Late Blooming Phases and Dose RateEffects in RPV Steels: Integrated

Experiments and Models

Second International Conference on Multiscale Materials ModelingUniversity of California Los Angeles

Los Angeles, CA Oct 12-15, 2004

Research sponsored by the US NRC

G. R. Odette1, B. D. Wirth2, T. Yamamto1

andM. K. Miller3

1University of California Santa Barbara2University of California Berkeley

3Oak Ridge National Laboratory

Outline• Overview multiscale modeling of embrittlement (DT)• Precipitate compositions in Cu-bearing alloys• Late blooming phases in low Cu and Cu free alloys• Summary and conclusions

DT

DT = f(synergistic combinations)

MetallurgicalCu, Ni, Mn, P, Si, heat

treatment, ….

EnvironmentalT, f, f(E), ft, history

AnnealingTa, ta

Reliable predictions andextrapolations require models

integrated with experiment

Frameworknm-scale features -> Dsy -> DT

0

100

200

300

0 100 200 300

LSF

DT = 0.7Dsy

model

DT ≈ C(…)DsyDsy ≈ Dsppt + DsmfDsppt = a(r)√fp …

DT o

(°C)

Dsy (MPa)

Examples of Models Leading Experiment• Cascade vacancy solute cluster complexes (1980)• Two-feature integrated multiscale model of DT by

accelerated precipitation of Cu from supersaturated solutiondue to radiation enhanced diffusion (RED) to form copperrich precipitates (CRPs) (1983)

• Mn and Ni partitioning to CRPs responsible for strong Nieffects on Dsy (1989)

• Mn-Ni rich precipitates (MNPs) and potential for severe DTin low Cu and Cu free steels due to late blooming Mn-Ni-Siphases (1992)

• Complex compositional structures of CRPs and MNPs (1995)• Many others

MSMP Modeling of Embrittlement

Multiscale Multiphysics Modeling of Irradiation Damaged Materials: Embrittlement of Pressure Vessel Steels, G. R. Odette, et al MRS Bulletin, March 2001

Modeling Hierarchy

• Defect production inaged cascades

• Long-range diffusionof defects & solutes

• Defect & soluteclustering

• Nanofeatures -> Dsy

• Dsy -> DT

• REVE RPV-1

Vacancy Solute Complex

CRP

Solute Atmosphere

Fine-Scale Hardening Features

Hardening Features• Cu rich precipitates• Mn-Ni rich precipitates• Matrix feature vacancy

cluster solute complexesor remnants (Cu notrequired)

MNP

Focus on modeling precipitatecomposition and LBP in low/no Cu alloys

Alloy Ni-Cu-Mn Interactions

0

0.1

0.2

0.3

0.4

0.5

DC c

u, D

C (w

t.%)

0.0 0.5 1.0 1.5 2.0Ni (%)

Cu ≈ 0.4

0.0

0.1

0.2

0.3

0.4 D

C cu,

DC

(wt.%

)

0.0 0.2 0.4 0.6 0.8 1.0 Cu (wt.%)

T13 CM 0.8 Ni

0

50

100

150

200

Ds

y (M

Pa)

0 0.25 0.5 0.75 1Cu (wt. %)

0

50

100

150

200

250

300

Ds

y (M

Pa)

0 0.5 1 1.5 2

Ni (wt. %)

0.2% Cu

0.4% Cu

0.0% Cu0.8 Ni

Ppt: Cu

Ppt: Cu + Ni + Mn

Pre-ppt limit

f p(%)

f p(%)

Thermodynamic Model• Model Cu, Mn, Ni, .., Fe chemical potentials -

µi = (∂G/∂ni)T,nj,k - in matrix (m) and precipitate (p) phases• Flow i to (µim>µip) or from (µim<µip) precipitates until µim= µip• G(XCu,XNi, XMn..,T) summed binary interaction contributions

from CALPHAD extended-regular solution model parameters• Composition and rp-dependent gpm contribution• Some approximations (e.g., no bcc Cu-Ni data) and

modest Mn-Ni parameter adjustments

µcu,ni,mn(Xcu,ni,mn)m

µcu,ni,mn(Xcu,ni,mn)p0.75% Ni

1.5% Ni

0.1% Cu, 1.0% Mn, 260°C

0.4% Cu,0.8% Ni,

1.25% Mn290°C

Initial randomdistribution

CRP growth & coarsening

Nucleation

LMC Simulations• Extract regular solution pair bond energies (eij) from

G(XCu,XNi, XMn..,T) at the ‘average phase’ composition• Kawasaki LMC free energy minimization• Provides insight on the complex compositional structures

Nano-Precipitates260°C, 0.4% Cu

0.0% Ni, 1.25% Mn260°C, 0.4% Cu

0.8% Ni, 1.25% Mn

290°C, 0.15% Cu1.2% Ni, 1.25% Mn

Cu Mn Ni

260°C, 0.4% Cu1.25% Ni, 1.25% Mn

CRP

MNP

LMC

Characterization of CRPs/MNPs•$ TD model predictions for Cu-bearing alloys are consistent with

observations from SANS, APT (minus Fe), AEM, combinedresistivity-Seebeck coefficient (RSC) and CDB PAS

Cu

Ni

Mn

LMCsnapshot

SANS

Nano-Precipitates

5 nm

Fe-0.2%Cu-1.6%Ni-1.6%Mn

Cu Mn NiNot color change

Ni + MnAPT

290°C, 0.15% Cu1.2% Ni, 1.25% Mn

260°C, 0.4% Cu1.25% Ni, 1.25% Mn

MNP

Cu

LMC

Cu Ni Mn

Late Blooming Phases?• Model predicts large fp of Mn-Ni(-Si-P) phases at low/no Cu• Lead to large DT due to Mn-Ni rich phases???• Steady-state cluster dynamics show slow nucleation rates at

low Cu - some Cu is a good catalyst - how much needed???• Late blooming phases (LBP) grow rapidly at high ft ???400

200

0

DT

( °C

)

1.5 0.75

f p(%

)

0.1% Cu, 1.0% Mn, 260°C

model

experiment

Mn-Ni-Cu-T

• Mn-Ni(-Si-P) LBPs - APT, SANS, RSC, PAS-CBD inno/low Cu special model alloys and Cu free RPV steels

• LBPs and Dsy ≈ 160 to 190 MPa in RPV steels whereexpected - lower T ≈ 270°C and f ≈ 3x1011 n/cm2-s, highNi = Mn = 1.6% - but only a moderate ft ≈ 1.8x1023 n/m2

LBPs Found!

0.0001

0.001

0.01

0.1

1

0 0.02 0.04 0.06 0.08 0.1

CM6 T18 - Apr'03

dS/d

W (c

m•s

ter)-1

q2 (Å-2)

Rg = 9.68 Å, b = .322, I o = 47, M/N = 1.14, N d =5.6e17, f v = 0.268%

CM7: 0.0Cu, 1.6Ni, 1.6Mn, 0.040PCM6: 0.0Cu, 1.6Ni, 1.6Mn, 0.005P

SANS

PAS-CBD

Fe

control

T18

NiSa

mpl

e/M

n [n

(pL)

/nm

n (p L]

APT-LEAP• CM6 Steel: 0.0Cu, 1.6Ni, 1.6Mn, 0.25Si, 0.005P, T ≈

270°C, f ≈ 3x1011 n/cm2-s, ft ≈ 1.8x1019 n/cm2

• OV3 Model Alloy; : 0.0Cu, 1.6Ni, 1.6Mn, T ≈ 290°C, f ≈8x1011 n/cm2-s, ft ≈ 1.3x1019 n/cm2

CM6 OV3

0

50

100

150

200

250

300

350

400

0 0.2 0.4 0.6 0.8 1 1.2

Dsys

vs fcp

1/2

Steel CM6 [0.2%Cu, 1.68%Ni, 1.50%Mn, 0.007%P, 0.17%Si]]Steel CM7 [0.0%Cu, 1.70%Ni, 1.55%Mn, 0.047%P, 0.17%Si]]Steel WP [0.04%Cu, 1.65%Ni, 1.43%Mn, 0.011%P, 0.5%Si]]Model OV3 [1.60%Ni, 1.60%Mn]Model OV4 [0.05%Cu, 0.80%Ni, 1.60%Mn]Model OV5 [0.05%Cu, 1.60%Ni, 1.60%Mn]Model OV6 [0.10%Cu, 0.80%Ni, 1.60%Mn]Model OV7 [0.10%Cu, 1.60%Ni, 1.60%Mn]Model OV17 [0.2%Cu, 1.60%Ni, 1.60%Mn]

fcp

1/2 (at%)1/2

Cu free RPV steelswith 1.6 Ni, 1.6 Mnand 0.005 and 0.04P

Model alloys0.8Ni, 1.6Mn and0.05 and 0.1Cu

Model alloys 1.6Ni, 1.6Mnand 0.0, 0.05, 0.1 and 0.2 Cu

Cu bearing alloys

Other low, Cu-free steels

Dsy vs. √fp (RSC)Fe-0% Cu-1.6% Ni-1.6%Mn

Dose Rate Effects• Strong f-effect on CRP (all these alloys in the IVAR

database*) and LBP hardening due to solute-vacancy traprecombination - for high Ni-Mn an effective fluence ftecollapses the data at a reference flux fr = 3x1011 n/cm2-s

0

50

100

150

200

250

300

Ds

y (M

Pa)

0 0.5 1 1.5 2

fte (1023 n/m2)

Onset LBP

0

50

100

150

200

250

300

Ds

y (M

Pa)

0 1 2 3 4

ft (1023 n/m2)

Low f CM6

Medium f CM6

High f CM6

Low f LD

Medium f LD

High f LDLD: 0.4Cu, 1.25Ni, 1.4Mn

CM6:0.0Cu, 1.25Ni, 1.4Mn

fte = ft(fr/f)1/2

* _ 1500 datapoints

Summary and Conclusions• Models have provided enormous insight on many complex

embrittlement mechanisms and phenomena that simplycould not be understood from experiment alone and viceversa

• Models have often led and guided experiment - e.g. theearly prediction and recent successful search for LBPs

• A marriage of modeling to basic experiments (mechanism,single variable property and microstructure) has enabledrobust predictive correlations of messy surveillance dataand perhaps avoiding nasty surprises like LBP

• But models are certainly neither perfect nor complete andthere remain a many open questions and opportunities