ASGSR/ISPS–Orlando,2013
EffectofConvectiononPrimaryDendrites:
ObservationsfromGroundbasedandSpace
StationProcessedSamples.
Masoud Ghods – ClevelandStateUniversity
MarkLauer– TheUniversityofArizona
Surendra N. Tewari – Cleveland State University
Surendra N.Tewari
ClevelandStateUniversity
RichardN.Grugel– MarshallSpaceFlightCenter
RobertE.Erdman– TheUniversityofArizona
D
id R P i i
Th U i
it
f A i
DavidR.Poirier– TheUniversityofArizona
CollaborativeEffortbetweenNASAandEuropeanSpaceAgency:Program
MICAST
(Microstructure Formation in Castings of Technical Alloys under Diffusive and
(MicrostructureFormationinCastingsofTechnicalAlloysunderDiffusiveand
MagneticallyControlledConvectiveConditions)
Outline
¾Introduction:Directionalsolidification(DS)
¾Dendritic array(why“lowgravity”?)
¾Previouslowgdendritic arrayDS
¾Purpose
¾Groundbasedandlowgravityresults
¾Primarydendritespacing
¾Primarydendritetrunkdiameter
¾Radialmacrosegregation
¾Surprises
Dendritic arraymorphologydependsuponDS
processing parameters: G
l
, R, C
o
,
Convection
??
processingparameters:G
l
,R,C
o
,
Convection
??
1 Primary dendrite arm
Primary dendrite
1. Primarydendritearm
spacing(
O)
:Extensive
literature(SCN/Metals)
2 Secondary/tertiary arm
Primarydendrite
spacing(
O)
2. Secondary/tertiaryarm
spacing:Extensive
SCN/Metals
3. Dendritetipradius:SCN/
p
/
limited(AlCu,PbAu,Pb
Pd)
4. Primarydendritetrunk
diameter(
))
:Limited
(Esaka:Thesis86,Grugel:
92/95)
Primarydendrite
trunk diameter (
))
trunkdiameter(
))
Primarydendritearrayduringdirectional
solidification in a positive thermal gradient
solidificationinapositivethermalgradient
•Thermalprofile:stabilizing •Solutal profile: d bili i ( l i h d d i E 0) Pb 6Sb) a. destabilizing(soluteenrichmentreducesdensity,Ec>0):Pb6Sb) ENaturalconvectionisinevitableduring
dendritic DSonearth
Th ll t bl Thermallystable Solutally unstable (Pb5.8Sb:40Kcm1,10Pms1)Felicelli Poirier Heinrich (1998)
Axialmacrosegregation
Radialmacrosegregation (Freckle)
Severalnumericalmodels
(Poirieretal.,Beckermann etal.) Felicelli,Poirier,Heinrich(1998)
Naturalconvectionisinevitableduring
dendritic DSonearth
Thermallystable Solutally stable (Al19Cu:81Kcm1,10Pms1) Radialmacrosegregation DendritesteeplingNoanalyticalornumericalmodels
Naturalconvectiondecreasesprimarydendritespacing
Al26.5wt%Cu
,30K cm1 4 2 m s1 (M.D.Dupouy,D.CamelandJ.J.Favier,Acta.Metall.Mater.Vol.37,No.4,pp.11431157,1989) Al26.5wt%Cu,30Kcm1, 4.2Pms1 cm1,4.2Pms1 Al26.5wt%Cu,25Kcm1, 4.2Pms1 Terrestrial:Solutally unstable, thermally stable mode Terrestrial:Solutally stable, thermallystablemode Microgravity: thermallystablemode 1540± 10Pm 340± 10Pm PrimaryspacingÎ 450± 20PmPurpose
MICAST
:Asystematicanalysisoftheeffectofconvectiononthemicrostructural
evolutionincastbinary,ternaryandcommercialAlSibasedalloys.
MICAST6
MICAST 7
RemeltandDSterrestriallygrowndendritic monocrystalsofAl7
wt% Si (9 mm dia 25 cm long) in
Pg
MICAST6
MICAST7
wt%Si(9mmdia,25cmlong)in
Pg.
Advantages: Minimize Thermo-Solutal Convection
Intent: Produce Segregation Free Samples Grown Underg g p Diffusion-Controlled Conditions
Purpose: Better Understand the Relationship between Processing and Microstructure-Development
1.Primarydendritespacing
2.Primarydendritetrunkdiameter
3.Radialmacrosegregation
Microgravityprocessing:
Partiallyremelt andthenDSfrom
terrestriallygrowndendritic monocrystalin
Pg.
MicrogravityprocessedsampleMICAST7
XrayradiographofMICAST7
EutecticMeltBack /Isotherm
MICAST6:ESALowGradientFurnace(1hrheatup,5hr
hold G
l~ 20 K cm
1): 3 8 cm at 5
Pm s
111 3 cm at 50
Pm s
1MICAST6:ESALowGradientFurnace(1hrheatup,5hr
hold G
l~ 20 K cm
1): 3 8 cm at 5
Pm s
111 3 cm at 50
Pm s
1hold,G
l20Kcm ):3.8cmat5
Pms ,11.3cmat50Pms
MICAST611
MICAST69
MICAST611
MICAST61
Terrestrialprocessing
Graphitecrucible(~9mmID,~19mmOD),104 torr vacuum
Thermal Gradient at the liquidus temperature
(Al-7%Si, Graphite Crucible, 3 TCs located along crucible length)
700 u re, o C 660 680 TC1_4.2Pm s-1 TC2_4.2 Pm s-1 TC3_85Pm s-1 Temp erat u 620 640 G 42 K cm-1 Translation Distance, cm 0 2 4 6 8 10 12 600 Gl= 42 K cm Gl= 41 K cm-1
Don’t have terrestrial samples which are processed in LGF or LGFQ equivalent
Don thaveterrestrialsampleswhichareprocessedinLGForLGFQequivalent
Comparisonofmicrostructures:Al7%Sidirectionally
solidifiedongroundandonISS(MICAST6)
5Pms1 50Pms1 MICAST6:20Kcm1 MICAST6SEED 41Kcm1,22Pms1 TerrestrialDS: 15 K cm1Î 15Kcm Î Convectioncausesdendrite clustering(steepling) andradial macrosegregation at low radialmacrosegregation atlow thermalgradientandgrowth speedsduringterrestrialDS.
Comparisonofmicrostructures:Al7%Sidirectionally
solidifiedongroundandonISS(MICAST7)
MICAST7SEED 41Kcm1,22Pms1 21Pms1 11Pms1 MICAST7:26Kcm1 TerrestrialDS: 24Kcm1ÎTheoreticalmodels:Primarydendritearmspacing
(Diffusioncontrolledgrowth)
(
g
)
•Trivedi (1984)
•HuntLu (1996)
Hunt Lu(1996)
J.D.HuntandS.Z.Lu,MMT,1996
Measured:, G,V
,
,
AlloyProperties:
T
o=sol’dn.temp.range,K
= Gibbs Thomson coefficient m K
= Gibbs-Thomson coefficient, m K
k = equilibrium partition ratio
TheoreticalModel:PrimaryDendriteTrunkDiameter( )
DendriteTipRadius RapidGrowth “Initial”TrunkDiameter(I0) Determination Initial Trunk Diameter,I0 p 12 0 5Ac Esaka(1986Ph.D.Thesis) SuccinonitrileAcetone “alloys” CoarseningRegime e r/ T ip Ra dius 8 10 0.5Ac 1.3Ac 3.6Ac 9.2Ac T runk Di amet e 4 6 FinalTrunk Diameter,I Dl Gl k/(ml R Co (k-1))1e-5 1e-4 1e-3 1e-2 1e-1 2
= 6 59 ± 1 3 r
Primarydendritetrunkdiameter(
model
After
I
0thetrunkdiameterincreasesvia
dissolutionofsecondaryarmsandredepo
iti
th t
k
til th
t
ti
sition onthetrunkuntiltheeutectic.
Kirkwoodripeningmodel(1985)isused.
Assumptions:
1. Secondaryarmmeltsbackbecauseof
itscurvature.
2. Mass of the melted arm deposits on
2. Massofthemeltedarmdepositson
trunksurfacewherethereisnegative
curvature.
Primarydendritearmspacingcomparedwith
HuntLuModel
ical), P m 1600 1800 Hunt-Lu calculations Terrestrial ISS a cing (empir i 1200 1400 Dendrite steepling d ri te arm sp a 800 1000 P rimary den d 400 600 ¾ISSDS:GoodagreementwithpredictionsfromHuntLumodel.Primary dendrite arm spacing (no convection), Pm
200 400 600 800 1000 1200 1400 1600 1800
P 200
¾ TerrestrialDS(“Notsteepled”):GoodagreementwithpredictionsfromHuntLumodel. ¾TerrestrialDS(“steepled”):Convectiondecreases primarydendritearmspacing.
Primarydendritetrunkdiametercomparedwith
trunk diameter model (using r (HuntLu))
trunkdiametermodel(usingr
t
(Hunt Lu))
ical), P m 300 m eter (empir i 200 250 ite tr unk dia m 100 150 rimary dendr 0 50
Trunk dia model calculations using Hunt-Lu tip radius
Terrestrial ISS
Dendrite steepling
¾ISSDS:Goodagreementwithpredictionsfromthetrunkdiametermodel.
Primary dendrite trunk diameter (no convection), Pm
0 50 100 150 200 250 300
P
r 0
(M.D.Dupouy,D.CamelandJ.J.Favier,Acta.Metall.Mater.Vol.37,No.4,pp.11431157,1989)
TrunkdiameterfromDupouy etal.Micrographs(Al26.5Cu)
Al26.5wt%Cu,30Kcm1, 4.2Pms1 Al26.5wt%Cu,30Kcm1, 4.2Pms1 ( p y, , , , pp , ) Al26.5wt%Cu,25Kcm1, 4.2Pms1Terrestrial: Solutally stable
120± 18Pm? 122± 18Pm? 92± 11Pm? TrunkdiameterÎ Terrestrial:Solutally unstable, thermallystablemode Terrestrial:Solutally stable, thermallystablemode Microgravity: 21 21