hotcorrosionof$ sicbased$ …12 joe hagan.pdfoctober 7-11, 2012 – pittsburgh, pennsylvania...

24
October 7-11, 2012 – Pittsburgh, Pennsylvania Hot Corrosion of SiCBased Ceramic Matrix Composite Materials Joseph Hagan, Elizabeth Opila University of Virginia – Materials Science Engineering

Upload: tranmien

Post on 29-Apr-2018

217 views

Category:

Documents


4 download

TRANSCRIPT

Page 1: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

Hot  Corrosion  of  SiC-­‐Based  Ceramic  Matrix  Composite  

Materials  

Joseph  Hagan,  Elizabeth  Opila  University  of  Virginia  –  Materials  Science  

Engineering  

Page 2: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

Mo?va?on  •  SiC/SiC  ceramic  matrix  composites  (CMCs)  are  now  being  introduced  into  

aircraB  turbine  engines  

•  NaCl  from  marine  environment  ingested  into  engine  combine  with  sulfur  

impuriEes  in  jet  fuel  resulEng  in  hot  corrosion  condiEons  

•  Hot  corrosion  of  monolithic  SiC  is  well  studied  and  rapid  aHack  rates  are  

known  to  occur  under  some  condiEons  

•  Hot  corrosion  of  composites  is  not  well  characterized  

•  SiC-­‐based  composites  will  likely  all  be  coated  with  Environmental  Barrier  

CoaEngs  (EBCs)  

•  Understanding  hot  corrosion  of  SiC-­‐based  composites  is  needed  in  case  

of  coaEng  imperfecEons  or  spallaEon  

Page 3: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

Hot  Corrosion  of  SiC  

• 

Page 4: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

•  Determine  the  effects  of  chemistry  on  the  corrosion  of  CMCs  as  

compared  to  monolithic  SiC  –  C  and  BN  interphases,  excess  Si  from  melt  infiltraEon  process    

•  Determine  the  effect  of  CMC  architecture  on  corrosion  

October 7-11, 2012 – Pittsburgh, Pennsylvania

Objec?ves  

Na2O·∙SiO2  Figure  12137,  ACerS  Phase  Equilibria  Diagrams Na2O·∙SiO2·∙B2O3  Figure  515,  ACerS  Phase  Equilibria  Diagrams

B2O3 SiO2 B2O3 SiO2

Na2O

Page 5: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

Materials  •  Substrate  Materials    

–  Hexoloy  -­‐  sintered  α-­‐SiC  (SASC)  with  C  and  B4C  sintering  aids  –  High  purity  CVD  SiC  –  Silicon  –  Real  composites  (MI  and  CVI  with  both  HNS  and  Sylramic  fibers)  

–  Coupons  are  ~  0.5”  x  0.5”  x  0.125”  

•  SiC  substrates  evaluated              uncoated  and  with    

         coaEngs  to  simulate    

         interphase  materials  –  C  coaEngs  ~0.8  µm  thick  

–  BN  coaEngs  ~1.0  µm  thick  

• fiber  

• matrix  

• interphase  

Opila  and  Boyd,  Unpublished

Page 6: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

•  Samples  loaded  with  Na2SO4  on  top  surface  

–  Salt  loading  of  ~2-­‐3  mg/cm2    

•  Samples  exposed  in  a  tube  furnace  in  pairs  –  One  sample  for  chemical  and  one  for  microstructural  

characterizaEon  

–  24  hour  exposures,  controlled  dry  0.1%  SO2/O2  atmosphere,  100  sccm  flow  rate  through  a  46  mm  tube  

•  Mass  change  measured  aBer  every  step  –  As-­‐received,  aBer  salt-­‐loading,  aBer  exposure  

•  Recession  measured  with  a  micrometer  

October 7-11, 2012 – Pittsburgh, Pennsylvania

Hot  Corrosion  Exposure  

~  1.25  cm

Page 7: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

Sample  Characteriza?on  

•  Scanning  electron  microscopy  (SEM)  and  energy  dispersive  spectroscopy  

(EDS)  used  to  determine  morphology  of  corrosion  products  in  plan  view  and  cross-­‐secEon  

•  Corrosion  products  removed  via  step-­‐wise  digesEon  procedure  –  H2O  to  remove  residual  Na2SO4  and  Na-­‐(B)-­‐Silicates  

–  HCl  to  remove  soluble  Na-­‐(B)-­‐Silicates  

–  HF  to  remove  SiO2  

•  InducEvely  Coupled  Plasma  -­‐  OpEcal  Emission  Spectroscopy  (ICP-­‐OES)  

used  to  analyze  composiEon  of  corrosion  products  removed  in  each  of  these  steps  

–  Atomic  concentraEons  (ppm  levels)  and  raEos  of  elements  

–  Analyzed  for  Na,  Si,  S,  and  B  

Page 8: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

-­‐10  

-­‐8  

-­‐6  

-­‐4  

-­‐2  

0  

2  

1000°C   900°C  

Mass  Ch

ange  (m

g)  

Hexoloy   Hexoloy   CVD  CVD  

Mass  loss  aMer  HF  diges?on  includes  loss  of  residual  

Na2SO4  

•  Dark  bars:  change  vs.  mass  aBer  salt  deposiEon  

•  Light  bars:  change  vs.  mass  aBer  salt  deposiEon  aBer  HF  digesEon    

•  More  mass  loss  at  1000°C  than  at  900°C  

•  More  mass  loss  in  Hexoloy  than  in  CVD  

Page 9: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

-­‐14  

-­‐12  

-­‐10  

-­‐8  

-­‐6  

-­‐4  

-­‐2  

0  

2  

4  

Uncoated   C-­‐Coated   BN-­‐Coated  

Mass  Ch

ange  (m

g)  

•  Greater  weight  loss  with  coated  samples  

•  C-­‐coated  sample  lost  more  weight  than  BN-­‐coated  

Mass  loss  aMer  HF  diges?on  includes  loss  of  residual  Na2SO4  

Page 10: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

SEM  of  Surface  •  Largely  silica  formaEon  

•  Pools  of  residual  Na2SO4  

remain  

•  Some  larger  silica  

features  present  

•  Typical  of  both  CVD  and  Hexoloy  samples  

SiO2   Na2SO4  

Hexoloy,  uncoated,  1000°C  for  24hr  

Page 11: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

SEM  of  Cross  Sec?on  •  Bubble  formaEon  on  leB  typical  at  1000°C  

•  Thick  layer  of  residual  sodium  sulfate  on  right  with  liHle  surface  aHack  at  

900°C  

CVD,  uncoated,  1000°C  for  24hr   Hexoloy,  uncoated,  900°C  for  24hr  

Silica  Bubble  

Na2SO4  

Substrate   Substrate  

Page 12: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

PiRng  Morphology  •  Pit  depths  on  the  order  of  20-­‐30  microns  

–  More  than  the  surface  recession  of  8-­‐10  microns  

•  Large  distribuEon  of  pit  sizes  

CVD,  uncoated,  1000°C  for  24hr   Hexoloy,  uncoated,  1000°C  for  24hr,  aBer  HF  

Substrate  

SiO2  in  Pit  

Page 13: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

ICP-­‐OES  •  Samples  digested  (dissolved)  in  a  known  

volume  of  liquid  

•  Liquid  pumped  into  a  spray-­‐atomizer  

•  Atomized  sample  injected  into  an  Ar  

plasma  

•  Atomic  emissions  are  analyzed  using  a  

spectrometer  

•  Trace  element  analysis  possible  –  DetecEon  limits  of  between  0.5  and  5  ppb  

•  ConcentraEons  in  normalized  by  volume  –  1  ppm  =  1  μg/mL  =  1  mg/L  

Page 14: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

ICP  Challenges  •  QuanEtaEve  results  rely  on  very  careful  procedures  

–  Since  trace  concentraEons  are  measured,  small  errors  or  impuriEes  can  affect  results  significantly  

•  Na  is  one  of  the  most  prevalent  contaminants  found  –  Very  difficult  to  ensure  complete  accuracy  with  Na  

•  Si  and  B  are  “sEcky”  elements  –  They  adhere  to  the  sample  introducEon  system  

–  An  acidic  rinse  must  be  used  to  remove  remnants  

–  Incomplete  rinse  can  result  in  inaccurate  calibraEons  and  measurements    •  Inaccurate  concentraEons  •  May  result  in  negaEve  values  in  some  cases  

Page 15: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

•  Corrosion  products  largely  water  and  HF  soluble  

Page 16: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

ICP  Results  –  By  Diges?on  Step  

•  ID-­‐3  and  Hex  are  

baselines  

•  ID-­‐6:  Hexoloy  at  1000°C  

•  ID-­‐8:  CVD  at  1000°C  

•  ID-­‐9:  Hexoloy  with  BN  at  1000°C    

•  ID-­‐21:  Hexoloy  at  

900°C  

•  ID-­‐22:  CVD  at  900°C  

•  ID-­‐25:  Hexoloy  with  C  at  1000°C  

Page 17: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

ICP  Results  –  Water  Diges?on  •  Most  of  the  products  

being  removed  are  Na  and  S  

•  Residual  Na2SO4  from  the  

tests  

•  Some  Si  removed  as  well  

in  the  1000°C  exposures  

•  Some  B  removed  from  the  

BN-­‐coated  sample  

Page 18: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

ICP  Results  –  By  Diges?on  Step  

•  ID-­‐3  and  Hex  are  

baselines  

•  ID-­‐6:  Hexoloy  at  1000°C  

•  ID-­‐8:  CVD  at  1000°C  

•  ID-­‐9:  Hexoloy  with  BN  at  1000°C    

•  ID-­‐21:  Hexoloy  at  

900°C  

•  ID-­‐22:  CVD  at  900°C  

•  ID-­‐25:  Hexoloy  with  C  at  1000°C  

Page 19: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

ICP  Results  –  HCl  Diges?on  •  Most  of  the  products  

being  removed  are  siliactes  

•  Rich  in  Na  and  S  •  Total  mass  removed  in  HCl  

is  four  orders  of  

magnitude  less  than  in  water  

Page 20: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

ICP  Results  –  By  Diges?on  Step  

•  ID-­‐3  and  Hex  are  

baselines  

•  ID-­‐6:  Hexoloy  at  1000°C  

•  ID-­‐8:  CVD  at  1000°C  

•  ID-­‐9:  Hexoloy  with  BN  at  1000°C    

•  ID-­‐21:  Hexoloy  at  

900°C  

•  ID-­‐22:  CVD  at  900°C  

•  ID-­‐25:  Hexoloy  with  C  at  1000°C  

Page 21: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

ICP  Results  –  HF  Diges?on  •  Most  of  the  products  

being  removed  are  Si-­‐based  

•  Hexoloy  had  over  three  Emes  as  much  silica  as  

CVD  

•  C  and  BN  coated  samples  

had  worse  oxidaEon  than  

uncoated  

•  Much  less  silica  at  900°C  

than  1000°C  

Page 22: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

Summary  and  Conclusions  •  CVD  SiC  is  more  resistant  to  hot  corrosion  than  Hexoloy  

•  Hot  corrosion  is  much  greater  at  1000°C  than  at  900°C  

•  Current  results  indicate  BN  and  C  degrade  SiC  corrosion  resistance  •  Hot  corrosion  presents  a  non-­‐uniform  aHack  on  the  sample  surface  

–  Bubbles  and  pits  

•  ICP  appears  more  robust  and  versaEle  than  measuring  mass  change  and  

morphology  characterizaEon  –  Allows  for  analysis  of  the  amount  of  corrosion  products  formed  

–  InformaEon  about  the  chemistry  of  corrosion  products  as  well

•  ICP  in  conjuncEon  with  tradiEonal  characterizaEon  will  allow  for  a  greater  understanding  of  the  hot  corrosion  of  SiC-­‐SiC  CMCs  

Page 23: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

Future  Work  •  Further  explore  the  dependence  of  corrosion  rate  on  temperature  

–  Explore  higher  temperatures,  up  to  1200°C    

•  InvesEgate  the  kineEcs  of  the  corrosion  and  how  they  develop  at  both  shorter  and  longer  Emes  

•  Gain  a  greater  understanding  of  how  B  and  C  affect  the  hot  corrosion  behaviour  of  SiC  

•  Determine  phase  equilibria  present  in  the  system  

•  Characterize  the  pits  and  measure  a  staEsEcal  number  of  pits  to  help  

inform  life-­‐predicEon  models  

•  Move  into  the  more  complicated  architecture  of  composite  materials  

Page 24: HotCorrosionof$ SiCBased$ …12 Joe Hagan.pdfOctober 7-11, 2012 – Pittsburgh, Pennsylvania HotCorrosionof$ SiCBased$ Ceramic$Matrix$Composite$ Materials$ JosephHagan,Elizabeth$ Opila$

October 7-11, 2012 – Pittsburgh, Pennsylvania

Acknowledgements  •  ONR  Award  No.  –  N000141110601    •  Program  Manager  –  Dave  Shifler,  Propulsion  Materials  Program  

•  Jayme  Curet  at  Thermo  Fisher  ScienEfic  

•  Elise  Poerschke