JAERI-Conf 96-010

6.8 EVALUATION OF LONG-TERM CREEP PROPERTIES OF HASTELLOY XR IN

SIMULATED HIGH-TEMPERATURE GAS-COOLED REACTOR HELIUM

Yuji KURATA, Yutaka OGAWA*. Tomio SUZUKI,

Masami SHINDO, Hajime NAKAJIMAand Tatsuo KONDO**

Japan Atomic Energy Research Institute.JAPAN

* now Research Institute for Metals, Tohoku University.JAPAN

** now Faculty of Engineering, Tohoku University.JAPAN

ABSTRACT

Creep properties are among the important basic items of material performance for

design of high temperature components of high-temperature gas-cooled

reactors(HTGRs). In order to evaluate creep properties of Hastelloy XR (a modified

version of the conventional Hastelloy X) developed for the High-Temperature

Engineering Test Reactor(HTTR), long-term creep tests were carried out in simulated

HTGR helium at 800, 900 and lOOO'C. The test results up to about 50,000h showed no

significant degradation in creep properties. The creep-rupture strength obtained through

the long-term tests was above the level corresponding to the design allowable creep-

rupture stress of the HTTR. Rupture lives could be estimated with sufficient accuracy

using Larson-Miller parameter. The values of the stress exponent were 4.5 to 5.7 when

the stress dependence of the steady-state creep rate was expressed in terms of the

Norton equation. It was judged that dominant creep process was dislocation creep. The

relationship between the steady-state creep rate and the rupture life was expressed in

terms of the Monkman-Grant equation. Carburization during creep in simulated HTGR

helium did not degrade creep properties of this alloy. Internally formed cavities and

cracks were initiated at sites of precipitates at grain boundaries, growing nearly

perpendicular to the stress axis. Creep fracture was caused by the nucleation, growth

and link-up of grain boundary cavities in long-term tests. Two phases, Cr-rich carbide

and Mo-rich carbide, co-exisred in specimens after long-term creep tests.

- 3 3 8 -

JAERI-Conf 96-010

1. Introduction

The HTTR, which uses helium gas as the primary coolant, is currently under

construction as the first HTGR in Japan(1). Creep properties are among the important

basic items of material performance to be considered in high-temperature structural

design of the HTTR. Although pure helium gas is inert, the primary coolant helium,

which passes through the large mass of hot graphite core structure, cannot avoid

containing slight amount of impurity gases, i. e., H2, H2O, CO, CO2 and CH4. It is

considered that these impurities can. cause oxidation and carburization or

decarburization of heat-resistant alloys and possible effects upon creep properties are

suspected. For this reason many research works on creep tests in flowing helium

containing a small amount of impurities have been carried out. We have also conducted

creep tests in simulated HTGR helium with special care on impurity control in helium.

An intermediate heat exchanger(IHX) of the HTTR is being manufactured using

Hastelloy XR. High temperature components such as the IHX were designed using High-

Temperature Structural Design Code^2) and Design Allowable Limits^. At the time of

the design of the HTTR a lot of creep data were used to generate the Design Allowable

Limits for which the allowable stress for long-time service was determined through some

reasonable extent of extrapolation from the test data obtained up to that time. The

allowable stress up to 105 h was determined using data up to about 25,000 Ir3 ' .

Therefore, it is necessary to confirm whether the allowable stress is sufficiently marginal

or not when creep data in longer tests are obtained.

We reported that there was no significant difference between helium and air in

the creep rupture lives at 800 to lOOO'C up to about 10,000h and that the test results

showed no significant degradation in creep properties^4'5). Since a series of creep tests

for Hastelloy XR in simulated HTGR helium have been accomplished, all results

including creep data up to about 50,000 h are analyzed in this study in conjunction with

carburization behaviour, microstructure observation and mechanism estimation of creep

deformation and fracture.

339

JAERI-Conf 96-010

2. Experimental procedure

2.1 Specimens

The material tested in this study is Hastelloy XR developed for high temperature

components of the HTTR. Hastelloy XR is an improved version of Hastelloy X. Table 1

shows the chemical composition of Hastelloy XR together with the specification of

Hastelloy X(ASTM B435). Hastelloy XR has a basal composition for the major

constituents common with that of Hastelloy X (i.e., nominally Ni-22Cr-18Fe-9Mo in mass

%), while contents of specific minor elements are optimized: Mn and Si are adjusted in

the optimum ranges and Al, Ti and Co are reduced to the possible lowest levels^. The

material was supplied in the form of bars of 15 mm in diameter. The solution treatment

temperature of the bars was 1180°C and the grain size was ASTM No.3-4. Specimens of

6 mm in diameter and 30 mm in gauge length were used for creep tests.

2.2 Test methods

Creep tests were carried out in helium environment designated as JAERI Type B

helium, which was one of the experimental specifications of the simulated HTGR

primary coolant environments. Impurity composition of JAERI Type B helium is shown in

Table 2. The chemical characteristic of this impure helium environment is low oxidizing

and slight carburizing for Hastelloy XR.

Single specimen type uniaxial creep machines'7' designed for a helium

environment were used. Special care was exercised in avoiding undesirable

perturbation of the local impurity composition at the test section due either to degassing

from or to reaction with the machine components.

Creep tests in the helium environment were carried out at 800,900 and lOOO'C,

and under 6.9 to 98.1 MPa. During the test, the temperature was monitored by platinum/

platinum-rhodium thermocouples attached along the specimen gauge section.

Microstructure observation was made for ruptured specimens using optical microscope ,

scanning electron microscope and EPMA.

- 3 4 0 -

JAERI-Conf 96-010

3. Results and discussion

3.1 Creep rupture properties

Figure 1 shows the relationship between stress and time to rupture for Hastelloy

XR in JAERI Type B helium. The solid lines are the regression curves obtained from

analysis in terms of Larson-Miller parameter, and the broken lines the design allowable

creep-rupture stress (SR) of the Design Allowable Limits^. The design allowable creep

rupture stress was determined from the minimum creep rupture time which was one tenth

of the mean values obtained by applying the time-temperature parameter (TTP) method

to available creep rupture data in air and in simulated HTGR hel ium^. The strength

level of the creep-rupture data including results of creep tests up to 48,557.5h is above

SRat each test temperature, and the long-term data above 10,000 h show no significant

degradation in creep rupture lives.

The TTP method is often used to analyze creep-rupture data and to predict

rupture life and stress^8-9). In the present study, the following parameters were

investigated.

Larson-Miller: LMP=T(C+log tR ) , (1)

Orr-Sherby-Dorn: OSDP=log tR - Q/19.1425T, (2)

where tR is time to rupture, T the absolute temperature, and C and Q parameter

constants. Equation f(o), which represents stress dependence of TTP, is approximated

using polynomials of logarithmic of stress, a, as

f(a)=b0+b1log(a)+b2(log(o))2+- • •+bk(log(a))k, (3)

where bQ, b̂ ,b2, • • • , b^ are regression constants. Parameter constants in equations

(1)-(3) are optimized to minimize standard error of estimate (SEE) of log tR.

SEE=Jl(Yi-Yi)2/(nd-np-k-1), (4)

where Yi is a measured value of log tR, Yi an estimated value of log tR nd the number of

- 3 4 1 -

JAERl-Conf 96-010

data points, np the number of parameter constant in TTP and k the degree of equation

(3). The optimization procedure was carried out following the study on standardization of

creep-rupture data evaluation of metals (9). Table 3 shows results of TTP analysis for

Hastelloy XR. It is found that the degree of polynomials is 2 for LMP and OSDP to have

good approximation. Furthermore, the application of LMP results in better fit than that of

OSDP since the SEE value of LMP is smaller than that of OSDP. On this basis, the

results obtained from analysis in terms of LMP are shown in Fig.1. These regression

curves fit experimental results containing long-term data. The final Larson-Miller

equation is shown as follows:

T(13.545293+logtR)

=23556.529+155.41729log(o)-1837.2623(log(a))2, (5)

The relationship between reduction of area and time to rupture at 800, 900 and

1000X: is shown in Fig.2. While rupture ductility generally decreases with increasing

rupture time, the results shown in Fig.2 have a tendency of leveling off in the decrease of

rupture ductility after a few thousand hours.

3.2 Creep curve and steady-state creep rate

Creep curves in long-term tests are shown in Fig.3. It was reported that normal

type creep curves consisting of transient, steady-state and accelerated stages were

observed at 800^ and that, for Hastelloy XR at lOOOX), some irregular creep curves

were often observed ('™. In the present case, while long-term creep curves at 800 and

900^ are normal type curves as shown in Fig.3, the creep curve at 1000°C is the

irregular creep curve where the transient stage is hardly recognized. Figure 4 shows the

relationship between the steady-state creep rate and stress. The creep rate at 3% strain

was adopted as the steady-state creep rate when the steady-state region was not

recognized clearly in the irregular creep curves.

The relationship between the steady-state creep rate, ES, and stress, a, is

generally expressed as a power law, often called Norton's Law.

- 3 4 2 -

JAERl-Conf 96-010

es = Ao n , (6)

where n is the stress exponent and A a constant. The stress exponent, n, obtained by

applying the least squares method to data at each temperature is 5.745 at 800T:, 4.686

at 900^ and 4.527 at lOOCC, respectively. The n value at each temperature does not

change within this experimental condition even if the stress decreases. It was reported

that n would be unity under the diffusional creep mechanism and that n would be greater

than 3 under the dislocation creep mechanism^'. The creep process of Hastelloy XR

within this experimental condition is interpreted to be dominated by the dislocation creep

mechanism since n values are above 3. Generally n value is about 5 for solution-

strengthened alloys, and it is often above 5 for precipitation-strengthened alloys^2 '.

Hastelloy XR is strengthened mainly by solution and additionally by precipitation. The

facts that the n value is above 5 at 800^) and that it decreases with increasing

temperature are explained by the decrease in the function of precipitation strengthening

due to the coarsening of precipitates at high temperatures.

An activation energy for creep, Qc can be derived from a plot of log eQ against

(1/T) at constant stress. The Qc values obtained were 340-450 kJ/mol. These values are

larger than the activation energy value, 285 kJ/mol'13', of lattice diffusion of Ni in Ni-20

at %Cr alloy. By these approaches, we can say that dislocation creep with n values of

about 5 and with Qc values larger than the activation energy for lattice diffusion of Ni in

Ni-20 at %Cr alloy is predominant under conditions applied in this study.

The following Monkman-Grant equation is useful to perform the prediction of life or

residual-life of high-temperature component.

log tR= c-mlog es , (7)

where c and m are constants. Figure 5 shows the relationship between the time to

rupture and the steady-state creep rate for Hastelloy XR. The regression curves at each

temperature or at all temperatures are also shown in this figure. Creep rupture lives for

Hastelloy XR can be predicted using the Monkman-Grant equation.

- 3 4 3 -

JAERI-Conf 96-010

3.3 Carbon analysis and microstructure observation

Since decarburization and carburization have appreciable effects on creep

behaviour 0 4 - 1 5 ) , carbon analysis was carried out for ruptured specimens. Furthermore,

microstructure observation for surface and cross sections of ruptured specimens was

performed.

Figure 6 shows results of carbon analysis. Carbon content was analyzed for

section including rupture portion where there are many cracks and for 8-mm-diameter

section where there is little creep deformation. Carbon content in the former is high as

shown in Fig.6. This tendency becomes significant with higher temperatures and longer

times. On the other hand, there is only a little carburization in 8-mm-diameter section

even after long-term exposure to the impure helium. It can be pointed out that carbon

Intrusion during the steady-state creep stage was limited to a negligible level. No

specimen experienced decarburization which caused significant decrease in rupture

lives(14). Carburization during creep in simulated HTGR helium did not substantially

degrade creep properties of Hastelloy XR.

The following characteristic of microstructures of specimens ruptured in long-term

creep tests can be described from Fig. 7. The depth of the surface crack was about 100-

150 urn at 800 and 900^, and about 200 |xm at 1000°C. It was found that creep fracture

was caused by the growth and link-up of cavities nucleated at sites of precipitates at

grain boundaries, growing nearly perpendicular to the stress axis. Grain boundaries

perpendicular to the stress axis are under tensile stress. Precipitates at grain boundaries

under tensile stress coarsened as shown in Fig.7. In order to obtain information on

precipitates, EPMA analysis was carried out. Figure 8 shows the results of EPMA

analysis for the specimen ruptured at 900^ in 48,587.5 h. It is clear that bright

precipitates co-exist with dark precipitates in both secondary and backscattered electron

images. The bright precipitates have high molybdenum, silicon and comparatively low

chromium contents, while the dark precipitates have high chromium, carbon and

molybdenum, and low silicon contents. Table 4 shows quantitative analysis of

precipitates and matrix by EPMA. The dark phase is Cr-rich carbide presumed to be

- 3 4 4 -

JAERI-Conf 96-010

2 3 g carb ide^ 6 ' " ' . The bright phase has high molybdenum and comparatively

low chromium contents. This bright phase enriched with Mo is presumed to be MgC type

carbide (1?).

4. Conclusions

Creep tests of Hastelloy XR were carried out in simulated HTGR helium up to

about 50,000 h at 800, 900 and 1000T:. The main results obtained are as follows:

(1)The test results up to about 50,000 h showed no significant degradation in creep

properties such as the rupture life, rupture ductility and the steady-state creep rate.

Creep-rupture stress is substantially above Spof the Design Allowable Limits.

(2)The stress dependence of the steady-state creep rate is expressed in terms of the

Norton equation. The values of the stress exponent is 4.5 to 5.7. On this basis, it is

judged that dominant creep process is dislocation creep.

(3) Rupture lives of Hastelloy XR can be estimated with sufficient accuracy using a

Larson-Miller parameter. The relationship between the steady-state creep rate and time

to rupture is expressed in terms of the Monkman-Grant equation.

(4)Some appreciable carburization was recognized in the specimens ruptured after

creep tests in simulated HTGR helium. Carburization during creep in the helium did not

substantially degrade creep properties of Hastelloy-XR.

(5)Creep fracture was caused by the growth and link-up of cavities nucleated at sites of

precipitates at grain boundaries, growing nearly perpendicular to the stress axis. There

were two phases, Cr-rich carbide presumed to be M23C6 and Mo-rich carbide presumed

to be MgC in specimens after long-term creep tests.

Acknowledgements

Authors are very grateful to staffs of Material Performance and Testing Laboratory

for their assistance during this long-term study.

- 3 4 5 -

JAERI-Conf 96-010

References

(1)Japan Atomic Energy Research Institute : Present Status of HTGR Research and

Development (1996).

(2)HTTR Designing Laboratory, Department of Fuels and Materials Research and

Department of High-Temperature Engineering : JAERI-M 89-005 (1989) [in Japanese].

(3)HADA, K, MOTOKI, Y, and BABA, O,: JAERI-M 90-148(1990) [in Japanese].

(4)KURATA,Y, OGAWA, Y. and KONDO, T. :Nucl. Technol... 66, 250(1984).

(5)0GAWA, Y, KURATA, Y .SUZUKI, T., NAKAJIMA, H. and KONDO, T. : Nihon-

Genshiryoku-Gakkai Shi (J. At. Energy Soc. Japan), 36, 967 (1994) [in Japanese].

(6)SHINDO, M. and KONDO, T.: Proc. Conf. on Gas-Cooled Reactors Today, Bristol/UK,

1982(British Nuclear Energy Society) Vol.2, p.179.

(7)0GAWA,Y and KONDO, T.: JAERI-M 8801 (1980) [in Japanese],

(8)VISWANATHAN, R.: Damage Mechanism and Life Assessment of High-Temperature

Components, ASM International, Metals Park, Ohio, (1989).

(9)VAMAS Data Evaluation Committee : Study on Standardization of Creep-Rupture

Data Evaluation of Metals, The Iron and Steel Institute of Japan, (1994) [in Japanese].

(10)YOKOI, S., MONMA, Y, KONDO, T,., OGAWA, Y. and KURATA, Y : JAERI-M 83-138

(1983) [in Japanese].

(11)FROST, H.J. and ASHBY, M.F. : Deformation Mechanism Maps, Pergamon Press,

London, (1982).

(12)SIDEY, D. and WILSHIRE, B.: Met. Sci. J., 3, 56 (1969).

(13)MONMA, K., SUDO, H. and OIKAWA, H. : Nihon-Kinzoku-Gakkai Shi (J. Jpn. Inst.

Met.), 28, 188 (1964) [in Japanese].

(14)KURATA, Y, OGAWA, Y. and NAKAJIMA, H. : Tetsu-to-Hagane (J. Iron Steel Inst.

Jpn.), 74, 380 (1988) [in Japanese].

(15)idem.: ibid., 74, 2185 (1988) [in Japanese].

(16)SABOL, G.P. and STICKLER, R.: Phys. Stat. Sol., 35,11(1969).

(17)TANABE, T, ABE, R, SAKAI, Y and OKADA, M. : Transactions ISIJ (J. Iron Steel Inst.

Jpn.), 26, 968 (1986) .

- 3 4 6 -

JAERl-Conf 96-010

Table 1 Chemical composition of Hastelloy XR (mass%)

Hastelloy X(ASTM B435)

Hastelloy XR

Hastelloy X(ASTM B 435)

Hastelloy XR

Cr

20.5~23.0

21.90

Mo

8.0~10.0

9.13

Fe

17.0~20.0

18.23

P

max. 0.04

0.005

C

0.05~0.15

0.07

S

0.030

0.005

Si

max. 1.00

0.27

W

0.2-1.0

0.47

Co

0.5~2.5

0.04

Al

0.03

Mn

max. 1.00

0.88

Ti

0.02

B

0.0003

Ni

Remainder

Remainder

Table 2

H2

200~210

Impurity

H2

0.8-

levels

0

'1.2

of JAERI

CO

100~11

Type

0

B helium

co2

2~3

(vol ppm)

CH4

5 ~ 6

Table 3 Comparison of standard TTP (time-temperature parameter) fit

for rupture lives of Hastelloy XR

Degree ofPolynomals

1

2

3

4

5

SEE(standard error of estimate)

Larson-Miller Parameter

0.223

0.160

0.164

0.159

0.156

Orr-Sherby-Dorn Parameter

0.278

0.173

0.175

0.171

0.168

Table 4 Quantitative analysis of precipitates and matrix of Hastelloy XR

ruptured at 10OO'C and 6.9 MPa in 13,014 h (rnass%)

Precipitate(Dark)

Precipitate(Bright)

Matrix

Ni

4.7

20.2

49.6

Cr

71.6

16.1

22.8

Fe

3.6

6.4

18.5

Mo

18.7

53.8

7.7

Mn

0.36

0.16

0.93

W

0.99

3.3

0.40

- 3 4 7 -

JAKRI-Conf 90—010

GO

84CO

200

100

so

20

10

5

2

•

•

1 '., 800°C

*^"-.^—J1000°C

Regression curve "*"*

800°C0

900°CA

1000°C

n .

Hastelloy XR

s

1

10J 105102 103 104

Time to rupture (h)

Fig. 1 Stress vs. time to rupture for Hastelloy XR In JAERI Type B helium. The solid

lines are the regression curves obtained from application of Larson-Miller

parameter, and the broken lines the design allowable creep-rupture stress

(SR) of the Design Allowable Limits (3).

100

80

£ 60

oI 404—>

o

1? 20

' ' ' ' " I

. Hastelloy XR

0l1 I 1 1 1 III

• " I

800°C

O900°C

A

1000°C

D

D

o

Iff102 103 104

Time to rupture (h)

Fig. 2 Reduction of area vs. time to rupture for Hastelloy XR in JAERI Type B helium.

- 3 4 8 -

JAERI-Conf 96-010

50

40

Q 30

Hastelloy XR_ JAERI Type B helium

900°C, 13. 7MPa _

0 10000 20000 30000 40000 50000

Time (h)

Fig. 3 Long-term creep curves of Hastelloy XR in JAERI Type B helium.

2 5 10 20 50 100 200Stress (MPa)

Fig. 4 Relationship between steady-state creep rate and stress for Hastelloy XR in

JAERI Type B helium.

- 3 4 9 -

JAERI-Conf 96-010

3

B

105

104

103

102

101

io-

: ' ' ' ' ' ' " 1

r ^

- 800°C: 900°C

• iooo°c• All

: 800°C0

900°CA

,

1000°C

D

i •

Hastelloy XR \

-

10~3 10~2

Steady-state creep rate (%/h)

10o

Fig. 5 Relationship between time to rupture and steady-state creep rate (Monkman-

Grant relationship) for Hastelloy XR in JAERI Type B helium.

0.10

0.08

0.06& 0.10

COCO

I 0.08

I 0.06.coo

0.14

0.12

0.10

0.08

0.O6

2ZZ

1000'C

Open : section includingrupture portion

Solid : 8-mnrdiametersection

ZZZ As receivedr ) I t I I I i I

10' 102 103 10<

Time to rupture (h)

105

Fig. 6 Results of carbon analysis on the specimens of Hastelloy XR creep-ruptured in

JAERI Type B helium.

- 3 5 0 -

JAERI-Conf 96-010

i- » . .V.: !i> -:"£',

x: x:o ioOJ 00IX) U)co" co"

23D.

0)

COQ.

co<6cocococu

COQ.

CO

toCOcu

oco"

2 oen a;.§ E

CO0.5enCD

COCU

w a>

P" P P§ § §co en °

UJ

CD

Q.CDCDi_O

ECD

i

cnc_o

CD

CO

CCX

o

COCO

X

§§

o ^.— CL

LL

- 3 5 1 -

JAERI-Conf 96-010

•ocns

Pooen

"aCD

ZJ

a.CD(D

bCCX

(0X

ocoa>01

> NnJ coV m

X ooo ^

V-* r"

x: coO T-

co

- 3 5 2 -