9cr-2W-0.2V Fusion Reactor

ISIJ International, Vol. 30 (1990), No, Il, pp. 937-946

Effects of Taand Nbon Microstructures

Activation Ferritic 9cr-2W-0.2V Steel

and Mechanical Properties of Lowtor Fusion Reactor

Kentaro ASAKURA,YOSyunYAMASHITA,1)TakemiYAMADA2)and Koji SHIBATA

Departmentof Metallurgy and Materials Science, The University of Tokyo, Hongo, Bunkyo-ku, Tokyo, 113 Japan. 1)UndergraduateSchool, The University of Tokyo. Nowat Chiba Works, Kawasaki Steel Corporation, Kawasaki-cho, Chiba, Chiba-ken, 260 Japan.2) Steel Research Center, NKKCorporation, Minamiwatarida-cho, Kawasaki-ku, Kawasaki, Kanagawa-ken,210 Japan.

(Received on January 12, 1990, accepted lh the final form on May18. 1990)

In order to develop low activation high chromiumferritic steels for fusion reactors, the effects of O05-0.16 wt'/• Taand0.05 wt"/• Nb alloyings, and heat treatments on elevated temperature strength, toughness and microstructures of

0.1C-9Cr-2W-0.2V-0.04N steels were investigated. Normalizing and tempering conditions were determined asheating at I 050'C for 30 min and at 800'C for I hthrough observing the dissolution of carbonitrides of Ta and Astemperatures, respectively. The optimumamountof Ta in consideration of creep-rupture strength and toughness wasapproximately 0.10 wt'/. Delta-ferrite formed in 0.16 wi"/• Ta steel of which Cr-equivalent wasapproximately 10 wt'/•.

Thesteels alloyed with Tashowedonly slightly lower creep-rupture strength than that of the steel alloyed with the optimumamountof Nb, 0.05 wt"/•. Toughnessof the steels alloyed with Ta was superior to that of the steel alloyed with Nb.Hence, Ta-addition can be recommendedover Nb-addition in view of the improved toughness.

KEYWORDS:tusion reactor materials; Iow activation steel; ferritic steel; creep-rupture strength; toughness: tantalum;

niobium; grain size; microstructure; precipitate; carbide.

~}

1. IntrQduction

During the course of the development of materialsfor the flrst wall componentsof fusion reactors, therehas been an increased emphasis placed on the con-ducting of researches on irradiation damages, in par-ticular, surface damagessuch as swelling, irradiation

creep and sputtering along with blistering.1~3) Re-cently, in addition to such damages, reduction ofinduced radiation has become to be important.4)Since the first wall of fusion reactors is subject to ir-

radiation by fast neutrons (14 MeV), it is estimatedthat the radiation level is extremely high in compari-

son to light water and fast breeder reactors. Themajority of residual nuclei produced in nuc]_ear reac-tions with neutrons are radioactive isotopes, and these

cause decay heat in materials and induced radiation.

Hence, the level of the induced radiation of' mate-rials used in the first wall of fusion reactors should bereduced from the viewpoint of the safety during the

reactor maintenance, repairing and waste treatments.Researcheshavc beenconducted o_n austenitic stain-

less steels (such as D-9, modified Type 316 stainless

steel) and high-Cr ferritic steels (such as HT-9, modi-fled 9Cr-1Mosteel) as pro_mising candidates for theflrst wall materials. However, stcels alloyed with Ni,

Moand Nb produce radioactive nuclei having alonger lifetime than Fe. It has been reported that

morethan 100 years are required to reduce the radio-activity to safe levels.5) Because the candidate aus-tenitic stainless and high-Cr ferritic steels contain Nior Mo, considerable radiation is induced. Therefore,

studies have begunto examinefusion reactor first wallmaterials in which low activating elements are sub-stituted ibr such elements which have high inductionproperties.

High-Cr ferritic steels have been reported to beinferior to austenitic steels in terms of high-tempera-ture strength. However, it has been shownrecentlythat certain high-Cr ferritic steels have high-tenlpera-

ture strength superior to austenitic steels at tempera-tures below 6OO'C, and equivalent to austenitic steels

even at 650'C.6-8) In addition, ferritic steels aremo_re suitable due to their smaller dimensional changesbrought about by swelling and irradiation creep.Moreover, ferritic steels have several additional su-perior properties such as high thermal conductivityand low coefl*rcient of thermal expansion which makeferritic steels extremely available as structural mate-rials used at high temperatures. Consequently, high-Cr ferritic steels are drawing an attention as candidatematerials for the first vval] of lhsion reacto_ rs.

Numerousresearches have als'o been conducted onconventio_ nal first wall candidate materials including

Mn-Craustenitic steels in which Mnis substituted for

Ni, and Cr-Wand Cr-V ferritic steels in which Vand Ware substituted for M0.9~12) On the otherhand, a high-strength ferritic steel was developedthrough optimizing amounts of carbide formers Nband V,13) However, Nbwhich produces radioactivenuclei having a longer lifetime than Fe has to beavoided for fusion reactors.

The objective of the present research is to investi-

gate the influences of Ta as an element in substitution

C1990 ISIJ 937

ISIJ International, Vol. 30 (1990), No. Il

for Nb, a powerful carbide former, in non-irradiationconditions. The addition of Ta has been restrictedin flssion reactors such as light water and fast breederreactors, because Ta produces radioactive nuclei infission reactors. Calculation, however, has suggestedthat the decay time of the induced radiation by Ta in

nuclear fusion reactors is shorter than 30 years.14)

There have been very few reports which examine theaddition of Ta to ferritic steels, while Kawai et al.,15)

Tamuraet al.16) and Tupholmeet al,17) have dealt

with this subject. Assuming the cooling method ofthe first wall materials of the fusion reactor to be watercooling, the wall surface temperature is estimated tobe in the vicinity of 400-450'C. In consideration ofthe safety factor, the working temperature was as-

sumedto be 500'C.

2. Experimental Procedures

2.1. Steels UsedandHeat Treatment Conditions

Ingots weighing 50 kg were melted in a vacuuminduction furnace. Thecompositions of the steels areshown in Table I .

The basic composition of this

series of steels was designed to be 0.1C-9Cr-2W-O.2V-O.04Non the base of the previous rescarch,18)

Steel NB20contains the optimum amount of Nb,0.05 wtolo'l3) In order to compare the effects of Nband Ta, the atomic weight ratio of' Ta and Nbwasapproximated to be 2: I and the optimumamountof

Ta was assumedto be in the vicinity of 0.lO wto/o'

Theamountsof Ta in the TAseries steels were variedwithin the range of 0.05-0.16 wto/o' In atomic per-cent, the amountof' Ta in Steel TA12and the amountof Nbin Steel NB20are nearly equal. As presentedin Table l, there is a little difference in the amountsof Cand N in this series of steels. Hence, the total

amount (C+N) should be considered in the evalua-tion of the properties. Steels TA11 and NB20con-tain the largest amounts of (C+N), whereas Steel

TA13has the smallest one.The normalizing temperature was determined by

estimating the solution of carbides and nitrides

through examining the changes in hardness of thesteels heated with 50'C steps from 900 to 1200'C.The tempering temperature was determined throughmeasuring the a'-r transformation start point (A.

temperature) using a dilatometer with the heating

rate of 3.3'Cjmin. Eachof the steels washeat treated

and then machinedinto testpieces.

The matrix of this series of steels is a single tem-pered martensitic phase except for Steel TA13whichcontains a ferritic phase on the order of I o/o'

2. 2. Mechanical Properties

Creep-rupture tests were performed at 500, 600and 650'C. The testpieces for these tests were 6mmin diameter and 30 mmin gauge length. Single-

specimen testing machinesof' the lever type were used.

The ductile-brittle transition temperature (DBTT)was determined using a Class 300.J Charpy impacttester and JIS No. 4 testpieces. This transition tem-perature was evaluated as the temperature at whichthe absorbed energy is equal to l/2 of the upper andthe lower shelf energies.

2.3. Precipitation Behavior of Carbides and Structural

Changes

Carbides were identified by X-ray diffraction anal-ysis of the residue which waselectrolytically separated

from the steels in an electrolytic solution of 10 o/o

acetylacetone, I o/o tetra-ammoniumchloride andme-thanol (current density was approximately 200 mA/cm2). In addition, weight changes of the residues

were measuredusing a high precision balance (mini-

mumscale unit was O.1 mg). Microstructures wereobserved using a transmission electron microscope(H-800) operated at 200 kV. Carbides were also

identifled by electron diffraction analysis and elementanalysis using an electron microscope (JEM-4000EX)equipped with an X-ray energy dispersive spectrome-ter (TN-5500).

3. Experimental Results

3.1. Determination of JVlormalizing and Tempering Tem-peratures

In order to investigate solution of carbides andnitrides, changes in hardness were measuredfor steels

normalized for 30 min with 50"C steps in the tem-perature range of 900 1200'C. Those results arepresented in Fig. l. In the case of Steel NB20to

which Nbhas been alloyed, hardness increases withnormalizing temperature. It is conceivable that Nb-(C,N) disso_ Ives completely and the hardness rise satu-rates in the vicinity of I 150'C.19) On the otherhand, in the case of Steels TAI 1-TA13 to which Tahas been alloyed, the increase in hardness saturates in

the vicinity of I 050'C, which shows that Ta(C, N)dissolves nearly completely at around this tempera-ture. Differences in hardness are observed in Steels

TAll-TA13 normalized at around I 050'C. Thisis attributable to slight differences in the amountsof CandNamongthe steels.

Onthe base of the results mentioned above, nor-malizi,ng temperaturc was determincd to be I 050'Cat which Ta(C, N) dissolves into the matrix. This

l

Table l. Chemical composition of steels. (wto/o)

C Si Mn P S Cr W V Ta Nb N C+NTAI lTA12TA13

O.097 O.

OOI O.0020O.

04 O.49

O.084 O.

002 O.0024O.06 O.51

O.083 O.

002 O.0025O.

06 0.51

8.99

9.189.18

2,32 O,

20 O.047

2.35 O,

20 O,082

2.36 O.

20 O, 164

O.043 O.

140

O.039 O.

123

0.035 O. 118

NB20 O.090 O.

002 O.OO18O.

07 O.49 9,16 2.36 O. 19 O.

05 O.048 O. 138

938


420

~- 400,1)

u)UJzOOE Q,in(,ovI

360

.,~,**'~----

;y/~~~~~~~~)}_ Jl~- ~:l~rlA/

tr //(c TAI1D/'A TAI2D TA13

o NB20

900 950 1000 1050 1100 1150

TEMPERATURE('C )Fig. 1. Effect of normalizing temperature on hardness.

Table 2. A* temperature of steels. ('C)

TAI I TA12 TAI3 NB20O'05 o/o 0'08 o/o 0'16 o/o 0'05 "A(Ta )( )(Ta )( ")

NbTa

500~:;o~~~~!~:{~~~~=~~~~~~~z~~~~~~~ic~==-=

~~:--E~~::T 5000c=L---t=1!~ ~ - - -~ 300l i

_ 200 *~-~~o)--_E- -~:~== 600ec-~c_~;~~~~t~~~}

=~~o~~_Q.__~~~)~lof) -iv)uJCE 100 TA11 6500c~ 7nv A TA12

D TA1350 e NB20

1ol I02 I051Oo IO3 410

RUPTURETIME (h)Fig. 2. Effect of tantalum and niobium on creep-rupture

strength at 500, 600 and 650"C.

Steels

As temperature('C) 829 832 838 835

temperatufe is generally adopted as the optimumnormalizing temperature for Nbbearing high strengthhigh-Cr ferritic steels.19) An excessively low nor-malizing temperature which causes undissolved car-bonitrides and an excessively high normalizing tem-perature which causes crystal grain growth result in

deterioration of toughness. However, in order to

compareconveniently the effects of Nband Ta, nor-malizing temperature was unified to be I 050'C for

all the steels.

With respect to tempering temperature, whentoughness is conceived to be more important, highertempering temperatures would be appreciated at the

expense of creep-rupture strength. The results ofmeasuring A* temperature using a dilatometer areshown in Table 2. In TAseries steels, A, tempera-ture rises slightly in order of Steels TAI ITA13, being roughly 830'C in the case of Steel

TA11 and 838'C in the case of Steel TA13. Thetransformation temperatures of Steels TA12 andNB20are 835'C. As a result, the tempering con-dition was determined to be heating at 800'C for

l h.

3.2. Creep-rupture Properties

The results of' creep-rupture tests at 500, 600 and650'C are shown in Fig. 2. Steel NB20shows thehighest creep-rupture strength at all the testing tem-peratures. Addition of Ta results in a little lowercreep-rupture strength, whereas there is only a slight

difference in creep-rupture strength amongthe steels.

Onthe short-time side at 500 and 600'C, Steel TAI lalloyed with 0.05wto/o Ta exhibits higher strength.

Onthe long-time side, Steel TA12alloyed with 0.08wto/o Ta shows higher strength. Steel TA13alloyedwith 0.16wto/o Ta exhibits a tendency of excessiveaddition of Ta and the lowest strength level among

500

~300o-Z 200

u)(f)UJ TOOt~* 70Ln

50

30

~~ ~'~~~ ' 500'c-1aooooh~7~~~:11"I I~ '~~~:~h~~hl~~'_~;~~~:'}!~~;~~:;~~\

i '

: _

6.0.0_;: :o_ooooh400'c - Iooocoh

Ic TA11

' '_\A TA12 ' \~'

D TA13

e NB20

24 26 28 30 32 34 36 38T (35 * iog t* )x 10-3

Fig. 3. Larson-Miller plots of creep-rupture strength.

the TAseries steels.

At 650'C, Steel NB20shows the highest creep-rupture strength which is higher by 20 MPathan thatof Steel TA12over short-time periods of l0-lOO h.

However, whenthe time period is extended to 10 OOOh, the difference in strength reduces to 10MPa.There seemshardly any large difference in the creep-rupture strength of the TAseries steels with the excep-tion of Steel TA13. The rupture elongation of the

TA series steels was roughly 20 o/o at 500'C androughly 25 o/o at 650'C. On the other hand, the

creep-rupture elongation for Steel NB20was slightly

smaller, being 18 o/o at 500'C and 20 o/o at 650'C.For both the Ta-addedsteels and the Nb-addedsteels,

no remarkable creep embrittlement wasobserved.Fig. 3 shows the interrelation between creep-rup-

ture strength and the LarsonMiller parameter (C=35). The assumedworking temperature of the nu-clear fusion reactor is in the vicinity of 400-450'C in

the case of the water cooling as stated previously.

Whenthe 105 h creep-rupture strength is estimatedfrom Fig. 3, Steel NB20shows the highest creep-rup-ture strength, being roughly 440 MPaat 400'C androug_hly 380 MPaat 450'C. The strength higherthan 290 MPais estimated even at 500'C and 105 h.

Amongthe TAseries steels, although the creep-rup-ture strengths ol' Steels TAI I and TA12are similar,

on the low temperature side at 400-450'C, Steel TAI lshows a little superior strength. Steel TA13whichcontains 0.16 wto/) Ta is most inferior amongall thesteels. If the cooling system of the nuclear fusion

reactor is assumedto be helium cooling instead of

water cooling, the working temperature increases to600-650'C. In this case, similarly to the case of

939

;,;i{I

';'1

;!I

ISIJ International, Vol. 30 (1990). No. Il

water cooling, Ta as a substitute for Nbshowsonly asmall negative factor with respect to creep-rupturestrength.

Based on the above results, as for creep-rupturestrength, the optimumamountof Ta is in the rang_eof O.04-0.lO wto/o' Although there is few data con-cerning the effects of Ta on creep-rupture strength,

Kawai el al.15) have investigated thc effects of Ta andN on the creep-rupture strength of 12Cr-1Mo-0.2Vsteel. According to their results, the combinedaddi-tion of Ta and N is effective in improving creep-rup-ture strength through increasing the intragranular

creep resistance due to the precipitation of fine

Ta(C+Nl-+)' However, this steel has shortcomingssuch as a small creep-rupture elongation.15) Fur-thermore, since the reduction of the induced radiationhas not been taken into consideration in the alloy

design process ; this steel contains I wto/o Mo. Ontheother hand, Tamuraet al.16) have reported that Taaddition improvcs creep-rupture strength by approxi-mately 20-30 Mpa(an improvement of roug_hly 20-30 o/o) by the addition of 0.04 wto/o Ta to 8Cr-2W-Vsteel (F-82H steel). According to their results based

on creep-rupture strength at 100 OOOh, the effect onthe strength is largest at 550'C, whereas small at500'C~". The effect becomessmall again at 600'C.

Whenthe creep-rupture strengths are compared to

those of the TA series of steels in the present work,the strengths of the TAseries steels are roughly 30-50 MPahigher. One reason for this fact can besurmised to be the difference in the amount of' N.Nitrogen content oi' the TA series steels is around0.04 wto/o' while that of F-82Hsteel is 0.002 wto/o'

Tupholme et al.17) have investigated the changein creep-rupture strength, high-temperature tensile

strength and toughness by the addition of 0.1 wt~~ of

Ta to I ICr-3W-0.25V and 9Cr-0.75W-0.25V steels

in which the amounts of N and/or Cwere reduced.As a result, it has been suggested that 9Cr-0.75W-0.25V steel, in which the Ncontent is reduced, showsnearly the samelevel of creep-rupture strength as thatof several conventional steels. For instance, althoughthe creep-rupture strength of this steel is inferior

to that of FV 448 (O.13C-l0.5Cr-0.75Mo-0.15V-0.45Nb-0.05N) at 500'C, in the temperature regionsof 550 or 575'C, this steel demonstrates superiorproperties to those of FV448. Such the high creep-rupture strength of this steel developed by Tupholmeet al.17) is mainly attributable to the low temperingtemperature. Hence, it is conceivable that the tough-

ness of this steel is inferior to that of the TAseries

steels cxamined in the present work.

3. 3. Toughness

Fig. 4 shows the ductile-brittle transition of thestcels in the NT (normalized and tempcred) condi-tion. The order of the transition temperature fromlow to high is Steels TA12, NB20followed by Steels

TAI I and TA13. The transition temperatures are in

the vicinities of -50'C for Steel TA12, -40'C for

Steel NB20and -30'C for Steels TAll and TA13.The upper shelf energy of Steel NB20which shows

,- 300>,J)

O~UJzuJ,:)

200UJaQocou)ao 100

~o~

:ru

Fig. 4.

o

.(~. -:- as temperlng:~/~1~f

~f( !l~

l riIl

ji

c TA11'1

A TA124 /f ~i D TA13

~..••'~/ ' NB20

80- -40 O 40 80 100TESTINGTEMPERATURE('C )

Ductile-brittle transition curves as tempered at

800'C.

reheated- 300 at 5000c_103h:>

__{r*~o ___~r~r/4(

llo(UJ I~zUJ 200 ji foUJ jilcn

o!i~ TA11oc

v) A TAI2oo 100 l! , D TA13"Io:

'~ !> / e NB20,L

~ '/'u o

- -40 O 40 eo 10080TESTINGTEMPERATURE('C )

Fig. 5. Ductile-brittle transition curves of steels after heat-

ing at 500'C for I OOOh.

the highest creep-rupture strength is roughly 200Jwhich is lower than that of the other steels.

Fig. 5 shows the ductile-brittle transition of thesteels heated for I OOOh at 500'C following NTtreat-

ment. In comparison to the results of the as-NTtreated steels, the TAseries steels exhibit hardly anyembrittlement after heating at 500'C. Toughnessis

improved and their DBTTshift to the lower tem-perature side. The primary cause of this improve-

ment is conceivable to be matrix recovery. In otherwords, as a result of the heating at 500'C, the internal

stress of the matrix is relieved without any remarkablecoalescence and coarsening of the precipitates whichwill be described later. Onthe other hand, a clear

tendency toward embrittlement is observed in Steel

NB20. That is to say, the upper shelf energy de-

creases further and DBTTshifts by approximately15'C to the high temperature side.

From the results mentioned above, it has becomeclear that the additio_ n of Ta as a substitute for Nbcauses an increase in toughness. According to theresults of Charpyimpact tests conducted by Tupho_Imeet al.,17) although there were no major differences be-

tween steels alloyed with and without Ta in the caseof tempering at 675'C, steels with Ta are superior in

terms of both room temperature absorbed energy andfracture appearance transition temperature (FATT)in the case of tempering at 750'C. As mentioned

=1

fl

940


previously, this improvement of toughness is attrib-

utable to the relief of the internal stress induced bynormalizing through tempering at higher tcmpera-tures. Therefore, in order to obtain adequate tough-

ness in steels alloyed with Ta, it is recommcndedthatthe tempering is carried out at a temperature as high

as possible.

3. 4. Changesin Hardness through Heating at 500'CChanges in hardness through the NT treatment

and heating at 500'C for 100-1 OOOh are presentedin Fig. 6. As described earlier, the hardness of Steel

TAI I containing a low Ta content is similar to thatof Steel NB20in the normalized state. Whenthese

are tempered at 800'C, Steel NB20shows the highestlevel of hardness. Hardness in the as-normalizedcondition decreases in order of Steels TAI I , TA12and TA13. The order of hardness of the steels

heated at 500'C for 100 h resembles the order of

creep-rupture strength. Although hardness ol' Steels

TA13and NB20decreases through heating for I OOOh, the decrease of Steel TA13 is particularly notice-able. The change in hardness of Steel TA12is verysmall.

Except for Steel TA13which contains the largest

amountof Ta, demonstrating sornewhat lower hard-ness values in comparison to the other steels, thechanges in hardness l~csemble the changes in creep-rupture strength described earlier. Although the

cause for this fact will be discussed later, Ta has alesser tendency toward the formation of carbides thanNb, and in the as-normalized state, the amount ofT~a(C, N) is less than that of Nb(C, N). In addi-tion, if the amountof Ta is optimumthe decreases inhardness through heating are suppressed due to slowrate of recovery and difficulty of' coalescence andcoarsening of Ta(C, N). The remarkable decreasein hardness of Steel TA13is attributable to an exces-sive amountof Ta and delta-ferrite. Precipitates atthe interface of martensitic phaseand delta-ferrite areliable to coarsen.

4. Discussion

In order to investigate the causes ibr the differencesin mechanical properties betweenTa-alloyed and Nb-alloyed steels, the interrelation between mechanicalproperties and microstructures ol' each steel will bediscussed.

4.1. Cr-equivalent, A* Temperature and Delta-ferrite For-mation

Fig. 7 illustrates the relationship between A, tem-perature and Cr-equivalent. The equation shownbelow wasused for calculating Cr-equivalent.20) Thisequation was originally obtained to estimate the ten-dency to form delta-ferrite in the condition of nor-malizing at I 050'C for 30 min. Element symbolsshown in brackets represent the weig_ht percent ofeach element. Concerning Ta, the eflbct is assumedto be roughly 1/2 of the effect of Nbin considerationof the atomic weight ratio of Nband Ta.

Fig. 6.

Fig.

VUJCCDhCCUJO-:~UJh

~

500

400

~LfD 300v)LLl

zcDo:

~200

100

(>

A~,i O

'~,, e*~,

~',t~

\'~

'*~\

~ ~~~~

NORMALIZINGTEMPERINeREHEATINO

TA11

TA12

TA13

NB20

1050'C•1/2h800'C• Ih500'C

IL*+*J_+as N. as T. IOOh 1000h

HEATTREATMENTChangesin hardness through tempering and heat-ing at 500'C.

840

830

820

TAI1

NB20TA13

TA12

9.585 10.09.0

Cr - EauIVALENT (wtey.)

7. Relationship between Cr-equivalent and Asperature.

tem-

Fig. 8. Delta-ferrite in

l 050"C.Steel TA13 as-normalized at

Cr-equivalent = [Cr] +6[Si] +II[V] +5[Nb]+2.5[Ta] + I.5[WJ - 4[Ni]

- 2[Mnl - 30[N] - 40[C]

Fig. 7showsthat the transformation point increases

as the Cr equivalent increases. The Cr-cquivalentfor Steel TA13is 9.96 wto/o which is highest in all thesteels used in the present work, Observation of mi-crostructures showed that Steel TA13was the onlysteel in which delta-ferrite was formed. Fig. 8 is anelectron micrograph of such a delta-ferrite observed

' ~ijl

941

ISIJ International, Vol

in the NT condition of Steel TA13. Coalescenceof carbides at the interface between the ferritic andmartensitic phases is particularly conspicuous. Whenferritic phase is formed on the order of 1-2 o/o' alarge cornposition gradient of alloy elements is formedat the interface between the both phases, which is

contributable to the promotion of coalescence andcoarsening of precipitates. As the amount of thedelta-ferrite increases to l0-20 o/o' the compositiongradient and the coarsening tendency decrease. Con-sequently, the formation of the ferritic phase on theorder of 1-2 o/o is a negative factor for toughness. It

has been observed21) that when ferritic phase is 10-20 o/o in volume, the effect on toughness is little andthat once ferritic phase exceeds 30 o/o' embrittlementagain becomesconspicuous.

4.2. TEMStructure Observation

Fig. 9presents the TEMstructure for the steels in

the NTcondition. Amongthe TAseries steels, Steel

TAI I exhibits the most noticeable intragranular pre-cipitation and coalescence of precipitates as shownin

Fig. 9(a). This is contributable to the high creep-rupture strength of this steel on the short-time side at500'C. As shown in Fig. 9(c), fine precipitates,

which maybe Ta(C, N), are observed in Steel TA13.In this steel, the total amountof precipitates is smalland the average distance between precipitating par-ticles is large. The highest degree of matrix recoveryof this steel is attributable to this fact. Onthe otherhand, the precipitates in Steel NB20are extremelylarge and numerousas can be observed in Fig. 9(d).

A Iot of precipitates of 100-200 nmin diameter areobserved at the prior-austenite grain boundaries.Because the strength of the grain boundary is con-ceivable to be increased by these precipitates, this

fact is contributable to the increased creep-rupturestrength of this steel on the short-time side at 500'C.However, precipitation at the prior-austenite grainboundaries is the largest cause for a decrease in tough-ness.

942

30 (1990), No. Il

Fig. 10 shows the results of observation by TEMafter heating for I OOOh at 500'C. As is observed inFig. lO(d) no remarkable changes in microstruc-tures arc observed in Steel NB20in comparison to the

NTcondition, although the amount of' precipitasesincreases slightly. In addition, the dislocation densityis high and matrix recovery does not progress signifl-

cantly. This is conccivable to be one of the rcasonswhy the creep-rupture strength oi' this steel is higherthan that of the TAseries steels. In the case oi' Steel

TA13in which the amountof precipitates in the NTcondition is smallest, coalescence of she precipitates

during heating for I OOOh at 500'C is remarkable asis observed in Fig. lO(c). Furthermore, matrix re-

covery is signiflcantly progressed. With respect toSteel TA12, fine precipitates which are presumedtobe Ta(C, N) are observed as shown in Fig. 10(b).

Coalescence of' the precipitates hardly occurs at all.

In addition, the rate of matrix recovery is observed tobe slowest amongall the TAseries steels. The fine

Ta(C, N) precipitates are contributable to suppressmatrix recovery. Lath structure of martensite wasobserved to remain. The superior creep-rupturestrength of this steel is attributable to these structuralcharacteristics.

4.3. Identification of Precipitates andPrecipitation Behavior

Table 3presents the results of identification of theprecipitates in the steels heated at 500'C for I OOOhfollowing the NT-treatment using an X-ray diffracto-

meter. Themajor precipitates are M23C6' Peaks for

Ta(C, N) were observed only slightly in Steel TA13which contained the largest amountof' Ta. The flne

intragranular precipitate presumed to be Ta(C, N)which was observed by TEMwas not able to bedetected in Steel TA12with an X-ray diffractometer.

Fig. 11 shows the results of electron diffraction

analysis of precipitates in Steel TA13 after heating

l OOOh at 500'C. X-ray spectra from the precipi-

tates in Fig. I I are shownin Fig. 12.

Fig. 13 shows the change in the amount of the

(a) TAll (b) TS12(c) TA13 (d) NB20

Fig. 9. Electron micrographs after tem-pering at 800'C fbr I h.

'l

ISIJ International, Vol. 30 (1990), No, Il

residue extracted electrolytically from the steels heatedat 500'C for 100-1 OOOh following the NT-treatment.Similarly to the results of TEMobservation, the resi-

due weight of Steels TAI I and NB2Ois heavier thanthat of Steels TA12and TA13in the NTcondition.

Evenwhenheated at 500'C" for I OOOh, there is hard-ly any change in the residue weight. In contrast tothis, although the residues weights of steels TA12and TA13is lighter in comparison to other steels in

the NTcondition, the residue weight increases to thelevels ol' the other steels through heating at 500'C for

1OOOh.

The amounsof' C in the steels is approximately0.09 wto/o' Ifall of this Cprecipitates as M23C6(M:

metallicCr, theof

Table

element)

amountand if all of M

of the precipitate

is assumedis on the

to beorder

Fig.

3. Results of X-ray diffraction analyses of ex-tracted residue.

SteelsPrecipitates

As tempered 500'C-1 OOOh'I'All (0.05 Ta)TA12 (0.08 Ta)TA13 (O. 16 Ta)

M23C6M23C6M23C6'Ta(C, N)

M23C6M23C6M23C6' Ta(C, N)

NB20(O.05 Nb) rlM23v6 M23C6

(a) TA11 (b) TA12(c) TA13 (d) NB20

lO. Electron micrographs after heat-

ing at 500'C fbr I OOOh.

llOe

llOe zoO

OOO eO llO

o

020e 13O

o

220

310e

(d)

~

s

~,~h

s

Fig. 11. (a) Electron micrograph of extraction replica for Steel TA13after

heating at 500'C for I OOOh. Selected area diffraction patterns(b) and (c) show that particles A and B in (a) are M23CsandTa(C, N), respectively. (d) showsindex diagram for the diffrac-

tion pattern of particle B, Ta(C, N).

Ta

Fig. 12.

Cu

(a)

lO

2 3 4 5 S 7 8 9ENERGY (k *_ V)

(a) For particle A, M23C6(b) For particle B, Ta(C, N)Energy-dispersive X-ray spectra fromparticles A and Bin Fig. 11(a).

Peaks of Cu come from a sample

support grid.

943


c TA11~ A TA12~2 o TA13 ~.;i,3 ' NB20 ra5

*)((~('

if

r~~;.•-

UJOC l'~

~lH

;~•'

z ~ NORHALIzING I050' c. I /2h:D TEMPERING 800'c . IhO= REHEATING 500' c

HJ=1O CISN asT 100h 1000hHEAT TREATMENT

Fig. 13. Changes in the amount of extracted

through heat treatments.

0.09x(1+1:x

23 x52)= 1.58 (wt'/")

12 T

residue

Nitrogen is also involved in the precipitate and me-tallic elements other than Cr are considered to mix in

M. Whenthis fact is taken into consideration, the

total amountof precipitates is estimated to be roughly

2wt~~• Therefbre, Fig. 13 suggests that nearly all

ofthe (C+N) in Steels TAI I and NB20are containedin precipitates in the NT condition. As a result,

during heating, the amount of precipitates does notincrease and a tendency for precipitates to coalesce

and coarsen is observcd. However, whcn the steels

are heated at a temperature at which precipitates

without involving Cor N, such as Laves phases, areformed, the residue weight mayincrease.22,23)

Onthe other hand, in the case of' Steel TA12whichshows favorable creep-rupture strength, some(C+N)remain in the matrix in the NTcondition. Theyarepresumedto precipitate gradually during heating. Inother words, it is conceivable that the stability against

coalescence of the precipitates in Steel TA12 is re-lated to the amountof (C+N) in solution.

Possible reasons for such differences in the precipi-

tation behavior amongthe TAseries steels are listed

below.(1) The amount of (C+N) in Steels TA11 and

NB20is slightly larger than that in Steels TA12andTA13.

(2) Calculation of the solubility shows that the

amount oi' Nb in undissolved Nb(C, N) in Steel

NB20is roughly 60 O/o' Such undissolved Nb(C, N)includes into the weight of' the residue oi' the steel in

the NTcondition.(3) Solute NbandTahavestrong attractive forces

with solute Cand N, and decrease thc activity of Cand N. Therefore, the amountof solute Nband Tamayplay an important role in the precipitation be-havior of carbon-nitridc.

Since the amount of Ta(C, N) and Nb(C, N) is

only a minor portion in the total precipitates, the

reason (2) is not reasonable.

4.4. Grain Size and Mechanical Properties

At low temperatures, grain boundaries becomebar-

riers against dislocation movementand crack prop-agation, and thereby contribute to improve strength

Fig. 14.

40

El30

UJN~; 20

z~:o(o 10

1050'C- Normalizing

e' TA11\ A TA12\ \ \ D TA13

\.\ o NB20

,\

,

O O.5 1_O 1.5

Ta - CONTENT(wt '/. )Relationship between Ta content and prior

austcnite grain size.

and toughness. However, at high temperatures, grain

boundaries becomesources of production or cites

of absorption for voids. Consequently, whengrainsize is small (whereby the grain boundary area perunit volume is large), dcfbrmations due to void flow

and dislocation climbing are promoted and the creeprate increases, which leads to weakening. Thcre-fore, there are numerous reports which claim that

within the rangc of suitable grain size, the lar~er

crystal grain size results in the higher crcep-rupturestrength.

Fig, 14 presents the relationship between the

amountof Taand the prior-austenite grain size of the

steels norrnalized at I O50"C. The prior-austenite

grain sizes of Steels TAI I, TA12andTAI3are rough-ly 20, 23 and 12 um, respectively. The grain size of

Steel NB20is roughly 13 um. In Fig. 14, it is ob-served tha_.t steels containing 0.08-0.lGwto/o Ta showsthe largest grain size and a further increase in the

amountof Ta results in a decrease in grain size. Thereason for the small grain size of' Steel TA11 (0.05

wto/o Ta) is conceivable to be due to the larger

amount of (C+N). Grain size can be predicted to

be roughly 27 umthrough linear extrapolation lromthe points for 0.08 and O. 16 wto/o Ta if the amountof

(C+N) of Steel TAll is assumedto be the sameasthat of other steels, and 33 umwheneither Ta or Nbis not alloyed.

With respect to the effects of grain size on creep-rupture strength, grain size dependency is observedonly in the TA series steels to which Ta is alloyed.

Creep-rupture strength of Steel TA13having smallergrain size was lower than those of Steels TA11 andTA12having larger grain size. However, althoughgrain size of' Steel NB20is as small as that of Steel

TA13, creep-rupture strength of Steel NB20is highest

amongall the steels examined. This fact contradicts

the intcrpretation stated earlier that creep-rupturestrength is improved with increasing crystal grainsize. This contradiction is attributable to the differ-

ences in the amountof (C+N), and the existence of

undissolved Nb(C, N) and precipitates at grainboundaries.

In regard to the effects of grain size on toughness

,'l

l

l

944


(DBTT), steels with smaller grain size do not neces-sarily show superior toughness. For instance, in the

case of Steels TAI l-TA13 in the NTcondition (Fig.

4), thc steel which demonstrated superior toughness(the lowest DBTT)was Steel TA12 which had the

largest grain size. The superior toughness of' Steel

TA12can be interpreted by the consideration of the

amount, distribution and size of precipitates in this

steel. Steel NB20of which grain size is smallest

amongall the steels, exhibits always low toug_hness.

The reason for this is also able to be explained bythe amount, distribution and size of precipitates in

this steel as mentioned above.

4.5. Strengthening Factors of EachSteel

The relationships between mechanical properties

and microstructures are summarizedas follows.

(1) In Steel TAI l, the amountof precipitates is

large in the NTcondition due to the high content of

(C+N). The crcep-rupture strength on the short-

time side at 500'C washighest lbr Steel TAI I amongthe TAseries steels. Whenthe amountof C is ex-cessive as in the case of Steel TAI I ,

it is conceivable

that the size of precipitates tends to increase and the

strength on the long-time side or hig_ h-temperatureside decreases. In order to reduce the carbide growthrate and maintain the strength of' the matrix, it is im-

portant to optimize the amount of (C+N). Whenthe amount is too small, strength will be low, and

even whenthe amount is too large, strength will bereduced. The toughness of Steel TAI I is inferior to

that of Steel TA12. This is attributable to the coale-

scence of precipitates.

(2) In the case of Steel TA12, the amountof pre-cipitates in the NTcondition is small, and precipita-

tion proceeds gradually during heating at 500'C.

The rate of matrix recovery is slowest amongall the

steels examined in this work. This slow matrix re-

covery is attributable to the suppression oi' dislocation

climbing due to the precipitation of fine Ta(C, N).

Since the precipitates are small in this steel, superior

toughness is obtained.(3) Steel TA13, similarly to Steel TA12, results

in a small amount of precipitates through the NT-treatment. However, the size of the precipitates andthe average distance between themare larger in Steel

TA13. Becauseof' such structural characteristics, the

rate of matrix recovery of Steel TA13is conceivable

to be most progressed. Delta-ferrite exists only in

this Steel TA13. As a result, precipitates such as

M23C6are liable to coalesce at the interface of the

martensitic and ferritic phases. Thc amount of Tais in excess in this steel.

(4) In Steel NB20, the amount of (C+N) is

roughly the sameas that in Steel TA11. Since the

amountof precipitates is also large in the NTcondi-

tion and the rate of matrix recovery is also slowest,

creep-rupture strength of this steels is highest at all

temperatures. However, comparedto the TAseries

steels, Steel NB20exhibits lo_wer to_ughness: Iowershelf energy and higher DBTT.

Creep-rupture strengths of Steels TAI I and TA12

were superior. In these steels the amountsofTa and(C+N)are suitable. In particular, Steel TA12formsfine Ta(C+N)and the total amountof solute (C+N)in thc rnatrix is relatively small in the NTcondition.

Thesestructural characteristics contribute to reducingthe embrittlement through decreasing the coarsening

rate of M23C6' Consequently, considering both creep-rupture strength and toug_hness, the optimumamountsof (C+N) and Ta are in the vicinity of O.12 and 0.lOwto/o' respectively.

5. Conclusions

The effects of addition of 0.05, 0.08 and 0.16 wto/o

Ta and 0.05 wto/o Nbon creep-rupture strength andtoughness were invcstigated using_ 0.lC-9Cr-2W-0.2V-0.04N steels in order to develop low activation

high-Cr ferritic steels for fusion reactors. The con-clusions are summarizedas follows.

(1 ) Carbonitride ol' Ta dissolves nearly complete-ly into the matrix through normalizing at I 050'Cfor 30 min.

(2) The optimumamountof Ta in consideration

of creep-rupture strength and toughness is approxi-mately O. 10 wto/o'

(3) Delta-ferrite formed only in 0.16wto/o Tasteel. The Cr-equivalent of this steel was approxi-mately 10 wto/o'

(4) Thesteels alloyed with Tashowedonly slight-

ly lower creep-rupture strength than that of the steel

alloyed with the optimumamountof Nb, 0.05 wto/o'

Toughnessof' the steels alloyed with Ta was superior

to that of' the steel alloyed with Nb.(5) Tantalum is superior to Nbfor low activation

steels in view of the improved toughness together with

a smaller induced radiation effect.

1)

2)

3)

4)

5)

6)

7)

8)

9)

10)

11)

12)

REFERENCESR. E. Gold, E. E. Bloom, F. W.Clinard, D. L. Smith, R. D.Stevcnson and V\,f. G.. ~,'Volfer: Nucl. Tech.jFusion, I (1980),

169.

N. Igata: J. JVlucl. Mater., 133-134 (1985), 141.

A. Kohyama,K. Asakura, Y. Kohno, K. Komamura,K.Suzuki, M. Kiritani T. Fujita and N. Igata: J. JV;ucl.

ladlater., 133-134 (1985), 628.

R. W. Connet al. : Panel Report on LowActivation Ma-terials for Fusion Apprications, UCLAReport PPG-728,(1983).

T. Nodaand F. Abe, H. Araki and M. Okada: J. JVlucl.

Mater., 141-143 (1986), 1102.

T. Fujita, K. Asakura and T. Sato: Trans. Iron Steel Inst.

Jpn., 19 (1979), 60J'.

T. Fujita: Met. Prog., 130 (1986), Aug., 33.

K. Asakura andT. Fujita: J. At. EnergySoc. Jpn., 28 (1986),

656.

R. L. Klueh, D. S. Gelles and T. A. Lechtenberg: J. JVlucl.

Mater., 141-143 (1986), 1081.

F. Abc, H. Araki, T. Nodaand M. Okada: J. JV;ucl. Ma-ter., 155-157 (1988), 656.

A. H. F~ott, F. B. Pickc.ring and G.J. Butterworth: J. JViucl.

M:ater., 141--143 (1986), 1088.

H. Kayano, A. Kimura, M. Narui, Y. Sasaki, Y. Suzuki andOhta: J. JV;ucl. Mater., 155-157 (1988), 978,

945

=,


13)

14)

15)

16)

17)

18)

T. Fujita, K. Asakura, T. Sawada.T. Takamatsuand Y.Otoguro: Metall. Trans. A, 12A(1981), 1071.

T. Noda, H. Araki, F. Abeand M. Okada: Trans. JV:RIM,30 (1988), 185.

M. Kawai, K. Kawaguchi, H. Yoshida. E. KanazawaandS. Mit0: Tetsu-to-Hagand, 61 (1975), 229.

M. Tamura, H. Hayakawa,M. Tanimura, A. Hishinumaand T. Kondo: J. JV:ucl. Mater., 141-143 (1986), 1067.

K.W. Tupholme, D. Dulieu and G.J. Butterworth: J.JV;ucl. Mater., 155-157 (1988), 650.

K. Asakura and T. Fujita: The 123th Comm.(Heat-Re-

19)

20)

21)

22)

23)

sisting Matels and Alloys), Japan Soc, the Promotion Sci.

(JSPS), 26 (1985), 271.

X. Y. Liu and T. Fujita: Tetsu-to-Hagand, 74 (1988), 1065.C.J. Boyle and D. L. Newhouse: Met. Prog., 87, (1965),

March, 61.

T. Fujita, K. Asakura and H. Miyake: Trans. Iron Steel

Inst. Jpn., 26 (1986), 1073,

M. Tamura, H. Hayakawa,A. Yoshitake, A. Hishinumaand T. Kondo: J. JVlucl. Mater., 155-157 (1988), 620.

K. Asakura, Y. Koubuti and T. Fujita: Tetsu-to-Hagard, 75(1989), 1209.

,'"I

946

9cr-2W-0.2V Fusion Reactor

Documents