..__ A ? tO H N I.OA It, NASA TM X-71895

! f

IX

_,-i (NASA-TM-X-71895) IN_E[_IM ANALYSIS O_ LONG N76-21318_a_ TIM_ CL_]:EP B_H_,VIOR OF COLUHBIUH C-IOJ ALLOY

(NASA) 21 p FIC $3.50 CSCL 11FOnclas

< G3/26 21539o.

,[

. .;'iINERIMANALYSISOFLONG-TIMECREEPBEHAVIOR

OFC,:''_;_"'tt_ r̂.-IO3ALLOY

by w. . _ndRobert H. TttranLewis Re_ _ .... ;or

Clevelaud_ Ohio 44135

March 1976

ii

' t

1976014230

https://ntrs.nasa.gov/search.jsp?R=19760014230 2020-06-22T14:05:38+00:00Z

1 lliqin++! N" _i. (+++iVlHllllilinl Al'cl+PSlil+bll Ikl,; ..++lllcl.llHl,llt,, +,++hl!I,GI t.

NAHA TM X-71fll)fi'l I+ltll} illl(I f]lllllllli' G Irilllilifl I'J,lllf

INTEIIIM ANAI.,YglH O1,1LONCI.TIME CREEP ........ ......Is. llilll(llllllllll I Illl;tlil/,llilill (;+iillllil4ttAViOlt OP COI,IIMIIII1M C-1(13 ALLOY

] AlllllOi(lil || llltl|Oiltiil'll} fillJ'ilihlillllill llll}lllll liil

+'++ "+- .................... ¢it1_, Wllllilin I), llopp lind thllit, rf II, Tilran lO. W,,+I Ih,li No "

lJ. tliifhltinll+lll OillilliilliliOll NiliilO Illld Alll|lli++

l+,lJW liY IIC Xciirch Colltor "+i+ "'c717,t;,]+7 +<li_71"+]]lt--I_'+'<,........................ ]Naliunai At;ronautlcs tu_.d Splice Administration

Clevehmd, Ohio 44135 13, 1"vimt,t il<,l,u,,,,,,+lv_,++-_;+-¢:_;v,i,_+7.....

17. Sl+onsoling Agtiric7 Name and Addtli$1

Nationtd Aeronttutics lid Space ldnlhlistration "_, si,oniorintlAgencyC,nlu

Washhtt4_on, D.C. 20546

15. Supplementary Notes _.

This information is being published in preliminary torsi in order to expedite ks early release.

I1U. Abstract

Analysis of 16 long-time creep testson columblum C-I03 alloy(Cb-l,0Hf-lTi-0.7Zr) indicates i

thatthe calculatedstresses to give.I percent creep strainillI00,000 hours at 12fi5K (1800 F) , I

are 7.93 and 8.96 MPa (I150 and 1300 psi)for fine-grained'andcourse-gralned material, re- i

spectively. The apparent activation energy and stress dependence for creep of this alloy are +

approximately 315 KJ/gmol (75, 300 cal/gmol) and 2.51, respectively, based on Dorn-Sherby j

types of relations. However, the 90 percent confidence limits on these values are wide because

ofthe limiteddata currently available, i

17. key Wolds (Suggl'ltod by Author(Ill 18. Distlibution Stlltllllll.ql!

Unt:hisslfied - unllnitted

+19 F_-_:tlrity Clilt++if (of this r+P_,iII 10. 17_,,_;ilrity Clils+it. (of this P,lltel 71 Nv _ll P,ttl,++ 22 hit ,

tint;hi+tiff led Unclassified

' Fol s,tle by the N<llion+llTcehilit',il liltoim,lhon,_elv,,i'. Sflllnl_ht,hl Villllill.I ;J7161

• ..., ...... ++ .] .. . _: ..........

" ......... " _ " " ' ..... • ....... " _'" _ " " " " " " "" _ i

1976014230-TSA03

INTERgM--A_YSIS OF LONG_TIME CREEPBEHAVIOR OF COLUMBIUM C-103 ALLOY

by Willinm D. Klopp a.d Rob.rt )l. T ltran

/ SUMMARY

An analysls is presented based on 16 creep tests from an on-golng !._

program to evaluate the long-time creep behavior of columbium C-103 a]- ,!loy. This interim analysis indicates that the calculated stresses to _Igive i percent creep strain in i00,000 hours at 1255 K (1800 F) are 7.93

t

and 8.96 MPa (1150 and 1300 psi) for flne-gralned and coarse-grained ma- '

terial, respectively. The alloy exhibits an accelerating creep rate at

strains of less than 1 percent which can be differentiated lhto periods

termed early tertiary creep (time dependence of 3/2) and late tertiary

creep (time dependence of 5/2). No periods of identifiable primary or

-- _ secondary creep were observed The tlmes-to-l-percent-straln and early

and late tertiary creep rates were correlated with stress and temperature _by Dorn-Sherby types of relations. The apparent activation energy and I

stress dependence applicable to all three relations are 315+._49KJ/gmol

(75,300+--11,700cal/gmol) and 2.51+._0.44,respectively. The creep rate forthe fine-grained material accelerates more rapidly with time than that for

the coarse-grained material.

INTRODUCTION

i.The columbium C-I03 alloy (Cb-10Hf-iTi-0.7Zr) was selected as the ma-

i terial of construction for the Heat Source Heat Exchanger of the Mini-Brayton Isotope Power System (ref. I) in July, 1974. This alloy was sub-

sequently also selected for the turbine scroll of the Mini-Brayton Power

System. The basis for these selections was a comparative evaluation ofC-103 and Cb-lZr by the authors of this report. At that time, it was es-

timated, on the basis of eight creep tests, that C-103 would have a strength

- of 20 MPa (2900 psi) for a creep strain of 1 percent in i00,000 hours at1255K (1800F).

Since that time, additional creep data have been generated on C-I03.

A re-analysls with these more extensive data (16 creep tests for times up

to 4897 hours) initiated in September, 1975, indicated that the original

correlation was erroneous with regard to the activation energy and stress

dependence of the creep rate. This re-analysls further indicated that the

extrapolated strength for ] percent creep in 100,000 hours at ]255 K

(1800 F) is somewhat less than one-half the 20 MPa (2900 psi) value orlg-

inally estimated.

In view of the need for accurate long-tlme creep strength predictions

for C-103 for the dezign and safety margin of the Mlni-Brayton Power System,

thls re-analysls was extended to include detailed s_ress and ttmlperature

effects, grain size effects, and shapes of the creep curves. The results of

this analys_s are reported herein.

1976014230-TSA04

I

1 k'+ I..... [I _. II _. I II " | J I II iiiiiiii iHi, ,,_I - ' ..i. -1 " I l ........ , ,.

" _ I: P.PlT(_I)UOFt.IILI'I.'7 0_' ']'tll_}'tl,'h_l_',TAl, PA(II_I lfl PO()I_,

Th_ analysis is to.rmod interim u:In_:_, l.h._u churnctori;,,,t-lon of the lon_:-_im_ cr_ep behavior of C-,103 is continu:l.llt;, it 'l,n m)ti(:Jl)ated thnt the re-

- Stilts of the completed study W,t,,1], b_ r_pm"tod tit n later date./

EXPERIMENTAl, PI(O(IEI)I.IRi_.

The C-I03 alloy was commerlcally pro_.ur_,das 0.076-cm (0.030-1n.)

thick sheet. Principal constituents were deform:[ned us follows:

Hafnium 9.75 wt pet

' Titanium i.I] wt pet

Zirconium 0.45 wt pet

Tantalum 0.31 wt pet

Tungsten 0.25 wt pet

Oxygen 214 wt ppm

... Nitrogen 62 wt ppmCarbon 37 wt ppm

Hydrogen 0.8 wt ppm

Creep specimens having a 0.635-cm (0.250-in.) wide by 2.54-cm (l.O0-in.) long gage section were machined from the 0.076-cm (0.030-in.) sheet.

These specimens were degreased, rinsed in alcohol and distilled water,

wrapped in tantalum foil, and annealed in a vacuum of 10-8 tort at 1600 to

2023 K (2420 to 3200 F) prior to creep testing. Weight changes which were

observed during annealing generally amounted to only a few milligrams, equlv-

alent to compositional changes of a few tens of ppm.

The grain size after annealing for 1 hour at 1600 K (2420 F) averaged

30 micro-meters, while annealing for 1 to 5 hours at 1700 to 2023 K (2600 to

3200 F) gave an average grain size of 90 mlcro-meters. These two structures

i are referred to below as flne-gralned and coarse-gralned, respectively.

i Creep tests were conducted in internally loaded high-vacuum creep units

: described earlier (ref. 2). A tantalum split sleeve resistance heater was

employed for heating the specimens. The pressure _s generally 10-8 tort at• the start of a creep test, decreasing into the i0-_ tort range after sev-

eral hundred hours. Strains were measured by frequent telescopic readings of

: fiducial marks at the ends of the gage sections during creep.i

Test temperatures ranged from i]00 to 1366 K (1520 to 2000 F) and stres-l. ses from 20.7 to 276 MPa (3 to 40 ksl). Tests were generally terminated af-

_ tar 1 to 3 percent strain. The duration of tht_ longest completed test was4897 hours.

RESULTS AND DISCUSSION

Analysis of Time-to-l_l'erc_mt-_train Dots

The creep data on C-I03 curranti_ avall_ible from this study are sum-

• marized In Table i. Tests current tv in pr,,l',,'ssare als,) ]isted to inform

_, the interested reader of add11.tonal dat.;J wh.ltit will t_oon be available.bI

¢.

1976014230-'[$A05

I

I

i I ............IIII I t "-'-'' i II ]H i II I lIj_p "F' I!IL____ IJ_.LI !.._lm I I J " , , ,,," ¢

R_pr_:_ntative cY_ep curve_ are _how_ Jn ,?:t_,,ur(_ 1.q

The creep behavior i_ seen to be unumml ._n ,2ompariH(m to thslf:. _f ,:

_ure metaln and of many alloyr_ Jn _:lml: q_nly ,_le(:_.,.l.,,rat,hl_; t:reep lt_ obn(_rw_.d ifor ¢]-103, Since the anlay_$a of C-103 c teop wns eomp.licatod further' by t

slightly differing behavJ.or between f:l.ne..._;rnined _ln(l conrse-gralned me- _,_tcrlals, a straight-forward analys.h} o.1: the l':[mo__, l:u ]. percent creep strainwas preferred over analyses of the creel) rates for the inltlal correlation .,and extrapolation to longer t_mes. ]'

e

Dorn-Sherby Analysis

For this initial correlation, the times to i percent strain from

Table 1 were correlated by a Dorn-Sherby type of relationship as expressed ,Iby Eq. (I), Table 2 (refs. 3,4).

It is seen in Figure 2 that a linear relationship exists on a log-

log basis between the temperature-compensated tlme-to-l-percent-straln and

stress for stresses between 20.7 and 82.7 MPa (3 and 12 ksl); above this

stress, the curve deviates upwards. This type of behavior is consistent

with that for most other metals and alloys when the temperature-compensatedlinear creep rate is plotted against stress. The upward curvature of this

line above a given stress indicates usually a change from a power stress

dependence to an exponential stress dependence (ref. 5).

The apparent activation energy (Q) and stress dependence (n) were

determined simultaneously for both the fine-gralned and coarse-grained

C-103 by a least-squares fitting of the data at 20.7 to 82.7 MPa (3 to

12 ksi) to Eq. (i). These values and their approximate 90 percent con-fidence limits were "determined to be 315+49 K3/gmol (75,300+11,700 cal/

gruel) and 2.51+0.44, respectively, as given in Table 2. The constant K1is also given in Table 2 for both the flne-grained and coarse-gralned ma-terials.

Figure 2 shows that the creep rates for the flne-gralned material are

about one-thlrd faster than those for coarse-grained C-I03. This is equiv-

alent to a strength advantage for the coarse-grained material of about 11

percent over the flne-grained material.

The creep strength of C-I03 for selected times and temperatures ofinterest to the M_ni-Brayton Power Syst_:m were predicted us I_, Eq, (I)

and the appropriate constants from Table 2. 'l'l,t,,'_,pred:icted strength

values and their approximate 90 pt_rcent c,mi!id,,nce llmlts arc given :InTable 3. Figure 3 shows the predlct,,d ,;tren[',th[or ] percent creep In

100,000 houre as a function of temperature. Tltem, new predicted 12.55 K

(1800 F) strength values of 7.93 to 8.q(, MPa (1150 to ]300 psi) art,40 to

i 45 percent of the 20 MPa (2900 psi) pr(,d]cted from th,, earlier ana]y6it; of

C-IO3 creep data. The wide con[Idenee ]im_t'; (i,,,., ;1 spread ,3f m.,arly a

[actor of 2) reflect the very limited amom_t of l_)l_;-tLme,'reep data ,ur-

rently available for C-I03 and the con._(_qu(,nlhlrh ]eve] (,I uncertainty in

_, thet_e predicted values.

!:" ,• ),

9760 4230-TSA06

i#

4

Th_ tlmo-to-l-p_rcent-_traln data w_r¢_ a],_.Jana]yz_d by the Lar,Hon-, TMll, le.r mot:hod (ref. 6) Th:[_ correlate.on I, f_hown in Figure 4. Hero,

] [r_ exprt_ssod in _ in hours, and t:h. com_tant :ln g_van its urals.1 wdu_,of 20. The be_. _._-_vo through tho,_e data _,xt:rnpolnt_m l:o a ere, el) _trengthfor l percent strain in 100,000 hours at 125% K (].HOO F) of 21 MPa (3050 *_,psi), _Jignificantly higher than those predicted by the Dorn-Sherby method.

Comparison of Dorn-Sherby and Larson-Miller Analyses le

It is obvious _hat the Dorn-Sherby and Larson-Miller analyses give

widely differing results for the extrapolated long-time strength of C-I03.

We eonsldet that, a_ least for the present analysls, the Dorn-Sherby

approach provides a more accurate relationship between creep rate, stress, ._and temperature and thus is bettersulted for data correlation and extrap- ielation than the Larson-Miller method for the following reasons: j

:: i. The Dorn-Sherby type of relationship includes terms for activation

,.;. energy and stress dependence which can be correlated with theoretical ex-

if': presslons for the rate-determlnlng dislocation reactions (at least for lln-

ear creep). In contrast, the Larson-Miller relationship is strictly empir-ical.

2. The Dorn-Sherby relationship can simultaneously correlate the creep

; behavior of a wide variety of pure metals when dlffuslvlty and modulus eor-rections are included. The Larson-Miller relation, in contrast, must be

individually fitted to each data set.

3. The Dorn-Sherby relationship provides a much better correlation of

the present data at 20.7 MFa (3 ksl), near the region of greatest interest,than does the Larson-Miller relation.

4. The Dorn-Sherby approach allows distinction between the fine-

grained and eoarse-gralned material while the Larson-Miller does not.5. Strength values derived from the Dorn-Sherby analysls are more con-

servative than those derived from the Larson-Miller analysis.

Thus, we recommend that data extrapolated by the Dorn-Sherby relation,

Eq. (i), be employed for design of those portlons of the Mini-Brayton

Power System made of C-I03.

Analysis of Creep Curve Shapes

Correlation of Early and Late Tertiary Creep Rates

As mentioned earlier, analysis of the creep behavior of C-I03 is mort:

complicated than that for most pure metals by the absence of periods of

Identifiable primary (parabolic or cubic) or secondary (linear) creel,. The

creep rate instead accelerates wlth time at strains of ](,_s than I percent

in a manner normally referred to as tertlary cr,,el,.

Similar accelerating creep has been ol,,_,,rv,.,lI,revt_m::ly In hl_,h wnuum

|or many columbium and tantahlm all,,y,'_(re[s. ,',7-12). ADalysts of the

1976014230-TSA07

I

• - ' .................. _ d #

5

e.urws for Ta-10W (rof. ii) nho_.d that ,.rc,p_rr_i. we. proportional totlm,_ (3/2) for strain, up to nc_v,_ral tmro,mt. I

The data from thn preso.nt nt-u(ly wore ntml,jz,_d t::o d(,tr.rm_.n(; the time

dopo.nd,_tice of creep u_;[ng the, method ,_t' l[,kil,. dll'Jot:,,tm_,N _1_1 ,h,ncrlbod

by Cr.uf_ard (ref. 13), The, time: dependence (m) of C-q.03 cro.o.p wan dete.r _

mitred on six specimens to rankle from 2..10 t,:, '2.8:], an .tad'leafed J.n Table 1. '.iTht_ average value was 2.41, which we, refit|deal o['f to 5/2.

We define creep which proceeds pr,l)ortto_m] t:o the. 5/2 powo_r of time I

as "late tertiary creep" to distlngulsh :It from "early tertiary creep".

which proceeds according to the 3/2 power ,'rl! time. The ]ate tertiary creep

rate, expressed in d_mensions of strain (2/5) see(-l) is termed _, whlle

the early tertla_y creep rate, with d lmenslons of strain (2/3) see(-l), is

termed _.

The term tertiary creep here refers to the shape of the creep curve

and not to the imminence of failure. Although the mechanisms which cause

accelerating tertiary creep are not well understood at present, it is

known that grain boundary voids nucleate and grow during this period of

creep. Thus, early and late tertiary creep behavior may be associated with

reactions involving grain boundary voids.

Both early and late tertiary creep rates as we]]. as the strain intercepts

at zero time for early tertiary creep were determined for each creep curve and

are included in Table i. These rates were correlated by Dorn-Sherby type

relations, Eqs. (2) and (4) in Table 2. The activation energies and stress

dependencies determined for early tertiary and late tertiary creep were verysimilar to those determined earlier for the tlme-to-l-percent-straln cor-

relations. In order to improve incercorrelatlons among the early (;) and

late (#) tertiary creep rates and the tlme-to-l-percent-straln data, the

activation energy and stress dependence determined for the tlme-to-l-percent-

strain data were employed -Iso for correlating the early and ].ate creep

rates. This approach prec±.ded the dete_raninatJon of separate confidence llm-

its for the constants in Eq. (2) and (4). These correlations are shown in

Figures 5 and 6, respectively, and the derived constants are given in Table 2.

The early and late tertiary creep rates arc seen to correlate fairly well by

these relationships and to exhibit an upswing at stresses greater than 82.7

MPa (12 ksi) as also observed for the tlme-to-l-percent-straln data in Fig-

ure 2.

Reconstruction of Creep Curves

It iS desireable to reconstruct as well as pos:+_Ible the avt_ra_w crt,ep

curves for fine-grained and coarse-gralned C-f03 :[t, t_rtlt,rt,, illustrate th,,

effects of grain size on the shapes of tile cret:p curves and to allow extrnp-

,,let[on of the curves to creep strains v:rt,ater than I percent. In order t_

reconstruct the creep curves, the ,4train [ntortt, pt ,t zero timt, (which rep-

resents ,+train on loading and any tra,:t,,,:_ t_f I*rlmar'y and .,_ect,t_dary creep) isneeded as well as the "transition t. tmt:," fl-'Om ¢.,;1_ly it, |;,t._, tt,rtt_ry ere,q).

It l.t; also necessary to assure, _ that the :dt.lu,tl ,,[ tl., ,',,_,,p ClllYVoH are un-

,'hanged (except for the effects of )_raln sl::(,) ,-_ver th,, tim,,, temp,,ratur,',

.... " ......................................... " .. a@]

1976014230-TSA08

I

i

'li'_;_lTODl.I(lll:_II,ITYOP THR6 _',_.I,:_."!AI,i!ACI¢.,]S POOP

and _traA_ tan8_ of _nt_rant hero,,- Although thi1_annumption in not true: ov_r wid. rnng_ of conditions, .lt J_ no.c_n_mry lmr_, In view of tlm ]tmltod

available data and-_._l for the. purponns of thl_ study,

I The attain intercept at. zo_o time In obl:_llned by correlation of the_mdata for early tertiary creo_ from Table 1. Assuming that the c_cep rep-

retreated by thin strain ha8 the _am_ activatJon energy as that for the other /_creep =atcs fo_ C-103, 315+.48 KJ/gmol (75,300+11,700 cal/gmol), a st¢o_s do-pandence _f 1.$5 was obtained by l east-squar,_s analysis of th. early tot =ttary c_eep, zero-time st_,at_ intercept, dnta. Th:Ls stre_s depvndence t_ t_lm-

' tlar _o the value of 2 previously observed between stress and initial creepstrain (r_f. 14). This correlation is seen in Figure 7 to be fair at best,probably reflecting the scatter inherent in the extrapolation to zero timeof experimental measurements of very small creep strains.

The transl_ion time from early to late tertiary creep 18 needed sincethe two rates must be employed sequentially to reconstruct th[ creep curve.This transition time is defined as that time at which the instantaneousearly and late tertiary creep rates are equal. The early tertiary creep re-lation and its dl£f_X-6ntlalwith respect to tlme (the instantaneous creeprate) are:

e - eo = 3/2 (5)

de/dt = (312)8312tI/2 (6)

The equlvalen_ ;e_a_ions for late tertiary creep are:

e - eo = 5/2 (7)

de/dt = (5/2)y5/2t3/2 (8)

The transition time is derived from Eqs. (6) and (8) as:

_.3 '3/2

t = (9)

The transition times end strains calculated from the experimentally ob-served early and late tertiary creep rates are given in Table 1.

Creep curves calculated for both flne-gra_ned and coarse-gralned C-IS3using Eqs. (2), (3), (4), and (9) are compared _n I"[gurt, 8 to representativeexperimental data. Here, the experimental times are compensated for crvel,stress and creep temperature so that the d_fterences between the two curvesrepresent only the effects of graln s_ze. It itscorse/dental that the ex-perlmental points lie below the calculated curw,s for both tests shown her,.lscatter both above and below the calculated curves was observed for othertests. [t is n¢'ted that the curve for the Itne.--gra:ln,,d material aceelerat,'Hmore rapidly than that for the coar_le-.K_;_Im,,l,natt,rl;_I.This dlff'_rencert,-fleets the transition to late tertiary rreep ai a low_r strain In the fine-grained than in the coarse-grained matt,r[a|,;_:_seen from the calculated

z. transition strains in Table i.

1976014230-TSA09

7

Th_ ealeul_tc_ e_o_p .tren_ths for f_tralni_ul_ to 5 percent _t _m'ioun

_Imas nnd _omp_ra_B_s of _n_ront at. glvon in Tob],_ 4. For c,xample, r_,_, ]axln_ the nllowablo .tr_$r_ in coarno_.gralm,d C_]O_ at ]255 K (1800 F) lilld

]00,O00 hou_s _Zom I po_eost _o 5 pf_reont:Increases the M1owablo ntr,,lu_

i from 9,17 to 12:7 _a (1,33 Co 1,84 kM), an J.erpnnr, of 38 p_renl_t, For

f_no-Br_nod ma_o_tl, _ho allowable otrcnn 'Incr,,as_'_ 29 percent, Timtlharpor acc.oloratlon of creep in the fi_to_grnlned material, :Is rospos_Ibl,, 'it'for the lessee £ncroas_ in allowable ntrcrm re,lnt_ve _o tho.eoar.o-Bra:l,n,.i

material, Similarly, the allowable sLrc_el Incroat_(,_with higher td.lowal),_e

strains are less _or C-103 in general than they wou_d bc for materia1_ _.x-I'

hlbltin_11neatcreep because o_ the accel_,ratJ.ng,creep exhlblted by C-f03.

.... SUMMARY OF RESULTS

Ma_or results _rom this interim analysis of the l_ng-tlme creep be-

havior of the columblum C-I03 alloy arc summarized as follows:

i_ 1. The stresses for 1 percen_ cre@p in i00,000 hours at 1255 K (1800 F),, are calculated as 7.93 and 8.96 MPa (1150 and 1300 psi) for line-gralned and

I_ coarse-_ralned c-IO_, respectively. These strengths are both substantlally

l: lower than the 20 MPa (2900 psi) predicted from an earlier analysis on fewer

data. Confidence llm£ts on these new strength values are wide because ofthe few data currently _vailable.

2. At ii0@ to 1366 K (1520 to 2000 F) and stresses from 20.7 to 276 MPa

(3 to 40 ksl), C-103 exhlblt-s accelerating creep which can be differentiated

into one period where strain is proportional to time (3/2) and a second per-iod where strain is proportional to time (5/2). These two periods are

termed early and late tertiary creep, respectively. No periods of identifi-

able primary or secondary creep were observed.

3. The tlmes-to-l-percent-straln and early and late tertiary creep

rates currently appear best correlated with stress and temperature by Dorn-Sherby types of relations. The apparent activation energy and stress de-

pendence applicable to all three relations are 315+49 KJ/gmol (75,300+

11,700 c_i/gmol) and 2.51+--0.44,respectively.4. Analysis of creep curves indicates that the creep rate accelerates

more rapidly wi_]t-_ime fo_ flne-gralned than for coarse-gralned C-I03.

REFERENCES

i. Mini-Brayton Heat Source Assembly Design Study.

Volume i" Space Shuttle Mission. (GESP-7]O3-VoI-I, General Electric C:o.;NAS3-16810) NASA CR-121223, 1973.

Volume 2' Titan 3C Mission. (GESP-7]L)3-Vol-2, (;enural E]ectrlc Co.;

NA83-16810), NASA CR-121223, 1973.

2. Hall, R. W.; and Tltran, R. H.: Creep Properties .of Co]umblum A11oy_ in

Very High Vacuum. Refractory Metals and Alloys Ill' Applied Aspects.

Robert I. Jaffee, ed., Gordon and Breath Pub]. Inc., 1966, pp. HBS-gOt).

I. Sherby, O. D.: Factors Affecting th_ High Itunpurature Stren_,,th,f ]'oly-

crystalline Solids. Acts Met., vol. i(),,u,. 2, Feb. 1962, pp. 135-]47.

" " 1976014230-TSA10

B

,.L

4. _nrofnla, _r.nk: Fundam_n_nln of Cr_p .nd Cr,_p-m._r_uro in M_nlt_.

M_Millnn, 196S, pp, 46_I01, /

5. Narofalo, Franks Fu_dnmen_alu of C_o_p and Cro_p_Rupture _n Me_a]rl.MacMlllan, i_65, pp. 50-5_.

i

• 6. Lateen, F. R.; and M_11(_r,J,| TJm_=Tomp_rature Rolatlon_hlp for Rapier{, ,?and _reop 8Creeses. Am. floe.Mesh. En_r,. rrann., vol. 74, no. 5,July, 1952, pp. 765"775. .!

/7. Titran, R. H.; and Hail, R. W.: Ultrahlgh-Vacuum Creep Behavior of

CQlumblum and Tantalum Alloys at 2000 _nd 2200 F _or T_mes Greater thani000 Hours. Re_ractory Metals and Alloys IV$ Research and Development,Vol. 2, Robert I. Jsf£ee, at al, Gordon and Breach Publ. Inc., 19%7, pp.761-774.

8. Titran# Robart H.; and Hall, Robert W.: High-Temperature Creep Behavior: of a ColumblumAlloy, FS-85. NASA TN D-2885, 1965.

9. Titian, Robert H.: Creep of Tantalum T-222 Alloy in Ultrahigh Vacuum ifor Times up to 1O,000 Hours. NASA TN D-4605, 1968.

_ I0. Sh_ffle_, K. D.; Sa_ye_,.J.C.; and Stelgerwald, E. A.: Mechanical -]: Behavior of Tantalum-Base T-Ill Alloy at Elevated Temperature. Am. I

]! Soc. for Metals Trans., vol. 62, no. 3, Sept. 1969, pp. 749-758.

ii. T_tran, Rober£ H.; and Klopp, William D.! Long Time Creep Behavior of

iTantalum-lO Tungsten in High Vacuum. NASA TN D-6044, 1970.

12. Titran, Robert H.: Creep Behavior of Tantalum Alloy T-222 at 1365 to1700 K, NASA TN D-7673, 1974.

13. Crussard, C.: Thirty Yea_s of Dislocation Theory and of Rheolo_y - AStudy of Transient Creep. Am. Soc. for Metals Trans., vol. 57, no. 4,D_c. 1964_ pp. 778-803.

• 14. Oarafalo, Frank: Fundamentals of Creep and Creep-Rupture in Metals.MacMillan, 1965, pp. 46-47.

2.,

19760" 4230'TSA11

I_!.:I'I.:._ilLIIg[P,II-JTY OP TItI_I, ,_^_'_i' IX POOR

l ')=

1976014230-TSA12

i

#

TABLE 2, _ CRF:F.P ItF'.LATIONf;IITI'f; M'II) (:()rtwrMTrf; FOR (:-10t

K]Orl_ _Q/I_T (1)R

A. 1<2¢jII#CI_/ItT (;J)

f-t_(_) '_ Klcln_. -q/I_'T ('t)

I ";' " _401t'tI_(]/RT el) I

Ftno-ar, ain¢_ _otorial, 2,tll_IO;! 'i,'_Oat6,(m'_ ,_,o(, ?,alKlO _' _ _10!Caarno,,grat.m,d mnt_rlal 1,4!tx]O;' ti,()Ol!j:4,()]'l h,_)O _,O/X]() _) I ;_lt) '1

IIt IIe

FLno-Lcrol.ned i,at:t, rt.ltl 2,51+0,44 l ,85 .:(,oar._o_-Bratned material 2o_(_O,td_ 1,85 ,!

_. : time, to 1 poretmt I_tr/lll-|o l_th_olidtl . .B eOrly tart:tory ereup rate, tJtr_|lt=2/3lluc.olid =!

Co(P) : _t, rain Intercept at ;;m'(, t[tlle lor. I_ ¢,rot.p 1| "l' _a[e tl]'_t'ial'_ _ Crl'll_) tart _) lit 1"I|].11 '1t[ (il|t_t]llllti =1

u = stre_s, HP_In ., utvu_a depcndunLy

Q ", apparent actiwitlon et_ergy for croup

= 315+49 RJ/gmolR ,', gas constant _ 8.314 d/R_gmoi

: T = temperature, K

K " cr_:p constant, (HPa)'n(uec)'lnr(HPa) "n

TABLE 3. - PREDICTED CREEP STRENGTII8 FOR C-I03

(BA_ED ON BORN-SIIERB¥ ANALYSIB)

Temperaturet / Time 8trees for 1 percent strain inindicated time antl approximate

90 percent confidence limitsK (Y) HPa (kilt)

:. Fine-grained C-103

_+.;; 11115 (1550) 5 yr 49.0+13,5 (7.]1+1.96)_" 7 yr 42,97112.4 (b. 22+1,81)

tOO,OOO hr 35.3+I 1.1 (5,12_1,51)1228 (175n) 5 yr 14.44 4.1 (2.09+0._0)

7 yr 13.3-+ _._ (1.83+o,5_)100,000 hr 10.4'+ 51.4 (1. 51+0. 50)

1255 (1800) 5 yr il,{_ *t,P (1,6ff120,477 yr 9,7'_ 'l.O (1.40+t),4t)

t IOO,OOo,he 7.9¢ 2.(, (1.1'=_'_0,38)Coarse- rained (:-:1o}

1116 (1550) 5 yr 55.3+14.4 (H,02+'.:',O9)7 yr 48, l+l '1. '} (7,014i .9'1)

lO0_O00 hr 'VI.')! II.') (!i. Tff{1.7:') 11228 (1750) 5 yr lb.3+ 4.3 (,!.:_b$t).b2)

7 yr 14.2( 4.0 (','.t)(,il/o.',8)IOO,OO0 In" 1}./4 3.h (1.704o.'_',!)

1255 (1800) 5 yr 1.2,4+ 3.3 (l..St)-liO.411)

7 yr IU.Ht. '_,l (1,',7+11,4'._)lO0,O()O iir '#.()+ 2./ (I.'Hl+(l.'Jq)

,#

1976014230-TSA13

1976014230-TSA14

• ,°

1976014230-TSB01

J,;+

........ .-L.dI I ,.L r I I I

1976014230-TSB02

'.. • .... -,

1976014230-TSBO3

I

'r _ .... L I "" I lit I j*I I _ III II II fill III III I lllll II I I i p, _ ' I

,J .

1976014230-TSB04

LDATE

0_ _ ,.-- -" ...... -7-7"- .....

_ f , ' :i

I¢

O,/Rr_°=........ i..........................-_.....................:.7L........

i_ I ' /- I "I

i0_,-- J....i

....i

- ' i -.I -_ .................

/

. , ?_--o

- /_ I

O. ,

- i. O -',NE-G,_N_

I

i10 i aO 3i0 _10 50 60 70 _'OClOlO0 150 aO0 a,50.:_)0I

5"rI_ess,mPAl I i I J [

1978014230-TSB05

l

_. JL_ _ t@

i

i_Oio t:)ar)-:... , [...........

I

t.,/

. 'r i ' I

t0q .I '

' , i r

- /

id) ...... ,i _. ,,_'........... , ...................... j .........._........:._=_'

106 i /.... I .....

- ! (: _.

' : 'j : ' i ; i :io_ I I , ",., ...... ___.,_..).....,.J.._'. , : ,_±_).....Jo aO, _o5_o 50 _o 7,0_0q01oo 15o _oo aso_o,:)

!( , : ', ':: ',,, ' I

] 9760] 4230-TSB06

1976014230-TSB07

. + I,+ ""!I

if i

I

: 19 ' . ii

,2,_ _"i'i '

: ..... i " I I,i i. ',,, J

+: _k_ , r, + " ' " '-,r '

............ i ',+ +'i ....... I .... I: + +':+- : + e++ -' ,.,,.

:, , I t_l .

!...... .:.: ....w+++:l,l+,.,,l _ :i i I _ , +,+,:+:+,++,.' ;+:+ :: N+b,LI I+-:,!,I ++ i I i . ,: f2+I:.+: .+- ........,,++ +......... + I,';',+ ++++;'+:'".... ;; ' ""I_I','i,.+...'+ :+ i '++ '.!'I:":"++,++";.... ' ...t ...... I _"J I_L.'I+II ' , ' . : |+_-_,,.'.,, _ ',

. .+_ + ..

+ " . , + ;_ " i + I + • I ' + 't ++ ' +'

:-+_: (--"-:......" -_;I' __ 1 :. '"r'"".;I''+++ ..... f..... _ - , , ,4 , , - • ,

u:___ ...J...,.. .j , , t !",l\ I++1 ! ].... _i

!:t: .i _ ml t,.. .; :.,_+.+_ .... : I i : +'_I\: . i .r:" '

, ;.:+1i I ...... _._ i, I

. :_- ...J ...._-.._.. :.......... i . ;.

I _ I i i' '' _ '+l"J.+, i ,-+ _t,+

"g

1976014230-TSBO8