N89-1292!

THERMAL BARRIER COATING LIFE PREDICTION MODEL*

B.H. Pilsner, R.V. Hillery, R.L. McKnight, T.S. Cook, K.S. Kim, and E.C. DuderstadtGeneral Electric

Aircraft Engine Business Group

INTRODUCTION

The objectives of this program are to determine the predominant modes of

degradation of a plasma sprayed thermal barrier coating system, and then to developand verify life prediction models accounting for these degradation modes. The

program is divided into two phases, each consisting of several tasks. The work inPhase I is aimed at identifying the relative importance of the various failure

modes, and developing and verifying life prediction model(s) for the predominantmode for a thermal Darrier coating system. Two possible predominant failure

mechanisms being evaluated are bond coat oxidation and bond coat creep. The work inPhase II will develop design-capable, causal, life prediction models for

thermomechanical and thermochemical failure modes, and for the exceptional

conditions of foreign object damage and erosion.

Currently, work is continuing in Task II of Phase I aimed at developing apreliminary TBC life prediction model. This model will be created by combining theresults of the analytical program, the thermomechanical experiments and the resultsof the failure mechanism examinations of Task I.

TBC SYSTEMS

The primary TBC system consists of a low pressure plasma-sprayed (LPPS) bond

coat layer of Ni-22Cr-lOAl-O.3Y, an air plasma sprayed yttria partially stabilized

zirconia (ZrO2-8%Y203) top coat, on a conventionally cast Rene' 80 substrate

alloy (Table l). This bond coat composition has been demonstrated to possess goodoxidation resistance and has a large data base as a TBC bond coat. The

ZrO2-8%Y203 top coat was chosen since numerous studies have shown thatzirconia partially stabilized with 6-8 wt.% Y203 is the best composition for

plasma sprayed TBCs (ref. l). The Rene' 80 substrate was chosen since a large TBC

data base exists for this substrate composition.

Four different TBC systems utilizing four different bond coats have been

evaluated in the experiment to evaluate the effect of bond coat creep strength on

TBC thermal cycle life (Table 2). These four TBC systems also utilize

ZrO2-8%Y203 top coats and Rene' 80 substrates. TBC system #l has the sameNiCrAIY bond coat utilized in the primary TBC system. TBC systems _2, m3, and _4have modified NiCoCrAIY bond coats with alloy additions to increase the bond coatcreep strength. An aluminide overcoat was useG in each of these systems (I-4) toreduce differences in oxidation resistance for the four bond coats. A comparison ofthe primary TBC system and its counterpart with an aluminide overcoat is shown in

Figure I.

*Work done under NASA Contract NAS3-23943.

z_47 _R_ PAGE BLANK NOT FILMED

https://ntrs.nasa.gov/search.jsp?R=19890003550 2020-04-13T07:31:03+00:00Z

THERMAL CYCLE TESTING

Thermal cycle testing is being performed in an automated Rapid TemperatureFurnace (Figure 2). The thermal cycles consist of ten minutes heat up, a 45 minuteexposure at I093°C (2UOOOF), and Ib minutes forced air cooling (Figure 3). Thisfurnace utilizes a lift which automatically cycles the specimens from the upperfurnace exposure zone to the lower cooling compartment where a fan provides forcedair cooling. Both air and argon pre-exposures have been used to create changes inboth bond coat and top coat prior to these thermal cycle tests.

BOND COAT OXIDATION EXPERIMENTS

In the bond coat oxidation experiments, pre-exposures in air or argon wereutilized. The goal of pre-exposures in air was to develop oxide scales prior tothermal cycling, while the goal of the pre-exposures in argon was to allow the otherthermally activated phenomena present in the air pre-exposures to occur withoutdeveloping the oxide scale. The intent was to isolate the effect of bond coatoxidation on thermal cycle life.

As reported previously (ref. 2), the specimens pre-exposed in argon failedbefore the specimens pre-exposed in air (Figure 4). The detrimental effect of argonwas believed to be associated with its effect on the type of oxide that forms on thebond coat surface and an in-house program (ref. 3) was performed to help understandthis phenomenon. This study indicated that a possible cause of the shortened lifewas the diffusion of Cr, Ta, W, and other substrate elements to the bond coat/topcoat interface during the argon pre-exposure prior to significant bond coatoxidation. The result was a less protective oxide scale.

To further evaluate this phenomenon, an experiment was run in which allspecimens (except one set of baselines) received a I0 hour air pre-exposure atI0930C (2000°F) prior to either air or argon pre-exposures, an----dthermal cycletesting. In this case, it can be assumed that all specimens developed the same(A1203) film in the initial air exposure and thaTthe effect of furtheroxidation of the bond coat would be seen only in those specimens exposed foradditional times in air. The results (Figure 5) clearly indicate that continuedexposure to air is more detrimental than a prolonged argon exposure (when both arepreceded by the air pre-exposure). The larger decrease in thermal cycle life forair exposures is attributed to the continued growth of oxide scales, whereas littleor no additional growth occurred in the argon exposures. These results demonstratethe importance of bond coat oxidation to the overall TBC failure mechanism.

Further evidence of the importance of bond coat oxidation is shown by thefollowing observations. Continuous oxide scales of approximately 4 _m weretypically observed at the bond coat/top coat interface for the "as-sprayed" and "airpre-exposed" specimens at failure after thermal cycle testing (Figure 6). Thisobservation of a "critical" oxide thickness being necessary to cause failure isconsistent with the work of Miller (ref. 4), who noted similar weight changes(oxidation) at failure of specimens with a CaSiO4/MCrAIY TBC, regaraless of testtemperature.

OF. PC'OR _UALFiY

448

BOND COAT CREEP EXPERIMENTS

The effect of bond coat creep strength on thermal cycle life was evaluated

utilizing four different bond coat alloys (Table II) that had significantly

different creep strengths. The modified NiCoCrAIY bond coats include variousadditions of Mo, Ta, W, Re, Hf, C, B, Si, Zr, and Ti and also received an aluminide

(Codep) coating (Figure Ib) as described earlier. All specimens were coated with

the same ZrO2-8%Y203 ceramic layer. Six specimens of each TBC system werethermal cycle tested. Two were exposed in argon for lO0 hours at I093°C (2000°F),

two were exposed in air for the same time and temperature, and two specimens

received no pre-exposure. The difference in thermal cycle lives was expected to bea function of bond coat creep strength and pretest conditions.

Fhe results clearly showed that the TBC specimens with the NiCrAIY + aluminide

bond coat, which has the lowest creep strength, resulted in the shortest thermal

cycle life for all pre-exposure conditions (Figures 7 & 8). However, the thermal

cycle life differences for the other TBC systems appears to be minimal. The smal_differences may indicate that the bond coat creep strength differences (Table If)

were not large enough to offset the effect of other failure mechanisms (NiCrAIY is

significantly lower in strength than the other three). Interestingly, the lO0 hour

air pre-exposure did not significantly affect the thermal cycle life of the systems

with "high strength" bond coats (Systems 2, 3, and 4). This indicates that, as thethermal cycle life increases (as a result of increasing the bond coat creep

strength), the relative contribution of the pre-exposure (oxidation) to the overall

failure mechanism is reduced.

In a recent additional study at GE, different bond coat creep strengths were

produced by applying various heat treatments to the same bond coat (System _4) toeliminate any differences that might have resulted from aluminide effects on thefour bond coats. The results of this study (ref. 5) indicate that TBC thermal cycle

life increases with heat treatment temperature (increasing creep strength), again

demonstrating that creep strength of the bond coat does indeed influence TBC life.

KEY PROPERTY DETERMINATIONS

Tensile strength, Poisson's ratio, dynamic elastic modulus, and coefficient of

thermal expansion for the bond coat were determined from room temperature to

approximately I093°C (20DO°F)- Standard testing procedures and test specimens wereutilized for the NiCrAIY bond coat specimens. The specimens were machined from 5.1

cm (2 inches) by 15.2 cm (6 inches) heat treated LPPS NiCrAIY billets of various

heights. The as-sprayed billets received a four hour vacuum heat-treatment atI_93:C (2000:F) to increase the machinability of the billets. The I093°C heattreatment was chosen since this is the soak temperature utilized in thermal cycle

testing. The test results are listed in Table Ill and IV, and Figures 9 and I0.

Dynamic elastic modulus, dynamic shear modulus, Poisson's ratio, and coefficientof thermal expansion for the top coat were determined from room temperature to

approximately I093°C. In all tests, free-standing air plasma sprayed specimens wereutilized and were produced by depositing the ceramic material on stainless steel

substrates and inducing a thermal shock to cause spallation of the intact ceramic

sheet. Some final machining was required to achieve the desired specimen

configurations. These specimens also received a four hour heat treatment in air atI093°C (2000°F) prior to testing. The test results are listed in Table V and Vl,

449

and Figure II. Interestingly, the average elastic modulus value determined at room

temperature from the bend test (Table V) is a factor of lO less than the values

measured by the resonant frequency method (Table VI). The difference is possiblyassociated with the presence of cracks, porosity, and splats which would tena to

decrease the apparent modulus in the bend test. These factors should play a smallerrole in the resonant frequency method.

TBC ANALYIICAL MODELING PROGRAM

Five different analytical tasks, each dealing witn a particular aspect of TBCfailure, are being investigated using finite element analysis. The first three

tasks involve an axisymmetric model (Figures 12 & 13) of a multilayer cylinder, thefourth task examines a disk model, and the fifth task is intended to combine finite

element models with simple crack and diffusion models. The specific conditions ofeach task are discussed below.

Task a. In this evaluation, the same temperature was assumed at the inner and

outer surfaces of the specimen (i.e. no gradient across the TBC coated tubular

specimen). The GE cyclic temperature rig's cycle (lU-minute heat up, 45-minuteexposure at 1093°C, 15-minute cooling, Figure 3) was modeled.

Task b. In this task, a temperature distribution generated by a gradient acrossthe TBC cooled tubular specimen is modeled. The work models the effect of the

large gradients (IUO-150°_) developed across the ceramic.

Task c. In this evaluation, cracks will be "placed" along the bond coat/top

coat interface, thereby producing a ring crack. One or more cracks

perpendicular to the free surface will then be added. The goal is to examinecrack tip driving forces to determine any changes resulting from accommodation

of displacements by the multiple cracks. Small submoaels involving a number of

cracks may be studied applying perturbation approaches (localized crackchanges). The conditions for this modeling will be based on the results of thefirst two tasks described above.

Task d. The importance of edge effects in multilayer disk specimens will be

evaluated. Since most TBC applications involve edge effects (coating "patches",component edges, cooling holes, etc.), it is important to examine how these

edges affect thermal cycle life.

Task e. In this task, the finite element model results from the four preceding

tasks, along with some simple elastic crack models, thermal mismatch strains, adiffusion model (e.g., L = Dt), and the effect of hydrostatic pressure to

further examine crack tip driving force. Since this is an elastic model only,

there are limits on its potential but it is hoped that some significant insight

on cracking in ceramics can be gained in this Task.

in the first four tasks of the modeling work, emphasis will be placed onextracting stress and displacement data as a function of time and location underchanges in geometry and boundary conditions. In cases where sufficient material

450

data is available (crack initiation, propagation, or failure data), quantitiespredicted by the models will be comparedto this data for failure information.

The axisymmetric finite element program (Figures 12 & 13) has been applied tothe first two analytical tasks. The bond coat stress free temperature was assumedto be 982°C (1800°F), while the top coat stress free temperature was assumedto be2U4°C(400°F). These are the temperatures of the substrate during application ofthese coatings. In the analysis, both elastic and plastic deformation wereincluded, but no plasticity developed for the temperature conditions selected (timeat temperature was not included). Analysis of the results for the first twoanalytical tasks is discussed below.

In Task a, the specimenwas assumedto undergo the themal cycle of 21°C -I093°C - 21°C in the cycling rig. Since this is a quasistatic test, the entirespecimenwas assumedto be at a given temperature. Effective, radial, axial, andhoop stresses versus distance in the radial direction are plotted in Figure 14 forfour different temperatures [21°C (70°F), 204°C (400°F), 982°C (1800°F), and I093°C(2000°F)]. As indicated, the stress free temperature for the top coat is 204°C(400°F), therefore, zero stress is found in the top coat at this temperature.However, since the top coat was applied to the bond coat, 982°C (1800°F) is nolonger the bond coat stress free temperature. Therefore, small stresses due to thetop coat application develop in the bond coat at this temperature.

In the Task b, a temperature distribution across the TBCsystem was modeled. Inthis examination, the surface of the ceramic was set at I093°C (2000°F), the bondcoat/top coat interface at 943°C (1730°F), the bona coat/substrate interface at941°C (1725°F), and the inner wall of the tube at 927°C (1700°F). These results(Figure 15) were plotted and comparedwith the results present whenthe system wasat 21°C (70°F). Interestingly, the largest effective stress is present in theceramic near the bond coat/top coat interface which is the typical failure locationfor thermal barrier coatings. Comparisonof the results of deformation behavior forTasks a and b (Figures 14 and 15) indicate how significantly the presence of thermalgradients can affect the stress state present in TBCs.

THERMOMECHANICALEXPERIMENTS

fhree different thermomechanical experiments have been planned to evaluate thethermomechanical characteristics of TBCs. The primary goal of these examinations isto measurethe strains induced during thermal cycling of TBCs, and to relate thesestrains to the observed failure modes.

In the first experiment, a thermal barrier coated LCF (low cycle fatigue) tubespecimen (Figure lO) will be thermally cycled using an induction heating system anda forced air cooling system. The thermomechanical nature of two different thermalcycles will be evaluated. The first thermal cycle will be as close to the GEcyclictemperature rig's cycle (Figure 3) as possible. This experiment is aimed atdetermining the magnitude of strains induced by thermal cycling of the TBCspecimenunder essentially zero mechanical loa_. This experiment will also attempt todetermine if any phasing exists between the strains observed for the substrate andthe ceramic, or if the ceramic simply follows the displacement of the metalsubstrate.

451

In the second experiment, a thermal barrier coated LCF tube specimen will againbe thermally cycled using an induction heating system ana a forcea air coolingsystem. Tensile, compressive, and zero loading will be applied to three TBCspecimens ana these specimens will be thermally cyclea to failure. The goal is toevaluate the effect of compressive and tensile loading on TBC thermal cycle life.

In the third experiment, the thermal barrier coating will be applied to thin

Rene' 80 (substrate) strips which will then be heated and cooled using inauction

heating and forced air cooling. It is anticipated that the thin Rene' 80 strips andtheir TBC coatings will bend measurably during thermal transients because of the

stresses induced by thermal expansion differences. Values of the curvature changes

during coating deposition and during subsequent thermal transients will be compared

to predicted curvatures based on natural properties of Rene' _0 and the coating

materials. This data, in conjunction with data from the uncoated Rene' 80 strip an,free-standing ceramic strip should provided insight into the behavior of coated

specimens, and thus contribute to a better understanding of the thermomechanicalcharacteristics of TBCs.

l •

1

.

REFERENCES

Stecura, S., "Effects of Compositional Changes on the Performance of a ThermalBarrier Coating System," NASA TM 78976, 1979.

Hillery, R.V. ana Pilsner, B.H., "Thermal Barrier Coating Life Prediction ModelFirst Annual Report," _ASA CR - 175010, 1985.

Saegusa, F., "Failure Mechanism Stuaies on Thermal Barrier Coatings," bEInternal Communication.

Miller, R.A., "High Temperature Protective Coatings," S.C. Singhal, Editor,AIME, page 293, 1982.

452

Table I

BASELINE THERMAL BARRIER COATING SYSTEM (WEIGHT PERCENT}

Substrate (Rene '80): Ni-14Cr-g.5Co-5Ti-4W-4Mo-3Al-O.17C-O.O3Zr-O.OlSB

Bond Coating : Ni-22Cr-lOAl-O.3Y (Low Pressure Plasma Spray)

Top Coating : ZrO2-SY203 (Air Plasma Spray)

Table II

BOND COAT CREEP EFFECTTBC SYSTEMS

Systems Substrate Bond Coating Over Coating Top Coating Bond Coat Creep

(Larson/Miller

Parameter @ 3

KSI - rupture

test)

1 Rene'BO Bond Coating l'

2 Rene'80 Bond Coating 2*

3 Rene'80 Bond Coating 3*

4 Rene'80 bond Coating 4*

Aluminide ZrO2-Y203

Aluminide ZrO2-Y203

Aluminide ZrO2-Y203

Aluminide ZrO2-Y203

39.0

45.7

47.0

48.4

' Ni-22Cr-lOAl-O.3Y

* Modified NiCoCrALY bond coats

453

TEST TEMPERATURE %

Ambient ZS* (77°F)1

538°C (IOUO°F) 2

76D°C (1400°F) 2

1038% (1800°F) 3

1U93°C (2000°F) 3

Table III

LPPS Ni-22Cr-10A1-O.3Y BOND COAT TENSILE PROPERTIES

ULTINATE STRENGTH MPa 0.2 YIELD MPa % ELONGATION

1320 (191KSI) ....

1240 (179 KSI) 1120 (162 KS]) 5.2

450 (65 KSI) 160 (23 KSI) 18.3

16 (2.3 KSI) 13 (1.9 KSI) 149.3

4 (0.6 KSI) 3 (0.4 KS[) 248.3

% REDUCTION IN AREA

-.

6.2

19.6

9S.b

92.4

1No measurable plastic deformation (1 specimen)

2Average of three test specimens.

3Average of two test specimens.

Table IV

ELASTIC MODULI AND POISSON'S RATIO OF LPPS Ni-22Cr-lOAI-O.3Y

Temperature "C E IAxiall GPa E (Oi_tral) GPa

20 (R.T.) 206 (29.9 NSI) 696 (100.8 NSI)

538 (IO00*F) 180 (26.1MSI) 602 (87.3 MSI)

760 (1400"F) 101 (14.7 MSI) 273 (39.6 MSI)

982 (1800*F) ....

1093 (2000eF) ....

Poisson°s

Ratio

0.30

0.30

0.37

* No linear portion to stress/strain curves.

454

Table V

I_ECHANICALTESTINGOF PLASMA-SPRAYEUZIRCONIABARS

Sample Width, Thlckness, Length,

I.D. cm cm cm

1-1 0.648 0.238 5.746

I-Z 0.648 0.235 5.747

i-3 0.648 0.232 5.746

BEND TEST

U1tlrMte Ultimate Strain to Elastic

Load, Stress Fallure Modulus

K 9 (Ib) MPa (ksi) in/inx10 -3 GPa (MSI)

7.03 (15.5) 53.3 (7.73) 2.89 19.9 (2.80)

6.21 (13.69) 56.8 (8.23) 3.07 20.5 (2.92)

5.26 (11.59) 49.3 (7.15) 2.59 21.8 (3.15)

Average Elastic Modulus 20.6 GPa (2.98 MS])

Table Vl

ELEVATEDTEMPERJ_TUREDETERMINATIONOFELASTICMODULUS,SHEARMODULUS,

AND POISSOg'S RATIO OF APS Zr0_-8Y_O_

E G

Elastic Shear

Temp., Resonant Frequency(Hz) Modulus Modulus"C Flexural Torsional GPa (MSI) GPa (MSI)

25" 1472 3697 210 (30.5) 91 (13.2)

25 1466 3672 208 (30.2) 90 (13.1)

100 1453 3630 205 (29.7) 88 (12.8)

150 1444 3610 202 (29.3) 87 (12.6)

200 1436 3573 200 (29.0) 86 (12.4)

300 1425 3443 197 (28.6) 79 (11.5)

400 1412 3343 193 (28.0) 75 (10.8)

450 1411 3325 193 (28.0) 74 (10.7)

SO0 1401 3299 190 (27.6) 72 (lO.S)

538 1395 3281 189 (27.4) 72 (I0.4)

600 1387 3265 187 (27.1) 71 (10.3)

700 1375 3209 184 (26.6) 69 (10.0)

800 1360 3160 179 (26.0) 67 (9.7)

900 1342 3135 175 (25.3) 56 (9.5)

982 1340 3122 175 (25.3) 65 (9.4)

10DO 1362 3147 179 (26.1) 66 (9.6)

1038 1374 3163 183 (26.5) 67 (9.7)

1093 1342 3185 175 (25.3) 68 (_.8)

Poisson's

Ratios

O.IS

0.16

0.16

0.16

0.17

0.24

0.29

O.31

0.31

0.31

0.31

0.33

0.34

0.33

0.34

0.36

0.37

0.29

* Specimen suspended on cotton thread, all others suspended on Pt wire.

455

OF POOR QUALITY

"NiCrAIY" Bond Coat TBC

Coat

i

AI203

i00 _mparticles I

Figure Ia

--Bond Coat

, --O.I.

I

"NiCrAIY + Codep" Bond Coat TBC

Top Coat

I

t

s

I i00 um r

aluminide

--Bond Coat

Figure Ib

456

RAPID TEMP FURNACE

OF POOR QUALITY

t_)

w

<r_L_CL:EL_JF--

Fi gure 2

1400

)4eQt Up]200-

lO00- /

'°- /i600_ / }

,oo-/i

RAPID TEMP THERmaL CYCLE,

q_ MinuteExPosureot lo_3°C (2oO0°F)',I

I I I 1 I I30 40 50MINUTES

)

I

)

, I6O

Figure 3

]5 Minute

Forced Air Cooling

I I I I I70 80 9O

457

Rene'BO / Ni-22Cr-lOAI-O. 3Y / ZrO2-B%Y2034O0-

358-

L_ 300-c_

._J

< 250-LL

CD

200-

w_J

>- 150-c_J

100-

50-

4OO

5Z 3O _8

_ F ...500 1O0 50 l 0

AIR190

Pn_-exposures

llO

r--

I0 10 50 100

PRE-EXPOSURES(HOURS)

Figure 4

AIR THEN AIR OR ARGON PRE-EXPOSURE

38

I III500

PRE-EXPOSURE< IENPERATORE:2000"F

273

%(_ 207

200 ,40

md

0No Pre- I0 Air 10 Air

Exposure

2000"F TEST

153

140

10 Air 10 Air 10 Air 10 Air

40 Air 40 Argon 90 Air 90 Argon 190 Air

Figure 5

133

10 Air1_0 ArGon

458

AverageOxideScaleThicknessatFailure(Mk:rons)

OXIDE SCALE AT FAILJRE

Thlckmm at Failure

/&

m

5--

4--A

3 - .,.,.,...,-1

2 - IJ'_''lr"i

/ Pre-F.xposureThickmmI

17

0 I IS4) 100

Pre-Exposure Tim (Hours)

Fi gure 6

%

LUr_

...J

LL.

C_I---

(f)LLJ

U>-

800-

700-

600-

5OO-

400-

300-

200-

lO0-

m

355

System

NO PRE-EXPOSURE

57Q

#1 System #2 System

Figure 7

54.5

#3

565

System #4

459

BOO-

I00 HOUR PRE-EXPOSURE

100 hour ARGON Pre- Exposure 100 hour AIR Pre- Exposure

700-- 630

--I

_ 41111- '

310 295

220

:t[i !1 " f0

System #i #2 03 #4 04 03 02 #i

2O

Fi gure 8

f-%

I_I

C

C

IQ

V

18-

16-I

]4-

c 12-0

t_c lO-0nXW 8-

0E 6-L _

_12 4-

0 2-

-d -0

Thermal Expansion Ni-22Cr-IOAl-O. 3Y //

J

I I I ! I I I I I I I I 1 I ! I I I I I

l O0 200 300 400 500 600 700 BOO 900 l OOO

Temporoture (o[)

Fi gure 9

460

E

Z(_.9

7O0

0

E0C

C'n

LJ

C

I

co

(n{-oc)..xw

oEL.

r"

0

220

200-

180-

140-

120-

I00 n{}

2O

18-

]6-

12-

I0-

8-

6-

4-

2-

0 u0

Oynomlc Modulus Ni-22Cr-]OAI-0.3Y

- SPECIMEN#I

o-_.. o SPECIWEN12.....

I I I I I I I I I I I l I I I I I I Il O0 200 300 400 500 600 700 800 900 1000

Temperoture (°C)

Figure I0

I

Thormo] Expons_ on ZrO2-8Y_ 03

I I I I- i i I J i i I i i] OO 200 300 4 O0 500 BOO ?OO

Tempor'ntL,re (°D

Figure II

i-i i l 'I' I800 gO0 1000

f

r]

I

461

_IS_¢'ETRIC I,___O_

Top Coat

Bond Coat

Substrate

R

d

L Av

Figure 12

FINITE ELEMENTS

Figure 13

462

v

II

tlO. 4

96.6 (i4.o)

27.6 (4.0)

z3.s (2.0)

oo.o (o.0

13.8 (2.0)

10.4 (Z.5)!

ORIGINAL PAGE IS

OF POOR QUALITY

WilIIAT|OH OF [FFKCY|UI[ STRKSS |N IIADZAL DIR[¢TIOH ¢VNqlDl[ MIALVSZS

Subs_rate _ B.C.-_--Top Coac "-_

&.

V--_12°c (xlK)oF)

6-*21% (70F)

• -1093°c (200OF)

x--21°C (7OF)

e.......

L _ • -- _

0.762 0.63.5 0.508 0.381 0.2.54 0.127 0.000 0.127 0.2.54 0.381 "0.508 0.63.5

(3o) (25) (207 (z57 (10) (5) (07 (5) (zo) (z57 (20) (25)

THICIKJIESS ms ( X 0.001 inch )

Figure 14a

U_MIATIOII OF RADIttL $11i(|$ IN IlltDIAL DllH[Cl'lOM CVMill)[ _WmLVSIM

: Substra¢e -- 6.C._--Top Coat'-'--'

6.9 (LOT

3.5 (0.5)%s

t

t_ 00.o (o.o)

t4

-J.5 (-o.57

-6.9

-10.4 (-1.5)'--

-13.8 (-2.0)

Iw.=.. _--

- -IA v --

! z/V-'9 2°C (lllOOF)__

o-..,zo,°c (,,,oo,> _ _ _ "A--.21°C (7OF)

e _XO_3°C (mOOV)--

II-- 21°C (701')

I i0.762 0.635 0.508 0.381 0.254 0.127 0.000 0.127 0.2.54 0.382 0.508 0.635

O0) (25) (20) (z.5) (zo) (.5) (0) (.5) (zo) (x.5) (20) (2.57

TIBI(2CqEss am X 0.001 inch )

Figure 14b

463

138.0 (20.0

103.5

t_NIIIAT|OI4 OF 4tXIAI, IITltlllll IN IIMDIAL DIMI©TIOII CYMIIDi N4MIL¥1IIfl

a L trail - B.C. -_Top Coat---'+

--4--_

bg.o (io.ol

34.5 (5.0)

III

oo.o (o.o)-

--34.5 (-5.o),

ol

-69.0

-z03.5 (-15.0).

-138.0 (-20.o)

....... 7

A

v _ Av

===-Jrv,-',,go2°c lira) _ j

mo.-eo4°c (_or)6--21°C (Poe')

O -IOI3°C tAmurs

..,-'21°c (7o,)I I

0.762 0.635 0.508 O.381 0.254 0.127 0.000 O.127 0.254 0.381 0.$08 0.635

00) (25) (20) (_) (XO) (5) (0) (5) (10) (15) (20) (25)

121ZCKMESS an ( X 0.001 inch )

Figure 14c

138.0

103.5

b9.0 (10.0

UNIIIIWXOII OF 14001 m IITNIlll IN NdlIIIAL |lMl[OT|Olt OYPMIII_ fllmLVlttl

• Stmtrate B.C. b-Top Coat "--'4

w -- --

34.5 (5.oi _ _

" 1oo.o (o.o). e : -, = _- _- :. :- :

-34.5 -- ....-,e-- ,.e..,..,_ ,,-,-.4t,--d

.-69+0 (-10. _p_-9IZ°c (floor)O.--It04°C (40Or)

6--._1°C (?OF)

-103.5 (-15.0 --- e_1o93Oc (zo_om,)--

m-- 21°c (7of)-138.0 (-20.0 I

t0.762 0.63.5 0.508 0.381. 0.254 0.127 0.000 0.127 0.2.54. 0.381 0.508 0.63.5

(3O) (25) (20) (XS) (10) (5) (0) (5) (ZO) (LS) (20) (25)

I['ISICKJIIKSS m ( X 0.001 Lnch )

Figure 14d

464

ORIGINAL PA_-E IS

OF. POOR QUALITY

138.0

103.5

69.0

I

34.s O.0

oo.o (o.o)ttq

-34.S (-5.0)In,1

_ -69.0 (-lO.O;Ihl

-103..5

-138.0 (-20.0

_WI|ATItW_ OF [FF[CTIUE STN[IIII - TEI_ (lltADllrl¢l' CVI_NIDIr _MqlALVIII|

21°C _70°F_ -_ - 2I°C (70oF) - 21°c (?O°F_

' t ' "_'

./

0.762 0.635 0.508 0.381 0.254 0.127 0.000 0.127 0.254 0.381 0.508 0.635(30) (25) (20) (_) (10) (5) (0) (5) (XO) (15) (20) (25)

THICKNESS mu ( X 0.001 inch )

Figure ISa

11

_q

8.3

6.9 (1.o

5.5 (0.8

4.1

2.8 (o.,_

1.4 (0.2

oo.o (o.o

-1.4 (-o.21

-2.8 (-o.4)

t_IATI(M Olr NADIAL JTNEI$ - YENP GNANDI£NT OYANIDI[ AIMLYI|8

: :",d)arrnte , E.C._"-Top Cost---"

\/r

//

,--_-_-,4,°c.o---_- zz°c (7o°e_ - 2z°,,- (m°e)

\.927Oc (17ooole) _ID

\/0.762 0.635 0.508 0.381 0.254 0.127 0.0OO 0.127 0.254 0.381 0.508 0.635

(30) (25) (20) (X5) (XO) (5) (0) (5) (ZO) 03) (20) (25)

I'ilIClOtESS am ( X 0.001 inch )

Figure 15bORIGINAL P_C_E .r:;

465

138.0

103. $

limltlAqrlolll olr ItXi4iL I_BI - 'l_lllP GR_ItI|NY

S_bstrate B.C.

CYMIM AlWiLYJl J

Top Coat;

, I

69.0

,j 34.5 (5.01I,

oo.o (o.o)

vl

-69.0

-103.$

-138.0 (-20.0

p

J

i i i I I. I i---8..._093°C (2000°r)-943°C (1730°F)-_l_leC [1725°F)-027°C (1700°F)

0--- llOC (70oF) . 21°(: (70°e_ - _1°c (m%) - 2x% rTo°e)

I I I l l I I J

0.762 0.635 0.508 0.382 0.254 0.127 0.000 0.127 0.254

(3O) (25) (20) (15) (10) (.5) (0) (5) (J.0)

0.381 0.508 0.635

(XS) (20) (ZS)

TUXCIUIF.SS mm ( X 0.001 inch )

Figure 15c

_ttldtTl| Of' NOOf IT'IIII - TEll I It_IfN'r ¢YMIIM II'IALYIII

• Substrate B.C. _..Top Coat ''_

69.0

oo.o (o.o)

-34.5

in

i-.40D3°C (2gOO°F)-_-94L°C (1725°f]-t27°(: (J.7OO°F)

e-- _1°c(7o*. - 2x°c -o°,_ -_I_--_

I i i I l l I I

0.762 0.615 0.$08 0.381 0.2.54 0.127 0.000 0.127 0.254 0.381

(3O) (25) (20) (_) (10) (}) (0) (5) (10) (15)

-69.o (-1o. o)

| I I-103.$ "-"

11/ICi_l_S " ( X 0.001 inch )

0.$08 0.635

(20) (25)

Figure 15d....... "Y" _.+_+ i J

PO(._R (¢UAC_iy

466

[(a)

LCF TUBULAR SPECIrIEN

Top Coating_7 _

II .. A_, //

Substrata ]

L(b)

Top Coating

__.,,,T,._'-_ __ .......Bond Coatins ]

(c)

Top Coerin8 7

(e) Type I - Uncoated Substrata on Cage Section(b) Type 2 - Uncoated Bond Coat Strip on CaSe Section(c) Type 3 - Fully Coated Gage Section

Figure 16

467