CREEP AND

SUPERPLASTICITY
CREEP
 REFERENCE

Engineering Materials 1
M.F. Ashby and D.R.H. Jones
Elsevier

Chapter 20 – 4th Edition

 INTRODUCTION
Creep is slow, continuous deformation that takes
place at high temperatures

Creep begins at a temperature T (in Kelvin)

T > 0.3 – 0.4 Tm METALS
T > 0.4 – 0.5 Tm CERAMICS
T > Tg POLYMERS

Tm = melting point (in Kelvin)

Tg = glass transition temperature
Figure 20.1
A tungsten lamp filament which has sagged under its own weight owing
to creep.
 CREEP DESIGN
In creep design we attempt to guard against creep “failure”
Definition of “failure” depends on application:
• Displacement-limited applications (precise dimensions
must be maintained – e.g. Turbine blades and discs)
• Fracture-limited applications (fracture must be avoided –
e.g pressure piping)
• Stress Relaxation limited applications in which an initial
stress relaxes with time (e.g. pretensioning of bolts and
cables)
• Buckling limited applications in which slender columns or
panels carry compressive loads (e.g. upper skin wing of
aircraft or externally pressurized tube)
Displacement
limited “failure”

Fracture limited
failure

Stress
relaxation

Buckling
 DISPLACEMENT LIMITED “FAILURE”

In such applications we would remove the

component from service before the creep strain
became too high
 MEASUREMENT OF CREEP DEFORMATION FOR
DISPLACEMENT-LIMITED APPLICATIONS
(CREEP CURVES)
STRAIN versus TIME
Experimentally measured at:
Constant Temperature
Constant Load (or Stress)

3 Regions to Creep Curve:

• Primary Creep
• Steady-State Creep (Secondary Creep)
• Tertiary Creep

Steady –state creep rate is the most important parameter in displacement-

limited applications
Gradient indicates speed
of creep
in displacement limited
applications
QUESTION

HOW DOES CREEP RATE VARY WITH

STRESS AND TEMPERATURE?
 STEADY STATE CREEP RATE (dε/dt)ss
The gradient of the curve in the steady state regime

A measure of the creep rate of a

component.

For many applications creep

fracture is unlikely, but
creep needs to be
displacement (i.e creep strain)
limited
i.e we need to know how fast
creep is taking place to predict
the amount of creep strain
after a given time Slope becomes steeper
(higher creep-rate) at
higher stress levels
EFFECT OF STRESS ON SECONDARY CREEP-RATE
MEDIUM – HIGH STRESSES

Experimentally found that:

 d 
   B
n

 dt  ss

At CONSTANT TEMPERATURE
Where:
n = creep exponent
(usually 3 – 8)

POWER LAW CREEP or

DISLOCATION CREEP
 POWER LAW CREEP (DISLOCATION CREEP)
AT MEDIUM-HIGH STRESSES

At Medium-High Stresses
 d 
   B
n

 dt  ss

at Constant T

 d 
log    log B  n log 
 dt  ss
 DIFFUSION CREEP
At LOW STRESSES

Occurs at Low Stresses

 d 
   k
 dt  ss
 EFFECT OF TEMPERATURE ON STEADY STATE
CREEP-RATE
At Constant Stress
 
d   Q 
   Ce  RT 

 dt  ss
 d  Q
log e    log e C 
 dt  ss RT

R = Universal Gas Constant

(8.314 J/K/mol)
Q = Activation Energy for Creep
HIGH LOW
TEMPS TEMPS
POWER LAW (DISLOCATION) CREEP EQUATION
Combining
 d 
   B
n

 dt  ss
and   Q 
 
d 
   Ce  
RT

 dt  ss

 
d   Q 
   A e  RT 
n
Gives:
 dt  ss

Where: A = Creep Constant

 MECHANISM OF POWER LAW CREEP
(DISLOCATION CREEP)
Mechanism at Medium
-High Stresses Dislocations
Creep occurs by dislocation
movement
For continuous dislocation
movement Climb is
necessary -which involves
Diffusion
Rate of diffusion (D) is given
by: Q 

RT 
D  Do e 
 REMEMBER: DISLOCATION MOVEMENT CAUSES DEFORMATION
When a dislocation moves it makes the material above the slip plane slides
relative to that below

(a) Initially perfect

crystal

(b) – (d) the passage

of the dislocation
across the slip plane
shears the upper part
of the crystal over
the lower part by the
slip vector b; when it
leaves, the crystal has
suffered a shear
strain
 MECHANISM OF DIFFUSION CREEP
Occurs at Low Stresses

At Low Stresses Creep is not dependent on the movement of

dislocations
Diffusion is solely responsible
Atoms move to grain boundaries normal to the stress direction
This results in elongation of grains

Grain size (d) increases → Creep rate decreases

 d   k   RT 
Q
    2 e
 dt  ss  d 
 WHAT IS DIFFUSION????

Diffusion occurs by atoms exchanging positions with

vacancies.
(vacancies are “holes” in the crystal structure)
 DIFFUSION IS FASTER AT GRAIN BOUNDARIES THAN
WITHIN CRYSTALS (GRAINS)

Figure 21.9
Grain-boundary diffusion.
Faster diffusion occurs at grain boundaries because of the more
open structure
Original grain shape =
Solid Line
After Creep - grain shape
is dashed line
Note Extension of Grain
in direction of stress
Creep Elongation of Grains (and therefore the Material) during Diffusion
Creep
 SUMMARY OF METAL DEFORMATION

At low stresses and temperatures materials deform by

Elastic deformation

At higher stresses materials deform by Plastic deformation

If the temperature is high enough by:

o Diffusional Flow (low stresses) or
o Power Law Creep (medium to high stresses)

This is summarised by Deformation Mechanism Maps

 DEFORMATION MECHANISM MAPS

These maps show the range of stress and

temperature in which we expect to find each of
the deformation mechanisms.

They also show the strain rate that each

combination of stress and temperature produces

Diagrams like these are available for many metals

and ceramics and are a useful summary of creep
behaviour helpful in selecting a material for high
temperature applications
A deformation mechanism map showing the regime in which each
mechanism operates
EXAMPLES OF REAL CREEP DEFORMATION MAPS
Example given in next two slides is that for pure nickel

Fig (a); 1 mm grain size

Fig (b) ; 10 um grain size

As the grain size becomes smaller the material has more grain
boundary area.
Boundary diffusion (transport) becomes more important
The area of the creep map taken up by Boundary Diffusion
(Transport) increases

The areas taken up by Bulk Diffusion (Transport) and by Power

Law Creep decreases
Power Law
Creep
Power
Law
Creep
As long as the operating conditions (stress-temperature) stay within the
Power law creep field the strain rate stays the same

However at small grain sizes as the power law field contracts and the
operating conditions move into the diffusion field the creep rate may change
dramatically.

Example

Consider Nickel 1 mm grain size

Take Shear Stress = 101
Temperature 600 oC
Strain Rate indicated (Power Law Creep) = 10-8 s-1

Consider Nickel 10 μm grain size

Shear Stress = 101
Temperature 600 oC
Strain Rate (Diffusional Flow) = 10-6 s-1
100X Higher Strain Rate
 APPLICATION OF CREEP MAPS TO TURBINE BLADES

FIGURE 13.23 The stress and temperature profiles for a turbine blade.
 COMPARISON BETWEEN PURE NICKEL AND NICKEL
SUPERALLOYS

Here we see the effect of a Nickel Base Superalloy (MAR M200) in

shrinking the power law creep field compared to pure nickel

The Superalloy has been developed so that it is resistant to both Power Law
Creep and Diffusional Creep

For Pure Nickel (Grain size 100 μm)

Typical Turbine Operation is in Power Law Creep Field with potentially very
high strain rates

For Superalloy MAR M200 (same Grain Size 100 µm)

Turbine operation is in Diffusional Flow field with much lower strain rates

For Superalloy MAR M200 (larger grain size of 1 cm)

Turbine operation remains in Diffusional Flow Field with almost zero Strain
Rate
FIGURE 13.24 (a) The stress–temperature profile of the blade plotted onto a deformation
map for pure nickel. (b) The same profile plotted on a map for the alloy MAR-M200.
CREEP FRACTURE
 CREEP DAMAGE AND FRACTURE

In some applications (particularly those that that

see extreme conditions) a component needs
to be designed against complete failure

Failure takes place by the formation and growth

of voids on transverse grain boundaries
The Creep Lifetime of Turbine Blades has been increased by removing all grain
boundaries (a source of damage at high temperatures)
 CREEP RUPTURE TESTS

Measure creep life

(Time-to-failure under Creep)
Stress–rupture time data for Type 304 stainless steel.
 EXTRAPOLATION OF CREEP RUPTURE DATA -
PREDICTION OF SERVICE LIFE UNDER CREEP CONDITIONS

If a component is to have a design life of 10+ years it is

impractical and costly to carry out creep tests for this
length of time (and have to wait 10+ years for the
creep data)

We need to carry out ACCELERATED TESTS

One approach by designers is the use of a Larson-Miller
plot
Larson-Miller plots of materials to be used for creep
applications are constructed from experimental
measurements
CONSTRUCTION AND USE OF LARSON-MILLER
PLOTS (SUMMARY)

TO CONSTRUCT THE LARSON-MILLER PLOTS

Experiments determine creep failure time at very high
temperatures (higher than service temperatures so as to give
an accelerated failure time) and for a given stress level is
measured
A Larson-Miller parameter may then be calculated (see equation
on following slide) at a given stress level from these high
temperature accelerated test experiments
 CONSTRUCTION OF LARSON-MILLER PLOTS
Larson Miller Parameter calculated from experimental creep
rupture data

Larson-Miller Parameter (is constant for a given material at a

given stress level)
T(C + log t) = CONSTANT (at a given stress)

T = Temperature (Kelvin)
t = Measured Time to failure at a particular temperature and stress level

USE OF LARSON MILLER PLOTS

Stress versus Larson-Miller Parameter
At the Service Stress level read off the Larson-Miller parameter
Then Calculate the failure time at the service temperature
 LARSON-MILLER PLOT
Obtained from accelerated tests at high Temps and given stresses

To Calculate Failure Life t at

700oC (973 K) and Stress of
100 MPa

Read off graph at 100 MPa

LMP = 25 (x 103)

T(20 + log t) = 25 x 103

973(20 + log t) = 25 x 103

Log t = 5.694

t = 494311 hrs

t = 56 years
FIGURE 13.20 Materials to resist creep, for each regime of temperature; below
room 0°C, the issue is brittle fracture, not creep.
 SUPERPLASTICITY
AND SUPERPLASTIC
FORMING
 INTRODUCTION
We have seen that for creep resistance at high temperatures we would
like our materials to possess large grain sizes

However, where the opposite of creep resistance is required (i.e. easy

plastic flow), such as in hot forming processes, fine grained processes
are preferred.

Under certain conditions certain fine grained materials can have,

what is known as Superplastic Behaviour.

In Superplastic Behaviour elongations may exceed 1000%

These large strains at low stresses drastically improve the formability of

certain alloys.

In this lecture, we will see that like Creep, Superplastic Deformation is

also controlled by Diffusion
  DUCTILITY IN TENSION

Under normal circumstances tensile elongation is usually limited by the

development of a neck

Let us look first therefore at what factors control the formation of a neck.

In considering superplasticity we are interested in tensile deformation at

high temperatures.

Consider high temperatures where T (in Kelvin) > 0.5 Tm

At such temperatures there is little work hardening and the flow stress is
more dependant on the strain-rate than the strain

The flow stress at high temperatures is given by:

= C eqn (1)
FEATURES OF A METAL STRESS-STRAIN CURVE FROM A
TENSILE TEST (DUCTILE TENSILE FAILURE)
AT ROOM TEMPERATURE
Stress –Strain curves at high temperature
Note the lack of work hardening
- Indicates that increasing the strain rate increases the stress level of plastic
flow
 Consider a specimen of uniform cross-section strained at a strain-rate
dε/dt.

The incremental strain dε in stretching the specimen is given by:

Since plastic flow (deformation) occurs at constant volume

(i.e. Volume = A.L = constant)
The strain is also equal to the fractional change in cross-sectional area
A

- (negative because area reduces)

Differentiate wrt time

 

And from equation 1: = C

Thus:

Next, let us express σ as F/A, where F is the force on the

specimen

Rearrange the equation:

 

As m approaches 1, then the rate of change of cross-section

with time dA/dt becomes less dependent on area

At m = 1: = A0 = 1

dA/dt is then independent of area A

All cross-sectional areas will shrink at the same rate

This means that necks cannot occur

More common is that m becomes close to 1 and the material

can then be given large plastic strains (over 1000%) before
necking
The closest m comes to 1 is in hot glass at high
temperatures

This accounts for the glorious artistry that can be exhibited

in glass blowing

Some hot polymer melts can also approach this condition

 SUPERPLASTIC DEFORMATION IN METALS

High values of m, and hence superplasticity, can be obtained

only with certain microstructures and good process control:

o Grains must be equiaxed (rounded)

o Grain size must be small (1 -10 µm)

o Deformation temperature must be high > 0.5 Tm

o m must be high

.
 HOW DOES SUPERPLASTICITY TAKE
PLACE ?

Superplasticity has some strange features

There is no evidence of deformation in the microstructure

afterwards – grains and other microstructural features do
not become elongated as in ordinary plastic deformation.

Several proposals have been made

Latest theories suggest that movement takes place by the

sliding of grains past each other controlled by diffusion
processes.

One researcher likened this to the movement of sand

grains in an egg timer
 PROCESSING CONDITIONS FOR
SUPERPLASTICITY

The ease of superplasticity depends on the

temperature and imposed strain rate

Next slide plots: LOG STRESS versus LOG STRAIN-

RATE

Note that m is the slope of this graph

The following slide plots: m versus LOG STRAIN RATE

The graphs show that maximum values of m (=0.9 in

this case) occur when Diffusion mechanisms dominate
Plot of log STRESS versus log STRAIN-RATE
m
The strain rate associated with superplasticity can be pushed to higher strain
rates by either decreasing grain size or increasing temperature
 SUMMARY

o Temperature must be above 0.5 Tm to allow Diffusion to

dominate

o Strain-rates must be neither too high or too low to

maximise m values

o The grains must be small (< 10 µm) to keep diffusion

distances small

o Grains must be equiaxed to allow easy grain “switching”

o Grains must not grow during processing

 SUPERPLASTIC MATERIALS

Superplasticity has mainly been exploited by the aerospace

industry and there are superplastic variants of all main
aerospace alloy classes:

Nickel Base Alloys; Aluminium alloys, Titanium alloys

The alert student may realise that the small grain required for
superplastic forming is not compatible with large grain sizes
required for creep at high temperature.

Superplastic alloys have been developed that can be

subsequently heat –treated to give a large grain size for high
temp service applications.
 SUPERPLASTIC FORMING APPLICATIONS

The relatively low strain rates required for superplastic forming mean that
unlike most other forming applications it cannot be used for mass
production
The often expensive tooling required means that it is also unsuitable for
one off parts.

Its main applications to date have been in aerospace to form complex

shaped panels.

High pressure inert gas is often used to carry out the shaping process

It becomes economical when used for single part products that

would otherwise have to be fabricated from multiple parts

It is possible to produce hollow panels

It is also possible to produce complex panels in which superplastic
forming is combined with diffusion bonding to form structures that
combine 2 or more original sheets
Here we see complex hollow panels produced
Combined Superplastic Forming and Diffusion(High Temperature)
bonding (between the green and brown sheets above) has allowed the
complex hollow panel to be produced