research on the controlling parameters of creep-thermal
TRANSCRIPT
Hindawi Publishing CorporationAdvances in Mechanical EngineeringVolume 2013 Article ID 406129 8 pageshttpdxdoiorg1011552013406129
Research ArticleResearch on the Controlling Parameters ofCreep-Thermal Fatigue Crack
Ming Yan Xiang-jun Zhu and Shi-jie Wang
School of Mechanical Engineering Shenyang University of Technology Shenyang 110870 China
Correspondence should be addressed to Ming Yan yanming7802163com
Received 2 August 2012 Revised 15 December 2012 Accepted 8 January 2013
Academic Editor Shandong Tu
Copyright copy 2013 Ming Yan et alThis is an open access article distributed under theCreativeCommonsAttribution License whichpermits unrestricted use distribution and reproduction in any medium provided the original work is properly cited
The opening process of creep-thermal fatigue crack (CTFC) is studied by finite element method considering the bilinear kinematichardening and creep characteristic of the materialThe crack is closed under the compressive thermal stress during the heating andinsulating processes The appearance of tensile stress during the late cooling process which is normal to the crack face makes thecrack gradually open At this time the temperature of material around the crack has become lower than the creep temperaturetherefore the creep fracture mechanism parameter Clowast is not applicable The CTFC is quite shallow and comparatively the plasticzone is rather large thus the method of stress intensity factor is restricted A modified J-integral method is put forward accordingto the stress analysis of CTFC which has been proved to be path independent and validThemethod is used for not only the CTFCbut also any unloading crack or the crack in the residual compressive plastic strain field Experimental results show that themodifiedJ-integral can be used as the controlling parameter of the CTFC propagation
1 Introduction
For damage-tolerant design and remaining life prediction ofhigh-temperature components it is important to investigatecreep-fatigue crack growth behavior [1 2] A large numberof studies on creep-fatigue crack growth have been reportedin recent years Most of them are on the crack under creepthermomechanical fatigue (TMF) conditions which refer toa thermal cycling loading with an additional mechanicalloading
Typical creep thermomechanical fatigue model workswith one side fixed and the other side under cycling externalload the temperature cycles with the external loadThe TMFcrack opens under tensile load and the crack opening dis-placement increases continuously with the combinationeffect of tensile holding load and temperature thus the TMFcrack propagation can be thought of as the combination insome way of crack propagation caused by TMF and creepThe crack propagation caused by TMF can be characterizedby using stress intensity factor or 119869-integral as a controllingparameter and crack propagation caused by creep can becharacterized by using creep fracture mechanism parameter119862lowast as a controlling parameter [3ndash5]
In addition there are cooling duct in nuclear power plantsolar panel of aircrafts in low earth orbit pulsed high powerlaser driver hot work die roller of the hot mill cylinder ofhigh power diesel engine brake disk of high-speed trains andother components working under creep-thermal fatigue con-dition which refers to a pure thermal cycling Comparativelythe study on the controlling parameters of CTFC is paid lessattention [6ndash8]
Creep-thermal fatigue model is fixed on both sides andthere is no external load During heating process compres-sive elastic strain generates in the specimen during insulatingprocess compressive creep strain generates in the specimenand the cracks keep closed in both heating and insulatingprocesses However how could creep-thermal fatigue crackgrow Actually the inelastic strain generating during theheating and insulating processes could not recover freelyTherefore the tensile stress which is normal to the crack facesand generated in the late cooling process makes the crackopen and grow Thus it can be seen that there is really quite adifference between the mechanism of creep-thermal fatiguecrack and that of the thermomechanical fatigue crack thecontrol parameter of the creep thermomechanical fatiguecannot be used directly on the creep-thermal fatigue crack
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
2 Advances in Mechanical Engineering
Table 1 Chemical composition (quality fraction) () of ZL111
Major elements ImpuritySi Cu Mn Mg Ti Fe Zn Sn Pb80 sim 100 13 sim 18 01 sim 035 04 sim 06 01 sim 035 lt04 lt01 lt001 lt005
Considering the bi-linear kinematic hardening and creepcharacteristic of the material the CTFC is analyzed by thefinite element method and its controlling parameters arestudied accordingly
2 Modeling and Material Properties
21 Modeling Assumption For convenience the actual com-ponent can be simplified to a rectangular plate (30mm times
10mm) fixed at both ends as shown in Figure 1(a) Assumethat the plate is in the plane stress state The temperature ofthe left side changes according to the curve shown in Figure 2and the temperature of the right side is always 30∘C A cracklies in the center of the left side and is enlarged in Figure 1(b)119877119860 and1198771198601015840 represent the crack faces (07mm)119877 is the cracktip
Generally in the finite element model the adjacentelements share one node to make the elements connect witheach other into a whole In the finite element model withcracks simulated crack elements have a fully coincidednode position on crack faces as is shown in Figure 1(b) theelements on both the up and down sides of the line AR Tosimulate the extrusion of the two crack faces during the heat-ing and insulating processes contact elements are used Tomeasure the singularity of the crack tip the elements aroundit should be quadratic with the midside nodes placed at thequarter points Such elements are called singular elements(Figure 1(c)) The mesh size at the crack tip is 003mm
22 Mechanical Properties of Material ZALSi9Cu2Mg (ZL111)is a kind of cast aluminumalloy and its chemical compositionis shown in Table 1 It is often used to cast the large dieselengine cylinder heads of naval vessels that are subject tothermal fatigue and creep damage in service
ZALSi9Cu2Mg (ZL111) has superior mechanical proper-ties under the room temperature and high temperature It isthematerial with bi-linear kinematic hardening [9] as shownin Figure 3 where OABC is the actual tensile curve and 119860is the yield point After yielding the relationship of stressand strain is nonlinear According to the bi-linear kinematichardening rule curve ABC can be replaced by straight lineAC and CDEF represents the unloading process
The creep equation of ZL111 is [10]
120576c= 119862112059011986221199051198623119890minus1198624119879
(1)
where 120576c is the creep rate (1119904)119862
119894(119894 = 1 2 3 4) are constants
From experiments 1198621= 05852 times 10
minus5 1198622= 2564 119862
3=
minus0764 and 1198624= 5196060 Other mechanical parameters
can be obtained by interpolation according to Table 2 119879represents the temperature (∘C)
119884
119883119874
(a)
119860(119860998400)119877
(b)
119874
119872
119873
119869
119870
119868 119875 119871
(c)
Figure 1 Finite element model
0 10 20 30 40 50 60 70 80 90 100Time (min)
230
210
190
170
150
130
110
90
70
50
30
Tem
pera
ture
(∘C)
Figure 2 Temperature curve
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 3
Table 2 Mechanical parameters of ZL111
Temperature (∘C) Linear expansionCoefficient (10minus6 sdot ∘Cminus1)
Elastic modulus(MPa) Poissonrsquos ratio Yield limit (MPa) Tangent modulus
(MPa)20 186 71500 031 180 10840100 189 69800 031 168 8670150 200 68000 031 155 6720200 215 65000 032 141 4740220 220 63000 032 136 2040
119860
119861
119862
119863
119864
119865
119874
120590
1205901
1205902
1205760 1205761 120576 1205762
Figure 3 The bilinear kinematic hardening model
3 Stress Analysis of CTFC
The commercial finite element package ANSYS and thesequential coupling analysis method are employed to com-pute the thermal stress First the initial conditions andboundary conditions for a thermal analysis are set to themodel to obtain the temperature field of the model at everydiscrete time Then put the nodal temperature as thecoupled-field loads and the structure boundary conditions tothe model for a stress analysis The stress and strain fields atevery discrete time can be obtained The temperature field isjust the thermal load for the stress analysis rather than theobjective of the paper and so it is out of discussion
Figure 4 shows the stress curve of the crack tip (119877 point)in one cycle it is obvious that the Mises equivalent stress isdominated by the 119884 component Its 119884 components of strainare shown in Figure 5
In the heating process (0ndash10min) the thermal strainincreases with the temperature and the model cannot expand
0 10 20 30 40 50 60 70 80 90 100
0
100
200
300
400
Stre
ss (M
Pa)
X stress componentY stress component
XY shear stress componentMises equivalent stress
Time (min)
minus200
minus100
Figure 4 Stress components of the crack tip
freely due to the restrictions at both ends so compressiveelastic strain and stress are produced the plastic compressivestrain appears when the stress reaches the compressive yieldpoint of the material In the insulating process (10ndash90min)compressive creep strain appears in the common role of com-pressive stress and time simultaneously the elastic compres-sive strain and compressive stress decrease and the thermalstrain and the plastic strain keep constant In the cooling pro-cess (90ndash100min) the thermal strain and compressive elasticstrain gradually disappear and the model trends to its initialshape Because the model is fixed at both ends and the com-pressive plastic and creep strain cannot get back freely thetensile stress which is normal to the crack faces is produced
If the tensile stress reaches the reverse yield point thematerial will yield reversely The crack opens gradually withthe increase of tensile stress
The pressure on points 119860 and 119877 is shown in Figure 6In the heating process (0ndash10min) the pressure on the crackfaces increases for the temperature of point 119860 is higher thanthat of point 119877 so the pressure on point 119860 is higher In theinsulating process (10ndash90min) the pressure on the crackfaces decreases slowly owing to stress relaxation the creeprate of point119860 is larger so its pressure decreasesmore sharplyIn the cooling process (90ndash100min) the pressure declinesrapidly the pressure on points 119860 and 119877 becomes zero at
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
4 Advances in Mechanical Engineering
0 10 20 30 40 50 60 70 80 90 100
0
2
4
6
8
10
12
Time (min)
Y co
mpo
nent
s str
ain
Elastic strainPlastic strain
Creep strainThermal strain
times10minus3
minus4
minus2
Figure 5 Strain components in the 119884 direction of the crack tip
0 20 40 60 80 1000
20
40
60
80
100
120
140
160
180
Time (min)
Pres
sure
(MPa
)
119877 point119860 point
Figure 6 The pressure on points 119860 and 119877
95min which proves that the two crack faces separate at thistime Figure 7 shows the displacements in the 119884 direction ofpoints 119860 and 1198601015840 as well as their relative ones
Figure 8 is the deformation of the crack region at 100min(the deformation is magnified to 10 times) The red areaaround the crack tip is the tensile plastic zone outside whichis a green area Some compressive plastic strain produced inthe heating process remains here in various extents so it iscalled compressive plastic zone
One point is picked up in each of the tensile plastic zoneand compressive plastic zone randomly marked as 119879 and 119862Figure 9 gives the strain components in the 119884 direction ofpoints 119879 and 119862 Figure 10 gives their 119884 components of stress
0 20 40 60 80 100
0
2
4
6
8
Disp
lace
men
t (m
m)
Relative displacement
Time (min)
times10minus3
minus4
minus2
119860 point119860998400 point
Figure 7 The displacements in the 119884 direction of points 119860 and 1198601015840as well as their relative one
MNMX
Figure 8 The deformation of the crack region at 100min
From the figures stress and the elastic strain of the two pointsget to zero at the moment of the crack opening which is justas the crack tip
According to the above analysis it can be deduced thatthe compressive plastic strains produced in heating andcompressive creep strain produced in insulating cannot comeback freely during cooling which will make the CTFC openIt can be also concluded that the more the compressiveinelastic strain is during heating and insulating the larger thetensile stress the tensile strain and the opening displacementof CTFC will be at the end of cooling
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 5
0
1
2
3
4
0 20 40 60 80 100
Time (min)
times10minus3
minus3
minus1
minus2
Elastic strain of 119862 pointPlastic strain of 119862 point
Elastic strain of 119879 pointPlastic strain of 119879 point
Stra
in co
mpo
nent
s in119884
dire
ctio
n
Figure 9 Strain components in the 119884 direction of points 119879 and 119862
4 Controlling Parameters of CTFC
The CTFC opens gradually in the late cooling process Atthis time the temperature of the material around the crackhas become lower than the creep temperature Therefore thecreep fracturemechanismparameter119862lowast ismeaningless to theCTFC Suppose the stress intensity factor or 119869-integral will bethe controlling parameter of the CTFC in the following
41 Stress Intensity Factor Method The stress intensity factorof a crack for a linear elastic fracture mechanics analysis maybe computed according to a fit of the nodal displacementsin the vicinity of the crack [11 12] In the small-scale yieldingcondition the factor can be modified according to [13]
Generally the small-scale yielding condition for a cracklength 119886 is [13]
119903119901
119886le 01 (2)
where 119903119901is the size of plastic zone here it is themean distance
from the boundary of the tensile plastic zone to the cracktip The CTFC is always quite shallow [6] comparatively theplastic zone is rather largeThe small-scale yielding conditioncan be satisfied only when the singularity of the crack issmall
42 Proposition of aModified 119869-IntegralMethod The classicaldefinition of 119869-integral is given by
119869 = int120574
119908119889119910 minus int120574
T120597u120597119909119889119904 (3)
where 120574 represents any path surrounding the crack tip 119904 is thedistance along the path T is the stress tensor vector u is thedisplacement tensor vector 119908 is the strain energy density inthe monotonic loading process
0
50
100
150
200
250
300
Stre
ss co
mpo
nent
s in
di
rect
ion
(MPa
)
minus200
minus150
minus100
minus50
0 20 40 60 80 100
Time (min)
119884
119862 point119879 point
Figure 10 Stress components in the 119884 direction of points 119879 and 119862
From the previous stress analysis of the CTFC thethermal stress near the crack does not loadmonotonously in atemperature cycle which does not meet the path-independ-ent conditionMoreover affected by the residual compressiveplastic strain field at the end of cooling the integral might benegative Thus the 119869-integral method must be improved toapplicable to be the CTFC
T120597u120597119909= (120590119909119899119909+ 120590119909119910119899119910)120597119906119909
120597119909+ (120590119910119899119910+ 120590119909119910119899119909)120597119906119910
120597119909 (4)
where 119899 is the unit outer normal vector to path 120574 It isobvious that the later term of the 119869-integral is naturally pathindependent and immune to the compressive plastic strainfield
According to the plastic deformation theory the strainenergy density is represented by
119908 = 119908119909119909+ 119908119909119910+ 119908119910119910
119908119894119895= int
120576
0
120590119894119895(119905) sdot 120576119894119895(119905) 119889119905
(5)
where120590119894119895 120576119894119895(119894 119895 = 119909 119910) are respectively the stress and strain
components Here only the improvement of 119908119910119910
similar tothat of 119908
119909119909and 119908
119909119910is discussed in detail
From the stress analysis of the CTFC the 119884 componentsof stress and elastic strain around the crack can be zero almostat the same timewhen the crack opensMoreover the coolingperiod can be regarded as amonotonic loading processThusonly the strain energy in this period should be computed
119908yy = int1199052
1199051
120590119910119910(119905) sdot 120576119910119910(119905) 119889119905 (6)
where 1199051stands for the moment of the opening and 119905
2is the
terminal time of cooling In order to eliminate the effect of
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
6 Advances in Mechanical Engineering
the residual compressive plastic strain field 120576119910119910(119905) is obtained
as
120576119910119910(119905) = 120576
119890
119910119910(119905) + 120576
119901
119910y (119905) minus 1205761199010
119910119910 (7)
where 120576119890yy(119905) 120576119901
yy(119905) are separately the elastic and plastic com-ponents of strain at the end of cooling 1205761199010
119910119910is the compressive
plastic strain produced in the heating processThe improvement of 119869-integral is put forward according
to the analysis of CTFC according to its definition Theintegral computed in this way is path independent Thus itcould be the controlling parameter of CTFC
In the linear elasticity small-scale yielding conditionsand plane stress conditions 119869-integral is related to stressintensity factor by the following relationship
119869 =1198702
Ι
119864 (8)
where 119864 is the elastic modulus Therefore the stress intensityfactor119870
Ιcan validate the modified 119869-integral cursorily
5 Examples
Take a group of concentric circles whose center is crack tipand radius is 119903 as the integral paths Figure 11 shows therelation between themodified 119869-integral and 119903119903
119901 Clearly the
modified 119869-integral is path independent as long as the integralpath is within the compressive plastic zone
The stress intensity factor and the modified 119869-integral arecalculated under different conditions The results are showninTables 3 4 and 5 In (8) the elasticmodulus119864 is 71175MPa
It is obvious that the stress intensity factor and the 119869-integral increases when the insulating time prolongs themaximum temperature increases and the crack grows Fromcalculating the stress intensity factor can be used as thecontrolling parameter of the CTFC only with a low heatingtemperature a short period of holding time and a deep crack
6 CTFC Propagation Test
The creep-thermal fatigue test machine is composed ofclamping mechanism infrared radiation heating devicehydropneumatic cooling system and control system Theoperational principle of the creep-thermal fatigue testmachine is shown in Figure 12 the specimens can be turnedupside down by the driving shaft which makes it easy andquick to heat and cool the specimensThe detailed size of thespecimen with cracks is shown in Figure 13The first secondand third specimens are used in the creep-thermal fatiguetest the temperature of the specimen cycles is from 30∘C to210∘C and the thermal cycle insulating time is 5 minutes at210∘C every time All the specimens undergo 1000 thermalcycles The length of the crack is detected for each 100 cyclesand the results are shown in Figure 14
The secant method [14] is used to measure the crackgrowth rate that is to calculate the gradient between two
0 1 2
1
04
05
0402
06
06
07
08
08
09
12 14 16 18119903119903119901
119869(J
mmminus2)
Figure 11 The relationship of 119869-integral and 119903119903119901
++
1 2 3 4 5 6 7 8
Figure 12 Operational principle diagram of the creep-thermalfatigue test machine 1 driving shaft 2 stock of the specimen 3unloading bolt 4 tightening bolt 5 infrared radiation heatingdevice 6 specimen 7 hydropneumatic cooling system and 8 armof the test machine
adjacent data points on the 119886-119873 curve to get the crack growthrate as
119889119886
119889119873asympΔ119886
Δ119873=119886119894+1minus 119886119894
119873119894+1minus 119873119894
(9)
where 119889119886119889119873 is the crack propagation under one thermalcycle Δ119886 is the crack propagation under Δ119873 thermal cycles119886119894and 119873
119894are the crack length and number of cycles of load
corresponding to the 119894th data pointAccording to the Paris formula
119889119886
119889119873= 119862119869119899
(10)
where 119862 and 119899 are constants related to material propertyenvironmental media and the geometry of the specimen
The creep-thermal fatigue 119869-integral value in correspond-ing conditions can be obtained according to the modified J-integral method referred to in the front of the paper the datapoints of ln(119889119886119889119873) and ln 119869 are shown in Figure 15 hence
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 7
Table 3 The crack length is 07mm the maximum temperature is 210∘C
Holding time(min)
Size of plasticzone (mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)0 00700 1944 05117 0531080 01050 09289150 01313 10750240 01570 11950
Table 4 The crack length is 07mm the insulating time is 80min
Maximumtemperature (∘C)
Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)180 0 239 00080 00082190 00630 1594 03440 03570200 00875 2115 05970 06285220 01778 14148
Table 5 The maximum temperature is 200∘C the insulating time is 80min
Crack length (mm) Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)05 00623 0463309 01025 0702912 01220 0808815 01493 2526 08568 08965
02
02
3plusmn01
25plusmn01
25plusmn01
4
6
2
2
2
AAA
A A
002
002
002
002
6h7
20plusmn01
13 1370
Figure 13 Specimen of creep-thermal fatigue
119899 = 251 119862 = 872times10minus6 can be obtained by some regressionanalysis of formula (10) using these data points
7 Conclusions
(1) The crack is closed under the compressive thermalstress during the heating and insulating processesBecause the inelastic strain generating in the two
0 200 400 600 800 10000
05
1
15
2
25
3
35
Number of thermal cycles
Crac
k pr
opag
atio
n ra
te (m
m)
1
2
3
Figure 14 Propagation length of the CTFC
processes cannot be back by itself the tensile stressduring the late cooling process which is normal to thecrack faces makes the crack gradually open
(2) When the CTFC is opening the temperature of thematerial around it has become lower than the creep
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
8 Advances in Mechanical Engineering
119869-integral100 101 102
10minus2
10minus3
10minus4
10minus5
1
2
3
Figure 15 (119889119886119889119873)mdash119869 data in double logarithmic coordinate
temperature therefore the creep fracture mechanismparameter 119862lowast is not applicable The stress intensityfactor and 119869-integral can be used as the controllingparameters The stress intensity factor can be usedonly with a short insulating time a low maximumtemperature and a deep crack The 119869-integral shouldbe modified in order to be applicable to the CTFCthat is to take the moment of opening as the startingof the strain energy density integral and eliminatethe effect of the residual compressive plastic strain by(7) It has been proved that the modified 119869-integralmethod is path independent and valid It is used fornot only theCTFCbut also any unloading crack or thecrack in the residual compressive plastic strain fieldThe stress intensity factor and 119869-integral increasewhen the insulating time prolongs the maximumtemperature rises and the crack grows
(3) Experimental results show that the modified 119869-integral can be used as the control value of the CTFCpropagation The research on the controlling param-eters of CTFC provides the theoretical basis for itsgrowth
Acknowledgments
This project is supported by the Natural Science Foundationof Liaoning Province (Grant no 201102167) Aeronautical Sci-ence Foundation of China (Grant no 20110450001) LiaoningExcellent Talents in University (Grant no LJQ2011012) andthe Science Program of Shenyang (Grant No F12-069-2-00)
References
[1] N Gao M Brown K J Miller and P A S Reed ldquoAn investiga-tion of crack growth behaviour under creep-fatigue conditionrdquo
Materials Science and Engineering A vol 410 no 25 pp 67ndash712005
[2] M F Geng ldquoAn idea for predicting crack growth time tofracture under creep-fatigue conditionsrdquoMaterials Science andEngineering A vol 257 no 2 pp 250ndash255 1998
[3] Y L Lu L J Chen P K Liaw et al ldquoEffects of temperature andhold time on creep-fatigue crack-growth behaviorrdquo MaterialsScience and Engineering A vol 429 no 1-2 pp 1ndash10 2006
[4] A T Yokobori and T Satoh ldquoThe characterization of creepcrack growth rate and its life of TiAl inter-metallic compoundwith full lamellar microstructurerdquo International Journal ofPressure Vessels and Piping vol 88 no 7 pp 435ndash440 2011
[5] G Marahleh A R I Kheder and H F Hamad ldquoCreep-lifeprediction of service-exposed turbine bladesrdquoMaterials Sciencevol 42 no 4 pp 476ndash481 2006
[6] V Maillot A Fissolo G Degallaix and S Degallaix ldquoThermalfatigue crack networks parameters and stability an experimen-tal studyrdquo International Journal of Solids and Structures vol 42no 2 pp 759ndash769 2005
[7] J Lansinger T Hansson andO Clevfors ldquoFatigue crack growthunder combined thermal cycling and mechanical loadingrdquoInternational Journal of Fatigue vol 29 no 7 pp 1383ndash13902007
[8] N Haddar A Fissolo and V Maillot ldquoThermal fatigue cracknetworks an computational studyrdquo International Journal ofSolids and Structures vol 42 no 2 pp 771ndash788 2005
[9] G T Yang Theory of Elasticity and Plasticity China BuildingMaterials Industry Press Beijing China 2005
[10] H Kraus Creep Analysis Wiley International Science Publica-tion New York NY USA 1980
[11] Z D Jiang A Zeghloul G Bezine and J Petit ldquoStress intensityfactors of parallel cracks in a finite width sheetrdquo EngineeringFracture Mechanics vol 35 no 6 pp 1073ndash1079 1990
[12] G Agnihotri ldquoStress analysis of a crack using the finite elementmethodrdquo Engineering FractureMechanics vol 44 no 1 pp 109ndash125 1993
[13] R Narasimhan and A J Rosakis ldquoA finite element analysis ofsmall-scale yielding near a stationary crack under plane stressrdquoJournal of the Mechanics and Physics of Solids vol 36 no 1 pp77ndash117 1988
[14] P Bensussan High Temperature Fracture Mechanisms andMechanic Mechanical Engineering Publications 2005
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
2 Advances in Mechanical Engineering
Table 1 Chemical composition (quality fraction) () of ZL111
Major elements ImpuritySi Cu Mn Mg Ti Fe Zn Sn Pb80 sim 100 13 sim 18 01 sim 035 04 sim 06 01 sim 035 lt04 lt01 lt001 lt005
Considering the bi-linear kinematic hardening and creepcharacteristic of the material the CTFC is analyzed by thefinite element method and its controlling parameters arestudied accordingly
2 Modeling and Material Properties
21 Modeling Assumption For convenience the actual com-ponent can be simplified to a rectangular plate (30mm times
10mm) fixed at both ends as shown in Figure 1(a) Assumethat the plate is in the plane stress state The temperature ofthe left side changes according to the curve shown in Figure 2and the temperature of the right side is always 30∘C A cracklies in the center of the left side and is enlarged in Figure 1(b)119877119860 and1198771198601015840 represent the crack faces (07mm)119877 is the cracktip
Generally in the finite element model the adjacentelements share one node to make the elements connect witheach other into a whole In the finite element model withcracks simulated crack elements have a fully coincidednode position on crack faces as is shown in Figure 1(b) theelements on both the up and down sides of the line AR Tosimulate the extrusion of the two crack faces during the heat-ing and insulating processes contact elements are used Tomeasure the singularity of the crack tip the elements aroundit should be quadratic with the midside nodes placed at thequarter points Such elements are called singular elements(Figure 1(c)) The mesh size at the crack tip is 003mm
22 Mechanical Properties of Material ZALSi9Cu2Mg (ZL111)is a kind of cast aluminumalloy and its chemical compositionis shown in Table 1 It is often used to cast the large dieselengine cylinder heads of naval vessels that are subject tothermal fatigue and creep damage in service
ZALSi9Cu2Mg (ZL111) has superior mechanical proper-ties under the room temperature and high temperature It isthematerial with bi-linear kinematic hardening [9] as shownin Figure 3 where OABC is the actual tensile curve and 119860is the yield point After yielding the relationship of stressand strain is nonlinear According to the bi-linear kinematichardening rule curve ABC can be replaced by straight lineAC and CDEF represents the unloading process
The creep equation of ZL111 is [10]
120576c= 119862112059011986221199051198623119890minus1198624119879
(1)
where 120576c is the creep rate (1119904)119862
119894(119894 = 1 2 3 4) are constants
From experiments 1198621= 05852 times 10
minus5 1198622= 2564 119862
3=
minus0764 and 1198624= 5196060 Other mechanical parameters
can be obtained by interpolation according to Table 2 119879represents the temperature (∘C)
119884
119883119874
(a)
119860(119860998400)119877
(b)
119874
119872
119873
119869
119870
119868 119875 119871
(c)
Figure 1 Finite element model
0 10 20 30 40 50 60 70 80 90 100Time (min)
230
210
190
170
150
130
110
90
70
50
30
Tem
pera
ture
(∘C)
Figure 2 Temperature curve
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 3
Table 2 Mechanical parameters of ZL111
Temperature (∘C) Linear expansionCoefficient (10minus6 sdot ∘Cminus1)
Elastic modulus(MPa) Poissonrsquos ratio Yield limit (MPa) Tangent modulus
(MPa)20 186 71500 031 180 10840100 189 69800 031 168 8670150 200 68000 031 155 6720200 215 65000 032 141 4740220 220 63000 032 136 2040
119860
119861
119862
119863
119864
119865
119874
120590
1205901
1205902
1205760 1205761 120576 1205762
Figure 3 The bilinear kinematic hardening model
3 Stress Analysis of CTFC
The commercial finite element package ANSYS and thesequential coupling analysis method are employed to com-pute the thermal stress First the initial conditions andboundary conditions for a thermal analysis are set to themodel to obtain the temperature field of the model at everydiscrete time Then put the nodal temperature as thecoupled-field loads and the structure boundary conditions tothe model for a stress analysis The stress and strain fields atevery discrete time can be obtained The temperature field isjust the thermal load for the stress analysis rather than theobjective of the paper and so it is out of discussion
Figure 4 shows the stress curve of the crack tip (119877 point)in one cycle it is obvious that the Mises equivalent stress isdominated by the 119884 component Its 119884 components of strainare shown in Figure 5
In the heating process (0ndash10min) the thermal strainincreases with the temperature and the model cannot expand
0 10 20 30 40 50 60 70 80 90 100
0
100
200
300
400
Stre
ss (M
Pa)
X stress componentY stress component
XY shear stress componentMises equivalent stress
Time (min)
minus200
minus100
Figure 4 Stress components of the crack tip
freely due to the restrictions at both ends so compressiveelastic strain and stress are produced the plastic compressivestrain appears when the stress reaches the compressive yieldpoint of the material In the insulating process (10ndash90min)compressive creep strain appears in the common role of com-pressive stress and time simultaneously the elastic compres-sive strain and compressive stress decrease and the thermalstrain and the plastic strain keep constant In the cooling pro-cess (90ndash100min) the thermal strain and compressive elasticstrain gradually disappear and the model trends to its initialshape Because the model is fixed at both ends and the com-pressive plastic and creep strain cannot get back freely thetensile stress which is normal to the crack faces is produced
If the tensile stress reaches the reverse yield point thematerial will yield reversely The crack opens gradually withthe increase of tensile stress
The pressure on points 119860 and 119877 is shown in Figure 6In the heating process (0ndash10min) the pressure on the crackfaces increases for the temperature of point 119860 is higher thanthat of point 119877 so the pressure on point 119860 is higher In theinsulating process (10ndash90min) the pressure on the crackfaces decreases slowly owing to stress relaxation the creeprate of point119860 is larger so its pressure decreasesmore sharplyIn the cooling process (90ndash100min) the pressure declinesrapidly the pressure on points 119860 and 119877 becomes zero at
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
4 Advances in Mechanical Engineering
0 10 20 30 40 50 60 70 80 90 100
0
2
4
6
8
10
12
Time (min)
Y co
mpo
nent
s str
ain
Elastic strainPlastic strain
Creep strainThermal strain
times10minus3
minus4
minus2
Figure 5 Strain components in the 119884 direction of the crack tip
0 20 40 60 80 1000
20
40
60
80
100
120
140
160
180
Time (min)
Pres
sure
(MPa
)
119877 point119860 point
Figure 6 The pressure on points 119860 and 119877
95min which proves that the two crack faces separate at thistime Figure 7 shows the displacements in the 119884 direction ofpoints 119860 and 1198601015840 as well as their relative ones
Figure 8 is the deformation of the crack region at 100min(the deformation is magnified to 10 times) The red areaaround the crack tip is the tensile plastic zone outside whichis a green area Some compressive plastic strain produced inthe heating process remains here in various extents so it iscalled compressive plastic zone
One point is picked up in each of the tensile plastic zoneand compressive plastic zone randomly marked as 119879 and 119862Figure 9 gives the strain components in the 119884 direction ofpoints 119879 and 119862 Figure 10 gives their 119884 components of stress
0 20 40 60 80 100
0
2
4
6
8
Disp
lace
men
t (m
m)
Relative displacement
Time (min)
times10minus3
minus4
minus2
119860 point119860998400 point
Figure 7 The displacements in the 119884 direction of points 119860 and 1198601015840as well as their relative one
MNMX
Figure 8 The deformation of the crack region at 100min
From the figures stress and the elastic strain of the two pointsget to zero at the moment of the crack opening which is justas the crack tip
According to the above analysis it can be deduced thatthe compressive plastic strains produced in heating andcompressive creep strain produced in insulating cannot comeback freely during cooling which will make the CTFC openIt can be also concluded that the more the compressiveinelastic strain is during heating and insulating the larger thetensile stress the tensile strain and the opening displacementof CTFC will be at the end of cooling
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 5
0
1
2
3
4
0 20 40 60 80 100
Time (min)
times10minus3
minus3
minus1
minus2
Elastic strain of 119862 pointPlastic strain of 119862 point
Elastic strain of 119879 pointPlastic strain of 119879 point
Stra
in co
mpo
nent
s in119884
dire
ctio
n
Figure 9 Strain components in the 119884 direction of points 119879 and 119862
4 Controlling Parameters of CTFC
The CTFC opens gradually in the late cooling process Atthis time the temperature of the material around the crackhas become lower than the creep temperature Therefore thecreep fracturemechanismparameter119862lowast ismeaningless to theCTFC Suppose the stress intensity factor or 119869-integral will bethe controlling parameter of the CTFC in the following
41 Stress Intensity Factor Method The stress intensity factorof a crack for a linear elastic fracture mechanics analysis maybe computed according to a fit of the nodal displacementsin the vicinity of the crack [11 12] In the small-scale yieldingcondition the factor can be modified according to [13]
Generally the small-scale yielding condition for a cracklength 119886 is [13]
119903119901
119886le 01 (2)
where 119903119901is the size of plastic zone here it is themean distance
from the boundary of the tensile plastic zone to the cracktip The CTFC is always quite shallow [6] comparatively theplastic zone is rather largeThe small-scale yielding conditioncan be satisfied only when the singularity of the crack issmall
42 Proposition of aModified 119869-IntegralMethod The classicaldefinition of 119869-integral is given by
119869 = int120574
119908119889119910 minus int120574
T120597u120597119909119889119904 (3)
where 120574 represents any path surrounding the crack tip 119904 is thedistance along the path T is the stress tensor vector u is thedisplacement tensor vector 119908 is the strain energy density inthe monotonic loading process
0
50
100
150
200
250
300
Stre
ss co
mpo
nent
s in
di
rect
ion
(MPa
)
minus200
minus150
minus100
minus50
0 20 40 60 80 100
Time (min)
119884
119862 point119879 point
Figure 10 Stress components in the 119884 direction of points 119879 and 119862
From the previous stress analysis of the CTFC thethermal stress near the crack does not loadmonotonously in atemperature cycle which does not meet the path-independ-ent conditionMoreover affected by the residual compressiveplastic strain field at the end of cooling the integral might benegative Thus the 119869-integral method must be improved toapplicable to be the CTFC
T120597u120597119909= (120590119909119899119909+ 120590119909119910119899119910)120597119906119909
120597119909+ (120590119910119899119910+ 120590119909119910119899119909)120597119906119910
120597119909 (4)
where 119899 is the unit outer normal vector to path 120574 It isobvious that the later term of the 119869-integral is naturally pathindependent and immune to the compressive plastic strainfield
According to the plastic deformation theory the strainenergy density is represented by
119908 = 119908119909119909+ 119908119909119910+ 119908119910119910
119908119894119895= int
120576
0
120590119894119895(119905) sdot 120576119894119895(119905) 119889119905
(5)
where120590119894119895 120576119894119895(119894 119895 = 119909 119910) are respectively the stress and strain
components Here only the improvement of 119908119910119910
similar tothat of 119908
119909119909and 119908
119909119910is discussed in detail
From the stress analysis of the CTFC the 119884 componentsof stress and elastic strain around the crack can be zero almostat the same timewhen the crack opensMoreover the coolingperiod can be regarded as amonotonic loading processThusonly the strain energy in this period should be computed
119908yy = int1199052
1199051
120590119910119910(119905) sdot 120576119910119910(119905) 119889119905 (6)
where 1199051stands for the moment of the opening and 119905
2is the
terminal time of cooling In order to eliminate the effect of
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
6 Advances in Mechanical Engineering
the residual compressive plastic strain field 120576119910119910(119905) is obtained
as
120576119910119910(119905) = 120576
119890
119910119910(119905) + 120576
119901
119910y (119905) minus 1205761199010
119910119910 (7)
where 120576119890yy(119905) 120576119901
yy(119905) are separately the elastic and plastic com-ponents of strain at the end of cooling 1205761199010
119910119910is the compressive
plastic strain produced in the heating processThe improvement of 119869-integral is put forward according
to the analysis of CTFC according to its definition Theintegral computed in this way is path independent Thus itcould be the controlling parameter of CTFC
In the linear elasticity small-scale yielding conditionsand plane stress conditions 119869-integral is related to stressintensity factor by the following relationship
119869 =1198702
Ι
119864 (8)
where 119864 is the elastic modulus Therefore the stress intensityfactor119870
Ιcan validate the modified 119869-integral cursorily
5 Examples
Take a group of concentric circles whose center is crack tipand radius is 119903 as the integral paths Figure 11 shows therelation between themodified 119869-integral and 119903119903
119901 Clearly the
modified 119869-integral is path independent as long as the integralpath is within the compressive plastic zone
The stress intensity factor and the modified 119869-integral arecalculated under different conditions The results are showninTables 3 4 and 5 In (8) the elasticmodulus119864 is 71175MPa
It is obvious that the stress intensity factor and the 119869-integral increases when the insulating time prolongs themaximum temperature increases and the crack grows Fromcalculating the stress intensity factor can be used as thecontrolling parameter of the CTFC only with a low heatingtemperature a short period of holding time and a deep crack
6 CTFC Propagation Test
The creep-thermal fatigue test machine is composed ofclamping mechanism infrared radiation heating devicehydropneumatic cooling system and control system Theoperational principle of the creep-thermal fatigue testmachine is shown in Figure 12 the specimens can be turnedupside down by the driving shaft which makes it easy andquick to heat and cool the specimensThe detailed size of thespecimen with cracks is shown in Figure 13The first secondand third specimens are used in the creep-thermal fatiguetest the temperature of the specimen cycles is from 30∘C to210∘C and the thermal cycle insulating time is 5 minutes at210∘C every time All the specimens undergo 1000 thermalcycles The length of the crack is detected for each 100 cyclesand the results are shown in Figure 14
The secant method [14] is used to measure the crackgrowth rate that is to calculate the gradient between two
0 1 2
1
04
05
0402
06
06
07
08
08
09
12 14 16 18119903119903119901
119869(J
mmminus2)
Figure 11 The relationship of 119869-integral and 119903119903119901
++
1 2 3 4 5 6 7 8
Figure 12 Operational principle diagram of the creep-thermalfatigue test machine 1 driving shaft 2 stock of the specimen 3unloading bolt 4 tightening bolt 5 infrared radiation heatingdevice 6 specimen 7 hydropneumatic cooling system and 8 armof the test machine
adjacent data points on the 119886-119873 curve to get the crack growthrate as
119889119886
119889119873asympΔ119886
Δ119873=119886119894+1minus 119886119894
119873119894+1minus 119873119894
(9)
where 119889119886119889119873 is the crack propagation under one thermalcycle Δ119886 is the crack propagation under Δ119873 thermal cycles119886119894and 119873
119894are the crack length and number of cycles of load
corresponding to the 119894th data pointAccording to the Paris formula
119889119886
119889119873= 119862119869119899
(10)
where 119862 and 119899 are constants related to material propertyenvironmental media and the geometry of the specimen
The creep-thermal fatigue 119869-integral value in correspond-ing conditions can be obtained according to the modified J-integral method referred to in the front of the paper the datapoints of ln(119889119886119889119873) and ln 119869 are shown in Figure 15 hence
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 7
Table 3 The crack length is 07mm the maximum temperature is 210∘C
Holding time(min)
Size of plasticzone (mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)0 00700 1944 05117 0531080 01050 09289150 01313 10750240 01570 11950
Table 4 The crack length is 07mm the insulating time is 80min
Maximumtemperature (∘C)
Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)180 0 239 00080 00082190 00630 1594 03440 03570200 00875 2115 05970 06285220 01778 14148
Table 5 The maximum temperature is 200∘C the insulating time is 80min
Crack length (mm) Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)05 00623 0463309 01025 0702912 01220 0808815 01493 2526 08568 08965
02
02
3plusmn01
25plusmn01
25plusmn01
4
6
2
2
2
AAA
A A
002
002
002
002
6h7
20plusmn01
13 1370
Figure 13 Specimen of creep-thermal fatigue
119899 = 251 119862 = 872times10minus6 can be obtained by some regressionanalysis of formula (10) using these data points
7 Conclusions
(1) The crack is closed under the compressive thermalstress during the heating and insulating processesBecause the inelastic strain generating in the two
0 200 400 600 800 10000
05
1
15
2
25
3
35
Number of thermal cycles
Crac
k pr
opag
atio
n ra
te (m
m)
1
2
3
Figure 14 Propagation length of the CTFC
processes cannot be back by itself the tensile stressduring the late cooling process which is normal to thecrack faces makes the crack gradually open
(2) When the CTFC is opening the temperature of thematerial around it has become lower than the creep
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
8 Advances in Mechanical Engineering
119869-integral100 101 102
10minus2
10minus3
10minus4
10minus5
1
2
3
Figure 15 (119889119886119889119873)mdash119869 data in double logarithmic coordinate
temperature therefore the creep fracture mechanismparameter 119862lowast is not applicable The stress intensityfactor and 119869-integral can be used as the controllingparameters The stress intensity factor can be usedonly with a short insulating time a low maximumtemperature and a deep crack The 119869-integral shouldbe modified in order to be applicable to the CTFCthat is to take the moment of opening as the startingof the strain energy density integral and eliminatethe effect of the residual compressive plastic strain by(7) It has been proved that the modified 119869-integralmethod is path independent and valid It is used fornot only theCTFCbut also any unloading crack or thecrack in the residual compressive plastic strain fieldThe stress intensity factor and 119869-integral increasewhen the insulating time prolongs the maximumtemperature rises and the crack grows
(3) Experimental results show that the modified 119869-integral can be used as the control value of the CTFCpropagation The research on the controlling param-eters of CTFC provides the theoretical basis for itsgrowth
Acknowledgments
This project is supported by the Natural Science Foundationof Liaoning Province (Grant no 201102167) Aeronautical Sci-ence Foundation of China (Grant no 20110450001) LiaoningExcellent Talents in University (Grant no LJQ2011012) andthe Science Program of Shenyang (Grant No F12-069-2-00)
References
[1] N Gao M Brown K J Miller and P A S Reed ldquoAn investiga-tion of crack growth behaviour under creep-fatigue conditionrdquo
Materials Science and Engineering A vol 410 no 25 pp 67ndash712005
[2] M F Geng ldquoAn idea for predicting crack growth time tofracture under creep-fatigue conditionsrdquoMaterials Science andEngineering A vol 257 no 2 pp 250ndash255 1998
[3] Y L Lu L J Chen P K Liaw et al ldquoEffects of temperature andhold time on creep-fatigue crack-growth behaviorrdquo MaterialsScience and Engineering A vol 429 no 1-2 pp 1ndash10 2006
[4] A T Yokobori and T Satoh ldquoThe characterization of creepcrack growth rate and its life of TiAl inter-metallic compoundwith full lamellar microstructurerdquo International Journal ofPressure Vessels and Piping vol 88 no 7 pp 435ndash440 2011
[5] G Marahleh A R I Kheder and H F Hamad ldquoCreep-lifeprediction of service-exposed turbine bladesrdquoMaterials Sciencevol 42 no 4 pp 476ndash481 2006
[6] V Maillot A Fissolo G Degallaix and S Degallaix ldquoThermalfatigue crack networks parameters and stability an experimen-tal studyrdquo International Journal of Solids and Structures vol 42no 2 pp 759ndash769 2005
[7] J Lansinger T Hansson andO Clevfors ldquoFatigue crack growthunder combined thermal cycling and mechanical loadingrdquoInternational Journal of Fatigue vol 29 no 7 pp 1383ndash13902007
[8] N Haddar A Fissolo and V Maillot ldquoThermal fatigue cracknetworks an computational studyrdquo International Journal ofSolids and Structures vol 42 no 2 pp 771ndash788 2005
[9] G T Yang Theory of Elasticity and Plasticity China BuildingMaterials Industry Press Beijing China 2005
[10] H Kraus Creep Analysis Wiley International Science Publica-tion New York NY USA 1980
[11] Z D Jiang A Zeghloul G Bezine and J Petit ldquoStress intensityfactors of parallel cracks in a finite width sheetrdquo EngineeringFracture Mechanics vol 35 no 6 pp 1073ndash1079 1990
[12] G Agnihotri ldquoStress analysis of a crack using the finite elementmethodrdquo Engineering FractureMechanics vol 44 no 1 pp 109ndash125 1993
[13] R Narasimhan and A J Rosakis ldquoA finite element analysis ofsmall-scale yielding near a stationary crack under plane stressrdquoJournal of the Mechanics and Physics of Solids vol 36 no 1 pp77ndash117 1988
[14] P Bensussan High Temperature Fracture Mechanisms andMechanic Mechanical Engineering Publications 2005
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 3
Table 2 Mechanical parameters of ZL111
Temperature (∘C) Linear expansionCoefficient (10minus6 sdot ∘Cminus1)
Elastic modulus(MPa) Poissonrsquos ratio Yield limit (MPa) Tangent modulus
(MPa)20 186 71500 031 180 10840100 189 69800 031 168 8670150 200 68000 031 155 6720200 215 65000 032 141 4740220 220 63000 032 136 2040
119860
119861
119862
119863
119864
119865
119874
120590
1205901
1205902
1205760 1205761 120576 1205762
Figure 3 The bilinear kinematic hardening model
3 Stress Analysis of CTFC
The commercial finite element package ANSYS and thesequential coupling analysis method are employed to com-pute the thermal stress First the initial conditions andboundary conditions for a thermal analysis are set to themodel to obtain the temperature field of the model at everydiscrete time Then put the nodal temperature as thecoupled-field loads and the structure boundary conditions tothe model for a stress analysis The stress and strain fields atevery discrete time can be obtained The temperature field isjust the thermal load for the stress analysis rather than theobjective of the paper and so it is out of discussion
Figure 4 shows the stress curve of the crack tip (119877 point)in one cycle it is obvious that the Mises equivalent stress isdominated by the 119884 component Its 119884 components of strainare shown in Figure 5
In the heating process (0ndash10min) the thermal strainincreases with the temperature and the model cannot expand
0 10 20 30 40 50 60 70 80 90 100
0
100
200
300
400
Stre
ss (M
Pa)
X stress componentY stress component
XY shear stress componentMises equivalent stress
Time (min)
minus200
minus100
Figure 4 Stress components of the crack tip
freely due to the restrictions at both ends so compressiveelastic strain and stress are produced the plastic compressivestrain appears when the stress reaches the compressive yieldpoint of the material In the insulating process (10ndash90min)compressive creep strain appears in the common role of com-pressive stress and time simultaneously the elastic compres-sive strain and compressive stress decrease and the thermalstrain and the plastic strain keep constant In the cooling pro-cess (90ndash100min) the thermal strain and compressive elasticstrain gradually disappear and the model trends to its initialshape Because the model is fixed at both ends and the com-pressive plastic and creep strain cannot get back freely thetensile stress which is normal to the crack faces is produced
If the tensile stress reaches the reverse yield point thematerial will yield reversely The crack opens gradually withthe increase of tensile stress
The pressure on points 119860 and 119877 is shown in Figure 6In the heating process (0ndash10min) the pressure on the crackfaces increases for the temperature of point 119860 is higher thanthat of point 119877 so the pressure on point 119860 is higher In theinsulating process (10ndash90min) the pressure on the crackfaces decreases slowly owing to stress relaxation the creeprate of point119860 is larger so its pressure decreasesmore sharplyIn the cooling process (90ndash100min) the pressure declinesrapidly the pressure on points 119860 and 119877 becomes zero at
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
4 Advances in Mechanical Engineering
0 10 20 30 40 50 60 70 80 90 100
0
2
4
6
8
10
12
Time (min)
Y co
mpo
nent
s str
ain
Elastic strainPlastic strain
Creep strainThermal strain
times10minus3
minus4
minus2
Figure 5 Strain components in the 119884 direction of the crack tip
0 20 40 60 80 1000
20
40
60
80
100
120
140
160
180
Time (min)
Pres
sure
(MPa
)
119877 point119860 point
Figure 6 The pressure on points 119860 and 119877
95min which proves that the two crack faces separate at thistime Figure 7 shows the displacements in the 119884 direction ofpoints 119860 and 1198601015840 as well as their relative ones
Figure 8 is the deformation of the crack region at 100min(the deformation is magnified to 10 times) The red areaaround the crack tip is the tensile plastic zone outside whichis a green area Some compressive plastic strain produced inthe heating process remains here in various extents so it iscalled compressive plastic zone
One point is picked up in each of the tensile plastic zoneand compressive plastic zone randomly marked as 119879 and 119862Figure 9 gives the strain components in the 119884 direction ofpoints 119879 and 119862 Figure 10 gives their 119884 components of stress
0 20 40 60 80 100
0
2
4
6
8
Disp
lace
men
t (m
m)
Relative displacement
Time (min)
times10minus3
minus4
minus2
119860 point119860998400 point
Figure 7 The displacements in the 119884 direction of points 119860 and 1198601015840as well as their relative one
MNMX
Figure 8 The deformation of the crack region at 100min
From the figures stress and the elastic strain of the two pointsget to zero at the moment of the crack opening which is justas the crack tip
According to the above analysis it can be deduced thatthe compressive plastic strains produced in heating andcompressive creep strain produced in insulating cannot comeback freely during cooling which will make the CTFC openIt can be also concluded that the more the compressiveinelastic strain is during heating and insulating the larger thetensile stress the tensile strain and the opening displacementof CTFC will be at the end of cooling
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 5
0
1
2
3
4
0 20 40 60 80 100
Time (min)
times10minus3
minus3
minus1
minus2
Elastic strain of 119862 pointPlastic strain of 119862 point
Elastic strain of 119879 pointPlastic strain of 119879 point
Stra
in co
mpo
nent
s in119884
dire
ctio
n
Figure 9 Strain components in the 119884 direction of points 119879 and 119862
4 Controlling Parameters of CTFC
The CTFC opens gradually in the late cooling process Atthis time the temperature of the material around the crackhas become lower than the creep temperature Therefore thecreep fracturemechanismparameter119862lowast ismeaningless to theCTFC Suppose the stress intensity factor or 119869-integral will bethe controlling parameter of the CTFC in the following
41 Stress Intensity Factor Method The stress intensity factorof a crack for a linear elastic fracture mechanics analysis maybe computed according to a fit of the nodal displacementsin the vicinity of the crack [11 12] In the small-scale yieldingcondition the factor can be modified according to [13]
Generally the small-scale yielding condition for a cracklength 119886 is [13]
119903119901
119886le 01 (2)
where 119903119901is the size of plastic zone here it is themean distance
from the boundary of the tensile plastic zone to the cracktip The CTFC is always quite shallow [6] comparatively theplastic zone is rather largeThe small-scale yielding conditioncan be satisfied only when the singularity of the crack issmall
42 Proposition of aModified 119869-IntegralMethod The classicaldefinition of 119869-integral is given by
119869 = int120574
119908119889119910 minus int120574
T120597u120597119909119889119904 (3)
where 120574 represents any path surrounding the crack tip 119904 is thedistance along the path T is the stress tensor vector u is thedisplacement tensor vector 119908 is the strain energy density inthe monotonic loading process
0
50
100
150
200
250
300
Stre
ss co
mpo
nent
s in
di
rect
ion
(MPa
)
minus200
minus150
minus100
minus50
0 20 40 60 80 100
Time (min)
119884
119862 point119879 point
Figure 10 Stress components in the 119884 direction of points 119879 and 119862
From the previous stress analysis of the CTFC thethermal stress near the crack does not loadmonotonously in atemperature cycle which does not meet the path-independ-ent conditionMoreover affected by the residual compressiveplastic strain field at the end of cooling the integral might benegative Thus the 119869-integral method must be improved toapplicable to be the CTFC
T120597u120597119909= (120590119909119899119909+ 120590119909119910119899119910)120597119906119909
120597119909+ (120590119910119899119910+ 120590119909119910119899119909)120597119906119910
120597119909 (4)
where 119899 is the unit outer normal vector to path 120574 It isobvious that the later term of the 119869-integral is naturally pathindependent and immune to the compressive plastic strainfield
According to the plastic deformation theory the strainenergy density is represented by
119908 = 119908119909119909+ 119908119909119910+ 119908119910119910
119908119894119895= int
120576
0
120590119894119895(119905) sdot 120576119894119895(119905) 119889119905
(5)
where120590119894119895 120576119894119895(119894 119895 = 119909 119910) are respectively the stress and strain
components Here only the improvement of 119908119910119910
similar tothat of 119908
119909119909and 119908
119909119910is discussed in detail
From the stress analysis of the CTFC the 119884 componentsof stress and elastic strain around the crack can be zero almostat the same timewhen the crack opensMoreover the coolingperiod can be regarded as amonotonic loading processThusonly the strain energy in this period should be computed
119908yy = int1199052
1199051
120590119910119910(119905) sdot 120576119910119910(119905) 119889119905 (6)
where 1199051stands for the moment of the opening and 119905
2is the
terminal time of cooling In order to eliminate the effect of
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
6 Advances in Mechanical Engineering
the residual compressive plastic strain field 120576119910119910(119905) is obtained
as
120576119910119910(119905) = 120576
119890
119910119910(119905) + 120576
119901
119910y (119905) minus 1205761199010
119910119910 (7)
where 120576119890yy(119905) 120576119901
yy(119905) are separately the elastic and plastic com-ponents of strain at the end of cooling 1205761199010
119910119910is the compressive
plastic strain produced in the heating processThe improvement of 119869-integral is put forward according
to the analysis of CTFC according to its definition Theintegral computed in this way is path independent Thus itcould be the controlling parameter of CTFC
In the linear elasticity small-scale yielding conditionsand plane stress conditions 119869-integral is related to stressintensity factor by the following relationship
119869 =1198702
Ι
119864 (8)
where 119864 is the elastic modulus Therefore the stress intensityfactor119870
Ιcan validate the modified 119869-integral cursorily
5 Examples
Take a group of concentric circles whose center is crack tipand radius is 119903 as the integral paths Figure 11 shows therelation between themodified 119869-integral and 119903119903
119901 Clearly the
modified 119869-integral is path independent as long as the integralpath is within the compressive plastic zone
The stress intensity factor and the modified 119869-integral arecalculated under different conditions The results are showninTables 3 4 and 5 In (8) the elasticmodulus119864 is 71175MPa
It is obvious that the stress intensity factor and the 119869-integral increases when the insulating time prolongs themaximum temperature increases and the crack grows Fromcalculating the stress intensity factor can be used as thecontrolling parameter of the CTFC only with a low heatingtemperature a short period of holding time and a deep crack
6 CTFC Propagation Test
The creep-thermal fatigue test machine is composed ofclamping mechanism infrared radiation heating devicehydropneumatic cooling system and control system Theoperational principle of the creep-thermal fatigue testmachine is shown in Figure 12 the specimens can be turnedupside down by the driving shaft which makes it easy andquick to heat and cool the specimensThe detailed size of thespecimen with cracks is shown in Figure 13The first secondand third specimens are used in the creep-thermal fatiguetest the temperature of the specimen cycles is from 30∘C to210∘C and the thermal cycle insulating time is 5 minutes at210∘C every time All the specimens undergo 1000 thermalcycles The length of the crack is detected for each 100 cyclesand the results are shown in Figure 14
The secant method [14] is used to measure the crackgrowth rate that is to calculate the gradient between two
0 1 2
1
04
05
0402
06
06
07
08
08
09
12 14 16 18119903119903119901
119869(J
mmminus2)
Figure 11 The relationship of 119869-integral and 119903119903119901
++
1 2 3 4 5 6 7 8
Figure 12 Operational principle diagram of the creep-thermalfatigue test machine 1 driving shaft 2 stock of the specimen 3unloading bolt 4 tightening bolt 5 infrared radiation heatingdevice 6 specimen 7 hydropneumatic cooling system and 8 armof the test machine
adjacent data points on the 119886-119873 curve to get the crack growthrate as
119889119886
119889119873asympΔ119886
Δ119873=119886119894+1minus 119886119894
119873119894+1minus 119873119894
(9)
where 119889119886119889119873 is the crack propagation under one thermalcycle Δ119886 is the crack propagation under Δ119873 thermal cycles119886119894and 119873
119894are the crack length and number of cycles of load
corresponding to the 119894th data pointAccording to the Paris formula
119889119886
119889119873= 119862119869119899
(10)
where 119862 and 119899 are constants related to material propertyenvironmental media and the geometry of the specimen
The creep-thermal fatigue 119869-integral value in correspond-ing conditions can be obtained according to the modified J-integral method referred to in the front of the paper the datapoints of ln(119889119886119889119873) and ln 119869 are shown in Figure 15 hence
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 7
Table 3 The crack length is 07mm the maximum temperature is 210∘C
Holding time(min)
Size of plasticzone (mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)0 00700 1944 05117 0531080 01050 09289150 01313 10750240 01570 11950
Table 4 The crack length is 07mm the insulating time is 80min
Maximumtemperature (∘C)
Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)180 0 239 00080 00082190 00630 1594 03440 03570200 00875 2115 05970 06285220 01778 14148
Table 5 The maximum temperature is 200∘C the insulating time is 80min
Crack length (mm) Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)05 00623 0463309 01025 0702912 01220 0808815 01493 2526 08568 08965
02
02
3plusmn01
25plusmn01
25plusmn01
4
6
2
2
2
AAA
A A
002
002
002
002
6h7
20plusmn01
13 1370
Figure 13 Specimen of creep-thermal fatigue
119899 = 251 119862 = 872times10minus6 can be obtained by some regressionanalysis of formula (10) using these data points
7 Conclusions
(1) The crack is closed under the compressive thermalstress during the heating and insulating processesBecause the inelastic strain generating in the two
0 200 400 600 800 10000
05
1
15
2
25
3
35
Number of thermal cycles
Crac
k pr
opag
atio
n ra
te (m
m)
1
2
3
Figure 14 Propagation length of the CTFC
processes cannot be back by itself the tensile stressduring the late cooling process which is normal to thecrack faces makes the crack gradually open
(2) When the CTFC is opening the temperature of thematerial around it has become lower than the creep
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
8 Advances in Mechanical Engineering
119869-integral100 101 102
10minus2
10minus3
10minus4
10minus5
1
2
3
Figure 15 (119889119886119889119873)mdash119869 data in double logarithmic coordinate
temperature therefore the creep fracture mechanismparameter 119862lowast is not applicable The stress intensityfactor and 119869-integral can be used as the controllingparameters The stress intensity factor can be usedonly with a short insulating time a low maximumtemperature and a deep crack The 119869-integral shouldbe modified in order to be applicable to the CTFCthat is to take the moment of opening as the startingof the strain energy density integral and eliminatethe effect of the residual compressive plastic strain by(7) It has been proved that the modified 119869-integralmethod is path independent and valid It is used fornot only theCTFCbut also any unloading crack or thecrack in the residual compressive plastic strain fieldThe stress intensity factor and 119869-integral increasewhen the insulating time prolongs the maximumtemperature rises and the crack grows
(3) Experimental results show that the modified 119869-integral can be used as the control value of the CTFCpropagation The research on the controlling param-eters of CTFC provides the theoretical basis for itsgrowth
Acknowledgments
This project is supported by the Natural Science Foundationof Liaoning Province (Grant no 201102167) Aeronautical Sci-ence Foundation of China (Grant no 20110450001) LiaoningExcellent Talents in University (Grant no LJQ2011012) andthe Science Program of Shenyang (Grant No F12-069-2-00)
References
[1] N Gao M Brown K J Miller and P A S Reed ldquoAn investiga-tion of crack growth behaviour under creep-fatigue conditionrdquo
Materials Science and Engineering A vol 410 no 25 pp 67ndash712005
[2] M F Geng ldquoAn idea for predicting crack growth time tofracture under creep-fatigue conditionsrdquoMaterials Science andEngineering A vol 257 no 2 pp 250ndash255 1998
[3] Y L Lu L J Chen P K Liaw et al ldquoEffects of temperature andhold time on creep-fatigue crack-growth behaviorrdquo MaterialsScience and Engineering A vol 429 no 1-2 pp 1ndash10 2006
[4] A T Yokobori and T Satoh ldquoThe characterization of creepcrack growth rate and its life of TiAl inter-metallic compoundwith full lamellar microstructurerdquo International Journal ofPressure Vessels and Piping vol 88 no 7 pp 435ndash440 2011
[5] G Marahleh A R I Kheder and H F Hamad ldquoCreep-lifeprediction of service-exposed turbine bladesrdquoMaterials Sciencevol 42 no 4 pp 476ndash481 2006
[6] V Maillot A Fissolo G Degallaix and S Degallaix ldquoThermalfatigue crack networks parameters and stability an experimen-tal studyrdquo International Journal of Solids and Structures vol 42no 2 pp 759ndash769 2005
[7] J Lansinger T Hansson andO Clevfors ldquoFatigue crack growthunder combined thermal cycling and mechanical loadingrdquoInternational Journal of Fatigue vol 29 no 7 pp 1383ndash13902007
[8] N Haddar A Fissolo and V Maillot ldquoThermal fatigue cracknetworks an computational studyrdquo International Journal ofSolids and Structures vol 42 no 2 pp 771ndash788 2005
[9] G T Yang Theory of Elasticity and Plasticity China BuildingMaterials Industry Press Beijing China 2005
[10] H Kraus Creep Analysis Wiley International Science Publica-tion New York NY USA 1980
[11] Z D Jiang A Zeghloul G Bezine and J Petit ldquoStress intensityfactors of parallel cracks in a finite width sheetrdquo EngineeringFracture Mechanics vol 35 no 6 pp 1073ndash1079 1990
[12] G Agnihotri ldquoStress analysis of a crack using the finite elementmethodrdquo Engineering FractureMechanics vol 44 no 1 pp 109ndash125 1993
[13] R Narasimhan and A J Rosakis ldquoA finite element analysis ofsmall-scale yielding near a stationary crack under plane stressrdquoJournal of the Mechanics and Physics of Solids vol 36 no 1 pp77ndash117 1988
[14] P Bensussan High Temperature Fracture Mechanisms andMechanic Mechanical Engineering Publications 2005
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
4 Advances in Mechanical Engineering
0 10 20 30 40 50 60 70 80 90 100
0
2
4
6
8
10
12
Time (min)
Y co
mpo
nent
s str
ain
Elastic strainPlastic strain
Creep strainThermal strain
times10minus3
minus4
minus2
Figure 5 Strain components in the 119884 direction of the crack tip
0 20 40 60 80 1000
20
40
60
80
100
120
140
160
180
Time (min)
Pres
sure
(MPa
)
119877 point119860 point
Figure 6 The pressure on points 119860 and 119877
95min which proves that the two crack faces separate at thistime Figure 7 shows the displacements in the 119884 direction ofpoints 119860 and 1198601015840 as well as their relative ones
Figure 8 is the deformation of the crack region at 100min(the deformation is magnified to 10 times) The red areaaround the crack tip is the tensile plastic zone outside whichis a green area Some compressive plastic strain produced inthe heating process remains here in various extents so it iscalled compressive plastic zone
One point is picked up in each of the tensile plastic zoneand compressive plastic zone randomly marked as 119879 and 119862Figure 9 gives the strain components in the 119884 direction ofpoints 119879 and 119862 Figure 10 gives their 119884 components of stress
0 20 40 60 80 100
0
2
4
6
8
Disp
lace
men
t (m
m)
Relative displacement
Time (min)
times10minus3
minus4
minus2
119860 point119860998400 point
Figure 7 The displacements in the 119884 direction of points 119860 and 1198601015840as well as their relative one
MNMX
Figure 8 The deformation of the crack region at 100min
From the figures stress and the elastic strain of the two pointsget to zero at the moment of the crack opening which is justas the crack tip
According to the above analysis it can be deduced thatthe compressive plastic strains produced in heating andcompressive creep strain produced in insulating cannot comeback freely during cooling which will make the CTFC openIt can be also concluded that the more the compressiveinelastic strain is during heating and insulating the larger thetensile stress the tensile strain and the opening displacementof CTFC will be at the end of cooling
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 5
0
1
2
3
4
0 20 40 60 80 100
Time (min)
times10minus3
minus3
minus1
minus2
Elastic strain of 119862 pointPlastic strain of 119862 point
Elastic strain of 119879 pointPlastic strain of 119879 point
Stra
in co
mpo
nent
s in119884
dire
ctio
n
Figure 9 Strain components in the 119884 direction of points 119879 and 119862
4 Controlling Parameters of CTFC
The CTFC opens gradually in the late cooling process Atthis time the temperature of the material around the crackhas become lower than the creep temperature Therefore thecreep fracturemechanismparameter119862lowast ismeaningless to theCTFC Suppose the stress intensity factor or 119869-integral will bethe controlling parameter of the CTFC in the following
41 Stress Intensity Factor Method The stress intensity factorof a crack for a linear elastic fracture mechanics analysis maybe computed according to a fit of the nodal displacementsin the vicinity of the crack [11 12] In the small-scale yieldingcondition the factor can be modified according to [13]
Generally the small-scale yielding condition for a cracklength 119886 is [13]
119903119901
119886le 01 (2)
where 119903119901is the size of plastic zone here it is themean distance
from the boundary of the tensile plastic zone to the cracktip The CTFC is always quite shallow [6] comparatively theplastic zone is rather largeThe small-scale yielding conditioncan be satisfied only when the singularity of the crack issmall
42 Proposition of aModified 119869-IntegralMethod The classicaldefinition of 119869-integral is given by
119869 = int120574
119908119889119910 minus int120574
T120597u120597119909119889119904 (3)
where 120574 represents any path surrounding the crack tip 119904 is thedistance along the path T is the stress tensor vector u is thedisplacement tensor vector 119908 is the strain energy density inthe monotonic loading process
0
50
100
150
200
250
300
Stre
ss co
mpo
nent
s in
di
rect
ion
(MPa
)
minus200
minus150
minus100
minus50
0 20 40 60 80 100
Time (min)
119884
119862 point119879 point
Figure 10 Stress components in the 119884 direction of points 119879 and 119862
From the previous stress analysis of the CTFC thethermal stress near the crack does not loadmonotonously in atemperature cycle which does not meet the path-independ-ent conditionMoreover affected by the residual compressiveplastic strain field at the end of cooling the integral might benegative Thus the 119869-integral method must be improved toapplicable to be the CTFC
T120597u120597119909= (120590119909119899119909+ 120590119909119910119899119910)120597119906119909
120597119909+ (120590119910119899119910+ 120590119909119910119899119909)120597119906119910
120597119909 (4)
where 119899 is the unit outer normal vector to path 120574 It isobvious that the later term of the 119869-integral is naturally pathindependent and immune to the compressive plastic strainfield
According to the plastic deformation theory the strainenergy density is represented by
119908 = 119908119909119909+ 119908119909119910+ 119908119910119910
119908119894119895= int
120576
0
120590119894119895(119905) sdot 120576119894119895(119905) 119889119905
(5)
where120590119894119895 120576119894119895(119894 119895 = 119909 119910) are respectively the stress and strain
components Here only the improvement of 119908119910119910
similar tothat of 119908
119909119909and 119908
119909119910is discussed in detail
From the stress analysis of the CTFC the 119884 componentsof stress and elastic strain around the crack can be zero almostat the same timewhen the crack opensMoreover the coolingperiod can be regarded as amonotonic loading processThusonly the strain energy in this period should be computed
119908yy = int1199052
1199051
120590119910119910(119905) sdot 120576119910119910(119905) 119889119905 (6)
where 1199051stands for the moment of the opening and 119905
2is the
terminal time of cooling In order to eliminate the effect of
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
6 Advances in Mechanical Engineering
the residual compressive plastic strain field 120576119910119910(119905) is obtained
as
120576119910119910(119905) = 120576
119890
119910119910(119905) + 120576
119901
119910y (119905) minus 1205761199010
119910119910 (7)
where 120576119890yy(119905) 120576119901
yy(119905) are separately the elastic and plastic com-ponents of strain at the end of cooling 1205761199010
119910119910is the compressive
plastic strain produced in the heating processThe improvement of 119869-integral is put forward according
to the analysis of CTFC according to its definition Theintegral computed in this way is path independent Thus itcould be the controlling parameter of CTFC
In the linear elasticity small-scale yielding conditionsand plane stress conditions 119869-integral is related to stressintensity factor by the following relationship
119869 =1198702
Ι
119864 (8)
where 119864 is the elastic modulus Therefore the stress intensityfactor119870
Ιcan validate the modified 119869-integral cursorily
5 Examples
Take a group of concentric circles whose center is crack tipand radius is 119903 as the integral paths Figure 11 shows therelation between themodified 119869-integral and 119903119903
119901 Clearly the
modified 119869-integral is path independent as long as the integralpath is within the compressive plastic zone
The stress intensity factor and the modified 119869-integral arecalculated under different conditions The results are showninTables 3 4 and 5 In (8) the elasticmodulus119864 is 71175MPa
It is obvious that the stress intensity factor and the 119869-integral increases when the insulating time prolongs themaximum temperature increases and the crack grows Fromcalculating the stress intensity factor can be used as thecontrolling parameter of the CTFC only with a low heatingtemperature a short period of holding time and a deep crack
6 CTFC Propagation Test
The creep-thermal fatigue test machine is composed ofclamping mechanism infrared radiation heating devicehydropneumatic cooling system and control system Theoperational principle of the creep-thermal fatigue testmachine is shown in Figure 12 the specimens can be turnedupside down by the driving shaft which makes it easy andquick to heat and cool the specimensThe detailed size of thespecimen with cracks is shown in Figure 13The first secondand third specimens are used in the creep-thermal fatiguetest the temperature of the specimen cycles is from 30∘C to210∘C and the thermal cycle insulating time is 5 minutes at210∘C every time All the specimens undergo 1000 thermalcycles The length of the crack is detected for each 100 cyclesand the results are shown in Figure 14
The secant method [14] is used to measure the crackgrowth rate that is to calculate the gradient between two
0 1 2
1
04
05
0402
06
06
07
08
08
09
12 14 16 18119903119903119901
119869(J
mmminus2)
Figure 11 The relationship of 119869-integral and 119903119903119901
++
1 2 3 4 5 6 7 8
Figure 12 Operational principle diagram of the creep-thermalfatigue test machine 1 driving shaft 2 stock of the specimen 3unloading bolt 4 tightening bolt 5 infrared radiation heatingdevice 6 specimen 7 hydropneumatic cooling system and 8 armof the test machine
adjacent data points on the 119886-119873 curve to get the crack growthrate as
119889119886
119889119873asympΔ119886
Δ119873=119886119894+1minus 119886119894
119873119894+1minus 119873119894
(9)
where 119889119886119889119873 is the crack propagation under one thermalcycle Δ119886 is the crack propagation under Δ119873 thermal cycles119886119894and 119873
119894are the crack length and number of cycles of load
corresponding to the 119894th data pointAccording to the Paris formula
119889119886
119889119873= 119862119869119899
(10)
where 119862 and 119899 are constants related to material propertyenvironmental media and the geometry of the specimen
The creep-thermal fatigue 119869-integral value in correspond-ing conditions can be obtained according to the modified J-integral method referred to in the front of the paper the datapoints of ln(119889119886119889119873) and ln 119869 are shown in Figure 15 hence
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 7
Table 3 The crack length is 07mm the maximum temperature is 210∘C
Holding time(min)
Size of plasticzone (mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)0 00700 1944 05117 0531080 01050 09289150 01313 10750240 01570 11950
Table 4 The crack length is 07mm the insulating time is 80min
Maximumtemperature (∘C)
Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)180 0 239 00080 00082190 00630 1594 03440 03570200 00875 2115 05970 06285220 01778 14148
Table 5 The maximum temperature is 200∘C the insulating time is 80min
Crack length (mm) Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)05 00623 0463309 01025 0702912 01220 0808815 01493 2526 08568 08965
02
02
3plusmn01
25plusmn01
25plusmn01
4
6
2
2
2
AAA
A A
002
002
002
002
6h7
20plusmn01
13 1370
Figure 13 Specimen of creep-thermal fatigue
119899 = 251 119862 = 872times10minus6 can be obtained by some regressionanalysis of formula (10) using these data points
7 Conclusions
(1) The crack is closed under the compressive thermalstress during the heating and insulating processesBecause the inelastic strain generating in the two
0 200 400 600 800 10000
05
1
15
2
25
3
35
Number of thermal cycles
Crac
k pr
opag
atio
n ra
te (m
m)
1
2
3
Figure 14 Propagation length of the CTFC
processes cannot be back by itself the tensile stressduring the late cooling process which is normal to thecrack faces makes the crack gradually open
(2) When the CTFC is opening the temperature of thematerial around it has become lower than the creep
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
8 Advances in Mechanical Engineering
119869-integral100 101 102
10minus2
10minus3
10minus4
10minus5
1
2
3
Figure 15 (119889119886119889119873)mdash119869 data in double logarithmic coordinate
temperature therefore the creep fracture mechanismparameter 119862lowast is not applicable The stress intensityfactor and 119869-integral can be used as the controllingparameters The stress intensity factor can be usedonly with a short insulating time a low maximumtemperature and a deep crack The 119869-integral shouldbe modified in order to be applicable to the CTFCthat is to take the moment of opening as the startingof the strain energy density integral and eliminatethe effect of the residual compressive plastic strain by(7) It has been proved that the modified 119869-integralmethod is path independent and valid It is used fornot only theCTFCbut also any unloading crack or thecrack in the residual compressive plastic strain fieldThe stress intensity factor and 119869-integral increasewhen the insulating time prolongs the maximumtemperature rises and the crack grows
(3) Experimental results show that the modified 119869-integral can be used as the control value of the CTFCpropagation The research on the controlling param-eters of CTFC provides the theoretical basis for itsgrowth
Acknowledgments
This project is supported by the Natural Science Foundationof Liaoning Province (Grant no 201102167) Aeronautical Sci-ence Foundation of China (Grant no 20110450001) LiaoningExcellent Talents in University (Grant no LJQ2011012) andthe Science Program of Shenyang (Grant No F12-069-2-00)
References
[1] N Gao M Brown K J Miller and P A S Reed ldquoAn investiga-tion of crack growth behaviour under creep-fatigue conditionrdquo
Materials Science and Engineering A vol 410 no 25 pp 67ndash712005
[2] M F Geng ldquoAn idea for predicting crack growth time tofracture under creep-fatigue conditionsrdquoMaterials Science andEngineering A vol 257 no 2 pp 250ndash255 1998
[3] Y L Lu L J Chen P K Liaw et al ldquoEffects of temperature andhold time on creep-fatigue crack-growth behaviorrdquo MaterialsScience and Engineering A vol 429 no 1-2 pp 1ndash10 2006
[4] A T Yokobori and T Satoh ldquoThe characterization of creepcrack growth rate and its life of TiAl inter-metallic compoundwith full lamellar microstructurerdquo International Journal ofPressure Vessels and Piping vol 88 no 7 pp 435ndash440 2011
[5] G Marahleh A R I Kheder and H F Hamad ldquoCreep-lifeprediction of service-exposed turbine bladesrdquoMaterials Sciencevol 42 no 4 pp 476ndash481 2006
[6] V Maillot A Fissolo G Degallaix and S Degallaix ldquoThermalfatigue crack networks parameters and stability an experimen-tal studyrdquo International Journal of Solids and Structures vol 42no 2 pp 759ndash769 2005
[7] J Lansinger T Hansson andO Clevfors ldquoFatigue crack growthunder combined thermal cycling and mechanical loadingrdquoInternational Journal of Fatigue vol 29 no 7 pp 1383ndash13902007
[8] N Haddar A Fissolo and V Maillot ldquoThermal fatigue cracknetworks an computational studyrdquo International Journal ofSolids and Structures vol 42 no 2 pp 771ndash788 2005
[9] G T Yang Theory of Elasticity and Plasticity China BuildingMaterials Industry Press Beijing China 2005
[10] H Kraus Creep Analysis Wiley International Science Publica-tion New York NY USA 1980
[11] Z D Jiang A Zeghloul G Bezine and J Petit ldquoStress intensityfactors of parallel cracks in a finite width sheetrdquo EngineeringFracture Mechanics vol 35 no 6 pp 1073ndash1079 1990
[12] G Agnihotri ldquoStress analysis of a crack using the finite elementmethodrdquo Engineering FractureMechanics vol 44 no 1 pp 109ndash125 1993
[13] R Narasimhan and A J Rosakis ldquoA finite element analysis ofsmall-scale yielding near a stationary crack under plane stressrdquoJournal of the Mechanics and Physics of Solids vol 36 no 1 pp77ndash117 1988
[14] P Bensussan High Temperature Fracture Mechanisms andMechanic Mechanical Engineering Publications 2005
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 5
0
1
2
3
4
0 20 40 60 80 100
Time (min)
times10minus3
minus3
minus1
minus2
Elastic strain of 119862 pointPlastic strain of 119862 point
Elastic strain of 119879 pointPlastic strain of 119879 point
Stra
in co
mpo
nent
s in119884
dire
ctio
n
Figure 9 Strain components in the 119884 direction of points 119879 and 119862
4 Controlling Parameters of CTFC
The CTFC opens gradually in the late cooling process Atthis time the temperature of the material around the crackhas become lower than the creep temperature Therefore thecreep fracturemechanismparameter119862lowast ismeaningless to theCTFC Suppose the stress intensity factor or 119869-integral will bethe controlling parameter of the CTFC in the following
41 Stress Intensity Factor Method The stress intensity factorof a crack for a linear elastic fracture mechanics analysis maybe computed according to a fit of the nodal displacementsin the vicinity of the crack [11 12] In the small-scale yieldingcondition the factor can be modified according to [13]
Generally the small-scale yielding condition for a cracklength 119886 is [13]
119903119901
119886le 01 (2)
where 119903119901is the size of plastic zone here it is themean distance
from the boundary of the tensile plastic zone to the cracktip The CTFC is always quite shallow [6] comparatively theplastic zone is rather largeThe small-scale yielding conditioncan be satisfied only when the singularity of the crack issmall
42 Proposition of aModified 119869-IntegralMethod The classicaldefinition of 119869-integral is given by
119869 = int120574
119908119889119910 minus int120574
T120597u120597119909119889119904 (3)
where 120574 represents any path surrounding the crack tip 119904 is thedistance along the path T is the stress tensor vector u is thedisplacement tensor vector 119908 is the strain energy density inthe monotonic loading process
0
50
100
150
200
250
300
Stre
ss co
mpo
nent
s in
di
rect
ion
(MPa
)
minus200
minus150
minus100
minus50
0 20 40 60 80 100
Time (min)
119884
119862 point119879 point
Figure 10 Stress components in the 119884 direction of points 119879 and 119862
From the previous stress analysis of the CTFC thethermal stress near the crack does not loadmonotonously in atemperature cycle which does not meet the path-independ-ent conditionMoreover affected by the residual compressiveplastic strain field at the end of cooling the integral might benegative Thus the 119869-integral method must be improved toapplicable to be the CTFC
T120597u120597119909= (120590119909119899119909+ 120590119909119910119899119910)120597119906119909
120597119909+ (120590119910119899119910+ 120590119909119910119899119909)120597119906119910
120597119909 (4)
where 119899 is the unit outer normal vector to path 120574 It isobvious that the later term of the 119869-integral is naturally pathindependent and immune to the compressive plastic strainfield
According to the plastic deformation theory the strainenergy density is represented by
119908 = 119908119909119909+ 119908119909119910+ 119908119910119910
119908119894119895= int
120576
0
120590119894119895(119905) sdot 120576119894119895(119905) 119889119905
(5)
where120590119894119895 120576119894119895(119894 119895 = 119909 119910) are respectively the stress and strain
components Here only the improvement of 119908119910119910
similar tothat of 119908
119909119909and 119908
119909119910is discussed in detail
From the stress analysis of the CTFC the 119884 componentsof stress and elastic strain around the crack can be zero almostat the same timewhen the crack opensMoreover the coolingperiod can be regarded as amonotonic loading processThusonly the strain energy in this period should be computed
119908yy = int1199052
1199051
120590119910119910(119905) sdot 120576119910119910(119905) 119889119905 (6)
where 1199051stands for the moment of the opening and 119905
2is the
terminal time of cooling In order to eliminate the effect of
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
6 Advances in Mechanical Engineering
the residual compressive plastic strain field 120576119910119910(119905) is obtained
as
120576119910119910(119905) = 120576
119890
119910119910(119905) + 120576
119901
119910y (119905) minus 1205761199010
119910119910 (7)
where 120576119890yy(119905) 120576119901
yy(119905) are separately the elastic and plastic com-ponents of strain at the end of cooling 1205761199010
119910119910is the compressive
plastic strain produced in the heating processThe improvement of 119869-integral is put forward according
to the analysis of CTFC according to its definition Theintegral computed in this way is path independent Thus itcould be the controlling parameter of CTFC
In the linear elasticity small-scale yielding conditionsand plane stress conditions 119869-integral is related to stressintensity factor by the following relationship
119869 =1198702
Ι
119864 (8)
where 119864 is the elastic modulus Therefore the stress intensityfactor119870
Ιcan validate the modified 119869-integral cursorily
5 Examples
Take a group of concentric circles whose center is crack tipand radius is 119903 as the integral paths Figure 11 shows therelation between themodified 119869-integral and 119903119903
119901 Clearly the
modified 119869-integral is path independent as long as the integralpath is within the compressive plastic zone
The stress intensity factor and the modified 119869-integral arecalculated under different conditions The results are showninTables 3 4 and 5 In (8) the elasticmodulus119864 is 71175MPa
It is obvious that the stress intensity factor and the 119869-integral increases when the insulating time prolongs themaximum temperature increases and the crack grows Fromcalculating the stress intensity factor can be used as thecontrolling parameter of the CTFC only with a low heatingtemperature a short period of holding time and a deep crack
6 CTFC Propagation Test
The creep-thermal fatigue test machine is composed ofclamping mechanism infrared radiation heating devicehydropneumatic cooling system and control system Theoperational principle of the creep-thermal fatigue testmachine is shown in Figure 12 the specimens can be turnedupside down by the driving shaft which makes it easy andquick to heat and cool the specimensThe detailed size of thespecimen with cracks is shown in Figure 13The first secondand third specimens are used in the creep-thermal fatiguetest the temperature of the specimen cycles is from 30∘C to210∘C and the thermal cycle insulating time is 5 minutes at210∘C every time All the specimens undergo 1000 thermalcycles The length of the crack is detected for each 100 cyclesand the results are shown in Figure 14
The secant method [14] is used to measure the crackgrowth rate that is to calculate the gradient between two
0 1 2
1
04
05
0402
06
06
07
08
08
09
12 14 16 18119903119903119901
119869(J
mmminus2)
Figure 11 The relationship of 119869-integral and 119903119903119901
++
1 2 3 4 5 6 7 8
Figure 12 Operational principle diagram of the creep-thermalfatigue test machine 1 driving shaft 2 stock of the specimen 3unloading bolt 4 tightening bolt 5 infrared radiation heatingdevice 6 specimen 7 hydropneumatic cooling system and 8 armof the test machine
adjacent data points on the 119886-119873 curve to get the crack growthrate as
119889119886
119889119873asympΔ119886
Δ119873=119886119894+1minus 119886119894
119873119894+1minus 119873119894
(9)
where 119889119886119889119873 is the crack propagation under one thermalcycle Δ119886 is the crack propagation under Δ119873 thermal cycles119886119894and 119873
119894are the crack length and number of cycles of load
corresponding to the 119894th data pointAccording to the Paris formula
119889119886
119889119873= 119862119869119899
(10)
where 119862 and 119899 are constants related to material propertyenvironmental media and the geometry of the specimen
The creep-thermal fatigue 119869-integral value in correspond-ing conditions can be obtained according to the modified J-integral method referred to in the front of the paper the datapoints of ln(119889119886119889119873) and ln 119869 are shown in Figure 15 hence
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 7
Table 3 The crack length is 07mm the maximum temperature is 210∘C
Holding time(min)
Size of plasticzone (mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)0 00700 1944 05117 0531080 01050 09289150 01313 10750240 01570 11950
Table 4 The crack length is 07mm the insulating time is 80min
Maximumtemperature (∘C)
Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)180 0 239 00080 00082190 00630 1594 03440 03570200 00875 2115 05970 06285220 01778 14148
Table 5 The maximum temperature is 200∘C the insulating time is 80min
Crack length (mm) Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)05 00623 0463309 01025 0702912 01220 0808815 01493 2526 08568 08965
02
02
3plusmn01
25plusmn01
25plusmn01
4
6
2
2
2
AAA
A A
002
002
002
002
6h7
20plusmn01
13 1370
Figure 13 Specimen of creep-thermal fatigue
119899 = 251 119862 = 872times10minus6 can be obtained by some regressionanalysis of formula (10) using these data points
7 Conclusions
(1) The crack is closed under the compressive thermalstress during the heating and insulating processesBecause the inelastic strain generating in the two
0 200 400 600 800 10000
05
1
15
2
25
3
35
Number of thermal cycles
Crac
k pr
opag
atio
n ra
te (m
m)
1
2
3
Figure 14 Propagation length of the CTFC
processes cannot be back by itself the tensile stressduring the late cooling process which is normal to thecrack faces makes the crack gradually open
(2) When the CTFC is opening the temperature of thematerial around it has become lower than the creep
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
8 Advances in Mechanical Engineering
119869-integral100 101 102
10minus2
10minus3
10minus4
10minus5
1
2
3
Figure 15 (119889119886119889119873)mdash119869 data in double logarithmic coordinate
temperature therefore the creep fracture mechanismparameter 119862lowast is not applicable The stress intensityfactor and 119869-integral can be used as the controllingparameters The stress intensity factor can be usedonly with a short insulating time a low maximumtemperature and a deep crack The 119869-integral shouldbe modified in order to be applicable to the CTFCthat is to take the moment of opening as the startingof the strain energy density integral and eliminatethe effect of the residual compressive plastic strain by(7) It has been proved that the modified 119869-integralmethod is path independent and valid It is used fornot only theCTFCbut also any unloading crack or thecrack in the residual compressive plastic strain fieldThe stress intensity factor and 119869-integral increasewhen the insulating time prolongs the maximumtemperature rises and the crack grows
(3) Experimental results show that the modified 119869-integral can be used as the control value of the CTFCpropagation The research on the controlling param-eters of CTFC provides the theoretical basis for itsgrowth
Acknowledgments
This project is supported by the Natural Science Foundationof Liaoning Province (Grant no 201102167) Aeronautical Sci-ence Foundation of China (Grant no 20110450001) LiaoningExcellent Talents in University (Grant no LJQ2011012) andthe Science Program of Shenyang (Grant No F12-069-2-00)
References
[1] N Gao M Brown K J Miller and P A S Reed ldquoAn investiga-tion of crack growth behaviour under creep-fatigue conditionrdquo
Materials Science and Engineering A vol 410 no 25 pp 67ndash712005
[2] M F Geng ldquoAn idea for predicting crack growth time tofracture under creep-fatigue conditionsrdquoMaterials Science andEngineering A vol 257 no 2 pp 250ndash255 1998
[3] Y L Lu L J Chen P K Liaw et al ldquoEffects of temperature andhold time on creep-fatigue crack-growth behaviorrdquo MaterialsScience and Engineering A vol 429 no 1-2 pp 1ndash10 2006
[4] A T Yokobori and T Satoh ldquoThe characterization of creepcrack growth rate and its life of TiAl inter-metallic compoundwith full lamellar microstructurerdquo International Journal ofPressure Vessels and Piping vol 88 no 7 pp 435ndash440 2011
[5] G Marahleh A R I Kheder and H F Hamad ldquoCreep-lifeprediction of service-exposed turbine bladesrdquoMaterials Sciencevol 42 no 4 pp 476ndash481 2006
[6] V Maillot A Fissolo G Degallaix and S Degallaix ldquoThermalfatigue crack networks parameters and stability an experimen-tal studyrdquo International Journal of Solids and Structures vol 42no 2 pp 759ndash769 2005
[7] J Lansinger T Hansson andO Clevfors ldquoFatigue crack growthunder combined thermal cycling and mechanical loadingrdquoInternational Journal of Fatigue vol 29 no 7 pp 1383ndash13902007
[8] N Haddar A Fissolo and V Maillot ldquoThermal fatigue cracknetworks an computational studyrdquo International Journal ofSolids and Structures vol 42 no 2 pp 771ndash788 2005
[9] G T Yang Theory of Elasticity and Plasticity China BuildingMaterials Industry Press Beijing China 2005
[10] H Kraus Creep Analysis Wiley International Science Publica-tion New York NY USA 1980
[11] Z D Jiang A Zeghloul G Bezine and J Petit ldquoStress intensityfactors of parallel cracks in a finite width sheetrdquo EngineeringFracture Mechanics vol 35 no 6 pp 1073ndash1079 1990
[12] G Agnihotri ldquoStress analysis of a crack using the finite elementmethodrdquo Engineering FractureMechanics vol 44 no 1 pp 109ndash125 1993
[13] R Narasimhan and A J Rosakis ldquoA finite element analysis ofsmall-scale yielding near a stationary crack under plane stressrdquoJournal of the Mechanics and Physics of Solids vol 36 no 1 pp77ndash117 1988
[14] P Bensussan High Temperature Fracture Mechanisms andMechanic Mechanical Engineering Publications 2005
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
6 Advances in Mechanical Engineering
the residual compressive plastic strain field 120576119910119910(119905) is obtained
as
120576119910119910(119905) = 120576
119890
119910119910(119905) + 120576
119901
119910y (119905) minus 1205761199010
119910119910 (7)
where 120576119890yy(119905) 120576119901
yy(119905) are separately the elastic and plastic com-ponents of strain at the end of cooling 1205761199010
119910119910is the compressive
plastic strain produced in the heating processThe improvement of 119869-integral is put forward according
to the analysis of CTFC according to its definition Theintegral computed in this way is path independent Thus itcould be the controlling parameter of CTFC
In the linear elasticity small-scale yielding conditionsand plane stress conditions 119869-integral is related to stressintensity factor by the following relationship
119869 =1198702
Ι
119864 (8)
where 119864 is the elastic modulus Therefore the stress intensityfactor119870
Ιcan validate the modified 119869-integral cursorily
5 Examples
Take a group of concentric circles whose center is crack tipand radius is 119903 as the integral paths Figure 11 shows therelation between themodified 119869-integral and 119903119903
119901 Clearly the
modified 119869-integral is path independent as long as the integralpath is within the compressive plastic zone
The stress intensity factor and the modified 119869-integral arecalculated under different conditions The results are showninTables 3 4 and 5 In (8) the elasticmodulus119864 is 71175MPa
It is obvious that the stress intensity factor and the 119869-integral increases when the insulating time prolongs themaximum temperature increases and the crack grows Fromcalculating the stress intensity factor can be used as thecontrolling parameter of the CTFC only with a low heatingtemperature a short period of holding time and a deep crack
6 CTFC Propagation Test
The creep-thermal fatigue test machine is composed ofclamping mechanism infrared radiation heating devicehydropneumatic cooling system and control system Theoperational principle of the creep-thermal fatigue testmachine is shown in Figure 12 the specimens can be turnedupside down by the driving shaft which makes it easy andquick to heat and cool the specimensThe detailed size of thespecimen with cracks is shown in Figure 13The first secondand third specimens are used in the creep-thermal fatiguetest the temperature of the specimen cycles is from 30∘C to210∘C and the thermal cycle insulating time is 5 minutes at210∘C every time All the specimens undergo 1000 thermalcycles The length of the crack is detected for each 100 cyclesand the results are shown in Figure 14
The secant method [14] is used to measure the crackgrowth rate that is to calculate the gradient between two
0 1 2
1
04
05
0402
06
06
07
08
08
09
12 14 16 18119903119903119901
119869(J
mmminus2)
Figure 11 The relationship of 119869-integral and 119903119903119901
++
1 2 3 4 5 6 7 8
Figure 12 Operational principle diagram of the creep-thermalfatigue test machine 1 driving shaft 2 stock of the specimen 3unloading bolt 4 tightening bolt 5 infrared radiation heatingdevice 6 specimen 7 hydropneumatic cooling system and 8 armof the test machine
adjacent data points on the 119886-119873 curve to get the crack growthrate as
119889119886
119889119873asympΔ119886
Δ119873=119886119894+1minus 119886119894
119873119894+1minus 119873119894
(9)
where 119889119886119889119873 is the crack propagation under one thermalcycle Δ119886 is the crack propagation under Δ119873 thermal cycles119886119894and 119873
119894are the crack length and number of cycles of load
corresponding to the 119894th data pointAccording to the Paris formula
119889119886
119889119873= 119862119869119899
(10)
where 119862 and 119899 are constants related to material propertyenvironmental media and the geometry of the specimen
The creep-thermal fatigue 119869-integral value in correspond-ing conditions can be obtained according to the modified J-integral method referred to in the front of the paper the datapoints of ln(119889119886119889119873) and ln 119869 are shown in Figure 15 hence
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 7
Table 3 The crack length is 07mm the maximum temperature is 210∘C
Holding time(min)
Size of plasticzone (mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)0 00700 1944 05117 0531080 01050 09289150 01313 10750240 01570 11950
Table 4 The crack length is 07mm the insulating time is 80min
Maximumtemperature (∘C)
Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)180 0 239 00080 00082190 00630 1594 03440 03570200 00875 2115 05970 06285220 01778 14148
Table 5 The maximum temperature is 200∘C the insulating time is 80min
Crack length (mm) Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)05 00623 0463309 01025 0702912 01220 0808815 01493 2526 08568 08965
02
02
3plusmn01
25plusmn01
25plusmn01
4
6
2
2
2
AAA
A A
002
002
002
002
6h7
20plusmn01
13 1370
Figure 13 Specimen of creep-thermal fatigue
119899 = 251 119862 = 872times10minus6 can be obtained by some regressionanalysis of formula (10) using these data points
7 Conclusions
(1) The crack is closed under the compressive thermalstress during the heating and insulating processesBecause the inelastic strain generating in the two
0 200 400 600 800 10000
05
1
15
2
25
3
35
Number of thermal cycles
Crac
k pr
opag
atio
n ra
te (m
m)
1
2
3
Figure 14 Propagation length of the CTFC
processes cannot be back by itself the tensile stressduring the late cooling process which is normal to thecrack faces makes the crack gradually open
(2) When the CTFC is opening the temperature of thematerial around it has become lower than the creep
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
8 Advances in Mechanical Engineering
119869-integral100 101 102
10minus2
10minus3
10minus4
10minus5
1
2
3
Figure 15 (119889119886119889119873)mdash119869 data in double logarithmic coordinate
temperature therefore the creep fracture mechanismparameter 119862lowast is not applicable The stress intensityfactor and 119869-integral can be used as the controllingparameters The stress intensity factor can be usedonly with a short insulating time a low maximumtemperature and a deep crack The 119869-integral shouldbe modified in order to be applicable to the CTFCthat is to take the moment of opening as the startingof the strain energy density integral and eliminatethe effect of the residual compressive plastic strain by(7) It has been proved that the modified 119869-integralmethod is path independent and valid It is used fornot only theCTFCbut also any unloading crack or thecrack in the residual compressive plastic strain fieldThe stress intensity factor and 119869-integral increasewhen the insulating time prolongs the maximumtemperature rises and the crack grows
(3) Experimental results show that the modified 119869-integral can be used as the control value of the CTFCpropagation The research on the controlling param-eters of CTFC provides the theoretical basis for itsgrowth
Acknowledgments
This project is supported by the Natural Science Foundationof Liaoning Province (Grant no 201102167) Aeronautical Sci-ence Foundation of China (Grant no 20110450001) LiaoningExcellent Talents in University (Grant no LJQ2011012) andthe Science Program of Shenyang (Grant No F12-069-2-00)
References
[1] N Gao M Brown K J Miller and P A S Reed ldquoAn investiga-tion of crack growth behaviour under creep-fatigue conditionrdquo
Materials Science and Engineering A vol 410 no 25 pp 67ndash712005
[2] M F Geng ldquoAn idea for predicting crack growth time tofracture under creep-fatigue conditionsrdquoMaterials Science andEngineering A vol 257 no 2 pp 250ndash255 1998
[3] Y L Lu L J Chen P K Liaw et al ldquoEffects of temperature andhold time on creep-fatigue crack-growth behaviorrdquo MaterialsScience and Engineering A vol 429 no 1-2 pp 1ndash10 2006
[4] A T Yokobori and T Satoh ldquoThe characterization of creepcrack growth rate and its life of TiAl inter-metallic compoundwith full lamellar microstructurerdquo International Journal ofPressure Vessels and Piping vol 88 no 7 pp 435ndash440 2011
[5] G Marahleh A R I Kheder and H F Hamad ldquoCreep-lifeprediction of service-exposed turbine bladesrdquoMaterials Sciencevol 42 no 4 pp 476ndash481 2006
[6] V Maillot A Fissolo G Degallaix and S Degallaix ldquoThermalfatigue crack networks parameters and stability an experimen-tal studyrdquo International Journal of Solids and Structures vol 42no 2 pp 759ndash769 2005
[7] J Lansinger T Hansson andO Clevfors ldquoFatigue crack growthunder combined thermal cycling and mechanical loadingrdquoInternational Journal of Fatigue vol 29 no 7 pp 1383ndash13902007
[8] N Haddar A Fissolo and V Maillot ldquoThermal fatigue cracknetworks an computational studyrdquo International Journal ofSolids and Structures vol 42 no 2 pp 771ndash788 2005
[9] G T Yang Theory of Elasticity and Plasticity China BuildingMaterials Industry Press Beijing China 2005
[10] H Kraus Creep Analysis Wiley International Science Publica-tion New York NY USA 1980
[11] Z D Jiang A Zeghloul G Bezine and J Petit ldquoStress intensityfactors of parallel cracks in a finite width sheetrdquo EngineeringFracture Mechanics vol 35 no 6 pp 1073ndash1079 1990
[12] G Agnihotri ldquoStress analysis of a crack using the finite elementmethodrdquo Engineering FractureMechanics vol 44 no 1 pp 109ndash125 1993
[13] R Narasimhan and A J Rosakis ldquoA finite element analysis ofsmall-scale yielding near a stationary crack under plane stressrdquoJournal of the Mechanics and Physics of Solids vol 36 no 1 pp77ndash117 1988
[14] P Bensussan High Temperature Fracture Mechanisms andMechanic Mechanical Engineering Publications 2005
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
Advances in Mechanical Engineering 7
Table 3 The crack length is 07mm the maximum temperature is 210∘C
Holding time(min)
Size of plasticzone (mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)0 00700 1944 05117 0531080 01050 09289150 01313 10750240 01570 11950
Table 4 The crack length is 07mm the insulating time is 80min
Maximumtemperature (∘C)
Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)180 0 239 00080 00082190 00630 1594 03440 03570200 00875 2115 05970 06285220 01778 14148
Table 5 The maximum temperature is 200∘C the insulating time is 80min
Crack length (mm) Size of plastic zone(mm)
Stress intensity factor(MPa sdotmmminus05)
Modified 119869-integral(J sdotmmminus2)
1198691015840
= (1198702
Ι119864)
(J sdotmmminus2)05 00623 0463309 01025 0702912 01220 0808815 01493 2526 08568 08965
02
02
3plusmn01
25plusmn01
25plusmn01
4
6
2
2
2
AAA
A A
002
002
002
002
6h7
20plusmn01
13 1370
Figure 13 Specimen of creep-thermal fatigue
119899 = 251 119862 = 872times10minus6 can be obtained by some regressionanalysis of formula (10) using these data points
7 Conclusions
(1) The crack is closed under the compressive thermalstress during the heating and insulating processesBecause the inelastic strain generating in the two
0 200 400 600 800 10000
05
1
15
2
25
3
35
Number of thermal cycles
Crac
k pr
opag
atio
n ra
te (m
m)
1
2
3
Figure 14 Propagation length of the CTFC
processes cannot be back by itself the tensile stressduring the late cooling process which is normal to thecrack faces makes the crack gradually open
(2) When the CTFC is opening the temperature of thematerial around it has become lower than the creep
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
8 Advances in Mechanical Engineering
119869-integral100 101 102
10minus2
10minus3
10minus4
10minus5
1
2
3
Figure 15 (119889119886119889119873)mdash119869 data in double logarithmic coordinate
temperature therefore the creep fracture mechanismparameter 119862lowast is not applicable The stress intensityfactor and 119869-integral can be used as the controllingparameters The stress intensity factor can be usedonly with a short insulating time a low maximumtemperature and a deep crack The 119869-integral shouldbe modified in order to be applicable to the CTFCthat is to take the moment of opening as the startingof the strain energy density integral and eliminatethe effect of the residual compressive plastic strain by(7) It has been proved that the modified 119869-integralmethod is path independent and valid It is used fornot only theCTFCbut also any unloading crack or thecrack in the residual compressive plastic strain fieldThe stress intensity factor and 119869-integral increasewhen the insulating time prolongs the maximumtemperature rises and the crack grows
(3) Experimental results show that the modified 119869-integral can be used as the control value of the CTFCpropagation The research on the controlling param-eters of CTFC provides the theoretical basis for itsgrowth
Acknowledgments
This project is supported by the Natural Science Foundationof Liaoning Province (Grant no 201102167) Aeronautical Sci-ence Foundation of China (Grant no 20110450001) LiaoningExcellent Talents in University (Grant no LJQ2011012) andthe Science Program of Shenyang (Grant No F12-069-2-00)
References
[1] N Gao M Brown K J Miller and P A S Reed ldquoAn investiga-tion of crack growth behaviour under creep-fatigue conditionrdquo
Materials Science and Engineering A vol 410 no 25 pp 67ndash712005
[2] M F Geng ldquoAn idea for predicting crack growth time tofracture under creep-fatigue conditionsrdquoMaterials Science andEngineering A vol 257 no 2 pp 250ndash255 1998
[3] Y L Lu L J Chen P K Liaw et al ldquoEffects of temperature andhold time on creep-fatigue crack-growth behaviorrdquo MaterialsScience and Engineering A vol 429 no 1-2 pp 1ndash10 2006
[4] A T Yokobori and T Satoh ldquoThe characterization of creepcrack growth rate and its life of TiAl inter-metallic compoundwith full lamellar microstructurerdquo International Journal ofPressure Vessels and Piping vol 88 no 7 pp 435ndash440 2011
[5] G Marahleh A R I Kheder and H F Hamad ldquoCreep-lifeprediction of service-exposed turbine bladesrdquoMaterials Sciencevol 42 no 4 pp 476ndash481 2006
[6] V Maillot A Fissolo G Degallaix and S Degallaix ldquoThermalfatigue crack networks parameters and stability an experimen-tal studyrdquo International Journal of Solids and Structures vol 42no 2 pp 759ndash769 2005
[7] J Lansinger T Hansson andO Clevfors ldquoFatigue crack growthunder combined thermal cycling and mechanical loadingrdquoInternational Journal of Fatigue vol 29 no 7 pp 1383ndash13902007
[8] N Haddar A Fissolo and V Maillot ldquoThermal fatigue cracknetworks an computational studyrdquo International Journal ofSolids and Structures vol 42 no 2 pp 771ndash788 2005
[9] G T Yang Theory of Elasticity and Plasticity China BuildingMaterials Industry Press Beijing China 2005
[10] H Kraus Creep Analysis Wiley International Science Publica-tion New York NY USA 1980
[11] Z D Jiang A Zeghloul G Bezine and J Petit ldquoStress intensityfactors of parallel cracks in a finite width sheetrdquo EngineeringFracture Mechanics vol 35 no 6 pp 1073ndash1079 1990
[12] G Agnihotri ldquoStress analysis of a crack using the finite elementmethodrdquo Engineering FractureMechanics vol 44 no 1 pp 109ndash125 1993
[13] R Narasimhan and A J Rosakis ldquoA finite element analysis ofsmall-scale yielding near a stationary crack under plane stressrdquoJournal of the Mechanics and Physics of Solids vol 36 no 1 pp77ndash117 1988
[14] P Bensussan High Temperature Fracture Mechanisms andMechanic Mechanical Engineering Publications 2005
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from
8 Advances in Mechanical Engineering
119869-integral100 101 102
10minus2
10minus3
10minus4
10minus5
1
2
3
Figure 15 (119889119886119889119873)mdash119869 data in double logarithmic coordinate
temperature therefore the creep fracture mechanismparameter 119862lowast is not applicable The stress intensityfactor and 119869-integral can be used as the controllingparameters The stress intensity factor can be usedonly with a short insulating time a low maximumtemperature and a deep crack The 119869-integral shouldbe modified in order to be applicable to the CTFCthat is to take the moment of opening as the startingof the strain energy density integral and eliminatethe effect of the residual compressive plastic strain by(7) It has been proved that the modified 119869-integralmethod is path independent and valid It is used fornot only theCTFCbut also any unloading crack or thecrack in the residual compressive plastic strain fieldThe stress intensity factor and 119869-integral increasewhen the insulating time prolongs the maximumtemperature rises and the crack grows
(3) Experimental results show that the modified 119869-integral can be used as the control value of the CTFCpropagation The research on the controlling param-eters of CTFC provides the theoretical basis for itsgrowth
Acknowledgments
This project is supported by the Natural Science Foundationof Liaoning Province (Grant no 201102167) Aeronautical Sci-ence Foundation of China (Grant no 20110450001) LiaoningExcellent Talents in University (Grant no LJQ2011012) andthe Science Program of Shenyang (Grant No F12-069-2-00)
References
[1] N Gao M Brown K J Miller and P A S Reed ldquoAn investiga-tion of crack growth behaviour under creep-fatigue conditionrdquo
Materials Science and Engineering A vol 410 no 25 pp 67ndash712005
[2] M F Geng ldquoAn idea for predicting crack growth time tofracture under creep-fatigue conditionsrdquoMaterials Science andEngineering A vol 257 no 2 pp 250ndash255 1998
[3] Y L Lu L J Chen P K Liaw et al ldquoEffects of temperature andhold time on creep-fatigue crack-growth behaviorrdquo MaterialsScience and Engineering A vol 429 no 1-2 pp 1ndash10 2006
[4] A T Yokobori and T Satoh ldquoThe characterization of creepcrack growth rate and its life of TiAl inter-metallic compoundwith full lamellar microstructurerdquo International Journal ofPressure Vessels and Piping vol 88 no 7 pp 435ndash440 2011
[5] G Marahleh A R I Kheder and H F Hamad ldquoCreep-lifeprediction of service-exposed turbine bladesrdquoMaterials Sciencevol 42 no 4 pp 476ndash481 2006
[6] V Maillot A Fissolo G Degallaix and S Degallaix ldquoThermalfatigue crack networks parameters and stability an experimen-tal studyrdquo International Journal of Solids and Structures vol 42no 2 pp 759ndash769 2005
[7] J Lansinger T Hansson andO Clevfors ldquoFatigue crack growthunder combined thermal cycling and mechanical loadingrdquoInternational Journal of Fatigue vol 29 no 7 pp 1383ndash13902007
[8] N Haddar A Fissolo and V Maillot ldquoThermal fatigue cracknetworks an computational studyrdquo International Journal ofSolids and Structures vol 42 no 2 pp 771ndash788 2005
[9] G T Yang Theory of Elasticity and Plasticity China BuildingMaterials Industry Press Beijing China 2005
[10] H Kraus Creep Analysis Wiley International Science Publica-tion New York NY USA 1980
[11] Z D Jiang A Zeghloul G Bezine and J Petit ldquoStress intensityfactors of parallel cracks in a finite width sheetrdquo EngineeringFracture Mechanics vol 35 no 6 pp 1073ndash1079 1990
[12] G Agnihotri ldquoStress analysis of a crack using the finite elementmethodrdquo Engineering FractureMechanics vol 44 no 1 pp 109ndash125 1993
[13] R Narasimhan and A J Rosakis ldquoA finite element analysis ofsmall-scale yielding near a stationary crack under plane stressrdquoJournal of the Mechanics and Physics of Solids vol 36 no 1 pp77ndash117 1988
[14] P Bensussan High Temperature Fracture Mechanisms andMechanic Mechanical Engineering Publications 2005
at PENNSYLVANIA STATE UNIV on May 9 2016adesagepubcomDownloaded from