&UR-93-3396

“r”: r-–--------–=--- ‘- ‘--–‘--‘-The Effect of Processing Parameters on PlasmaSprayed Beryllium for Fusion Applications

Richard G. Castro, Paul W. Starwkand Loren A, Jacobson

MaterialScience and Technology DivisionLOE.AlamosNational Laboraloty, LOS Aiamos, NM 87545 USA

Author(s):

Donald F. CowgillPhysics of Matedsla Depsrtmont

Sandia National Laboratow, Liverrnore,CA S4551 USA

Subnmad

LosAlamosNATIONAL LABORATORY

to:

Lance L. Sneac’Metals and CerarnlceDlvieion

Oak Ridge National Laboratory, Oak Ridge, TN 37831 USA

MASTIIIWorkshop on Be~llium for Fusion Appllcatlons

Karlsruhe, Germany

AA

/

October 4-5, 1993

About This ReportThis official electronic version was created by scanning the best available paper or microfiche copy of the original report at a 300 dpi resolution. Original color illustrations appear as black and white images.

For additional information or comments, contact:

Library Without Walls Project Los Alamos National Laboratory Research LibraryLos Alamos, NM 87544 Phone: (505)667-4448 E-mail: lwwp@lanl.gov

The Effect of Processing Parameters on Plasma SprayedBeryllium for Fusion Applications

Richard G. Castro, Paul W. Stanek and Loren W. JacobsonMaLerial Science and Technology Division

Los Alamos NaLional Laboratory, Los Alamos, NM 8’7544 (USA)

Donald F. 03wgi11Physics of Materials Department

Sandia National Laboratory, Livermore, CA 945!11 (USA)

ABSTmCT

Plasma spraying is being investigat.ecl as a poLential ~’111[~IICj t C’clln. iqti(!tor applying thin (0.1-5mm) laver~ of beryllium [M) ~)l,isma l’iac~:lcj

surfaces of blankeL modules i]l I’JER and also as an illsitu ~’epai~”Lerhnique for repai~-ing eroded tx?ryllium surfaces in );,igllneat tlllxdiverter regions. High density spray deposits (:.98% (.>1 t.lleoreLi~”:i1d(?nsity) of beryllium will be ~equired in ordel [O maximii’.e [.11(’Lherma] cmnductivi t.y c)f“ Ll]t! be]-yllium rcmt ings, A }~]l?limi ]lal”yI.llvcstigal-ion was dorle to deLerminc tile el’!f’cL c)! v,iI”ItMls ~~roc~s::illqpa]amcters (}3a]”Licle size, pdrticlc morpholt>qy, :;(!r(.)ndcll”yg,l:;addit-.ions a]ld ~“educed chamber p],.easll]”e)011 Lhc ti~-dt?pll:;il.t?(id,lllsit.yt)llx’ryllium, The deposits were made ~lsing sphe~.ic:dl lmt’ylliurn lt!cdsl~l(.kpowdel which was pxoduced by cellt~”ifugal at.omizat.il)l~,~t,I,t):;AIAIm.J::National Laboratc)l.y (LANI,). lmp]:ovements il) t 11{’ ilS tl(!po:; iI(’(id~nsiti[~~ and deposit. cfficicdcies (“)f“ the b~~]ylli~ml :;p]dy d{+l~t~:.+it::will be disrussed illo]~~ with Itl{?~-{n]].t:s}]nllditlg1.11(’1 ’[11,1I (“(11 ](111(”1 i vi I ‘,7illld OULg~lllfJj’lCj k)mv i CJI 0! t Imse dep.):; i t.s .

1.0 Introduction

F.ltll’I tl.’(’)1;1I[1111’1’lliilillllill ‘I’llf ’I’lll(lllllr. ” 1I’i+t I’;xllt’liml’tll(II Nf,,l[’1,11

1

I-P’1~1-vlldm 111’filll Ilkll Il+ul,M@k’1 II ,>*1),,111-I II P” II

II-F l~lm ik-”Pk’ tl l-’ Idr-1

I-II,* r++-plllll+l 1. +1111 1111)1

qllll%llbimlll11.11 “MI 1=’ .I,’1lll’’. ’-l111111 Ill! l,F~M\-l ]#l. -P-h ll=”l’p’l

m.1 lA1’,1~ n/* 1“11’ltimll~’ 1’ 1,11-1,+1 ,“,rltlll~lll’1 ,, I 1P, Ilp.l, ,.ll m-l,.,

m

r

- transferred arc.heating/cleaning of the substrate whichimproves deposit densiLy and adhesion,

- broad spray pattern}; for coating large surface areas,

and Lhe ability to spray reactive materials under aprotective atmosphere,

A large number of processing variabl.e~; can ultimately affecl: Lk?quali,ty cjf Lhe berylli~lm spray deposits when spraying under a rt?duccdpressure. Operating parameters which have a direct influe’lr.wcm Lhcdwell time of the injected hnrylliurn powders will :Jignificanl.lyill!“ocIt:he degl-ee of mell:illg OL Lhc l>c~”yl.lium powdel fucd!:l.ock.lnvestiqat ion~ have shown t:lmt~]owder ~lize, powder morpi~olfqy, powderinjec~iun, particle velcx:ity, and the heat c!ont_enL(enLhalpy) nt Lllcplasma, c’i=III substantially rharlqe the melting behavior of th(? illj~c:tedpowders [10],

2.0 Experimental Details

1111:1 AI;.1111111: NtIlimml ]M’l~J(llI’11’)J”y ]1,1:1 I“ls(slsllt Iy ])fti~ll iliv,,l;l i,l,ll ill,! 1]11~I’t~lll l.i lIItl, I] ,It omi Zilt” 11111 ~)1’(Ir’I~lll; tl)l I)tx)rlil(’ilill 1111111’I i{’[ll 11111yl IIII(II,111(1lII,lyl I illtn :11l(ly ~Il)wdl*l”l:ill: Il.fl[llll(l(’p,1111.111~1iill 1,,1 ,I(lvclllf’!’fiI’1111::1)1i(lrlt illtl lIIINVqIIIIIBIl i .II, , 11111 It:():ll,lt if’ I)trl:l:lill,l (1111’),111(1111,1 ::111,1:Il}l”,lyillr]. ‘1’11,, I’I,l\tiij1111,11,Iltlllli;..;llillll 1,1,,(’,!,!::ill!:r,l~t.::“,~,11“1111111

illtl\lf’1itlll1111111illtl~11 iI lIIIIylIillm 111111.11 rIII. IIYlt I ill ,1 M{lII ,’111,’i],],, w]li ],,

,Iitlll’1 11111 t III! mfIll IsIl lIIIIyl I iilm 111111,11 I 11111111111 ,1 Xto:! t 1,1111:1111 I 11111. (1111II

I 111~ 1:111 1,11’1! 1)1 !1 tlllli(lly Il]lillttillllWIIIIIII wtiittI ii: (Iti’11’11III;,111 .Iilllllltlllf~, },’iflm ] , ‘[’]11, I itllli,l ]1111,~1 ]i~llll Il,,,f,ll i:l lu,~(”tl.llli,”.11 1’,’ .Il,,llli;l,(i

illtl) Iilll’ly[Ii’.fitl,.,lIll!qlll,!:: !11 ltIIIlIIItil,ll,,lyIII 1111~::~lltlllitlllWlIIII’1 .‘1’111’ [11111111’ 111 Fill’ l:{]li!llll,~tl Ill 111,1111 itlllllll.11111,Ilt]llli;”ill,l

1,,’,,,rII‘;:;;;:’’;’:,’,:,,:::’,1 f III:II:VIIII:IIIlIw III 11111ilml11.1::t11111.II::F, ,1,11 I i~ll: I II,.111,,,1111’1 illlf, .I l’)11’l,llps ::..l,,ll ,11..1 , ‘I’lIi:: lItIII’1,::::II II II II II:I,:! ::11111,1 i,’lll

lIIIlylIilml lIIIWIII.I willl ::,lli,llli,.ltIi III 1,111,:: ,111 llll. ,lIrlt. I ,,1 Ill” Ill”

,

degrees c/see . With Lhe current equipment. configuration, theberyllium powder is collected in a cani3Ler below the cyclone

separaLor, valved off, and transferred to an inert-gas dry box tor

powde I’ classification. Parameters which can influence the resulting

beryllium powder size distribution are given in Table 1 .

Table 1,

Parameters which can influence the size distribuli.on

of centrifugally atomized beryllium powders.

I Melt temperature 1

I Pour temperature ISuperheat

Disk speed

Nc=zle diameter

I A[mospheue I

A. 400 nl(?till (4 3f) pm)t!. ;! ’/u +400 nlf~t]ll ( IH pm 5.! pm)

(’ . ;! ()() 4 :J’1(~ 11W?L411 ( !) 3 pm ‘/‘1}1111)

r). I 4() I ;!01) mI.~[:}1 ( “/!J ~lm 1oh 11111)

I’(llli(”lt, t:i;!t’ f ),1(’I i(.)tll; A,l], l), Wl,l’(! 1:(11 (1(’1 {I(I 11)1’ I 11

lM*(’ll\lj:fI I l)t~\J [1111 ,Iilllt(l t 1111 l,ll’1 [(,1:1 ft,.lr”l i[)ll,ll yi(ll(lt;

l)(Jw(iIIi }; ,IIItl .Ilj:(, lr,])tf,:ll,t)tf}(l ,1 wi(if, t(il)r]iI 01 Il,llt if”lt~ t

L: illvl’1:1 i~jlll il)ll1)1 t Ill’ fli lmli ;’,(?(i

ixt’i;,

:il~trly (Ir’ll(,l!il :: (II IMII}’I 1 illltl wtIl~I 11111{111 I]l:ill,l I III, I,IW 111, I; I: II III I)l!llml,l:;111 #l\’ I’l I, IIIIt M. I ,11 I, AN I,, K’i,l. ;!, ‘1”11111(’111111111111I’1 1111,Iitl:: ,1 {’(,flml,ll{’i,~l :;!;111[) 1)1,11:111,1 llylll, 1111~.11 Wlli[’11 i:l lII(IllIIIIS~l IIVI*I ,1 I I, II II I], It 111{] (IiJ]IIIIII I’1111111~1

{’III)!:! l,!l!~,,, AI’( 1111,1111 1-111111111 !Il t Ill’ Iltt)t’1’f:l:itl,l 11,1::,11: 11~11111 it) I 1)1~

pl,llllll.1 ::1 11, I\’ ;II(I{’11::1: WI II: ,ll”l’1 1111111i::t](’[1 Il:lilltl .I 1111111 I 11111: I lIIW 1111111It)l:;) ’:11,,111, 141,1\’1 ] IIJIII lIIIwf{I.1 Will: I ,’,’!1 itllt, I Ill, Illrl::lll!l ::])1 ,1}’ 11111’11 11)’II:litlll ,1 wr.illlll 111:::1 ,O!,lltl,,l 1:’,’:;1 ,1111 wllit’tl illtt,ilt,llt~:: ,1 l’t!lllllll.t(’i,ll]1! )WIII, I ll!,.ll, ,1 Wll)l ,1 “1’,,] !,,:,, W,l!llll ;: (’,111, l,! II II,, l;: IIII1 !11]11 l,(!llt 1,11 I Ill.

1,,,’11 1111,~ ,11 I Ill! 111’t’;llilllll I!, w,llll Il,l,(!l:l, )(,p, filll III,! I 111. 1:11111)’

.

opcrat ion, Pl,lsma-spraying of beryllium using thi~done u;~der eiLher a reduced pressure environmentatmosphere .

To understand Lhe effect of processing parameters on

faciliLy can bc?

or in Can at-gon

the as-deposited

densities and deposit efficiency of plasma-sprayed beryllium,parameters were changed from an initial si.andard operating conditionwhich was ~sLablished using commercial S-65 beryllium powder 1’loomBrush Wellman II~c.,Table 2.

Table 2.Standazd Opc~”~~LingConditiol)

Spray torch Plasmadyne SG-1OOCurrent 700 ampsVoltaqe 30 V(-)ltf:

Primary arc gas (AL) 30 ::”lm

Powder carrier gas 2.!)f;lnl(AI’) —

Powder feed ].at.e .5 lk3/111’Spray dist.ancp 76,2 mm

substrate Lran:l]al.ic>ll 39 ipm

At.mcx!pher.w ilrqollA1]ode/Catilod~! 14’1/;?’Jl)

::~ll,ly (~9i}lt)l;itf+ (1I lxIIylli(ml WIIIII Inl IIlrS (III ( !. l’/’1 111111) Ilii,.k (’ll~qll,l:;11}1::11,1111}:wlli(’11wl~t”llI 1,1111 ;1!111’[1 IIlll”k ,111(1 11)1111 111111111 11111 ]ll!ll :111.1

:~])l-l’; 101( ’11, A 1111]I I:]h I] II~(4 (~l~ll(tl;i I

::1111::1 1,11 c~:l

Wl,:l ~),l)lill(’,,(1 II,, 11,1, ‘,qq)i~!

,111~’1 f:l]t.lyilltl 1111 ‘I Illillllll’l:m llll~,t,::ll t Ilil’klll,l:l:,,::I.111111111I1!1111“/111111III 1(1111111 ,11 I ]111 I Iii,’k{.::f l.fl[~if,ll III I Ill. ,1,,]1, ,::11 wit II fll’~llv:il1~’11{111111 1111 I Ill! 11111111!11 ‘1[)111111, IJ,I:II,,V.11 ,,1 1]1,! 111’tylliiml ,I,ll)f,::i.:i 1111111I 11,. “’’1’1’”1 !:lll )1111,1! 1’:: [1)1 1:111)::1, (1111~1)1 l“1llllll,’ 1111 l;:lll ir~ll W.1::Ilfllll’11”,’I,,llltlilvlIIll, ::11)1::1 I ,It 1, 11111 i I I 1111 llt~l)(~l:il ;: fill! ,1,,11,.11, A:: tllll,,,::ill’rl.Itlll::itit,ll1,1 11111lIISIylIillm I:l,li Iy [1,~11,.::il:: WIIII, 111,II I::lit,,,l Ill:ilifl ,1 w.111,1

imm,l ::i,lllt,,,”tltli~lllt~(Att’llim’111’1:IIIilll’l]llr’), MI,, I:: II II 1111111II I; W,. 1,. 111,1,1,. IIIt Ill, I Ilif’ht.l:t 1,.,[illll III 1111. ::]It,i\’ ,I,l,,,l:ill: ,Ilt I,l I’1111 itt,llf~t ill,lill,l l’,I 1.111!1!:, ! IIPI IUIIISt II illlll”::III1111’111111::IIOI]IISII::lIIiIy IIttl,,,::il::,

1,

,

, ,

Deposit” t?fiicit!nci.cs (Lht? iI“arl.iol)of nla Lc?rial depos iLed On Lk?Subsl.rate) were detennj ned hy rncasuring the weigh L of the substratehef.ore and after depositing bel-yllium. ‘rhe differenc’c in weighl. wascompared 1:0 Lhe tota1 amount: 01” hc~yl 1ium powder dispensed during theplasma spray run . Since Lhe lx:l”y].1ium powdel.-feed-rate is controlledby a weighl /1.es::sy~t.em whirh places Lhe pow(ie]-feed hopper cm aweight scale, an accurate amounl. of powdcl p?-csent.in Che I]opperbefore and aftel the spray operation could be dct~rmined.

2.3- Llepziil ..CIMxac.lxxtiaL im

:!,3.1 Mi c.1”0::1I“u(’tural and L’hi+mic-~llAIIal ysi :;

lJIIIyl 1ilml :i, ImlIl[I Ii [,mnl ill(li,lmc~lttt IIy !lIIIIII 1lUl\l . M(’,l~:lllv*ml~llt~; WIII”IIIIItI(it*.11 (),lk ]/ i (j(~(! N,II i[)lhll II, IIM”)I”.11 111 }’ Il}:ill,l ,1 111111111,1I ]Iulx(I Ir’t’lllliqlli’ill

wlli(”llI~tKSI.III(s !lt t III” fIINI(IlmISII W, II; i Illmlill,~tIv+ wi I11 ,1 X(’II(UIII.11:11IiIllIII

.111(1 I Ills 111111’1 WlII: mllllil l)IIIIi wil II ,111 [1{ lll.(ll(’llll, ‘1’1]1 I lmqhll”, ll 111”11 ,11; ,1

11111,’1 iljll 1)1 I iml~ if. lll(llll~li ,111(1 I Ill’ I tlt’1 ’111,11 (tilllll:ivity i:; ~l,lll’IIl,ItIItl

I1,1111I111~ 11’1:1111 !Illl 11%1)111 illll,lll ,11 ])1 111.I I,t,mil]l.(1 1 III, IIIIL II I Imllll:il*lll, ‘1’11,1

1111’1111111IIillll:lii’il}’11~~W.I::I’.IIIm111,IlII~l{Illill,lI IIIS 11)1 1,%’111,1 ,~x]ll,,};l:it,ll

WIIIOI,I x i:: III!, 1:1)1’t’imttll llli[’kll,~::f: ,11111 I , il: ltIII I im~’ ttillllilr’(i 1111 I Ill’1101(’P, 1,11’11 III I IIIS ::jh,i’itlvll III 111,11111 11,111 “ II :: 111.lxlllllllll I I’IIIIN” II II 1111~, ‘1’111*

I 111,1111,11 I’111111111’I ivil ;’ (h) fill I !11, lIIIt\”lIillm 1:]! 1!1)’ ,I,ll),,l:il I: W. IllI’. II I’IIIIIII. (I Il::ill,l II ICI Illll(,will,l,Oxlll,,::l:iilll

h [\(’l,/~

1,

,

,

where p is the density and CI)is the heat capacity .

2 .3,3 Vacuum Outgassinq

Thermal resorption spectra were obtained on several samples of plasma-sprayed beryllium at Salldia National Laboratory, Livclmore,California, to determine their vacuum outgassing charac’Leristics. Theoutgafising system consisted of a Lurbo-pumped, ultra-high vacuumquartz tube furnace with a UTI 100°C residual gas analyzer(RGA). ‘l’hemeasurements were made on small block-shaped samples (100mg, 5x5x2 mm)by first loading them into an unheated portion of Lhe furnace und pre-outgassing the tube to BOO°C to remove the contaminants resulting fromair exposure. The samples were then remotely Luansferred inLo Lheheated zone for Lhermal resorption. Several samples were run at thesame time to gain sufficient signal for accurate measurements. Thesample temperature was ramped liaearly from loom temperature Lo 600°Cdt. 10 OC/mill using a temperature p~ogrammer. For one experimental run(#2)# Lhe temperature ramp was paused at 3500C for 90 minutes toexamine the effect of vacuum bak.illgat 350°C’. RGA signal amplitudeswe1-1‘ converted to gas partial pressures by calihral.ing against. carbcnme)’~xide and lly’dl-ogenstandard leaks,

3.0 Results and Discussion

‘t

,

The highest deposit densities were achieved using the berylliumatomized powders which were below 38~m. The spherical morphology ofthe atomized powders, Fig. 3a. allowed for better feeding of powdersinto the spray torch than commercially impact ground beryllium powderswhich are more angular and difficult to feed below 45pm Fig, 3b.Beryllium atomized FGwders in the size range of S3-.38pm also showedrelatively good deposit densities (91-92% T.D.) when spraying with thehelium gas addition at a reduced pressure of 35o torr. The largerberyllium feedstock powders (106-4511m) were difficult to melt usingthe standard operating conditions and required an increase in theoperating current from 700 amps to 800 amps in order to melt theberyllium and adhere to the copper substrates. The deposit densiLies(c60%) were considerably lower and did not significantly change withthe introduction of the helium secondary gas and the low pressurespray environment . No further analysis was done on these spraydeposits which were made from large diameter powders.

A graph of the as-deported density, deposit efficiency and the levelof porosity of plasma-sprayed -3Ei}lmberyllium powder under the variousprocessing conditions is given in Fig. 4. The highesL deposit density(94,9%) was achieved using the standard operating parameters given in,‘rable 2, with the addition of 15-standard liters pe]: minute (SLP4) ofhelium as a secondary plasma gas while operating under a reducedpressure (350 torr). ‘rhis operating condition also resulted in thehighest deposit efficiency (approximately 60%) with a porosity levelon the order of 4 percent. These observed increases welw attributed toLhe higher heat. conr.ent (enthalpy) of the plasma jet wi~h the addiLionof the helium ~econdal”y gas, and the higher pal.tirluf;velocities that.result when spraying under a reduced pressure, increased particlevelocities imprc]ve the impacting al~d splatLing of tile melted atldpartially melted beryllium LeedsLock powders I-esulting i11 heL 1.P t-bonding and mmmlidat.im of LIN &?pr]GiL,

Analysi: of the -4OO mesh (38pm) beryllium atomized powders, berylliumspray deposits and Lhe over-sprayed beryllium powder (which wascollected on the bottom of Lhe spray chamber) was periomed todetermine Lhe oxygen level and other impurity elements. The oxygenlevels of these samples were compared to commercial SP-65 berylliumpowder produced by Brush Wellman, Inc., and beryllium plasma-sprayeddeposits produced by Battelle, Columbus, Fig. 6.

The oxygen level of Lhe plasma sprayed beryllium deposiLs (.3S%) wasapproximately half that of the starting atomized beryllium powder(.65%) and much lower t-hen the over-sprayed beryllium powder (1.15%) .In addition, the oxygen content of the spray deposits was lower uhaLwhat was previously reported in earlier investigations done by theUCSL and AWRE [III. Oxygen levels in the beryllium spray deposits inthese investigations were reported to have increased fromapproximately 20 to 140 percent over the starting beryllium fecdstockpowder.

Metallic powders which were deposited using low pressul”e plasmaspraying have shown oxygen levels in the spray deposits at least ashigh as the levels presenL in the starting feedstock powders. Whenplasma-spraying copper using hydrogen as a secondary plasma gas,oxygen levels in the spray deposits were shown to have decreased belowthe starting powders [71. 111this investigation, helium was used asthe secondary plasma ~as and uhould not have affected Lhe oxygml luvelin Lhe spray deposits.

The lower oxygen level in Lhe beryllium spray deposits is 11oL well~mder~tood but may be attributed to the plasma/particle? interactinllsthat occur when spraying beryllium powders which contain a sur’” qlayel c>i 1300, During melting, LIICpowders may tend to segregate illbet-y]litun and heryll ium uxick particles due to their differvnves inmelLil](~1>[.)i]lLF+and dell~itivs (BeO-3 .03 q/cw13, M.P. -2823K ;1!1[113P1.H’>g/cm3’ M.P..156OK). This segregation may cause dif!:eruxlt pdrtirlvIr~ljc~.’Lol.ii.:tiof the 13P and llc![)ptlrtirlr,s A:: t:t]cy exit. thf.!~Jltl::maI:ol”c.’ll.m’ 1]11(7 pa~t:i(=leti may solidity in f li(llll. before inlp,lc.t.ill~j I IN-:;lli)l; t- 1“,.lt.r~ ,111(1 I;ulx:I Ic]\IcIIILl\ (11!1 t’c!l r“)f c 1.11(1 :;lli): it. l..ilt 1? , A(tditi(ul:llly,

till?111’1)]1.111 il.’llJ#; nhly IN’ :; iq]]ilic:alltlytimclll(!l.t.llloIl tll,,I I+4s~Mll i(’1~’::

:.\lld111.!(’(IIIIIS ,?llo,lill,,cj i 11 1.111? prc)cc?~luirlg qaI; (2G wllicll IIlft (lIIf JI.!Iml I!(l

,1l. Y)\lIl(i 1 11(, I;[lh:t.l”cll t’. Tllt”?sj.” Iwo po}; };iili 1 iLi(~f: may d(:(:oul)t 101 t.111’]Ii(l]l,,] (lxy. Jl~ll II! VIJ]S ~)l”l,l;?~lll itl l.llt’ ove I Hpl”, iy!!d ]M”)w(i:-’l ”!; , Fill” !.tlc’l

illvt’t;l i(~lll i(lll:+ ,.11 (.? i)~fill(j dIItII. t !1 1111( 1(!1.1,(:111(1 Llll?}; l! l’(’!; 1111 :; .

I.:llJ\r,ll,,(l ],,v,~l:; ,,1 }1’1~, Ni, I’) WI!),) ,11[;() Ill, tl, f’ll?(i ill I 111* ,il(mli::,l[l11~”’lyllillll: ]JIIWIII II::, ,Itl(l 11111 ::IIIIt:II(lIII.*llt I;~)I, IyII(l ,jt I] M,i Iii:: WtI,SII f’fmI])lllIIIlIll l“”lmlll~’1”, .1,1 I illlllill’l lllll\lllfl ::11 1,1, 1)1.l”yll 111111 II,,lhll.1 ill, llllll:lt”lllll *(l I)y

‘1

.

Brush Wellman, Inc. These elevated levels are a resulL of the erosionthat occurs in the i.nner walls of the stainless steel cycloneseparator during the beryllium powder production. Efforts areunderway to coaL. the inside surfaces of the cyclone separator withplasma sprayed beryllium in order to mininize the contamination of theberyllium powders through impact with the stainless steel walls.

The observed quantitic J of outgassed species per gram of samplematerial are given in Fig. 7. The major gas species are water vapor(18 amu) and hydrogen (2 amu). Measurable quantities of meLhane (16amu) , carbon monc de or nitrogen (2EIamu), and ammonia (1-7 amuj arealso observed. Here, the NH3 and CH4 values are tile residualamplitudes for 17 and 16 amu obtained after subtracting off thefragmentation con~ributions from H20 and NH3, respectively. Neitherthe interruption of the temperature ramp performed in exp[?~-imen~ (#2)nor the size difference of the two sample sets affec’ted the totalquantities released per gram The hydrogen released is roughly thesame for the twil experimental xuns; however, more was released in theforms of NH3 :Itd CH4 in run #1. Typical qualities ouLgass,’d from 85%and 95% dense, S-65 beryllium samples from Brush Wellman, Inc., arealso given for comparison.

Much of the outgassing behavior of the plasma-sprayed samples isdominated by the presence of the large H20 signal. It is l~ot knownwhether this signal is typical of plasma sprayed material (resultingfrom post-processing exposure to air) or is an artifact of L hespecific sample pretreatment. Prior to outgassinq, poL”oHiLy anddensity measurements were done on these samples by immerfiinq each inwater. l’he H20 probably resides in t-he oxide present. on the surfaceC)f the spray deposits. BeO is know;lto be very hyg~.o:;copic Andaccommodates several waters of hydration, f~>rminq Be0.xH20.

m .

concluded that this low temperature peak is N2, whereas the hightemperature 28 amu peak is CO. Apparently, it is the presence ofweakly bound N2 which gives rise to the formation of NH3. Thisnitrogen may result from post-process absoption from air or may be dueto nitrogen present in the initial powder feedstock. It also may bepresent in the processing gases as was described in reference [11].If present in the powder, nitrogen may be removed by outgassing theberyllium powder feedstock prior to spraying. Other weakly-boundcontaminants, including water, shoul,d also be removable by a pre-outgassing step.reduce much ofinvestigating airrapidly return tomust be prevented

Water outgassing

As mentioned above, removing the water ‘may alsothe detected hydrogen. However, experiments

exposure of outgassed S-65 have shown both N2 and H2near their pre-outgassed levels. Thus air exposurefollowing such a pre-outgassng.

from the PS material continued throughout each

experiment, but varied somewhat with sample temperature, T. Evenafter baking at 600 ‘C for 90 minutes, substantial H20 outgassingremained. Figure 11, gives the H20 outgassing as a function of timefor the PS and 85% dense S-65 materials, Each da~.a set can be fit wiLhan exp(-tl/2) function indicating diffusion-controlled release. Thus ,

although the water is weakly bound, probably in the oxide on the

surface of each deposit, it appears to follow a very tortuolls path torelease. Outgassing from the PS material is substantially slower andwill require a much longer pumpdown time.

3ti 1 r(~~ d-u-u

Results of the room temperature thermal conductivity measuremenl,s aregiven in ‘rable 4.

Table 4.Room temperature thermal conductivity of plasma-sprayed beryllium

Sample Mean Thermal Specifir Densitr

The rma1Diffusivlt.y Heat (q/cm3 (~rJIl[~ll~t:jvi [.y

(cmZ/a) (J/kqiK) (W/m-K)m

A 0,22(;9 1750 1 .750U (,CJ./ll)—14 0,2145 1“?50 1 .7141 r)4#”!4

c! 0.158(; 171io 1.6702 4fi63f1

1) 0.1442 1750 1.6653 42.0:!A

11 0.156’) 1“/50 1 ,66(;B 4!?,’/1AF o.124f) 1“/50 1 .(199’/ 1’/ .0(!/ >

11

.

“.

The thermal conductivity of the beryllium spray deposits weresignificantly lower than that of pure beryllium which has a thermalconductivity of 21B W/m”K [12]. These results were in generalagreement with previous thermal conductivity data taken for plasmasprayed beryllium produce by inert gas plasma-sprayinq at Battelle,Columbus [8]. An increase in the thermal conductivity alsocorresponded to an increase in the deposit density, except for sampleF. The low thermal conductivity values in all cases can be attributedto the porosity and layered microstructure throughout the bulk of thedeposit, Fig.5. The presence nf interfaces created by impactingberyllium liquid particles will be a controlling factor in maximizingthe thermal conductivity of the spray deposits. Improvements in thethermal conductivity of plasma sprayed beryllium can result byminimizing the splat interfaces through better melting and deposition,and also through post heat-treatments. Spray deposits of berylliumwhich were produced by Battelle, Columbus were heat-treated at 9000Cfor 1 hour to promote diffusion bonding across splat interfaces. ~increase in the thermal conductivity from 25-200% over the as-sprayedberyllium resulted [8]. Since heat-treatments or consolidation by l~otisostatic pressing will not be applicable for ITER, post-depositionsurface conditioning techniques such as laser surface treatmentsshould be investigated.

4.0 Conclmuoions

- The spherical nature of the beryllium centrifugal atomized powders,in comparison to the angular impact ground powders, allowed forbetter feeding of powders below 38pm into the plasma-spray torch.

- 3B~Lmspherical beryllium powder, made by centrifugal atomization,produced, the highest deposit densities under the inve~t.igatedconditions,

- Increases in the deposit density and deposit efficicmccy o! l)l~~milsprayed berylllum renulted when spraying under a low prPHfiuI:r?condition (350 tori-)with helium ati iJ sccondil~”y pldmd qA~J,

. oxyg(+]l ]c\’f!~tJ III t.1.r+ b~ryl]lum f]pray dl?pm3it-fJ pl”(uiut”~’d by low

pl”efifilll”t’ plmlm,i l~pl”ayillg Wt?l.e lowel” (by a fil(”t l’” I 1! I WI’)) I Ilrlll 1 Ill,Ht n]”l. i II(1 ,11.r)mi :td hr?l-yll lUIII pcrwri(~rfx,

TIIII I Ilr)l”nlill (’011(111(’1 ivi t y III ])lttllllhl Il])t’tlyr’d 1)111yl 1 ilm WlI::[Iiqlll i.i(’.tlll. ]y lf)Wr*l. I 111111 ~)111 rl 1)111’y] ] illlll, “:11( ’l!)llt ]’111’t 111,11 11,,11 lllc~::}111[”11 ,1,1 II 1)1 iit. illt.(’]!lll’{’tl mlIy Iv! ,1 rIWIt 1111 I ilhl I ,IIIt IJI ,

] ;!

,

Acknowladgmenta

This investigation was funded through Sandia National Laboratory,F’usiGn Technology Department under the guidance of Dr. Robert D.Watson. The authors would like to acknowledge Oak Ridge NationalLaboratory for the thermal conductivity measurements, K. Elliott. forthe beryllium plasma-spraying and J. MonLoya for sample preparationand metallography.

1.

2.

3. .

4.

5.

cl.

‘1.

H.

g.

Refarencsa

J.B. Whitley, K.IJ. Wilson and D.A. Buchenauer, ~. of NUC.

Mat., 155-157 (1988) 5!2.

R.F. Mattas, D.L. Smith, C.H. Wu, l’. Kuroda and G. ShaLalov,

J. of Nut. Mat., 191-1.94 (1992) 139.

K.L, Wilson, R.A. Causey, W.L. HSU, 13.IZ. Mills, M.Fand J.B. Whitley, J. of Vat?-Sci Tech A 8(3) (1990)

K.J. Diets and Lhe JeL Team, Fusiol] TecJmology, 19

D. WebsLer and G.J. London. , Ed. llBeqlliUm Science

Smith

1“/50.

19’)1) 2031.

and

Technology”, Vol 1., D.R. Floyd and J.N. Lowe, Ed. Vol 2.,(I>lenum Press, New York, 19’/9)

M.F. Smith, R.El. WdLfiDil, R.-l’.McGl”aLh, C.Il. C1-OeSslllall~llld

J.B, WhiLley, J. of Nut. MaC.., l“JI(1990) 158.

M.F. Smith, C.I). C!r’oessman, F.M. Hcwkinq, 1/.1),Wil[solldll(i

J.A. Koski, 211d Pldsmd-Twh]]ik Symposium, Lucel”]lt’,

SwiLzel’land,Vol .1. (1991) 43.

1~1’ival.ecommu]lica[,iol]with l)r. R,I), Wal.uoIIuI” Siilldi:lNilt.if)llal

I,abol-story, F’usioll‘J’c?cluloloqyDivi:;ic?l],Ali>uqu(~~’c{llt?,N.M.IJSA,

l;:, Muchlberqer, P1.or, ‘/tllIIIL. ‘1’llt’f”mcll Spl”dy (.lx]f , l’1”( )(’. 1,011(11111,

I ‘1

Sfopperod

Crucibl

Moltenalloy

s

Nozzl exit

lets

DAirturbine

“.

Fill, .:, !!Iql~l I 1111111, i%’~lll. I: I;l II II 111 ,11,111,1 :;l II, IJ; 1,1111 I ii y ,11 I,ANI,,

I 1,

,,, , !,,i’ ;!

,, ,! ’:, ‘“ :.:’

%

D9mlty Dspo811EfflclOncy % Poro81ty

mBatldlo M81h Q Stsndwd Opmthq a Hollum ws adctltlonCoWtions

] }!

,

.

1.2

1

0,8

w%Oxygen 0’6

0,4

0.2

0

,

SP46 Battolle Atomlzod LANL Ovm-cprayoddeposits powd,m doposltn powchm

Fig. 6. (“tmI]IiII i 111111 1)1 ,IxytlIIll I{IVI’1:: ~)! [I(llmo’t(’ir.ll :;1’ II’, !Jlltyllilllll] )1IWl 1!, 1 t 111111 I!lll:lll ‘W(JI 1111!111 1111’, , 1) 1,1:1111,1 ::~Jl,lyt’[1 11111}’I I illlll fl,~~j(jllit IIIJII,IIIII,,.(1 ,It lliIl I t’1 III, t’IIIIIIIIl m::, 111111 t ilibllll ,IIIIIIIl;:IIIi lJIIIyl I illltl )JI~w{lI,Ifl

1,1(, (111( ’!1(1 ,11 [, AN I,, ], II Iy]l illlll 1:1II, Iy ,!IIII(J!IIII: ~ll~,IIIl?”JItl .It IIANI, ,111~1 I 1111I)vl,l ll~llclyl. ll })l)wlbll l:,

;!()

h

.

40

30

20

10

0

■ C0,N2

■ H20

● NH3

❑ CH4

❑ H2

Ll la-95%S65 85% S65 93% Ps

.

50

40

30

20

10

0

~ N2,C0

o 100 200 300 400 500 600 700

Temperature (“C)

Fig, H, (;,1}1 (l!*llot I! L: ,311 :l~)ll(”t 1,1 I l’[ )111‘l!? [1111]:;11 }1 I il:lllltl :; I) I’(IYI, (I I) f*l’yl I ~1111):Irlllll)l! l!: ,

.). ),1. .

●

✎ ✎

la

a

6

4

2

0

10

8

6

4

2

~ PSI 93%

~ S65, es%

~ S65, 95%

I

Y/NH3

o 100 200 300 400 500 600 700

mo0 100 200 300 400 500 600 700

. 1

10

8

6

4

2

00 100 200 300 400 500 600 700

Temperature(°C)

1,’i(~. lo, (’tmll],lt-i:;t>ll ~~f ~iII:IIJt’}11 if~tl ::]11’( Itll 1~11 (1’:;) 111,11111 illl ]1111 iJ,Ik,I{l

!11 11,()”(’ 1(11 ‘to milllil 1111 wil II IIllilllkl’ll 111111(ll’illl , I’ll’ IJ,lkilltl II IIUIIVII(I 1111’

will 1,1 lMl, Ik I,(II II,Jt 11111 WIII III (III, (],ltl:lill(l,

.

10:

1’

Water Releaseat 600°C

93°/0dense PS:exp ~ @2.8 hr)]

85% dense S65:

.1

10 20 30 40 50 60 70

time”2(8”2)

rig, 11. wilt, r 1’ l“l?J,f7iirlr2rat. f? ill f)o(l(+’.

;! 1,

.

The Effect of Processing Parameters m PlasmaSprayed Berylhium for Fusion Appkica$kms

Richard G. Castro, Paul W. Stanekand Loren A, Jacobson

Material Science and Technology DivisionLos Alamos National Laboratory, Los Alamos, NM 87545 USA

Donald F. CowgillPhysics of Materials Department

Sandla National Laboratory, Llvermore, CA 94551 USA

Lance L. SneadMetale and Ceramics Dlv1810n

Oak Ridge National Laboratoq, Oak Ridge, TN 37831 USA

Workshop on Beryllium for Fusbn AppiicatlonsKartsruhe, Germany - 0

Otto’be’f4-5, 1993

Facilities

Applications

Requirements

Preliminary Results

Future Activities

Centrifugal Atomizer Schematic

stopperrod —

Crucible-

Molten-_alloy

Ill Imduction0+ -- -.0 cdlo

11 45 chamber

Qassn5#:

Transfer

\\i I

I.,/ \

quench gas..d”~”a”

Nozzle ~\ Powder exi’

Alloy Atomized dropictsstream

Atomizer whoei

I[1

I

L-1Air

turbine

Beryllium Powder Morphohgy

LANL: ry~ XSRcentrifugal dtomiz~d

.. .

--_.--

+

i,

z

XSR Powder Size Distributions60

S(I

40

30

20

10

0

IIRun 9Run 10Run 11Run 12

177

Size (yin)

--=--=

/----

%+-

i- -==5sia

I Plasma Spraying of Beryllium for ITER I

In-situ repair of sputter eroded and disruptiondamaged beryllium armour tiles in high heatflux regions.

Fabrication of large area (1000 my berylliumcoatings (1-2mm) over stainless steel orvanadium first wallsurfaces.

● Requirements:

high density

high deposit efficiency

good therms/ conducfivlfy

good bond strength between coatings andsubstrate materla/s (Be, S.S etc.)

enhanced mechanical bahavlor urjder pulsefusion conditions

others ......

/-”/_/

+’

.>‘/-’2

M!l.um/’-”-

/“‘---*.

,/”

/

‘-

Monobkxk ArmoredCirwhr Tube Rectangular Ttdiw

dBUM

Mcl)onnelf lkqykw Aerospace

.,

*

o

r ———. .—— —— ..— _—. —

Optimlzlng Spray Depiiits of 13wylllurn> \

● Ifwwse f)wbide metting:

pwtfcie Iwphdogy

particle size distribution

whstmte ternpenxtufe

puttic}edwdlthw:

lowvokxxty kwnharflow

h@h IYa?m-U(9

——— ..—.——..-. ————— .-. .-—.— 1

Layered Assembly of Impacting Discs

(&

-..

Wul

W, .,

!,, .

‘-r’ ‘“’TM”

m “ ““’““’k “’it’” p “ “!”IH. Herman, 1988

#

Operating Parameters for Plasma SprayingBerylllum

Primarygas I (Ar) - slpm 30.0Powder gas I (Ar) - slpm I 2.5

Powder feed rate I grams/min I 3.8Spray distance I cm 7.6Translation speed I cmlmln 99.0Chamber pressure I torr I 500.0Oxygen level ] ppm ! 100.0,Current I 700.0Leak-up rate ! mHltorr/mln I 5.0Substrate (Cu) I mm 3.2No, of passes I 140.0Spray time I min 5.0Total powder sprayed I grams 9.0

Deposit Density and Deposit Efficiency of Plasma

Sprayed Beryllium Under Various Conditions-400mesh centrifugal atomized powders

.. .

., .

.. .

,,,

,.,

,,.

.

,,, ,

,.

,., ,

,.

m Dc)posit Density ❑ Deposit Etticloncy % Porosity

~ A Comparison of Oxygen Levels in BerylliumSpray Deposits and Beryllium Powders I

1.2

1

0.8

‘t”’o 0.6Oxyoorl

0+4

0,2

01

0,6

IISP=66

0.84

0,65

13attollodoposlts

Atomlzodpowdors

0.35

mLANL

doposlts

1.15

Over,

sprayedpowdors

.. —-—— .—..—-. -.-—-——-—.————--.,——— . . .. . . . ..——--- . ..—. .—_.. ,..-—... . . -..—.——.-_

Outgassing Behavior of Plasma Sprayed and S-65Beryllium

40

30

Z()

10

0 lE%El--9950/0 S65

il. I/ T“-850/o S65 930/0 Ps

—. . ...... . —.- -.—___ .—-—. .,.. ..- ——-.. . .. . . . . ........ ... .. . .

Water Release Rate at 600 ‘C for PlasmaSprayed and S-65 Beryllium

10

1

1

-— ——. - .- -.. -.—

Water Ftolonso~t 600”C

I Plasma Spraying of Beryllium for ITER

● Critical Research Areas:

- Optimize plasma spray parameters to producehigh densitylhigh thermal conductivity deposits

of beryllium.

- Optimize centrifugal atomization process to

produce high yieids of low oxide, -400 mesh

spherical beryl/ium powder.

- Investigate sudace preparation techniques on

the bond strength of piasma-spr8yed beryilium

- Evaluate the performance of plasma-sprayed

beryllium coatings under pulse fusion conditions.

- Fabricate a roboticaily controlled plasma-

spray test cell to evaluate remote manipulation

and in-situ repair.