MODIFIED 9Cr-lMo STEEL FOR ADVANCED

iSTEAM GENERATOR APPLICATIONS* f,'prr>3> T'! . '"[\')'\

MAY 0 4 1990

C. R. Brinkman, D. J . Alexander, and P. J . Maziasz

CONF-901026—1

DE90 010497

Oak Ridge Nat iona l Laboratory

Oak Ridge, TN 37831-6154

ABSTRACT

Results are reported from several types of mechanical property tests

conducted on a number of commercial heats of modified 9Cr-lMo steel. Data

from long term creep-rupture tests conducted on base and weldment material

were compared with an analytical model which has been shown to give good

agreement between measured and predicted values. Weldment material had

somewhat inferior creep-rupture strength in comparison to base material due to

a soft zone at the edge of the HAZ.

mmfflSTWBBTHMI OF THIS MCUMEffT S BWUWTBI

Data are presented from elevated temperature tensile and creep-rupture tests

conducted on material thermally aged for periods of up to 75,000 h (8.6

years). Some reduction in strength was shown to occur in comparison to unaged

material. Models were developed for predicting the reduction in short term

elevated temperature tensile and yield strength for material thermally aged in

the temperature range of 482 to 704°C. Results from Charpy impact tests

conducted on material thermally aged at 538°C for periods of up to 75,000 h

showed an increase in the ductile-brittle transition temperature. Finally,

results from transmission electron microscopy studies were presented to

explain changes in mechanical properties due to thermal aging. These

observations showed that Laves phase precipitation and recovery occurs on

prolonged exposure of this alloy in this temperature range.

INTRODUCTION

Modified 9Cr-lMo steel, or grade 91, was developed as a ferritic steel

with improved mechanical properties (Bodine et al., 1983, Sikka et al., 1983,

Sikka, 1984, and Roberts and Canonico, 1988). It's high thermal conductivity,

low thermal expansion, high strength, and resistance to corrosion (Banks, 1984

and Iseda, 1988) make it an outstanding candidate for many steam generator

applications, including components such as tubing, piping, and headers (

Haneda, et al., 1988). The material has been installed as tubing and as other

components in a number of superheaters and reheaters in steam generators

operating at temperatures of 538 to 593°C in fossil fired steam power plants

in several countries including the United States, United Kingdom (Townsend,

1987), Canada, and Japan (Masuyama et al., 1988). Periodically samples of

tubing from these insertions have been removed for visual and metallurgical

examination. Results reported to date have been from relatively short term

exposure studies e.g. approximately 30,000 h (Ellis et al., 1990); however, no

instances of tube sagging, excessive corrosion, or marked changes in

mechanical properties have been reported. However, because the material is

relatively new and its microstructure is initially fully martensitic, some

concern exists as to possible changes in mechanical properties due to

prolonged exposure to elevated temperatures. Hence, it is the objective of

this paper to report results of prolonged exposure to elevated temperatures or.

the mechanical properties of this steel. These mechanical properties include

creep, tensile, and toughness properties of base and weldment materials where

available. Microstructural observations concerning changes due to thermal

exposure will also be reported.

*Research sponsored by the U.S. Department of Energy, Office of Technology

Support Programs, under contract DE-AC05-84OR21400 with Martin Marietta Energy

Systems, Inc.

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United StatesGovernment. Neither the United States Government nor any agency thereof, nor any of theiremployees, makes any warranty, express or implied, or assumes any legal liability or responsi-bility for the accuracy, completeness, or usefulness of any information, apparatus, product, orprocess disclosed, or represents that its use would not infringe privately owned rights. Refer-ence herein to any specific commercial product, process, or service by trade name, trademark,manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recom-mendation, or favoring by the United States Government or any agency thereof. The viewsand opinions of authors expressed herein do not necessarily state or reflect those of theUnited States Government or any agency thereof.

COMPOSITION AND HEAT TREATMENT

The allowed range in chemical ccmpcsi t ion of t h i s a l l o y i s compared to

t h a t of s tandard 9Cr-lMo s t e e l in Table 1. The comparison shows t h a t Grade 91

Table I . Chemical analysis of modified9 Cr-1 Mo steel and i t s comparison

with standard 9 Cr-1 Mo steel

Content range, wt X

Element Modified Standard9 Cr-1 Mo 9 Cr-1 Mo

(Grade 91) (Grade 9)

Carbon 0.08-0.12 0.15 maxManganese 0.30-0.60 0.30-0.60Phosphorus 0.020 max 0.030 maxSulfur 0.010 wax 0.030 maxSilicon 0.20-0.50 1.00 maxChromium 8.00-9.50 8.00-10.00Molybdenum 0.85-1.05 0.90-1.10Nickel 0.40 maxVanadium 0.18-0.25Niobium 0.0&-0.10Nitrogen O.O30-O.O70Aluminum 0.04 max

contains an addition of vanadium and niobium and a more closely specified

range of chemistry for each element in particular nitrogen in comparison to

Grade 9 or standard 9Cr-lMo steel. The heat treatment consists of

normalization at 1040°C, holding for 1 h per 25-nun of thickness, followed by

air cooling.

Subsequent tempering is done in the range of 738 to 76O°C which produces a

fully martensitic microstructure as shown in Figure 1 .

*Yu'~"•>-',•'•. -V-.-.-I •/-•* Photo Y 185741

Fig. 1. Typical martensitic microstructure of Modified 9Cr-lMo steel.

Grain sizes are typically fine (ASTM 8-9) and transmission electron microscopy

reveals a complex microstructure consisting of a high dislocation density and

subboundaries decorated with carbides. The subboundaries are stabilized by

the precipitation of MC and M23 C6 precipitates which accounts for the alloy's

increased strength.

CREEP-RUPTURE STRENGTH

Stress-rupture plots covering the creep behavior of base material are

given in Figure 2 for temperatures ranging from 427 to 704°C. The data are

from material obtained from several heats, product forms, and melting

processes including air induction, argon-oxygen decarburization (AOD),

electroslag remelting (ESR), and combinations as shown in Figure 2.

O - XA36O2 CE AIF: INDUCTION- X A 3 6 I 8 CE AIR INDUCTION

INDICATES TEST IN PROGRESSI LJ-U I 1_UJ I I I I I I I I I I I I I I<O

.HFAT. MflTFW. O - FSJ49 OUAKER AOD

A - 3 0 1 8 2 CARTECH AOD/ESR- • - 3 0 1 7 6 CARTECH AOD/ESR

O - 30383 CARTECH AOD20 h 7-30394 CARTECH AOD/ESR

C> —10148 ELECTRALLOY AOD

10° to' IO J 10* IO ! 10°TIME TO RUPTURE Ihl

ORNL-DWG 83-12160R

<o* 10' io4

Fig. 2. Stress-rupture plots for Modified 9Cr-lMo steel at several

temperatures.

The data shown in Figure 2 were fit by an equation developed by Booker

et al., (1983), which is as follows:

log tr = Ch - O.231CT - 2.385 log a + 31,080/T,

where

tr = rupture life (h),

a = stress (MPa),

T = temperature (K).

Logarithms are in base 10. The parameter Ch is a "lot constant" that

reflects the relative strengths of different lots of material, assuming that

the stress and temperature dependence is the same for all lots. The average

value of Ch was -23.737. The analysis yielded an overall standard error of

estimate (SEE) of 0.324 and a minimum (average minus 1.65 SEE) Ch of -24.272.

Recent data added to the plots given in Figure 2 show good agreement between

measured and predictad values as indicated.

Long term creep-rupture ductility is also of interest as an indication

of the resistance of the material to creep-fatigue interaction (intergranular

cracking) and to such creep phenomena as stress-relaxation induced cracking.

Figure 3 is a plot of creep-rupture ductility for multiple heats at several

temperatures as a function of time. The plot shows that the material has

excellent short term creep-rupture ductility. There is some indication of

decreased ductility beyond about 20,000 h, but no values below 10% in terms of

reduction of area have been reported,eo

50

% 30

3< 20o

0

100

90

*? ao

<UJ

§ 70u.OO 60

~i i—rrj r

Va o

• ° V O D

i i i I I [ i i i I

CO

Da

i i i I i i i i I i ; i

30

(a)

Fig. 3.

DA

otf>A 9A

D

TESTTEMPERATURES (°C)

O-482 0 -649A-538 A-677Q-593 V-704

HEATSF5349301823017630383303941014891387XA360<>14361

I, I l_!_ 1 1 1 1 1

i i l

a

D

(0° 101 <02 1O3 10TIME TO RUPTURE (h)

Creep-rupture ductility data as a function of rupture time at

various temperatures for commercial heats.

Base material was also taken from two heats (25.4 mm thick plates) and

aged for periods up to 75,000 h (8.6 years) at temperatures ranging from 482°

to 649°C prior to creep-rupture testing at these same temperatures. The

objective of this effort was to determine the extent to which rupture strength

behavior is changed by pre-thermal aging. Figure 4 is a plot of the above

rupture equation showing average and minimum lines for unaged material as

determined by the above equation. These lines are compared with data points

obtained from material that had been pre-aged either 50,000 or 75,000 h prior

to testing.

D

•THEF

— AV!:RAGE FOR HE ATSTEUNAGED CONDITION

MINIMUIN UNA

1

M FORGED C

1 1

f 1HEATS TESTi

DNDITION——1

1 l1

PREAGED 50,000 hPREAGED 75,000 hTWO HEATSMAI ARlhlft A

WERE THE !649 °C .

!

SAME AND TE!NO VA

T TEMI

STED 1

nit:D

ERATLRIEO FROM 48

M

RES2 TO

1

s s

•40 - 3 2

log (I) - 31,080/T

Fig. 4. A comparison of the creep equation predicted rupture strengths

values and data obtained from tests conducted on pre-aged material.

The comparison shows that pra-aging does reduce rupture strength somewhat and

that for 75,000 h (8.6 years) exposure rupture strengths can fall below

minimum values for unaged material.

However, in the case of the 75,000 h pre-aging tests, the temperatures were

either 593 or 649°C such that substantial changes in microstructure (as will

be shown) were exptcted to occur.

Creep-rupture tests were also conducted on specimens taken from

weldments. Butt joint weldments of plates were prepared using a variety of

processes including gas tungsten arc (GTA), submerged arc (SA), and shielded

metal arc (SMA) welding. The filler metal types for these weld deposits were

standard 9Cr-lMo or modified 9Cr-lMo steel. Following welding, a post-weld

heat treatment of 732 or 760°C for 1 h was given.

Weldment specimens were taken with their major specimen axis transverse

to the fusion line such that the gage section contained base, heat-affected

zone (HAZ), and weld metal. All weld metal specimens were taken with their

major axis parallel to the fusion line. Creep-rupture tests were conducted at

538, 593, and 649°O. An example stress-rupture plot for test data obtained at

593°C is given in Figure 5.

10

[defined as average behavior minus 1.65 multiples of the standard error of the

estimate (SEE)]. Minimum rupture life of weldment behavior is similarly

defined and is also plotted in Figure 5.

Figure 5 shows that the average strength of weldment is somewhat less

than that of the average strength of base material. This is due to a weakened

or soft (overaged) region at the edge of the HAZ. An example of this soft

zone is shown in Figure 6 for a GTA weldmert by a hardness profile taken .

across the weldment.

ORNL PHOTO 1442-83

ORNL-Pioro 1442-83

360

200

160

- BASE METAL | HAZ | WELD | HAZ | BASE METAL

I I I I i I I I I I • 1 I I I I12 16 20 24

INDENTATION NUMBER28 32

Fig. 6. Hardness profile across GTA weldment of Modified 9Cr-lMo steel.

12

Other investigators such as Haneda, (1988), have similarly reported a soft or

preferential zone for creep-rupture failures of weldirents of Modified 9Cr-lMo

steel. Hence, reduction factors in weldment strengths, subject to prolonged

loads in service at elevated temperatures appear to be justified when failure

by creep-rupture is a possibility. Using the above equation and appropriate

lot constants, ratios of average weldment to average parent metal strengths

can be calculated for various times and temperatures. As an example, at 593°C

and at 300,000 h this ratio was determined to be 0.84. A complete table of

these ratios or reduction factors has been calculated from 454 to 649°C and to

various times and submitted to the appropriate ASME Code groups for

incorporation into the Code.

TENSILE STRENGTH

The influence of prolonged exposure on the elevated temperature short

term tensile properties is also of interest to steam generator designers.

Accordingly, material from three heats (25.4 mm thick plate) was aged for

periods up to 50,000 h (5.7 years) at temperatures ranging from 482 to 704cC.

Following aging the material was tensile tested at a strain rate of

6.7 x 10"5s"1 and at the temperature at which it was aged. The data were then

plotted in a parametric form for yield and tensile strength values as shown in

Figures 7 and 8, respectively, in order to develop equations that would permit

extrapolation of the data.

13

QUJ

O

z

CO

Q

UJ

IHI

o<C3LU

rrCO

aUJ

1.2

1.1

1.0

0.9

0.8-

0.7-

0.6-

•

0.5-:

0.4-1 C

X . A

•

A

•9

> 1 1

A

A A

•

HEAT 30176

HEAT 30383

HEAT 30394

1 ?

R =

A

e

14.

A

143* PA-1.1029

P = T(logt+ 10)/1,000

T = Temperature (K)

t = time (h)

A *

A • A

Aging And Test TemperatureWere The Same

i —i , 1 .

ORNL DWG 89-7801

Fig. 7. Ratio (R) of yield strength of aged to yield strength of unagedmacerial as a function of time and temperature in a parameterized (P) form forthree heats of Modified 9Cr-lMo steel.

oz3

UJ

ccCO

UJ

2

3QUJ

a<i

atuccCO

1.2

1 i ™I . I

1.0-

0.9-

0.8-

0.7-

0.6-

0.5-

0.4-

R =

a aa

• 755 K (482 C)

0 811 K (538 C)

• 866 K (593 C)

A 922 K (649 C)

A 977 K (704 C)

89-7802

12.418 *PA-1.0584

P = T(iogt + 10)/1,000

T = Temperature (K)

t = time (h)

* *

^ " S . A AA

Aging And Test TemperatureWere The Same

10 11 12 13 14 15

tig. 8. Ratio (R) of ultimate strength of aged to ultimate strength ofunaged material as a function of time and temperature in a parameterized (P)form for three heats of Modified 9Cr-lMo steel.

14

Figures 7 and 8 contain plots of ratios of yield or ultimate tensile

strengths for aged to unaged material as a function of a time-temperature

parameter P, where P is given by the following expression:

P = T(log C + 10}/1000,

T = temperature (K),

t = time (h).

Limited tensile data from a single heat are also available from tests

that were conducted on material aged for 75,000 h (8.6 years) at temperatures

ranging from 482 to 649°C. This permitted comparison between measured and

predicted values to be made as shown in Table 2. Predicted values are based

on the parametric expression (R equations) given in Figures 7 and 8.

Agreement between predicted and measured values is good considering that the

parametric expressions were developed from data taken from three heats and

considerable scatter is inherent as shown in Figures 7 and 8.

15

Table 2. Measured and predicted values of yield strength (YS) and ultimate

and tensile strength (UTS) at several temperatures.

Aging andtesting

temperature

CO

482

593

649

649

Unaged

Measured8

YS(MPa)

427

303

208

209

UTS(MPa)

490

324

233

243

YS(MPa)

428

256

163

170

Aged 75

Measured3

UTS(MPa)

486

275

187

185

,000 h

Predicted11

YS(MPa)

418

256

164

165

UTS(MPa)

470

269

181

189

aAt strain rate of 6.7 x 10"5 S"1.bUsing parametric expressions.

The parametric expressions were then used to estimate the reduction in

yield and tensile strengths that would occur following prolonged service at

elevated temperatures. Comparisons between unaged values for yield and

ultimate strength and value- estimated after thermal exposure for periods up

to sixty years are shown in Figu.es 9 and 10 respectively.

16

ORNL DWG 90-8162

soa«3

600

500

400

300

200

100

Average UnagedA

Aged 10 Yearsi i

Aged 30 Years

Maximum Stress Allowable,I I I 1 I

1 0 0 2 0 0 3 0 0 4 0 0 5 0 0

TEMPERATURE CO6 0 0

Fig. 9. Yield strength as a function of temperature comparing unaged to

estimates of aged material.

ORNL DWG 90-8163

£SX

Iasco

U

co

I125

soo

700

600

500

400

300

200

100

0

1

Avera•e U

-

t

—i • - —

i

i , _

—Aged 10 Years,

Aged

Maximum Stress Allowable, S,.

60 '^ears

• - — .

— .

IN^>N

100 200 300 400 500

TEMPERATURE (°C)6 0 0

Fig. 10. Ultimate strength as a function of temperature comparing unaged to

estimates of aged material.

17

Also shown in Figures 9 and 10 is the maximum stress allowable, So, for this

material as found in the ASME Code. The comparison indicates that prolonged

thermal exposure will not degrade tensile properties below So values.

TOUGHNESS

(Charpy Impact)

Charpy impact data are available from a single heat preaged for periods

up to 75,000 h (8.6 years) at 538°C. Results of these tests are summarized in

Table 3. Prolonged thermal exposure is seen to increase the transition

temperature and decrease the upper shelf energies somewhat in comparison to

unaged material.

18

Table 3. Charpy test results for unaged and aged heat 30394 (0.4 wt % silicon)

Agingtemperature

CC)

538

538

Agingtime

(h)

0

50,000

75,000

Hardness

(RB)

98

99

96

40.7 J(30 ft/lb)

energy

-30

15

20

Transition

67.8 J(50 ft/lb)energy

-4

39

52

temperature

0.89-mm(35-mil)lateral

expansion

-4 .

42

43

(°C)

50%shear

fracture

38

UO

49

19

MICROSTRUCTURAL OBSERVATIONS

Techniques of analytical electron microscopy (AEM) were used to study

both microstructural and microcompositinal changes that occurred in material

aged for 50,000 h (5.7 years) at either 482, 538, or 593°C. Figure 11 shows

representative transmission electron micrographs taken from thin-foil

specimens in which aged and unaged material is compared. Figure lla shows

details of the normalized and tempered microstructure in which high-

dislocation-density subboundaries within the matrix are apparent. Also seen

are spherical and lenticular shaped precipitates of MC (primarily vanadium-

and niobium-rich carbides) and M23 C6 (primarily Cr and Mo rich carbides)

alor.g subboundaries a'ld within subgrains. Thermal aging generally can cause a

number of changes in the microstructure including increased precipitate

densities of several phases e.g., M23 CB, MC, and Laves, and recovery

depending upon the time, temperature, and specific composition of the heat

involved. Laves phase (primarily Si-, Mo-, Fe-, and in some cases P-rich)

precipitation which was not found in unaged materials occurred over the

temperature range of 482 to 593°C, examples of which are shown in Figures lib

and lie. Laves phase is normally thought of as an embrittling component, the

presence of which can reduce room temperature toughness and long term creep-

rupture ductility. Recovery i.e. - reduction in dislocation density and

sharpening of subgrain boundaries, also occurred particularly with longer

times and at the higher temperatures as shown in Figure lid. Some fine

precipitate dissolution may have also occurred in the process of growing

larger precipitates. These processes of recovery and precipitate dissolution

would be expected to cause strength at the aging temperatures to decline

somewhat with time as reported herein.

20

Photo V 181757

(a)

LAVES FILMSAND COATINGS

Photo YE-13967

(b) Aged at 482°C

Fig. 11. Transmission electron microscopy comparisons of the microstructure

of aged and unaged material. Aged material was exposed for 50,000 h at the

indicated temperatures.

21

SUMMARY

Modified 9Cr-lMo steel has been under development as a steam generator

alloy for approximately 15 years. Currently there is world wide interest in

this alloy and it is expected to find use in many applications involving

elevated temperature service to about 593°C.

Results were reported herein from a number of both long and short term

mechanical property tests conducted on material exposed to elevated

temperatures for prolonged periods of time. The following are specific

conclusions.

1. Predictions of rupture behavior of base material using previously

developed rupture equations are very good over the temperature range of

427 to 704°C and to test times about 80,000 h.

2. Plots of creep-rupture ductility for base material measured as total

elongation or reduction of area indicate that these values decrease

somewhat with test times in excess of 10 to 20,000 h in the temperature

range of 593 to 649°C.

3. Specimens taken from weldments in the transverse direction such that the

gage length contained base, HAZ, and weld metal that had been given the

standard heat treatment showed reduced stress-rupture lives in

comparison to base material similarly heat treated. Failure occurred in

a soft zone at the edge of the HAZ. A model was given allowing rupture

23

strengths of weldments and base materials to be compared.

4. Results of elevated-temperature tensile tests were reported from work

conducted on three heats of previously thermally aged material at

various temperatures from 482 to 704°C and at aging times to 50,000 h.

The yield and tensile strengths were analyzed in terms of a time-

temperature parameter in order to make estimates of the elevated-

temperature changes following prolonged periods of thermal exposure in

service.

5. Exposure to a temperature of 538°C for periods up to 75,000 h increased

the ductile-to-brittle transition temperature.

6. Results from analytical electron microscopy were presented which showed

that prolonged thermal aging causes some alteration in the

microstructure which can account for the changes observed in the

mechanical properties.

24

Ellis, F. V., Henry, J. F., and Roberts, B. W., "Welding Fabrication,

and Service Experience with Modified 9Cr-lMo Steel," to be published in an

ASME/MPC Special Publication, Proceedings of ASME Pressure Vessel and Piping

Conference, Nashville, TN, June 17-21, 1990.

Haneda, H., Masuyama, F., Kaneko, S., and Toyoda, T., "Fabrication and

Characteristic Properties of Modified 9Cr-lMo Steel for Header and Piping,"

pp. 231-41 in Proceedings of the International Conference on Advances in

Material Technology for Fossil Power Plants, September 1-3, 1987, Chicago, IL,

ASM International, 1988.

Hasuyama, F., Haneda, H., Kaneko, S., and Toyoda, T., "Applications of

Super 9Cr Steel Large Diameter and Thick Wall Pipes," Mitsubishi Technical

Bulletin 182, Mitsubishi Heavy Industries, Ltd., July 1988.

Roberts, B. W. and Canonico, D. A., "Candidate Uses for Modified 9Cr-lMo

Steel in an Improved Coal-Fired Power Plant," pp. 5-55 to 5-82, Conference

Proceedings: First International Conference on Improved Coal-Fired Power

Plants, EPRI Report CS-5581-SR, Electric Power Research Institute, Palo Alto,

CA, 1988.

Photo YE-13987

LAVES PHASEPARTICLES

( c ) Aged a t 538°C

GENERAL RECOVERY *

Pho to YE-14035

(d ) Aged a t 593°C

22