with steel see Fig. 2.5 below;
in some applications, e.g. electrical motors, the use of aluminium windings increases the
volume of the equipment, so copper is used instead despite the disadvantage in weight.
Figure 2.5 Aluminium overhead power line cable, reinforced with steel wire cores
These are minimally alloyed and generally used in the annealed condition with yield strengths
y 10 MPa. Applications: electrical conductors, chemical equipment, foil and architecturally.
2xxx
Al-Cu precipitation hardened alloys, and used extensively in civil aircraft due to their high
strength to weight ratios. These alloys generally are not fusion weldable, and aircraft are
riveted. Various alloy additions can be made to optimise their properties. Most recently, the
addition of Li has been investigated.
Lithium as an alloy element in aluminium alloys is unique in that it, unusually for most alloying
additions, significantly improves stiffness: each 1 wt.% of lithium reduces the density of an ALLi alloy by about 3% and increases the stiffness by ~5%.
The alloy 2090 has a composition of 2.7 wt.% Cu, 2.2 wt.% Li and 0.12 wt.% Zr.
3xxx
These are the AlMn or AlMnMg alloys with moderate strength ductility and excellent
corrosion resistance. The strength, at about y 110 MPa, comes from dispersoids which
form in the early stages of solidification. The Mn concentration is restricted to about 1.25 wt%
to avoid excessively large primary Al6Mn particles. Magnesium (0.5 wt%) gives solid solution
strengthening and the AlMnMg alloy is used in the H or O conditions. Beverage cans
represent the largest single use of either aluminium or magnesium alloys. A typical alloy has
the chemical composition Al0.7Mn0.5Mg wt%.
4xxx
AlSi is a simple binary eutectic with eutectic composition ~11 wt.% Si. Used only for castings
(e.g. aluminium car engine blocks with hyper euctectic composition of ~18 wt.% Si) or for
brazing other aluminium alloys (based on eutectic composition).
5xxx
The magnesium concentration is usually maintained to less than 34 wt% in order to avoid
Mg5Al8. The strength is in the range y 40 160 MPa with rapid work hardening during
deformation. Work hardened aluminium alloys tend to soften with age because the
microstructure is not stable even at ambient temperature. Therefore, it is better to excessively
work harden and then to anneal to the required strength and stability. The alloys, which have
excellent corrosion resistance, are used to make the bodies of boats or vehicles. They are
readily welded.
6xxx
This Al-Mg-Si family of alloys relies on Mg2Si precipitation hardening. These alloys are readily
extruded and also anodized. Hence they are used both architecturally as well as for sports
equipment.
7xxx
High strength alloy based on MgZn2 precipitation hardening. Tendency to stress corrosion for
greater than (Mg + Zn) > 6 wt.%.
8xxx
Various other alloys including Li based with Li3Al pptes. Low density, increased stiffness
leads to aircraft applications. Problem with poor toughness.
Dispersion strengthened alloys: Molten aluminium is broken up into droplets which are immediately
oxidised on their surfaces. On compaction, the surface oxide breaks up into highly stable
dispersoids in an aluminium matrix. There are alloys with up to 20 wt% of alumina. Other
particles such as SiC can be added in large quantities these are often called metal matrix
composites.
rob.wallach@msm.cam.ac.uk
Page 15 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
2.5 Precipitation
See Part IA and Part IB lecture notes which cover the following key points.
(a) Age hardening involves the rapid cooling of a solid solution from a high temperature to one at
which it becomes supersaturated so that ex-solution of solute (precipitation) begins on ageing.
Note that it is not just the solute concentration which is supersaturated but also the vacancy
concentration and the vacancies act as heterogeneous nucleation sites.
(b) The need to age in a manner to avoid precipitatefree zones. The latter form either due to either
vacancy or solute depletion in the vicinity of grain boundaries.
(c) The role of metastable precipitates in the development of precipitation hardening (more coherent
metastable precipitates, with lower surface energies, decreases the nucleation activation energy).
2.7 Fatigue
There are two major difficulties. Coherent precipitates are cut by dislocations; each passage of a
dislocation shears the particle, producing steps at the entry and exit sites, thereby reducing the
particle crosssection on the slipplane (Fig. 2.6). This makes it easier for a subsequent dislocation
to cut the particle. Slip then tends to focus on particular planes, leading to stress concentrations
which promote fatigue. It is better therefore to have a mixture of fine, coherent and bigger semi
coherent precipitates so that the danger of inhomogeneous slip is reduced.
Fatigue is also initiated at pores in thick aluminium components. This can only be controlled by
careful processing, and by rolling deformation where this is permitted.
rob.wallach@msm.cam.ac.uk
Page 16 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Titanium Alloys
3.1 Pure Titanium
Density 4.51 g cm
Pure titanium has excellent resistance to corrosion and is used widely in the chemical industries.
Ti forms a passive oxide film which makes it particularly resistant to corrosion in oxidising solutions.
The corrosion resistance can be further improved by adding palladium (0.15 wt%), which makes
hydrogen evolution easier at cathodic sites so that the anodic and cathodic reactions balance in the
passive region.
Many chemical plants use steel vessels which are protected by titanium alloys; titanium sheet (0.5
mm thick) is frequently a loose liner or may be roll or explosion bonded to the steel. Titanium
condenser tubes are used in power plants and in desalination plants.
Titanium appears ideal, given its high melting temperature, for use in components which operate at
elevated temperatures, especially where large strength to weight ratios are required.
Figure 3.1 Specific strength (yield stress divided by density) versus temperature
However, if titanium alloys rub against other metals at elevated temperatures, the titanium alloy
oxidises extremely rapidly and causes severe damage. This limits its application, in the harsh
environment of aero engines, to regions where the temperature does not exceed 400C.
The world production of titanium (currently electrolytically) is, therefore, relatively small, hundreds of
thousands of tonnes, which compares with steel at 750 million tonnes per annum. Some 80% of all
the titanium produced is used in the aerospace industries. Car suspension springs could easily be
made of titanium with a great reduction in weight but titanium is not available in the large quantities
needed at the price required for automobile applications. The target price for titanium needs to be
reduced to about 30% of its current value for serious application in massmarket cars.
This now is potentially possible by replacing the traditional Kroll process and subsequent refinement
by electrolysis of fused salts see Section 1.6.2.
The crystal structure of titanium at ambient temperature and pressure is closepacked hexagonal ()
with a c/a ratio of 1.587. Due to the distortion from the ideal c/a ratio of 1.67, slip is possible on the
pyramidal, prismatic and basal planes in the closepacked directions. At about 890C, titanium
undergoes an allotropic transformation to a bodycentred cubic phase which remains stable to the
melting temperature.
3.2 Alloying of Ti
All elements which are within the range 0.851.15 of the atomic radius of titanium alloy
substitutionally and have a significant solubility in titanium as per Hume Rothery rules.
Elements with an atomic radius less than 0.59 that of Ti occupy interstitial sites and also have
substantial solubility (e.g. H, N, O, C).
The ease with which solutes dissolve in titanium makes it difficult to design precipitationhardened
alloys. Boron has a similar but larger radius than C, O, N and H; it is therefore possible to induce
titanium boride precipitation. Copper precipitation is also possible in appropriate alloys.
rob.wallach@msm.cam.ac.uk
Page 17 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Figure. 3.2. Plot of atomic radius versus Pauling electronegativity for elements. Notice many elements of similar
size to titanium, and B, H, N, O and C all fall in interstitial range of HumeRothery rules.
Alloying elements can be categorised according to their effect on the stabilities of the and phases:
Al, O, N, Ga, Zr and Sn are stabilisers;
Mo, V, Si, W and Ta
are stabilisers.
Cu, Mn, Fe, Ni, Co and H are also stabilisers but form the eutectoid. The eutectoid reaction is
frequently sluggish (since substitutional atoms involved) and is suppressed.
Molybdenum and vanadium have the largest influence on stability and are common alloying
elements. Tungsten is rarely added due to its high density. Cu forms TiCu2 which makes the alloys
agehardening and heat treatable; such alloys are used as sheet materials. It is typically added in
concentrations less than 2.5 wt% in commercial alloys.
Interstitials
Interstitials inevitably cause changes in the lattice parameters. Body centred cubic Ti has three
octahedral interstices per atom whereas hcp Ti has one per atom. The latter are therefore larger, so
that the solubilities of O, N, and C are much higher in the phase.
Hydrogen is the most important interstitial. Titanium is capable of absorbing up to 60 at% of
hydrogen, which can also be removed by annealing in a vacuum. Hydrogen enters the tetrahedral
holes which are larger in b.c.c. than c.p.h. Thus the solubility of hydrogen is larger in . The enthalpy
of solution of hydrogen in Ti is negative (H < 0).
H2 2H
0
G = G + RT ln
so that
a H2
pH2
a H2
S
H
= exp exp
pH2
R
RT
The solubility actually decreases with temperature (Figure. 3.4). This contrasts with iron which shows
the opposite trend.
rob.wallach@msm.cam.ac.uk
Page 18 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Because of this characteristic, titanium had been considered as a candidate material for the first wall
of magnetically confined fusion reactors. The hydrogen-based plasma is not detrimental since at
500C and 1 Pa pressure, the Ti does not pick up enough hydrogen for embrittlement. An additional
feature is that Ti resists swelling due to neutron damage.
A large enough concentration of hydrogen induces the precipitation of hydrides. TiH1.52.0 has a
cubicF lattice and its precipitation causes embrittlement due to a volume expansion of about 18%.
There are regions of hydrostatic tension at crack tips where it forms preferentially, leading to large
increases in the crack growth rate, some 50fold during fatigue.
The hydride reaction can also be used to store hydrogen reversibly:
2.2 FeTiH1.04 + H2
2.2 FeTiH1.95
The energy to weight ratio for such a cell is about a tenth that of petrol. However, one problem with
this method of hydrogen storage is that hydride formation is accompanied by a considerable volume
expansion, which in turn can embrittle the alloy. Amorphous alloys of titanium are better in this
respect, since they do form hydrides and yet reversibly accommodate large quantities of hydrogen by
an expansion of the nearest neighbour distance.
The ZrTi Laves phase Ti0.24Zr0.76(Ni0.55Mn0.3V0.065Fe0.085)2.1 has been found to reversibly
1
accommodate nearly 1.5 wt% of hydrogen, with a battery rating of some 440 mA h g .
3.3 Specific alloys
alloys
Aluminium is the main alloying element. The combined effect of the stabilisers is expressed as:
aluminium equivalent, wt% = Al + 1/3 Sn + 1/6 Zr + 10 (O + C + 2N)
If this exceeds about 9 wt% then there may be detrimental precipitation reactions, generally by the
formation of Ti3X which has an ordered h.c.p. structure.
The presence of a small amount of the more ductile phase in virtually all alloys is advantageous
for heat treatment and the ability to forge. The alloys may therefore contain 0.5 - 2 wt% of Mo, e.g.
Ti 6Al 2Sn 4Zr 2Mo
where the Zr and Sn give solid solution strengthening.
A near alloy has been developed, with good elevated temperature properties (T < 590 C):
Ti 6Al 4Sn 3.5Zr 0.5Mo 0.35Si 0.7Nb 0.06C
The niobium is added for oxidation resistance and the carbon to allow a greater temperature range
over which the alloy is a mixture of + , in order to facilitate thermomechanical processing. This
particular alloy is used in the manufacture of aeroengine discs and has replaced discs made from
much heavier nickel base superalloys. The final microstructure of the alloy consists of equiaxed
primary grains, Widmansttten plates separated by the phase.
+ alloys
Most + alloys have highstrength and formability, and contain 46 wt% of stabilisers which
allow substantial amounts of to be retained on quenching from the + phase fields, e.g.
Ti 6Al 4V
rob.wallach@msm.cam.ac.uk
Page 19 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Al reduces density, stabilises and strengthens while vanadium provides a greater amount of the
more ductile phase for hotworking. This alloy, which accounts for about half of all the titanium that
is produced, is popular because of its strength (1100 MPa), creep resistance at 300C, fatigue
resistance and castability.
One difficulty with the phase, which has a bodycentred cubic crystal structure, is that like ferritic
iron, it has a ductilebrittle transition temperature. The transition temperature tends to be above room
temperature, with cleavage fracture dominating at ambient temperatures.
Burnresistant alloys
Titanium fires can occasionally occur in aero engines or in titanium based heat exchangers used in
the chemical industries.
The addition of chromium in concentrations exceeding 10 wt% helps improve the burnresistance of
titanium alloys. The alloy Ti35V15Cr wt% has sufficient chromium to resist burning in an
aeroengine environment to temperatures up to about 510C. Chromium is not found to be effective in
binary TiCr alloys where it does not encourage the formation of a continuous film of protective oxide.
Quenching from
Quenching the phase leads to the formation of h.c.p. martensite. This is not particularly hard and
there are increasing quantities of retained in the microstructure as the solute concentration
increases and the MS temperature decreases (Fig. 3.5). The habit plane of the martensite is close to
{3 3 4} and the crystallographic relationships between the original and the martensitic are:
(1 1 0) || (0 0 0 1)
and
Ms
Mf
[1 1 1] || ([1 1 2 0]
Transformation to
The phase is metastable and forms from in alloys based on titanium, zirconium and hafnium. It is
important because its formation generally leads to deterioration in mechanical properties. There is
also an increase in the electrical resistance as forms and, in TiNb alloys, its formation influences
superconduction.
The to transformation is diffusionless, occurs below the T0 temperature and frequently cannot be
1
suppressed even by quenching at 11,000C s . It is reversible and diffusionless but is not
martensitic in the classical sense since there is no invariantplane strain shape deformation.
However, it does involve the coordinated motion of atoms.
Titanium aluminides Ti3Al and TiAl
The most successful of the aluminides has a lamellar structure made up of alternating layers of an
ordered hexagonal (D19) Ti3Al and 2 compound and ordered f.c.t. (L10) TiAl or . The phase forms
with its most closely packed plane parallel to the basal plane of the 2:
{1 1 1} || {0 0 0 1}2
< 1 1 0 > || < 1 1 2 0 > 2
The lamellar microstructure is a direct consequence of this orientation relationship.
Ductility is about 4- 6% at ambient temperature and the aluminide tends to be more ductile. Their
2
densities are about 4.5 g cm and the aluminium makes them resistant to oxidation or burning. The
alloys have been extensively studied for aerospace and automotive turbochargers because of their
high strength, low density and creep resistance.
rob.wallach@msm.cam.ac.uk
Page 20 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
3.4 Superplasticity
Superplasticity is ability of polycrystalline material to deform to very large strains without failure.
- strains greater than 200% common and several commercial materials can show over 1000%
while development alloys can exceed 4000%
- need two-phase microstructure which remains stable at processing temperature (T 0.5 Tm)
- grain size must be (and has to remain) very small, few m diameter
-4
-2
-1
= k [ ]
Can also incorporate diffusion bonding (solid-state joining) as part of a combined manufacturing
operation, e.g. to form stiff honeycomb structures. Apply pressure at elevated temperature for time to
promote bonding and then use inert gas at high pressure to cause superplastic shape change.
pressure
stop off so
no bonding
3.4.1 Theory
a) T < 0. 4 Tm
n
True stress - true strain described by = k
Load (P) instability, i.e. dP = 0, occurs when
d T
=
d T
Note if
d T
>
d T
rob.wallach@msm.cam.ac.uk
Page 21 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
b) T > 0.4 Tm
Additional time dependence, hence
= k' n [ ]
= k n
m 0.03 and
= k [ ]
P
A
also
1/m
1/m
= 1
A
hence
1 l
t l
P
k
1 l
l t
- 1 A
A t
- dA = A
dt
1
m
1
A
P
k
P
k
1
m
1
m
(1-m)/m
Measurement of m
m =
d log
d log
log
ln 2 / 1
log
ln 2 / 1
log
10
-5
10
-3
10
rob.wallach@msm.cam.ac.uk
-1
log
strain rate
Page 22 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
!
Figure 3.6 Superplasticity model of Ashby-Verrall
[Ashby M.F. and Verrall R.A, Acta Met., 21 p29, 1973]
Figure 3.7 Deformation map for copper with grain size of 100 m
rob.wallach@msm.cam.ac.uk
Page 23 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
high stress: dislocation core diffusion and glide leading to dislocation climb;
low stress: grain boundary diffusion (Coble creep) or bulk diffusion (Nabarro Herring);
rate controlled by vacancy creation & annihilation at grain boundaries.
c. Dislocation glide
dislocations glide in a grain and reach the grain boundary and hence dislocation pile up;
pile-up stress results in a higher stress in the adjacent grain and hence slip initiates in it;
dislocations will also climb resulting in more slip;
dynamical recovery and recrystallisation also needed to prevent work-hardening.
rob.wallach@msm.cam.ac.uk
Page 24 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
4. Magnesium Alloys
4.1 Introduction
-3
Magnesium is the lightest of the structural metals (density of only 1740 kg m , although most
magnesium alloys have slightly higher densities due to the alloying additions needed.
The world production of Mg is relatively small compared to other structural metals such as Al and
steel. However, the alloy is being promoted strongly in China which now produces ~ 50% of the
worlds Mg using the Pidgeon process (see Section 1.6).
About half Mg usage is as an alloying addition to aluminium alloys, e.g. the 5xxx series of alloys such
as 5083 with 4.5 wt.% Mg.
Other metallurgical uses include:
-
pressure die castings, e.g. parts for the car industry to aid fuel economy including
transmission casings cast in AZ91D resulting in 25% weight saving over Al alloys
steering components using AM50A & AM60B alloys (more ductile)
instrument panels, intake manifolds, cylinder head covers, inner boot lid sections
GM Savana & Express vans in the USA use up to 26 kg of Mg alloys.
sacrificial anodes to protect steel structures from corrosion, e.g. ships, oil and gas pipelines.
Uses
AZ91
General casting
Good castability,
good mechanical properties at
T<150C.
AM60
AZ31
Wrought magnesium
Good extrusion alloy.
products
ZE41
Specialist casting
AS41
General casting
There are many other specialist alloys containing zirconium, expensive rare earths and even silver.
Current research aims to improve corrosion resistance and high temperature creep resistance of Mg
alloys.
1 http://mg.tripod.com/mggen.htm
rob.wallach@cam.ac.uk
Page 25 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
improves castability
Mn
controls Fe content by
n
Fe-Mn compound form
Zn
precipitation hardening
can tend to brittleness
Si
decreases castability
weak grain refiner
rob.wallach@cam.ac.uk
Page 26 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
5. Anisotropy
5.1 Introduction
The response of many materials, natural and man-made, are not uniform when used in service, i.e.
the material behaves in an anisotropic manner. This is a consequence of the dependence of
properties on a materials
atomic structure
chemical composition
defects
microstructural
and
crystallographic.
The former can apply to all materials including natural materials such as wood, while the latter clearly
applies only to crystalline materials whether natural (e.g. minerals such as quartz) or man-made
(many metallic alloys).
Image processed
rob.wallach@cam.ac.uk
Page 27 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
Fracture
stress
(MPa)
Fibre orientation ()
Figure 5.3 Variation of fracture stress with fibre orientation for a composite
[http://aluminium.matter.org.uk/ Anisotropy module, page 4]
For composites, the analysis is relatively straightforward using the rule of mixtures, i.e.
L = Vf fb + (1 Vf) mb
Certain types of microstructural anisotropy in metallic alloys can be treated rigorously, e.g. the
variation of yield strength Y with the mean grain size d by the Hall-Petch equation:
Y = i + k d 1/2
where i is the lattice friction stress (i.e. stress to move dislocations other than near gb pile-ups)
k is a constant.
However, since microstructural anisotropy in metallic alloys generally is less easy to describe than
that in composites, predictions of properties arising from it are more difficult than for composites.
5.3 Crystallographic anisotropy in single crystals
5.3.1 Transport properties such as diffusion, thermal and electrical conductivities
Single crystals have regular long-range arrangements of atoms/ions. However, the properties will not
be the same in every direction, unlike a gas, an amorphous solid, or even a polycrystalline solid.
Mechanical and physical properties generally show an angular variation with respect to the major
crystal axes in a single crystal and also reflect the inherent symmetry of the crystal. This is the basis
of the piezoelectric, pyroelectric and ferroelectric effects in perovskites, as was introduced in Part IA.
The variation in behaviour is described by Neumanns principle1 (of symmetry), namely:
the symmetry elements of any physical property of a crystal must include the symmetry
elements of the point group of the crystal.
The variation of a property such as diffusion or thermal conductivity in a single crystal is described
using a second rank tensor. Diffusion in a material is described using Ficks first law in which the flux
Ji is given by:
dC
Ji = - Dij
dx j
where
dC
is the concentration gradient and Dij is the diffusion coefficient.
dx j
Page 28 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
-
A representation surface for hexagonal or tetragonal lattices is shown below in Fig. 5.4.1 2
Ji
dC
or grad C
dx j
Dx
1
Dz
Figure 5.4 Representation surface for self-diffusion in magnesium
[modified from DoITPOMS TLP Anisotropy www.doitpoms.ac.uk/tlplib/anisotropy]
If the concentration gradient is in the same direction as one of the principal axes (x, y or z) then the
resulting flux will be in the same direction. However, if it is an any other angle to the principal angles,
the consequence of the differences in the magnitudes of Dy and Dz, is that the resulting flux is not in
the same direction as the concentration gradient.
The direction and magnitude of the flux may be calculated using direction cosines but also are shown
conveniently using the above representation surface. Using the representation surface, a line is
drawn to show the direction of the concentration gradient relative to the two principal axes as then:
-
the direction of the resulting flux is given by the normal to the tangent to the ellipse at the point
where line for the concentration gradient crosses the ellipse;
1/2
Consider diffusion in, say, an hcp Mg single crystal. A unit cell has two lattice parameters a and c,
and the ratio c/a ideally is (8/3) or 1.633. This suggests that diffusion within the close-packed basal
plane (0001) should be higher than that normal to the basal plane, i.e. in the c direction or [0001].
-12
-1
This has been confirmed experimentally; the measured values were Dx or Dy = 1.85 x 10 m s and
-12
2 -1
Dz = 1.0 x 10 m s at a temperature of ~425C.3 Hence the representation surface would be
similar to that in Fig. 5.4 above.
5.3.2 Mechanical properties: crystal orientation and texture hardening
Plastic deformation by dislocation movement occurs when the applied force/s F on a body of crosssectional area A results in a shear stress in a slip system which exceeds the critical resolved shear
stress CRSS.4 A slip system is a combination of a close (or closest )packed plane and a close
packed direction in that plane.
The resolved shear stress on a close packed plane and in a close packed direction is given by the
expression
F
=
cos cos
A
where is the angle between the tensile force direction and the normal to the slip plane and is the
angle between the tensile force direction and slip direction in the slip plane. The product cos cos is
known as the Schmid factor.
1 As introduced in Pt II Course C4 Tensors in Section 8.1 The representation surface for second rank tensors.
2 DoITPoMS TLP Anisotropy www.doitpoms.ac.uk/tlplib/anisotropy, section Anisotropy ellipsoid.
3 McKie D and McKie C, Crystalline Solids, Thomas Nelson, p 367 1974.
4 Part IA Course D and Pt II Course C6 Crystallography in Section 6 Deformation and texture.
rob.wallach@cam.ac.uk
Page 29 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
If a tensile stress is applied to a single crystal, two approaches can be used to find on which slip
system the resolved shear stress will be maximum: OILS rule or Diehls rule.1 The latter uses a
stereographic projection as shown in Fig. 5.5 for an fcc or bcc single crystal on which twenty-four slip
systems are shown. The approach to finding the slip system with the maximum resolved shear stress
1,
is implicit in the figure and has been covered elsewhere. ,2
Note that the three corners of each stereographic triangle are of the form 100, 110. 111.
Figure 5.5 Stereographic projection for fcc or bcc crystal showing 24 triangles corresponding to the 24 slip
systems. [From Pt. II Course C6 Crystallography, Section 6]
The actual location of the tensile axis with respect to the particular (111) plane normal and [110] slip
direction affects the Schmid factor, and also can affect the number of slip systems that can operate
initially. Accordingly, the magnitude of the applied force F required for plastic deformation, varies with
the location of the tensile axis as represented on the stereogram.
From the expression for the Schmid factor, the load needed for plastic deformation will be minimum if
both angles and are 45. Conversely, a higher load would be needed if the tensile axis was at
some other orientation, e.g. if it was near the ,110> or <111> corners of a stereographic triangle.
This is texture hardening and applies to all crystal systems, not just in fcc which was used above to
introduce the concept.
(b) hcp single crystals and work softening
Texture hardening is very evident in hcp single crystals in which there are only three slip systems,
namely the three close packed directions in the only close-packed plane, the basal plane (0001).
Hence if a force is applied at 90 to the (0001) plane, there can be no resolved shear stress in the
plane and so the yield stress would be infinitely high (unless slip took place on a different plane).
As the angles between the tensile axis and slip plane normal and slip direction both are decreased
from 90, the load needed for plastic deformation would decrease until it reached its minimum value
when both angles and were 45. As the angles decreased further below 45, the load would again
increase and would become infinite if the applied force was in a direction parallel to the (0001) plane.
This can be observed in practice and the decrease in the necessary load is called work softening. An
example for a cadmium single crystal is shown in Fig. 5.6. Initially, the Schmid factor is high due to
the orientation of the tensile axis with respect to the single crystal. As deformation proceeds, the
tensile axis rotates towards the slip direction or, if the axis along which the load is applied is fixed, then
the slip system rotates during deformation. In either case, the Schmid factor can decrease and so the
load for plastic flow decreases.
1 DoITPoMS TLP Slip in Single Crystals www.doitpoms.ac.uk/tlplib/slip, section Slip geometry.
2 As introduced in Pt II Course C6 Crystallography in Section 6 Deformation and texture.
rob.wallach@cam.ac.uk
Page 30 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
Figure 5.6 Schematic load-extension curves for tensile deformation of two cylindrical cadmium crystals.1
In Fig. 5.6, the solid line shows the behaviour of a crystal showing geometric softening, where the
load decreases once plastic deformation commenced. The dotted line is for a sample in which no
geometric softening occurs, i.e. the tensile axis is at a different angle to the operative slip system. The
rise in load in both cases as the extension increases is due to normal work hardening.
(c) Variation of other mechanical properties, e.g. Youngs modulus
Given the different packing densities in different directions in single crystals, it is not surprising that
certain properties will reflect this. As an example, the anisotropy of Youngs modulus for an aluminium
single crystal has been estimated by calculation to increase by ~ 15% when the tensile direction is
rotated from the [100] to the close-packed [110 direction, as is shown below.
Figure 5.7 Effect of single crystal orientation on the Young modulus of Al.2
H
Figure 5.8 Magnetocrystalline anisotropy: variation in applied .field H required to achieve the same induced field B
for two different crystallographic axes, i.e. easy and hard directions.1
1 DoITPoMS TLP Slip in Single Crystals www.doitpoms.ac.uk/tlplib/slip, section Slip in HCP metals 4.
2 http://aluminium.matter.org.uk/ - choose anisotropy section
rob.wallach@cam.ac.uk
Page 31 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
Magnetocrystalline anisotropy occurs because the ease of aligning electron spins is affected by the
crystallographic direction in a unit cell. The terms soft and hard directions are used to differentiate
between the magnitudes of the applied magnetic field H required to induce a field M of a given
magnitude. This is shown in Fig. 5.8. Two consequences arise.
Firstly, magnetocrystalline anisotropy energy can be minimised by forming domains such that the
electron spins are aligned in easy crystallographic directions. However, in any domain walls, there
must be a change in the direction of the magnetisation and the electron spins will not all be aligned
along easy axes. Hence magnetocrystalline energy is minimised in materials with large domains and
correspondingly few domain walls.
Secondly, the hysteresis loop on reversing an applied field will be larger if magnetisation takes place
along hard directions rather than soft. There will be a higher energy associated with the larger
hysteresis loop and this means lower efficiencies (and more heat evolution) in applications such as
motors or transformers see Fig 5.9.
B
Figure 5.9. Hysteresis loops for hard and soft magnetic materials
[www9.dw-world.de/rtc/infotheque/electronic_components/inductors.html]
Hard and soft directions in iron, nickel and cobalt are as follows:
Easy
Intermediate
Hard
Fe
bcc
<100>
<110>
<111>
Ni
fcc
<111>
<110>
<100>
Co
hexagonal
<1000>
<10 0>
Page 32 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
annealed
dislocation density
high
low
grain shape
elongated
equiaxed
yield strength
high
lower
texture
strong
compounds used as permanent magnets, Journal of Magnetism and Magnetic Materials, 173, pp 20-28, 1997.
rob.wallach@cam.ac.uk
Page 33 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
During recovery of a heavily worked alloy, dislocations realign to form subgrains with low angle grain
boundaries within the original but now deformed, strongly elongated grains. During subsequent
recrystallisation, the subgrains with favoured orientations grow and so the resulting annealed alloy
has a more intense and different texture than the original deformed alloy.
Examples:
- fcc alloys with high stacking fault energies include {100} <001>
- bcc alloys {111} <1 0>
The variation with yield stress and texture in a recrystallised Al-Mg alloy is shown in Fig. 5.10. Note
that the highest value of the yield stress is for the sample cut parallel to the rolling direction. The true
strain to failure (elongation) is much greater for the 45 which has a value approximately 25% greater
than for the sample cut parallel to the rolling direction.
[1 1 0]fcc || [1 0 0]bcc
[1 1 0]fcc || [0 1 0]bcc
However, the experimentally observed orientation relationships are irrational and not as simple as
,
above1 2.
Figure 5.11. Orientation relationships between parent (fcc) and martensite (bcc) phases for (a) Bain, (b)
NishiyamaWassermann and (c) KurdjumovSachs paths. Blue atoms indicate a bcc unit cell. The red arrows
indicate part of the motion initiating the transformation. The dashed arrows indicate the invariant direction which
2
is shared by the parent and martensite phases.
Page 34 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
Figure 5.12. (a) Stages in the drawing and ironing of a drink can.
(b) minimal and maximum 'earing' in can bodies as a consequence of anisotropy in the sheet deformation.
(d) Iron silicon transformer steel
The ferrous alloy used for sheet from which electrical transformers are made has to be magnetically
soft in order to minimise energy losses as the magnetic field reverses. This is achieved by controlling
both the composition of the sheet and its crystallographic texture.
An ideal alloy is grain orientated silicon sheet (Goss texture) Fe-3wt.% Si iron since the addition of Si
modifies the equilibrium diagram so that there are no phase transformations on heating, as well as:
- lowering the anisotropy constant K (a measure of the magnetic anisotropy);
- increasing the electrical resistivity which reduces eddy current losses;
- enabling one of two optimal textures (see Fig. 5.13)
cube on face or {100}<100> texture in which a {100} plane lies in the sheet plane, or
cube on edge {011}<100> Goss texture in which a {011} plane lies in the sheet plane which
can increase the magnetic flux density by up to 30% relative to a steel without this texture.
The silicon modifies the Fe-C phase diagram to that shown in Fig. 5.14. This stabilises the ferrite
phase and suppresses the transformation to austenite even at the high processing temperatures.
1 http://aluminium.matter.org.uk/content/html/eng/default.asp?catid=84&pageid=-1941055071
rob.wallach@cam.ac.uk
Page 35 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 5: Anisotropy
Figure 5.13. Cube-on-edge texture and the cube texture in grain oriented silicon steel.1
This avoids phase changes during processing which otherwise result in too small grains in the final
sheet and poorer soft magnetic properties; the final optimal soft magnetic grain size is ~ 5 mm.
A disadvantage of Si additions is that Si raises the ductile-brittle transition to room temperature for
alloys with 4 wt.% Si as is shown in Fig. 5.14. To maintain the single ferrite phase field at lower Si
contents, e.g. between 2 - 3 % Si, the C content needs to be low. Hence a typical composition (in
wt.%) is:
3.2 %Si, 0.03 %C, 0.08%Mn, 0.02%S.
The magnetically favoured Goss texture is produced by secondary recrystallization during high
temperature anneals in controlled atmospheres of hydrogen.
Texture formation is aided by fine particles of MnS which inhibit normal grain growth; the number is
subsequently reduced to avoid domain wall pinning in usage.
A typical manufacturing sequence will be of the form:
- hot roll at 1300C to 2 mm and then remove oxide;
- cold roll to 0.2 mm in two steps with intermediate softening anneal at 800 - 1000C;
- decarburise at 800C in moist H2 which also allows recrystallisation during which the presence of
MnS particles helps to stop excessive grain growth;
- anneal in dry H2 at 1100 - 1200C for several days to form the Goss texture by grain growth.
MnS, having fulfilled its role in the earlier recrystallisation step, also is reduced and this avoids
domain wall pinning in service; the resulting Mn goes into solution;
- shaped by cutting or punching followed by a stress relieving anneal at 800C in dry N2.
Sheets may be coated with MgO before the anneal in dry H2 as this then produces a surface coating
of magnesium silicate which keeps the sheet in tension and so minimises stress magnetostrictive
losses.
1 http://softmagneticalloy.com/soft_magnetic_materials.htm
rob.wallach@cam.ac.uk
Page 36 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 6: Steels
Steels
Stainless steels
Cast irons
---- ferritic
---- austenitic
---- martensitic
--- white
-- whiteheart
-- blackheart
-- HSLA
-- bainitic
-- dual phase
-- martensitic
-- TRIP
-- TWIP
---- ultra high strength
-- secondary hardening
-- maraging
6.2 Alloying in steels - a brief guide (note most elements result in hardenability increasing)
Element
Influence on
ferrite
austenite
Effects on carbides
formation
tempering
Principal function
limited
solubility
stabilises
essential
coarsen
Mn
strengthens
markedly
A3 stabilised
less than Cr
little effect
Si
strengthens
stablises
restricts
forming
Ni
toughens &
strengthens
stabilises
Cr
strengthens
stablises
forms
carbides
resists
softening
Mo
stablises
creep props
promotes
carbides
[> than Cr]
resists
softening
- reduces coarsening
- minimise temper embrittle
ndry
- higher T usage (2
harden)
V, Nb
slight solid
solution
very strong
resists
softening
- reduce coarsening by
ppte pinning of gbs
rob.wallach@cam.ac.uk
carbide formers
- reduce S embrittlement
- cheap way hardenability
- deoxidiser
- oxidation resistance
- use in transformer steels
- stabiliser (stainless steels)
- graphite former
- improve ferrite properties
Page 37 of 61
sion
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 6: Steels
TS (MPa)
Iron whiskers
Scifer
Ausformed (hardened) steel
Martensitic steel
Bainitic steel
Pearlitic steel
Cold-worked C steel
Low C steel
Pure, single-crystal iron
13000
5500
2930
2070
1380
1200
690
340
10
Scifer is a wire made by drawing 10 mm diameter rods of a dual-phase steel with a microstructure of
martensite and ferrite into fine wires of ~ 10 m diameter, a very high true strain in excess of 9. The
very high strength arises from the very fine grain size and a dislocation cell size of ~ 1015 nm.
While commercial Scifer wires have exceptionally high strength, components for many applications
require much larger dimensions and also have to be simple to fabricate. To achieve this:1
- the material must not rely on perfection to achieve its properties: strength can be generated by
incorporating the large number density of defects such as grain boundaries and dislocations, but
the defects must not be introduced by deformation if the shape of the material is not to be limited;
- defects can be introduced by phase transformations, but to disperse them on a sufficiently fine
scale requires the phase change to occur at large undercoolings (large free energy changes);
- a strong material must be tough in order to fail in a safe manner if it should do so;
- recalescence (heat evolved during a phase transformation) limits the undercooling that can be
achieved: therefore, the product phase must be such that it has a small latent heat of formation
and grows at a rate that allows the ready dissipation of heat.
These design criteria are helpful to bear in mind when considering the range of steels shown below.
IF
Interstitial free
(Section 6.3 3)
HS
High strength
HSLA High strength low alloy
(Section 6.4 a)
DP-CP Dual phase-complex phase
(Section 6.4 c)
HSS
High speed steel
ation
TRIP Transform
induced plasticity
(Section 6.4 d)
TWIP Twinning induced plasticity
(Section 6.4 e)
MART Martensitic
Figure 6.1. Strength and elongation for major classes of steel (categories as above)2
1 Bhadeshia H.K.D.H., Bulk nanocrystalline steel, Ironmaking and Steelmaking, 32 (5), pp 405- 410, 2005.
2 www.keytometals.com/page.aspx?ID=CheckArticle&site=kts&NM=207
rob.wallach@cam.ac.uk
Page 38 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 6: Steels
2.
Quench and temper as greater hardenability (more uniform microstructure and less quenching strains)
than plain C steels.
Usage:
Cr-Mo or Cr-Mo-V ferritic creep resistant steel used in power plants (steam generation)
Quench and temper to give alloy carbide precipitate distribution that remains stable at the service
temperatures for which the steel has been designed so many variants.
C 0.1% max., high Ni 5-9% cryogenic applications as good toughness at low temperatures.
3.
Controlled low levels of carbon and nitrogen give very high levels of formability used to fabricate car
body panels.
Interstitial concentrations are <0.003 wt.% C and <0.004 wt.% N such that there is no C or N in solid
solution and this is achieved principally by:
- control of the melt chemistry during steelmaking;
- Ti and/or Nb additions to form precipitates, control of which during hot mill processing ensures grain
size of hot rolled sheet is as small as possible. Precipitates include Ti2CS (carbosulphide), TiC,
TiN, NbC, NbN.
To optimise formablitiy (including deep drawability):
- ensure small grain size in the final cold rolled and annealed product;
- control crystallographic texture and so penultimate rolling of austenite is at temperatures just above
A3 to avoid rolling of ferrite at temperatures where it would recrystallise;
- cold roll ferrite and subsequently anneal to enhance drawability.
Solid solution strengthening is from Mn, P and Si. The absence of interstitials means deformation is
continous, i.e. no yield drop, and no Lders bands which are detrimental to surface appearance.
Page 39 of 61
2013-14
Materials Science Pt II
University of Cambridge
4.
C9 Alloys
Lecture 6: Steels
(a) High strength low alloy (HSLA) or microalloyed e.g. 0.1% C, 0.3% Si, 1.4-2.0% Mn, 0.05% Nb
Microalloyed with small amounts of carbide and/or nitride from elements such as Nb, V.
Precipitates (form in at 1450C) help stabilise fine grain size, plus precipit hardening.
n
delay
final
rolling
Time
Texture strengthening
initial
rolling
Temperature
565
A3
535
Rapid recrystallisation of
Yield stress
505
MPa
Partial recrystallisation of
475
No recrystallisation of
445
Work hardening
Precipitation hardening
Calculated base strength
A1
800
750
700
650
600
Figure 6.2. Fabrication route for high-strength low-alloy HSLA steels and
effect of finish rolling temperature on the final yield stress
(b) Bainitic steels compositions similar to HSLA plus 0.002% B and 0.5% Mo
A wide range of bainitic steels now are available.1 Their strengths can be attributed to their small
ferrite grain size (bainite laths), uniform fine carbide dispersions, high dislocation density, and
solid-solution hardening.
These include high strength and toughness carbide-free bainites, whose microstructures comprise
1
fine plates of bainitic ferrite separated by C-enriched austenite. The advantages include :
- absence of cementite which can initiate fracture so the bainites are more resistant to cleavage
failure and void formation;
- the bainitic ferrite is almost C free, which would otherwise strenghen and hence embrittle it;
- the bainite ferrite plates, and so the effective grain size, are < 1m thick this fine grain size
accounts for the high strength and toughness;
- the ductile austenite films separating the ferrite plates have a crack blunting effect and
additionally can transform locally to martensite as a consequence of the stress field associated
with an advancing crack this is the basis of TRIP steels and also partially stablised zirconia;
- stress corrosion resistance is higher as hydrogen diffusion is slower in austenite than ferrite.
These carbide-free bainites do not require expensive alloying. Sufficient Si is needed to suppress
carbide formation; a typical composition is 0.3 C, 1.5 Si, 1.5 Cr, 0.25 Mo, 3.5 Ni, balance Fe.
1 Bhadeshia H.K.D.H., Bainite in Steels, 2nd ed., Institute of Materials. 2001. Chapter 13 Modern bainitic steels
www.msm.cam.ac.uk/phase-trans/newbainite.html
Page 40 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 6: Steels
Dual-phase steels have a soft ferrite matrix containing islands of martensite as the secondary
phase, the volume fraction (20 - 50%) of which determines the tensile strength.
Produced by holding / deforming at temperatures in the A1 - A3 region followed by quenching.
The properties of dual-phase steels include:
-
low yield to tensile strength ratio (yield strength / tensile strength = 0.5)
strain rate sensitivity (unusually the faster it is deformed the more energy is absorbed)
Due to the strain distribution in the two-phase alloy, there is no yield drop on plastically deforming,
i.e. no Lders bands which are detrimental to surface appearance.
Dual-phase steels are used for body panels, wheels, and bumpers.
(d) TRIP steels
TRIP steels have excellent formability due to their high work hardening rates which can continue to
higher strains (up to 35%) than for dual-phase steels allowing more stretch forming. Their retained
austenite also provides energy absorption by its transformation to martensite in crashes when used for
cars.
TRIP steels have a matrix of ferrite with a significant volume fraction (> 5%) of retained austenite
and small amounts of martensite and bainite. The Si and C contents are higher than in dual-phase
steels to assist stabilising the austenite to below room temperature.
During their heat-treatment, an isothermal hold at an intermediate temperature is necessary in order to
form the bainite. The Si (together with Al) helps to suppress carbide precipitation as the bainite forms
and this also is helped by the austenite which retains C in solution.
During subsequent deformation, the strain at which the retained austenite starts to transform to
martensite is controlled by adjusting the C content:
-
at low C contents, retained austenite begins to transform almost immediately to form martensite
and this increases both the work hardening rate and formability;
at higher C contents, retained austenite is more stable and transforms to martensite only at
strain levels beyond those produced during forming. Hence the retained austenite is part of the
microstructure of the final steel. It is beneficial as it then transforms to martensite if later
deformed, which can be beneficial as this absorbs energy, e.g. in a car crash.
rob.wallach@cam.ac.uk
Page 41 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 6: Steels
solvus
G
ferrite
martensite
Md
Ms
Mf
gamma
Mf
Ms
Md
ln (time)
temperature
temperature
Figure 6.4. Martensite formation at a temperature below the martensite deformation temperature Md
but above the normal martensite start temperature Md.
Ms Martensite starts to form when free energy difference between metastable austenite and
martensite is sufficient to allow transformation (some energy needed to overcome strain and
surface energies).
Mf
Martensite transformation complete: the amount (%) of martensite when Ms < T < Mf depends
only on T, not time.
Md If the temperature is slightly above Ms, the transformation can be initiated by mechanical
deformation as this provides an additional contribution (strain energy).to the driving force for
the transformation which is sufficient to allow martensite to form.
(e) Twinning induced plasticity steels (TWIP)1 15 30% Mn, 3% Si, 3% Al, balance Fe.
The high Mn contents ensure the steel is fully austenitic at room temperature. When plastically
worked, extensive twinning occurs which results in a high work-hardening rate (i.e. a high value of the
work hardening exponent n, see Section 3.4) as the microstructure becomes finer and finer and the
resultant twin boundaries act like grain boundaries to strengthen the steel.
TWIP steels have both extremely high strength (> 1GPa) and high formability (elongations of 50%).
The n value increases to a value of 0.4 at an engineering strain of ~ 30% and then remains constant
until the total elongation reaches 50%. TWIP steels can show elongations of up to 90%
The ductility of TRIP steels is ~ 35% as a consquence of the collective shear of the austenite to
martensite but this then limits further ductility in that the martensite has a rigid bct structure.
The higher Mn, Si and Al in the TWIP steels provides additional ductility since two martensitic
transformations can now occur first the austenite twins to an hexagonal martensite, and this
subsequently transforms to the bct martensite. The twinning is the cause of the additional ductility.
5.
1 www.keytometals.com/page.aspx?ID=CheckArticle&site=kts&NM=207
rob.wallach@cam.ac.uk
Page 42 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 6: Steels
(b) Maraging steels (0.02% C max., 18% Ni, 5% Mo, 7.5% Co, 0.4% Ti)
High alloy content allows martensite to form on air-cooling without the need for a high quench
rate and so residual stresses are minimised.
The low C content means the martensite is relatively ductile and can be formed, and with a low
work-hardening rate there is a low extent of uniform ductility.
Strength is then achieved by precipitation of Ni3Mo (and Ni3Ti) on ageing at 700-800C.
6.
Ferritic creep resistant steels e.g. 2% Cr, 1% Mo, 0.25% V and 0.1% C.
Ferritic steels based on Cr-Mo-V have been developed over many decades for use in power
generation applications where creep resistance is crucial.
Typical compositions range from 1 11% Cr, 0.5 1.5% Mo, ~0.25% V, 0.1 0.2% C.
The steels with lower alloy content typically have allotriomorphic ferrite/pearlite microstructures.
The higher chromium steels have been developed to meet the increased demands of operating the
turbines at higher temperatures to improve their efficiency.
The basis is to form a martensitic structure and temper it extensively to provide a stable microstructure
with a fine distribution of precipitates of the form M23C6, NbC, VN and similar types of precipitate.
7.
Temper embrittlement
Temper embrittlement can occur particularly in some quenched and tempered steels steels and is a
marked reduction in impact toughness (generally shown by an increase in the ductile-brittle transition
temperature) and fatigue properties after heat-treatment; ambient tensile strength is not affected.
There are two principal forms of temper embrittlement characterised by temperature:
(a) tempered martensite or 350C embrittlement or blue
brittleness is irreversible and occurs in the range of
approximately 250400C;
(b) reversible temper embrittlement may occur when steels
are slowly heated or cooled through a temperature range
of 450650C.
Figure 6.6. Temper embrittlement: temperature dependence.1
(a) Tempered martensite embrittlement arises from the nature of the cementite precipitation on prioraustenite grain boundaries or interlath boundaries.
The prior austenite grain boundaries remain during martensite formation and so are where
carbides will heterogeneously nucleate preferentially during tempering. As the carbides grow and
coarsen, cracking and/or void nucleation occurs which leads to a reduction in toughness.
(b) The higher temperature embrittlement is associated with just segregation of the impurity elements
(P, Sn, As, Sb) to the prior austenite grain boundaries. This leads to decohesion at the
boundaries, and again lower toughness subsequently. Since it is segregation, this embrittlement
is reversible by reheating to a temperature at which the impurity elements go back into solution
just below the A1 and then cooling again but at a faster rate so there is insufficient time for
diffusion to the grain boundaries.
The extent of embrittlement is affected by the prior austenite grain size and the hardness of the steel,
a fine-grained and softer steel will be less susceptible. Also, alloying elements such as Cr, Mn and Ni
1 http://steel.keytometals.com/articles/art102.htm
rob.wallach@cam.ac.uk
Page 43 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Lecture 6: Steels
tend to promote temper embrittlement and so it can be prevalent in the ferritic creep resistant steels.
Small additions of Mo are beneficial in reducing embrittlement due to P.
8.
y (MPa)
UTS (MPa)
elongation (%)
1400 - 1800
up to 2400
450 - 800
600 - 1200
HSLA
550
650
10
Dual-phase
370
670
20
150 - 300
400 - 450
up to 25
Maraging
Bainitic
Mild steel
rob.wallach@cam.ac.uk
Page 44 of 61
2013-14
Materials Science Pt II
University of Cambridge
9.
C9 Alloys
Lecture 6: Steels
Stainless Steels
Schaeffler diagram
(a) Ferritic (15-30% Cr, 0.1% C)
Good ductility (so formable) and stress corrosion resistance, cheaper than austenitic stainless
steels but high-temperature mechanical properties poorer. Toughness may be problematic at low
temperatures and in heavy sections.
(b Martensitic (12-17% Cr, 2% Ni max, 0.1-1.0% C)
High hardenability results in martensite on air-cooling from phase.
Often tempered to optimise strength and toughness.
(c) Austenitic (18-25% Cr, 8-20% Ni, <1% C)
Austenitic (fcc) at all temperatures with good corrosion resistance and ductility over wide
temperature range, depending on precise composition.
No ductile-brittle transition so used for cryogenic applications.
(d) Duplex (22% Cr, 5% Ni, 3% Mo, 0.1-0.2% N, 0.03% C)
Formed in / region and controlled rolled to give very fine grain size (can be superplastic).
Similar corrosion resistance to austenitic alloys but better stress corrosion cracking resistance.
Higher tensile and yield strengths, but poorer toughness, than austenitic stainless steels.
(e) Controlled transformation (14-17% Cr, 3-5% Ni, 3% Mn, 0.1-0.3% C)
Precipitation hardened stainless steels with metastable at room temperature so that martensite
forms during fabrication or service deformation (e.g. TRIP).
As martensite forms during deformation, work-hardening rate increases so uniform ductility occurs
before plastic instability.
(f)
rob.wallach@cam.ac.uk
Page 45 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
7. Copper alloys
Pure copper
Good electrical and thermal conductivities, reasonable corrosion resistance electrical applications.
Brasses
<36 wt% Zn
solid-solution strengthening
5 wt%
15
30
34
>36 wt% Zn
cap brass
gilding brass
cartridge brass
yellow brass
plumbing, jewellery
very ductile, deep drawing
major brass, cheapest, workable, screws, tubes
Muntz metal
leaded brass
free machining brasses contain upto 1.5 wt.% Pb to aid cutting without long metal swaths
high tensile
manganese bronze / + ~ 2 % Mn
Bronzes
Sn
<10 wt%
9 Sn
19 Sn
leaded bronze
phase solid-soln strength, corrosion resistant, ductile e.g. 3.5 Sn 1.5 Zn coins
gun metal workable
casting alloy
< 4% Pb
~30% Pb
Bronzes: non Sn
aluminium bronze 4 - 7 Al
7 - 9 Al
Nickel alloys
[see nickel sheet] cupro nickels,
monels,
nickel silver
precipitation hardened (solution heat treat ~850C, age at 450C) nonmagnetic springs,
nonsparking tools
rob.wallach@cam.ac.uk
Page 46 of 61
2013-14
Materials Science Pt II
University of Cambridge
rob.wallach@cam.ac.uk
C9 Alloys
Page 47 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
8. Nickel alloys
8.1 Range of nickel alloys
pure nickel
electrodeposit, electroformed, sintered powder (catalysts, filters, fuel cells)
ferrous alloys
stainless steels
stabiliser
ultrahigh strength
marageing steels
cast irons
Ni-hard (5 Ni, 2 Cr)
non-ferrous alloys
solid-solution strengthening
> 50% Cu
cupro nickels
corrosion resistant
< 50% Cu
monels
improved properties
Ni Cu Zn
c.f. brasses but stronger
decorative, springs, zips
high-temperature alloys
corrosion/oxidation resistant
Ni Cr
Nichrome 80
oxidation resistant
up to 1000C
Ni Fe Mo
Hastalloys
Ni Fe Cr
Inconels
C tolerant
Ni Cr Al Ti
superalloys
ODS
Nimonics
use <1000C
aerospace alloys
magnetic alloys
Ni Fe
40 80% Ni
soft
Al Ni Co
Al Ni Fe
hard
low
Invar (60% Ni)
high
72 Mn, 18 Cu, 10 Ni
superelastic
rob.wallach@cam.ac.uk
constant modulus
Fe-Ni-Cr and Fe-Ni-Co
50% Ti - 50% Ni
shape memory
Page 48 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Historically: the range of alloys available to Whittle in 1937 for his gas turbine was relatively restricted:
- austenitic steel ("Stayblade"), originally developed for steam turbines, used for blades and discs
0.22 C
20.0 Cr
8.5 Ni
1.2 Ti (weight %)
- problems with yielding in service led to replacement by the more complex alloy G18B
0.4 C
13.0 Cr
13.0 Ni
10.0 Co
1.8 Mo
2.5 W
3 .0 Nb
- used for discs in Rolls-Royce Derwent engine (powered Meteor fighter aircraft in early 1940s).
Developments to find alloys stable at higher temperatures resulted in the awareness that:
- face centred cubic (fcc) alloys were more stable than body centred cubic (bcc) alloys;
- larger amounts of solid-solution alloying additions could be accommodated in fcc alloys;
- precipitation or age-hardening (initially interest focused on carbides) was readily achieved.
rob.wallach@cam.ac.uk
Page 49 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Figure 8.1 Increase in turbine entry temperature (TET), by more than 700 K,
as seen in Rolls-Royce civil aero engines.
rob.wallach@cam.ac.uk
Page 50 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Modern aircraft engines are turbofans in which some of the energy produced is used to power the fan
blades which can move a larger mass flow and so provide an increase in thrust and efficiency and a
similar temperature versus entropy diagram is shown in Fig 8.3.
Energy changes are denoted by TdS on the T versus S curves. For both cycles (as well as for the
idealised Carnot cycle where there is no heat exchange with the surroundings), the area enclosed by
the lines on the T versus S curve is a measure of the amount of work energy W. The efficiencies are
given by expressions of the form:
Brayton efficiency
T TC
= H
TH
P
T TC
= 1 C
Turbojet efficiency = H
TH
PH
( 1)/
where is the heat capacity ratio, the ratios of the specific heats at constant pressure and constant
volume = Cp /Cv
Hence the higher the operating temperature, the more efficient is the engine and CO2 emissions also
are reduced. This provides the rationale for increasing the operating temperatures of gas turbine and
the need for materials which can operate at the higher temperatures.
1 http://en.wikipedia.org/wiki/Brayton_cycle
2 http://en.wikibooks.org/wiki/Jet_Propulsion/Thermodynamic_Cycles
rob.wallach@cam.ac.uk
Page 51 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Nickel has a high melting point and is face centred cubic, and so has the potential for
good tensile and creep properties at both room and high temperatures (compared with
bcc alloys in which, for instance, diffusion rates will tend to be higher).
Ni-Cr alloys.
Nichrome (80%Ni - 20%Cr) is still used extensively in heating elements and is resistant
up to approximately 1000C. Alloys of this composition were originally used in a forged
condition in early turbines.
Ni-Cr with Ti and Al. Work on steels led to awareness that improved strength achieved by:
- gamma prime ' ordered precipitates which have a nominal composition Ni3(Al)
- up to 65% of the Al can be replaced by Ti and so ' is normally written as Ni3(TiAl)
- ' precipitates are both
ordered:
providing additional strengthening
coherent:
minimising coarsening at elevated temperatures.
- adjustment of the Ti-Al ratio also affects the coherency with the nickel matrix
- two competing requirements:
increased coherency for temperatures > 0.6 Tm in order to minimise coarsening;
decreased coherency for lower temperatures to maximise the local strains and so
improve strength.
Optimisation of ' leading to Mo, W and Al additions.
- reductions in the Cr content lead to higher volume fractions of '
- but also lead to reduced solid-solution hardening and oxidation resistance
- to maintain
strength: other solid-solution alloying additions added (e.g. Mo and W)
oxidation resistance: Al (4%) increased to form Al2O3 surface oxides.
Additions of Co and C led to higher creep resistance.
Co raises the ' solvus (temperature at which precipitates go into solution in the nickel).
In addition, Co is hcp and can diffuse to and stabilise the stacking faults which exist
between partial dislocations. This Suzuki locking improves creep resistance by making
cross-slip of dislocations more difficult. The same phenomenon occurs in brasses by
Zn segregation to the annealing twins during recrystallisation.
C forms carbides e.g.MC, M23C6 and M6C where M is a metal such as Ti, Nb, Cr, W, Mo.
These are very stable and form at high temperatures both within grains and at grain
boundaries. Boundaries then are pinned so creep resistance is improved.
Since the C is added principally to pin grain boundaries, it is not added to blade alloys but
is used in disc components.
Precise composition depends on the component as this determines both
the required properties and the fabrication practice.
rob.wallach@cam.ac.uk
Page 52 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
rob.wallach@cam.ac.uk
Page 53 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Requirements:
- high tensile strength so fine grain size & optimised precipitate size/spacing
- good ductility (relaxation at stress concentrators)
- high fracture toughness
- fatigue resistance
- oxidation resistance
Factors to consider - operating temperatures ~ 600-700C are lower than turbine blades so
creep is less of an issue
- design based on tensile properties and not creep resistance
- toughness very sensitive to:
defects (need non-destructive NDT evaluation and quality control)
inclusions (minimise by use of clean melts)
-1
The contours on the maps indicate creep rates (strain rate in units of s ). Alloying and the absence of
grain boundaries dramatically reduce the creep rate for the operating conditions under which turbine
blades in an aircraft engine are likely to be used (indicated by the rectangular box in the figures).
Diffusion, and hence creep rates would be further reduced in single crystal turbine blades by the
elimination of grain boundaries.
1 Composition of MARM200 in weight % is 5 Al
2 Ti
rob.wallach@cam.ac.uk
Page 54 of 61
9 Cr
12 W
10 Co
1 Nb
0.15 C
balance Ni
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
9. Non-destructive testing
9.1 Introduction
An increasing number of techniques can be used to examine, without causing damage, the integrity
of an individual component or an assembly, e.g. a completed printed circuit board on which there are
various components. It is carried out both to ensure that standards are being adhered to as well as
for the detection defects within components, either to monitor or to prevent jeopardising service life.
Common NDT methods, apart from visual, for examining individual metallic components include:
liquid dye penetrant
magnetic particle
eddy current
radiography (X-ray & neutron)
ultrasonics
acoustic emission.
When selecting a technique, factors to consider include:
defect location
type of defect
orientation of defect
ease and speed of testing
information required.
Figure 9.1 Dye penetrant. 1. Clean surface. 2. Apply dye penetrant and leave to penetrate any cracks.
3. Remove excess penetrant. 4. Apply developer into which dye from cracks bleeds out. 1.
Advantages:
cheap minimal investment
sensitive to fine cracks
defect location clear
Disadvantages:
surface defects only
need relatively smooth surface
non-quantitative
1 http://brennanndt.co.uk/#/dye-penetrant-inspection/4547126320
rob.wallach@cam.ac.uk
Page 55 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Adapted from1
The disruption of magnetic flux lines where there is a defect is shown in (a). Hence it is ideal to magnetise in two
perpendicular directions, as shown in (b) and (c), to ensure defects in different orientations are detected.
Advantages:
cheap and easy to use
low equipment costs
defect location clear
Disadvantages:
ferromagnetic materials only
defect alignment is critical
non-quantitative
1 www.milinc.com/services/nondestructive-testing-ndt
rob.wallach@cam.ac.uk
Page 56 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
When an alternating current flows through a wire coil, designed as a probe, it generates an oscillating
magnetic field. The probe is then brought near to a metal specimen and so an eddy current is induced
in the metal in its near-surface region, the depth of which is material and test frequency dependent.
The eddy current in the metal metal generates its own magnetic field, and this interacts with the probe
coil through mutual inductance. Any change in the metal at or near its surface, e.g. defects,
thickness, conductiviity, alters the nature and amplitude of the local induced eddy current and the
resulting magnetic field. This feeds back to change the impedance amplitude and phase angle within
the probe coil alters and so allows identification of the changes in the metal; it can be quantitative.
The induced current density is highest very near the surface of the metal and then decreases. The
depth of penetration is usually defined as the depth at which the eddy current intensity I is 37% of
its surface value is Io when using a signal frequency f. This can be calculated from:
I = Io exp
i.e.
The signal intensity, test sensitivity, resolution, and penetration are affected by the materials
magnetic permeability and electrical conductivity , and also by the nature and geometry of the
probe coil. As an example, a signal depth of ~1 mm can be achieved in a mild steel when using a
simple probe and a signal frequency of 500 Hz.
Higher signal frequencies increase near-surface resolution but decrease the depth of penetration.
The probe coil type, and size are selected to optimise test depending on the material being tested
and the purpose of the test.
The magnitude of the signal change can be calibrated using known standards and then this allows
the sizes of defects or coating thicknesses to be determined. The technique also can be used for
high speed applications such as wire drawing or inspecting the surfaces of plates, as the responses
to changes are very rapid. In such applications, differential probe coils are used, i.e. the coils are
wound in opposite directions and so give a null signal except when a discontinuity, such as a crack,
occurs in the metal being examined.
Advantages:
easy to use and automate
-1
high speed usage (up tp 50 ms )
quantitative for some applications
Disadvantages:
limited depth of inspection
orientation of defects critical
1 www.eurondt.com/EDDY%20CURRENT.html
2 www.milinc.com/services/nondestructive-testing-ndt
rob.wallach@cam.ac.uk
Page 57 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
The extent dI by which a material absorbs X-rays depends on its (mass) absorption coefficient .
Io
t
x
dI
= dt
I
I = Io exp t
x
Ix
= Io exp t
If there is a region of thickness x within the sample with a different absorption coefficient x, the final
intensity for X-rays having passed through that region is given by:
Ix = Io exp (t x) x x
Ix I
~ 2%
I
and hence cracks in metals often are difficult to detect. This is especially problematic if fatigue is an
issue as well as for fracture toughness assessments.
To detect a difference, the contrast difference on the film (or recording medium)
X-ray radiography can be used to inspect almost any material for surface and internal defects. The
location of the defects will be shown but not their depths within the sample; the overall thickness of
the sample can be measured.
Advantages:
good for larger defects
inspect virtually all materials
minimal sample preparation
Disadvantages:
thin defects (cracks) not detected
radiation hazard potentially
equipment cost relatively high
Neutron radiography
Neutron and X-ray radiography are complementary techniques as different information is provided
due to the different nature of signal attenuation. X-rays are scattered and absorbed by electrons, and
so higher atomic number atoms interact more strongly. In contrast, neutrons interact with atomic
nuclei. Due to the complexity in generating a neutron signal, the technique is primarily used for
research, especially neutron diffraction for accurate stress analysis.
1 www.ndt-ed.org/GeneralResources/MethodSummary/MethodSummary.htm
rob.wallach@cam.ac.uk
Page 58 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
9.6 Ultrasonics
Basis:
Observe ultrasonic waves reflected from interfaces within a sample.
A piezoelectric crystal transducer is used to both generate and receive back ultrasonic sound waves
which are transmitted through a specimen. Reflection of part of the sound waves occurs from the
interface of any discontuity encountered (including the front and back surfaces of the specimen).
The reflected wave signals are displayed electronically in real time as a graph of reflected signal
intensity versus time elapsed from signal generation to when the reflection is processed. The
reflected signal travel time is directly proportional to the distance travelled by the sound wave in the
specimen. This provides information about the location, size, orientation of any defect encountered.
The velocity v of a sound wave is related to the density of the material by the acoustic impedance Z
Z=V
and ultrasonic signals are a consequence of the reflection at a boundary betwen two materials with
different acoustic impedances.
Ultrasonic testing has excellent penetration and so can be used on thick samples and also even hairline cracks give strong reflections. The signal from the transducer has to be coupled into the
specimen and so relatively good surface finishes are required moreover, a front surface signal will
always be seen so that surface defects cannot be detected.
Where a good surface finish is not possible, e.g. on many fusion welds, a shear wave transducer can
be used as this does not have to be placed directly above the possible location of any defect as is
required for a longitudingal wave transducer (such as is shown in Fig. 9.5 above).
Relating a signal to the nature of a defect is extremely difficult in that reflections from boundaries are
similar regardless of the nature of the defect. A comparision between ultrasonics and X-ray
radiography is shown in Fig. 9.6.
a) incomplete penetration
b) slag inclusions (linear)
c) porosity cluster
d) crack parallel to sidewall
1 www.virtualengg.com/ultrasonic.html
rob.wallach@cam.ac.uk
Page 59 of 61
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
Advantages:
easy to automate
can inspect virtually all materials
single-sided access only needed
hair line cracks detected
Disadvantages:
skilled interpretation needed
thin samples difficult to inspect
2. Consider (a) and treat Pi and Pr as forces (by considering unit area on which they act)
Pi
Pt
Pr
hence Pt = Pi + Pr
[equation X]
and
t=i r
Pt
P P
= i r
Z2
Z1 Z1
Pr
Z Z1
= 2
Pi
Z2 + Z1
Ir
Ii
rob.wallach@cam.ac.uk
P
= r
Pi
and
Pt
2Z2
=
Pi
Z2 + Z1
Z Z1
= 2
Z2 + Z1
Page 60 of 61
[equation Y]
2013-14
Materials Science Pt II
University of Cambridge
C9 Alloys
energy does not have to be supplied to operate acoustic emission; it is a passive device waiting
to record any energy change within a component;
it can monitors any changes occurring in a component in live time rather than measuring the
extent of any change after the change has occurred, providing that the changes are of sufficient
magnitude to be detected.
The technique is not quantitative. Also, given the relatively low signal intensity for many situations,
signal discrimination and noise reduction are required to ensure the signal is valid.
Other similar approaches are now being used and/or are under developed. These include:
-
carbon nanotubes mixed into paints3 changes in electrical resistance can be monitored by
passing an electrical current through the paint. Cracking of the substrate material changes the
nanotubes electrical resistance and the information is transmitted wirelessly and then recorded.
This allows early detection of crack propagation and hence any cracks can be repaired.
1 www.ndt-ed.org/EducationResources/CommunityCollege/Other%20Methods/AE/AE_Intro.htm
2 www.ncl.ac.uk/press.office/press.release/item/989496353
3 www.strath.ac.uk/civeng/research/students/phd/davidmcgahon
rob.wallach@cam.ac.uk
Page 61 of 61
2013-14