Structural Integrity Assessment Considering Creep, Stress Rupture and Creep Crack Growth Mechanisms

Download as pdf or txt
Download as pdf or txt
You are on page 1of 76

Structural integrity assessment considering creep,

stress rupture and creep crack growth mechanisms

Dr. Mahendra Kumar Samal


Reactor Safety Division
Bhabha Atomic Research Centre, Trombay, Mumbai
Professor (Engineering Sciences), Homi Bhabha National Institute (HBNI), Mumbai

13th March 2034

HBNI-CEA joint course on ‘Structural Integrity Assessment’


Outline of the lecture

• Introduction
• Characterization of creep deformation
• Micro-mechanism of creep
• Creep testing and extrapolation techniques
• ASME Section-III, Sub-Section NH
• Creep fracture mechanics
• Example of calculation of creep crack growth rate
Introduction
• Creep is defined as the tendency of a solid material to slowly
move or deform permanently under the influence of stresses

P
load

load
t=0 t>0

displacement displacement
P

Instantaneous Time dependent


elastic creep
deformation deformation
Nature of Creep
• Creep occurs as a result of long term exposure to stress < YS

• Creep is insignificant below 0.3Tm

• The rate of creep deformation is a function of:


-material microstructure
-exposure time
-exposure temperature
-applied stress

• Unlike brittle fracture, creep deformation does not occur


suddenly upon the application of stress

• Creep strain accumulates as a result of long-term stress.


Creep is a "time-dependent" deformation
Nature of creep deformation
• Total strain as a function of time is important for components
requiring dimensional tolerance and stability
• For shell-nozzle junctions, creep rupture or crack growth may
be more important
• Secondary creep rate or steady state creep rate is the limiting
factor for creep life design

Creep rupture Primary creep


Tertiary creep

Strain rate (/sec)


Tertiary creep
Total Strain

Secondary creep
Secondary creep

Primary creep

Instantaneous creep
Time (Hrs) Time (Hrs)
Grain boundary and inter-granular cavitation

Vector of deformation due to


dislocation based creep is
along different directions near
grain boundary  Grain
boundary sliding
Grain boundary sliding
Micro-structural aspects of creep deformation
• Creep damage is mainly inter-granular
• At low temperature, the grain boundary is stronger compared to
the grain
• At higher temperature, grain boundary is weak
• Weakening of grain boundary and grain boundary sliding leads to
creep void formation
• Crack advance takes place intermittently by the linking up of
micro-cracks and voids that are generated on grain boundaries
ahead of the main crack tip, to the crack tip
Micro-mechanism of creep deformation
• Homologous stress is defined as the ratio of applied stress to
shear modulus
• Homologous temperature defined as the temperature in Kelvin
to the melting point of the material
• Deformation mechanism map helps to identify the dominant
mechanism of creep deformation and hence to develop new
alloys for modern applications
1.0e-1
Plastic deformation
1.0e-2
Homologous stress

Dislocation
creep Bulk
diffusion
1.0e-3
Elastic

Bulk diffusion
deformation
1.0e-4 Boundary
diffusion

1.0e-5
0 0.5 1.0
Homologous temperature
Dislocation creep
• Movement of dislocations becomes easy at higher temperature and
high stresses due to thermal and athermal type mechanisms
• Vacancy-assisted dislocation motion becomes the dominant at higher
temperatures
• Dislocations tend to move through climb process rather than glide

Shear stress 

 b tan 
 Glide plane
b

Precipitate

Precipitate-assisted
Shear stress  dislocation climb
Diffusion-assisted creep
• Diffusion of atoms becomes dominant at higher temperatures
• Atoms move in the direction of application of tensile stress
• Vacancies move in opposite direction of tensile stress
• Boundary diffusion is proportional to 1/D2

• Bulk diffusion is proportional to 1/D 3

Boundary
diffusion

Nabarro-Herring model for  Bulk 


diffusion
diffusion creep


Creep damage through time-fraction rule
• The creep damage can be modeled by a parameter, which is the
ratio of effective area to the initial area at any time of creep
deformation.
• The effective area reduces because of creep-cavitation and
cracking.

Creep damage is defined as : 1  

Aeff

A0

 0  F A0  t  F Aeff   0 
Derivation of time-fraction rule
Derivation of time-fraction rule (cont..)
Derivation of time-fraction rule (cont..)
Extrapolation of creep lives from short-term tests
• It may not be possible to determine creep rupture life in actual
stress and temperature conditions by laboratory test
• Accelerated tests are conducted
• Creep life is extrapolated from short-term lab-based tests
• It should be ensured that the mechanism of creep deformation
does not change

Larson-Miller parameter
Polpular parameters for extrapolation
• Orr-Shorby-Dorn parameter
• Manson-Harferd parameter
• White-Lemay parameter
• Monkman-Grant law
Comparison of creep resistance of materials

Larson-Miller
parameter for
comparison
Example-1
• A high temperature component initially designed for 3E4
hours of life at 1000oF with a stress of 12000 Psi. After 5E3
hours of operation, a malfunction subjects the component to a
stress of 15000 Psi at 1100oF. What is the predicted remaining
life?
• If the malfunction is corrected after 100 hours and the
operation is returned to 12000 Psi and 1000oF, what will be the
remaining life? Use Larson-Miller parameter. Use a suitable
material for the components.
Solution:
Let us take the material as ASME A335 P22 (2.25 Cr 1Mo Steel)
From the data:
LMP(avg)=R0+R1*σ+R2*σ^2+R3*lnσ+R4*exp(-σ)
R0=4.56112e1; R1=-2.82477e-2; R2=2.6754e-4; R3=-3.65748;
R4=-9.05122; C_LMP=20
Example-1 (cont..)
Rupture time=10^c1
c1=1000*LMP/(TMP+460) – C_LMP
TMP=1000 oF
σ = str = 12 ksi
LMP=R0+R1*str+R2*str^2+R3*log(str)+R4*exp(-str)
= 36.2222
c1=1000.0*LMP/(TMP+460.0) – C_LMP
= 4.8097
Tr=rupture time=10^c1= 6.4525e4 hrs
The designer has designed it for 3e4 hrs.
Factor of safety = 2.1508
Example-1 (cont..)
Damage accumulated after 5e3 hrs
= 5e3/ 6.4525e4 = 0.0775

Abnormal operation:
TMP=1100 oF
σ = str = 15 ksi
LMP=R0+R1*str+R2*str^2+R3*log(str)+R4*exp(-str)
= 35.3430
c1=1000.0*LMP/(TMP+460.0) – C_LMP
= 2.6558
Tr=rupture time=10^c1= 452.6829 hrs
Example-1 (cont..)
Damage accumulated during abnormal operation
= T_op/ Tr = 100/ 452.6829 = 0.2209
Previous damage = 0.0775
Total damage = 0.0775 + 0.2209 = 0.2984
Remaining life fraction= 1- 0.2984= 0.7016

Remaining life calculation (operation with 1100 oF)


Allowable time with factor of safety
= rupture time * remaining life fraction / factor of safety
= 452.6829 * 0.7016/ 3.0
= 105.8674 hrs

Total life = 500 + 100 + 105.8674 = 705.8674


Original designed life was 3e4 hrs. % reduction = 97.65 %
Example-1 (cont..)
Remaining life calculation after returning to 1000 oF:

Damage accumulated during abnormal operation


= T_op/ Tr = 100/ 452.6829 = 0.2209
Previous damage = 0.0775
Total damage = 0.0775 + 0.2209 = 0.2984
Remaining life fraction= 1- 0.2984= 0.7016

Operational parameters restored back:


TMP=1000 oF
σ = str = 12 ksi
LMP=R0+R1*str+R2*str^2+R3*log(str)+R4*exp(-str)
= 36.2222
c1=1000.0*LMP/(TMP+460.0) – C_LMP
= 4.8097
Tr=rupture time=10^c1= 6.4525e4 hrs
Example-1 (cont..)
Allowable time with factor of safety
= rupture time * remaining life fraction / factor of safety
= 6.4525e4* 0.7016/ 3.0
= 1.5090e4 hrs

Total life = 500 + 100 + 1.5090e4 = 15690 hrs

Original designed life was 3e4 hrs. % reduction = 47.7 %


Codes for high temperature design
ASME B&PV Code, Section III, Subsection NH
• Subsection NH of ASME B&PV Code, Section III, provides
high-temperature design rules for construction of Class 1
components having metal temperatures exceeding those
covered by the rules and stress limits of Subsection NB
Modes of failure considered in design
1) Creep rupture under sustained primary loading
2) Excessive creep deformation under sustained primary
loading
3) Cyclic creep ratcheting due to steady primary and cyclic
secondary loading
4) Creep-fatigue due to cyclic primary, secondary, and peak
stresses
5) Creep crack growth and non-ductile fracture
6) Creep buckling
•First step: To ensure that components are safe for low
temperature operation.
•Introduction of additional time-dependent failure modes.
•Rules to protect against these failure modes: elastic or
elastic-plastic analysis techniques
Creep rupture data
• Stress vs. time to rupture data for Alloy 800H
• Adequate safety margin is used to derive the minimum line
Primary Stress Limit for Base Metal
•In addition to the time-independent Sm, the code introduces a
temperature- and time-dependent quantity St to account for creep
effects
•For each specific time t and temperature T, St for the base metal
is defined as the least of the following three stresses:
(1) 100% of the average stress required to obtain a total
(elastic, plastic, primary creep, and secondary creep)
strain of 1%;
(2) 80% of the minimum stress to cause initiation of tertiary
creep; and
(3) 67% of the minimum stress to cause creep-rupture
Definition of Smt
• A basic primary stress limit for high temperatures is Smt, which
is the lesser of Sm and St
• Smt is a function of both time and temperature
• The definition of St assumes: material has a classical creep
curve
• For materials having non-classical creep curve (e.g., nickel-
based alloys) with no clear primary creep or secondary creep
regime, this definition is not applicable
Allowable stress in ASME for 2.25Cr1Mo steel
Limits for primary membrane stress
• Service Level A and B loadings
Pm  Smt
• Service Level C loadings
In addition, the time-based
1.2Sm
Pm   usage-fraction sum
 St associated with the general
primary membrane stresses
• Service Level D loadings for all levels of loadings shall
 2.4Sm be less than 1
 0.7 S (For variables stress loading

Pm   u
conditions)
 0.67 Sr
0.6 RSr
Limits for Primary Membrane and Bending
• Service Level A and B loadings
PL  Pb  KS m
Kt accounts for relaxation of
PL  Pb K t  S t
extreme fiber bending stress
• Service Level C loadings due to creep and is known as
PL  Pb  1.2 KS m creep bending factor
PL  Pb K t  S t
In addition, the time-based
• Service Level D loadings usage-fraction sum
 3.6 S m associated with the general
PL  Pb  
1.05S u primary membrane stresses
 0.67 S r for all levels of loadings shall
PL  Pb K t   be less than 1
0.8 RS r
(For variables stress loading
conditions)
Creep bending shape factor
• Bending stress relaxes due to creep at the extreme fibres
• More the value of creep stress exponent ‘m’, faster the relaxation
• Assuming a simple Norton's power law creep, it can be shown that
Kt is conservative for solid rectangular sections as long as the creep
exponent is greater than 3.

K t  0.5  K  1
K (≤1.5) represents the
ratio between the loads
to cause fully plastic
section and initial
yielding in the extreme
fibers of the section.
For shells and solid
sections, K = 1.5.
Remarks on conservativeness of factor Kt
• For the uniformly loaded circular plate, the ratio K can be as
high as 2.66. Hence, Kt is 1.83. Code allows a maximum value
of Kt as 1.25
Damage due to creep fatigue interaction
• The combination of Level A, B, and C loadings shall be
evaluated for accumulated creep and fatigue damage,
including hold time and strain rate effects.
1.0

SS 304 and SS316

 n   t  Creep damage
p q

       D
j 1  N d  j k 1  Td  j
2.25Cr-1Mo Steel and
0.3 Alloy 800H

0.1
0
0.1 0.3 1.0
Fatigue damage
Elastic analysis route in NH
• NH uses isochronous stress-strain curve instead of strain-time
creep curve
• The isochronous stress-strain curve is basis to the elastic
analysis route
Summary of elastic analysis route
Creep deformation of cracked bodies
Steady state creep and concept of C* integral

Strain energy
 .  release rate
 ui 
C *   W *dy  Ti  ds
 x 

 
where W *is the stress power
.
and given by W    ij  ij
*

Ti is the traction vector along


the contour  of differential
length ds and ui is the
velocity vector
Crack tip stress and strain field in terms of C*

Small scale creep


Characteristics of C* integral
• It is a path independent integral, which can be computed
along contours remote from the crack tip

• It can be measured experimentally from load-velocity


diagram

• It can be uniquely related to the magnitude of crack-tip stress


and strain rate

Thus, it is an attractive parameter for characterizing the


creep crack growth rate under conditions of extensive
steady-state creep and there is considerable interest in
methods for determining its value
Energy Rate Interpretation of C*
Analogous to energy release rate interpretation of J-integral
except the introduction of velocity instead of displacement
Methods of Determining C*
(i) Experimental method
(ii) Numerical solutions
(domain integral)
(iii) Estimation schemes
(EPRI)
Experimental method
Expressions for C* integral
By analogy to the method for estimating the J-integral,
expressions for estimating C* can be obtained
Compact
tension
specimen

Centre
cracked
specimen

Similar expressions can be used for other types of


specimens or components such as pipes based on analogy
of the J-integral expression
Example-1
Pr: A 50mm-wide and 25mm-thick CT specimen of SS304 is
subjected to a constant load of 18kN at 594oC. If the A
and n values of this material at 594oC are 2.13 e-18 and 6,
respectively (for stress in MPa and strain rate in 1/hr),
calculate the value of C* when a/W = 0.5 and the
measured load-line deflection rate is 2.5e-6 m/hr.
Solution: The value of C* can be obtained by the use of
semi-empirical method as we know the steady state
deflection rate and the load which is constant.
C* solution (EPRI)
Such equations are necessary for estimating C* in components where no
measurements of the load-line deflection rate are available. As with the
J-integral, these solutions are restricted to either plane-stress or plane-
strain conditions.

EPRI
solution

Analogy of
C* with Jp
solutions
Example-2
Pr: Show that (for CT specimen)

Find its accuracy for a/W values of 0.5 and 0.625 for n of 7.
Solution:

a/w=0.5
a/w=0.625
Reference stress estimation of C*
Reference stress

Evaluation of C* integral from analogy of J-integral


J-integral: Energy release C*-integral: Stress power
rate ahead of crack tip release rate ahead of crack tip
Stress intensity factor (SIF)
Relation of C* with reference stress

K2
J G  Yref2  ref  ref a  href  ref  ref a where
E

This expression is used when information on strain rate


is available at the loading condition
Small scale and transient creep
• The validity of the C*-integral is limited to extensive
steady-state creep conditions.

• In practice, this condition may not always be realized as


the components contain stress and temperature gradients
and are designed to resist widespread creep deformation.

• The crack tip stress fields for these conditions of small-


scale creep (SSC) and the transition creep (TC) will not
be governed by C*.

• elastic stress redistributions due to creep deformation


needs to be considered.
C(t) integral
Bassani and McClintock recognized that the crack tip stress
fields under SSC can also be characterized by a time-
dependent C-integral, whose value is determined along a
contour taken very close to the crack tip.

•Note that C(t) is the same as C* but its value is determined


close to the crack tip within a region where the creep
strains dominate over the elastic strains.

•Thus, determining the C(t)-integral requires accurate


solutions of stress and strain near the crack tip.
C(t) and C* integral
Steady state creep crack growth rate

a  D0C *
where D0 is a material constant and is a constant slightly
less than 1 (for universal creep crack growth law, it is 0.85)
While stress redistribution due to creep is taking place at a
crack tip, C* is not path independent and it is more
appropriate to describe the local stress field in terms of the
parameters C(t) or Ct integrals
C(t) is evaluated from equation similar to that of C*, but for
a contour that has shrunk to the crack tip. Prior to stress
redistribution C  t   C * , but steady state is reached when
C t   C *
Estimate of stress redistribution time
A conservative estimation of stress redistribution time is:

G G is elastic strain energy release rate


tred 
 n  1 C * n is the creep strain rate hardening
exponent
Ct integral has been developed in literature for dealing
with small-scale creep

Creep crack growth can be estimated in case of transient or


small scale creep as:

a  D0 C (t ) or a  D0 Ct 


 
Experimental data

For 304 SS at 594 deg. C


Universal creep crack growth law

Plain stress Plain strain

Creep crack growth in an Creep crack growth in a carbon steel


aluminium alloy using DCB using circumferentially cracked
specimen cylinder
Experimental determination of C* integral
ASTM E-1457-92:
standard test method for determination of C*
integral and creep crack growth rates
Use of standard compact tension specimen
High temperature furnace with temperature control
Crack mouth opening displacement is measured using high
temperature clip gauges
Crack growth is measured using a suitable technique such
as alternating current potential drop (ACPD), direct current
potential drop (DCPD) techniques, ultrasonic methods,
compliance method or optical method
Furnace with
temperature
Experimental determination of C* control
instrumentation
Load P

Furnace with
temperature
control
instrumentation
Experimental determination of C*

Displacement in mm
Load P

Displacement  Crack length a in mm


Calculation of creep displacement rate
The creep displacement rate is calculated from the total
displacement rate as follows.
. . .
  e  c
.
Pa . dC e
where  e 
EBW d  a / W 
C e is the compliance of the specimen
(function of crack growth)
.
 e can also be written in terms of SIF "K" as
.
2 a BK 2
.
e 
EP
Calculation of creep displacement rate

If there is significant plastic deformation:


the effect of plastic zone size has to be taken
into account in the calculation of K
(using effective crack length)
.
Creep deflection rate  c is calculated as:
. . .
c    e
Analysis of data to obtain C*
Calculation of C* integral is analogous to J-integral (ASTM
method)
.
P c
C 
*
f a /W 
Bn W  a 
where Bn is the net thickness of the side-grooved specimen.
n
f a /W   p
n 1
where  p =1 for Centre Cracked Tensile specimen
=2 for Three point bend specimen
=2+0.522 1-a/W  for compact tension specimen
Determination of crack growth constants

a  D0  C 
*  The material parameters to be determined
by regression analysis of the data set
ASTM validity criteria for
The parameter C* is valid only if certain ASTM validity
criteria is satisfied
Attainment of steady state creep condition
during the experiment
Elastic deformation rate should be negligible
compared to the creep deformation rate
.
c
ASTM validity criteria if .
 0.8, the data is fully valid
. 
c
 0.5 to 0.8
. .
(i ) c c
.
if  0.5 and  0.5
 . .
 
it not fully valid and additional
criteria has to be satisfied
These are given below.
ASTM validity criteria for C*
K 2 1  2 
(ii) t 
E  n  1 C *
This condition means the time at which C* is being
determined should be sufficiently greater than the transition
time so. that steady state is achieved
a  ref
2

(iii) *
 0.5
EC
This condition is to ensure that the elastic deformation
owing to crack growth is small compared to the creep
deformation
This condition is analogous to the condition stated in
criteria-(i)
ASTM validity criteria for C

(iv) Bn , W - a  , a  25

This condition is to ensure that plane strain conditions exist


ahead of the crack tip
However, if the component thickness is same as the
specimen thickness, this criterion can be neglected
Example-1
SENT specimen P / BW  300MPa

W  100mm
a  20mm
Temperature : 360 deg.C
Find the time for the crack to a
propagate through thickness B
W

Estimate reference stress


 a a a 
2 3

PLc  1.155 y BW 1   1.232      


 W  W   W  
 a     
a
2
a
3
Handbook
 ref  0.866  P / BW  / 1   1.232      
 W  W   W   solution
Example-1 (Cont..)
 y  400MPa
 ref  342MPa

Calculate reference strain rate from Norton’s power law


for creep
  
n

 
0   
 0  6.25e  5 h -1
n  16
Calculate C* integral: C*  9.0e  5 MPa.m.h-1
Calculate universal creep crack growth rate:
a  4.4e  3 mm.h -1
Example-1 (Cont..)
Verify for steady state creep condition
a ref
2

*
 0.03  0.5
EC

Hence steady state creep condition is reached and C* is


valid. Update crack growth at some time interval and redo
the analysis till crack reaches through thickness
Example-2
Pr: A 25.4mm thick and 50.8mm wide CT specimen of lCr-
lMo-0.25 V steel is tested under constant load conditions.
The initial crack size after pre-cracking is 26.25mm and
the applied load is 26.7kN. The measurements in Table-1
were made during the test. The output voltage from
electric potential is represented by U and is expressed in
micro-volts. The thermal voltage during the test was
constant and was found to be 3mV. The deformation
constants for this material at 538°C are given as follows.
Calculate the crack growth rate and compare with
experimental data.
Example-5 (cont..)
displacement
Example-5 (cont..)
Example-5 (cont..)
Example-5 (cont..)
Thank You

You might also like