COMPUTATIONAL MATERIALS DESIGNFOR

ACCELERATED IMPLEMENTATION

G. B. Olson Northwestern University / QuesTek Innovations LLC

Evanston IL

DOE SC-NE Workshop on Advanced Computational Mat. Sci.

March 31, 2004

A) Cybersteel 2020: Ultratough Plate Steels (ONR; CAT)B) HT Carburizing Steels (DOE-OIT; GM, P&W)

C) Superalloys (AF-MEANS, DARPA-AIM; RMCI)D) Bulk Metallic Glasses (DARPA-SAM)

MTL/SRG

GOVERNMENT UNIVERSITY INDUSTRY

NAWC/AD A

Lee

ARL/WMD B

Montgomery

AFRL C,D

Woodward Miracle

NORTHWESTERN A,B,C,D

OlsonAnkenmanAstaBrinsonDunandFine

FreemanGhoshIsheimLiu

MoranVoorhees

High Resolution Microanalysis

WPI/CHTE B

Apelian

PURDUE-CALUMET B

Abramowitz

DREXEL C

Doherty

KTH (Stockholm) C

Agren Sundman

OHIO STATE C

Fraser Mills

CSM C

Eberhart

WISCONSIN-MAD C,D

Perepezko

MIT D

Argon Parks

IIT D

Nash

VIRGINIA D

Poon Shiflet

QUESTEK A,B,C,D

KuehmannHuangJou

QiuRathbunScharer

CATERPILLAR A,B

ChenHsieh

JohnsonYang

ALLVAC STEEL A,B

Lippard Stevenson

INLAND STEEL A

Bhattacharya

GM B

Mishra Sachdev

PRATT & WHITNEY B,C,D

Fowler Schirra

REFERENCE METALS C

Carneiro

HOWMET D

Wolter Wright

BOEING D

Bowden

TangTuftsWright

LEHIGH

Harlow

C

PROCESSING PROPERTIESSTRUCTURE

MATRIX

TEMPERING

SOLUTION TREATMENT

HOT WORKING

SOLIDIFICATION

DEOXIDATION

REFINING

Lath Martensite Ni: Cleavage Resistance Co: SRO Recovery Resistance

STRENGTHENING DISPERSION

(Nb,V)C x

Avoid Fe 3C, M6C, M23 C6

GRAIN REFINING DISPERSION

Microvoid Nucleation Resistance

AUSTENITE DISPERSIONStability (Size, Comp.) Amount Dilatation

GRAIN BOUNDARY CHEMISTRYCohesion Enhancement Impurity Gettering

STRENGTH

TOUGHNESS

RESISTANCEHYDROGEN

]

](Mo,Cr,W,V,Fe) 2Cx

PERFORMANCE

d/ f

K (∆γ)

SAM

TC/MART

CASIS, MAP

GBFLAPW

DVMRW-S

SANS, XRD

APFIM, AEM

σy , H TC(Coh)/DICTRA - KC

ABAQUS/EFGPPT-H

LM, TEM

JIC, γiABAQUS/SPO

TC, ∆V

LM, TEM

MQD

DSC

Transformation DesignTransformation Design

Micromechanics Micromechanics DesignDesign

NanoNano DesignDesign

Quantum DesignQuantum Design

1.0 1.0 µµmm

0.1 0.1 µµmm

1.0 nm1.0 nm

0.1 nm0.1 nm

Solidification DesignSolidification Design

10 10 µµmm

LM

SEM/EDSDICTRA

TC/∆ρL

6

S53 Nanostructured UHS Stainless Results

S53 Nanostructured UHS Stainless Results

15-5PH13-8Mo

Custom 465

Ferrium S53

4340/300M

AF 1410

AerMet 100

Multiscale Ductile Fracture Simulator 50µm

Microvoidingmatrix +primary particles

2µm

50nm

γ

debonding

shear test

micromσ

microij

micromicrop

ij σφκε

∂∂

= && )(macroε

macroσ

macro

macromacro

ij

pijE

Σ∂Φ∂

Κ= && )(

Multi-scaleConstitutivelaw

Iron matrix +secondaryparticles

σY= 1.1GPadσ/dε = 0.6 GPaδIC = 120 µmγi=0.2fTiN= 0.052%dTiN= 1-10 µmfs= 0.015%ds = 0.003-3 µm

CAT SteelSubatomic scale

Fracture toughness

Heterogeneous Precipitation of Austenite on Copper Particles

Toughness - Strength Combination

13

ARCHITECTURE DESIGNARCHITECTURE DESIGN

KBE

Work Flow

Analysis

Information Repository

LSF/Java

Geometric Modeling

Structural Modeling

Thermal Modeling

Gamma Prime Model

Microsoft Excel

Gamma Prime Model

PrecipiCalc

eMatrix

Oracle

Microsoft Access

Analysis Components

Models provided by Pratt & Whitney, General Electric, Questek, and others. Integrated by Engineous into the DKB architecture via iSIGHT

DRM *Java/HTML

Exploration Tech.

* Distributed Resource Management

Networking

3RD Party tools to extend iSIGHT’s integration capabilities

Integration Infrastructure

Core Utilities

iSIGHT framework provided by Engineous Software

Documentum

14

PrecipiCalc™ TimelinePrecipiCalc™ Timeline

Software/Hardware Improvement

1/02 6/02 1/03 6/03 1/04

hierarchical sequencing

24 hours* on Pentium III

600MHz

Applications/Demonstrations

2 hours* on Pentium III 1GHz with

improved numerical nucleation treatment

0.6 hours* onPentium IV

2.2 GHz withcompiler

optimization

0.3 hours* onPentium IV

2.2 GHz withoptimizedparameters

0.05 hours* on Pentium IV 2.2 GHz with cluster hardware

diffusivityscaling

coherencytransition σ(R)

interfacialdissipation

iSIGHTintegration with grain size and

APB energy models

1/01 6/01

Birth

Base Start

Option Start

3D/2Dmapping

multiphaseinteraction

averagecoring coalescence

PWA1100F117 diskcenter

γ/γ’ Latticeparameters

spatialminidisk

prior heat

treatment

spatialV2500

uncertainty error

analysis

R88 tensile

samplesR88 coupons γ’/carbide

grain pinning

minidisk selected locations

* single IN100 PWA1100 simulation

Composition Profile (at.%) across Matrix Channel in between Secondary Precipitates w. Tertiary Precipitate in IN100 - Center 1st Disc

Secondary γ’ precipitates

Diffusion field in Secon-dary γ’ precipitates

Tertiary γ’ precipitate(not fully resolved in this representation)

γ Matrix channel

Secondary matrix inclusionwithin secondary γ’precipitate

(not fully resolved in this representation)

Dieter Isheim, 8Dec 2002

Northwestern UniversityBDA C

DA CB

yield strength, ksi145 150 155 160 165 170 175

prob

abili

ty

0.001

0.0050.010

0.0500.100

0.2500.5000.7500.9000.990

1150F, integral data, sample size = 701RT, integral data, sample size = 1291150F, simulation 110303, sample size = 377RT, simulation 110303, sample size = 377

chemistry

Weibullprobabilitypaper

Impact of DARPA AIM Initiative• Supply chain impact on material capability

capturedEnables versatile processing for smaller lot sizes

• Supply chain impact on material capability captured

Enables versatile processing for smaller lot sizes

14629801366

1309

298

1065

444

PrecipiCalc YS model

Inputs Inputs Inputs

Yield Strength

Network

Heat TreatProcess

Oracle Database

RSSqlApplication

DHET ManagementApplication

Environment/Event Data

Heat Treat Parameters

17

Minidisk Microstructure Prediction with PrecipiCalc

Minidisk Microstructure Prediction with PrecipiCalc

Exp. PpC Exp. PpC Exp. PpC24 23.5 23.1

25.2 25 25.71.23 1.181.27 1.2

Fraction (%) 32.4 35 34 34.6

109 132 120 103129 157 135 114

14618 19.7

20.8 21.821.4 20.7Tertiary γ’ Size (nm) 21.5 21.4

Secondary γ’Size (nm) 107.9 84.2

23.3

Size (µm) 1.28 1.29 1.32 1.31Primary γ’

Fraction (%) 22.6 23.5

Minidisk Comparison

Bore Rim Attachment

Impact of DARPA AIM Initiative

• Material behavior intimately linked and participating in the design process

4 months to improved capability

• Material behavior intimately linked and participating in the design process

4 months to improved capability

ANSYS

Modeling

Analysis & Post-

processing

subrotor.exp

UGsubrotor_orig.prt

ugbatch.bat

mach.inp

mach.prtforge.iges

ht_zones.data

Output_deform.prl

DEFORM

Modeling & Analysis

Post-processing

forge.inp

mach.ecn

mach.sye

N_Burst

baseload.inp BC fileANSYS

Modeling

Analysis & Post-

processing

ANSYS

Modeling

Analysis & Post-

processing

subrotor.exp

UGsubrotor_orig.prt

ugbatch.bat

mach.inp

mach.prtforge.iges

ht_zones.data

Output_deform.prl

mach.prtforge.iges

ht_zones.data

Output_deform.prl

mach.prtforge.iges

ht_zones.data

forge.iges

ht_zones.data

Output_deform.prl

DEFORM

Modeling & Analysis

Post-processing

DEFORM

Modeling & Analysis

Post-processing

forge.inp

mach.ecn

mach.sye

N_Burst

baseload.inp BC file

iSIGHTiSIGHT

32

32.5

33

33.5

34

34.5

35

35.5

36

PWA1100 - Tool PWA1100 - Meas Mod HT - Tool Mod HT Meas

Bur

st S

peed

(x10

3 rpm

)

+ 4.6%

+ 1.6%

+ 2%

+ 4.2%

Rim hole Bore

19

S53: Example Variation Analysis

S53: Example Variation Analysis

Compositional Variations(wt%, ±6σ):C ± 0.01 Cr ± 0.2 Mo± 0.1 W ± 0.1 Co ± 0.3 Ni ± 0.1 V ±0.02

Variations of:Structure — carbide solvus Ts, martensite

Ms, precipitation control ∆G’sProperty — hardness HRc, toughness CVN

CMD/iSIGHT

Results of 1000 runs (12 minutes on a Pentium IV 2.2GHz CPU)

20

S53A Scale-up PropertiesS53A Scale-up Properties

0

70

140

210

280

350

YS [ksi] UTS [ksi] El. [%] RA [%]

S53A (300 pound heat)S53A (3,000 pound heat)

ksi %

60

30

45

15

225 225

285 285

16 15

6359

Processing Structure Properties

Oxidation ResistanceHigher DAl/DO

Low Oxygen Solubility

Creep StrengthDispersion

Stability

Ductility/ Embrittlement

ResistanceMetallic Matrix(with low O2

solubility)

Thermal Barrier Coating

Oxygen Barrier Coating

Bond Coat

Oxide Scale ( Al2O3, YAG)

Film Interface

Dispersed Phases: PdAl,YAl(Pd,Pt, Ru)2(Hf,Zr,Y,Nb)Al

Lattice misfit

High Stability

Coherent Interface

Low Coarsening Rate Constant

Coating

Surface Treatment

Aging

Solution Treatment

Solidification + Shaping

Melt Refining

Solid SolutionLattice Parameters

Solid Solution Strengthening

Oxygen Activity

Control Diffusivities

PERFORMANCE

Tie-tetrahedra in the Nb-Pd-Hf-Al quaternary system at 1200˚C

Oxide scale in Alloy A’ oxidized at 1300oC

A: Al2O3+HfO2B: HfO2

MSc390 Materials Design Projects

Spring 2003

I. Blastalloy II: LC160 Martensite

Client: ONR, Dr. Julie Christodoulou Advisors: Arup Saha; Yana Qian Team: Dan Cogswell, Joe Dudas, Ken Liu

II. Blastalloy III: PH-TRIP Austenite

Client: ONR, Dr. Julie Christodoulou Advisors: Dr. Su Hao; Zhe Liu Team: Danijel Gostovic, Sai-Pong Leung, Derek Norton

III. Dragonslayer II: Carburizing

Stainless Bearing Steel (CS62+) Client: DOE-OIT, P&W, QuesTek Advisors: Dr. Jay Gao; Ben Tiemens Team: Loren Darling, Thor Gudmundsson

IV. MX4: Ni Aeroturbine Blade Alloy

Client: NSF-FRG (OSU), P&W, GEAE Advisors: Dr. Gautam Ghosh; Chandler Becker Team: Travis Harper, Mike McCarren, Paul Von Dollen

V. Noburnium: Nb Superalloy

Client: AF-MEANS, Dr. Craig Hartley Advisors: Abhijeet Misra; Dave Bryan Team: Erhan Altinoglu, Jennifer Bolos, Nora Colligan

VI. Terminator 4: FrankenSteel Goes to Mars

(Biomimetic Self-Healing Alloy Composite) Client: NASA-Houston, Dr. Brad Files Advisors: Jin-won Jung; Michele Manuel Team: Wendy Cheng, Steve Knapp, Richard Scheunemann

VII. HT Aluminum/Bulk Metallic Glass

Client: DARPA-SAM, Boeing, P&W, QuesTek Advisors: Ryan Rathbun; Keith Knipling

Team: Bryan Harder, Nik Hrabe, Alison Markowitz

V-Cr-Ti Alloys

Solubility Limits of C and O in V-4.1Cr-4.3Ti at 1000°C

Isothermal Section at 600°C

Paradigm Shifts: MSE Integrationa) discovery based → design based

- downstream cost of discovery

b) empirical → mechanistic/predictive

c) statistical (eng.)deterministic (sci.)

- prediction of multiple properties from defect distribution functions

- designed variation (predictive robust design: performance/variation tradeoff)

d) computational mat. sci. (toys) → computational mat. eng. (tools)

e) reductionist analytical → holistic (systems) synthetic

} → probabilistic

Optimal Integration:Tactical science in support of strategic engineering