computational models in the materials world · computational models in the materials world - we are...

Computational Models in the Materials World - We are nearly there….

David Furrer Pratt & Whitney

Rollie Dutton Air Force Research Laboratory

Not Subject to the EAR per 15 CFR Part 734.3(b)(3) ©2013 United Technologies Corporation

AIAA Conference

April 10, 2013

Boston, Mass.

Engineering Materials Today

• Materials are critical for every engineered product

• Traditionally materials were developed by trial and

error processes, separate from application

requirements

• Materials are currently defined by static

specifications based on empirical data Design

Mfg Mtls

• Challenge and opportunity of

Computational Materials Engineering

is the linking of Materials,

Manufacturing Processes and

Component Designs No Technical Data per the EAR and ITAR ©2013 United Technologies Corporation

2

Thrust Specific Fuel Consumption (lb/hr/lb)

707 (JT3C)

B727 (JT8D)

MD-80 (JT8D-200)

747 (JT9D)

777

PW4000 A380

737 A320

(V2500)

Turbojet

Low Bypass Turbofan

BPR 1.1 to 1.8

Engine Certification Date

767

PW4000 94”

High Bypass Turbofan

BPR 4.5 – 5.5 BPR > 6

Evolution of System Efficiency

1950 1960 1970 1980 1990 2000 2010

©2013 United Technologies Corporation

3 Not subject to the EAR per 15 C.F.R. Chapter 1, Part 734.3(b)(3)

Propulsion History

J58 Powered SR-71

Propulsion Innovations Enabled by Materials and Processing Technology

DS blades, Cast &Wrought

disks, 1st Gen Thermal

Spray TBC coatings

1st Gen SC blades, 1st Gen

PM disk, 1st Gen EB-PVD

TBC

2nd Gen SC blades, Aluminide

coatings, 2nd Gen PM/fracture

tolerant disk

LFW Ti IBR, Dual Property Ni

Disk, TBC blades, Burn resistant

Ti, CatArc Metallic Coatings

Dual Property 3rd Gen PM disk,

High modulus blade, 2nd Gen TBC coating

4th Gen PM disk alloy,

Hybrid metallic airfoils,

3rd Gen TBC

F100 Powered F-15 / F-16

F119 Powered F-22 F135 Powered F-35

JT9D powered Boeing 747

PW1133G Powered A320neo



Ni Superalloy Turbine Airfoils: Significant

Advances in Alloys and Casting Processes A

lloy T

em

pera

ture

Capabili

ty

(oF

)

Base

1960 1970 1980 1990

50

100

150

200

250

2000

Equiaxed

Columnar

Single

Crystal

PWA 1426

PWA 1484 PWA 1487

PWA 1497

PWA 1422

PWA 1480

MM247 B1900

IN100



Key Technology Advances for Turbine Airfoil Materials

Time

200

400

600

800

1000

1200

1400

1600

Turb

ine T

em

pera

ture

Incre

ase

Base

1800

Convective Cooling

Film Cooling

Thermal Barrier Coatings

Advanced Casting

Material Technology

ICME is another technology to “break the curve”



Materials & Product Engineering

Mechanical Properties = fn (chemistry and

microstructure)

Microstructure = fn (chemistry and processing)

Processing = fn (component geometry)

Design

Mfg Mtls

Materials, Manufacturing Methods

and Component Design are

Strongly Coupled

ICME -Integrated Computational Materials Engineering


7 No Technical Data per the EAR and ITAR

MIL-HBK-5H

What a Tensile Test Looks Like…..

To a Materials Engineer To a Mechanical Engineer



Properties

True material capability

and property distributions

are controllable and

reproducible; not “random

variability”.

Properties are a fn

(chemistry, microstructure,

stain, cooling rate, etc.);

i.e. pedigree.

From D.Furrer,

OPTIMoM Conf., 2010

Materials Capability Definitions

Materials properties are path dependent and are often

“location-specific”. Engineering specifications often treat

entire material volume as single, homogeneous property

capabilities. Modeling and simulation can help enhance

component property capability definitions



Traditional Materials Definitions

• Design Curves – Empirical; Data Driven

• Specifications

• Prints Notes

• Fixed Process Requirements

Requires Defining Material Equivalency and Methods to

Differentiate Material of One Control Pedigree from Another



The Challenge: Need Models and Computational Infrastructure

Current materials definitions for design limit design

flexibility and final component capabilities

There is a need for:

Model-Based Materials Definitions

Goal is prediction and control of capabilities

Model-based material definitions enable location-

specific prediction, analysis and optimization

Model-based materials definitions enable greater

material, process and component definitions



ICME Involved Linkage with Other Discipline Activities

Material Definition

MfgProcessDefinition

Component Modeling &Prediction

LifingAnalysis

Life-CycleCostAnalysis

Mechanical Design Customer

Requirements

Product

Definition

Holistic Design

Optimization



Materials Technology Enablers Computational Models and Advanced Data Management

Spec Min

LCL

UCL

Failure below control limit Failure due to trending

X X

Goal is prediction and control of capabilities

Properties

Predicted Material and Component

Properties

Traditional empirical property

measurement and analysis migrating to

materials models

Properties

Capture of developmental and production data to

support model development

Predicted Material Properties

Predicted Property Variability

Pro

pe

rty

Serial Number



Material & Process Modeling Goals

• Develop Simulation Tools that Emulate Reality

• Develop Analytical Tools that Provide Insight in

Material - Process - Property Relationships

• Implement Tools for Design and Manufacturing

Benefits

• Model-based Decisions

• Tangible Improvements obtained based on

Decisions



Holistic Integration: Digital Thread

Design

Mfg Mtls

Example of Integrated Computational Materials & Mfg Engineering

DataMechanical Properties

Chemistry

Microstructure

Meta-Data


Chemistry

Microstructure

Meta-Data


Chemistry

Microstructure

Meta-Data


Chemistry

Microstructure

Meta-Data


Chemistry

Microstructure

Meta-Data


Chemistry

Microstructure

Meta-Data


Chemistry

Microstructure

Meta-Data

log Life (e.g. Cycles or TACs)

Usag

e (e.

g. S

tress

)

TypicalMean

Lower

Bound

System Requirements

PMDO – Preliminary

Multi-Disciplinary

Optimization

Location-Specific Optimized

Component Definition

Model-Based

Manufacturing Definition Data Capture /

Knowledge Generation

Microstructure &

Process-Sensitive

Materials Definition

Materials Genome Initiative



Example of ICME Application

From R. Shankar

Cost Benefit

System Benefits

Case Study Heat Treat Forging Part Forge Wt Part Wt Burst Speed Comments

1 Constant Variable Variable -18% -15% +6% Current State of the Art

2 Variable Variable Constant -11% n/a +12% Final Part shape constrained

3 Variable Variable Variable -21% -19% +19% Full impact of tool

Final disk shape before and after

Variation in Weak Pairing Curve - Strong Pairing Curve Intersection

by changing APB Energy and Volume Fraction Independently

0

200

400

600

800

1000

0 10 20 30 40 50 60 70 80 90 100

Gamma Prime Size (nm)

Yie

ld S

tre

ng

th C

on

trib

uti

on

(M

Pa

)

Decreasing APB Energy

Increasing APB Energy

Reference Point:

APB = 0.2 J/m^2, Fraction = 0.34

APB = 0.10 J/m^2

APB = 0.16 J/m^2

APB = 0.08

APB = 0.26 J/m^2

APB = 0.31 J/m^2

Fraction = 0.02

Fraction = 0.06

Fraction = 0.18

Fraction = 0.50

Fraction = 0.75

Decreasing Volume Fraction

Increasing Volume Fraction

ys fp (T )Ni3Ald

dCiCi

i

+MftAPB

b

M 1 fp 0.43Gb

fs1 fp

1 / 2

d

2.56d APB

Gb2 1

1/ 2

fTo

T

d

dCi

1/ 2

i

Ci1/ 2

(1 fp)ky d1 / 2

fpky'd

1 / 2

Yield of Primary ’ Shearing of Secondary ’ (Pairs)

Shearing of Tertiary ’ Solid Soln StrengtheningHall-Petch

PhaseHall-Petch

Primary ‘

• Feasibility of full design

integration demonstrated Over 600 design loop runs with coupled part

geometry and material capability driving design

evolution

Realistic case studies

>50% Reduction in Design

Cycle Time

Geared Turbofan

Engine



Disciplines Touched by ICMSE

• Materials

• Manufacturing

• Design

• Structures

• Quality

• Supply-Chain

Integration of Computational Materials

Science and Engineering is Complicated



Challenges to Effective ICME Deployment

• Accurate computational models

• Efficient simulation software tools

• Data and databases for model application

• Industry standard methods and protocols

• Computational methods for design linkages

• Well trained interdisciplinary workforce

Unique engineering skill sets are required to support each challenge



Computational Supply-Chain

A series of well-established, capable and viable

organizations that provide necessary portions

of the ICME Value Chain

Software Companies

Industry

Government Labs

Academia

– Fundamental Model

Development

– Model Integration into Software

Packages

– Maintenance of Software Tools

– Database Generation

– Application Engineering

– Customer Approval and

Certification

– Education and Training



Engineering

Societies

Conclusions

• ICME: Potential for dramatic changes to

development time, cost, and product capabilities

• Computational materials engineering enables

virtual manufacture and component testing for

optimization and risk mitigation

• Application of ICME has several challenges:

trained practitioners; tools and methods; and

computational infrastructure



21

Any Questions?

computational models in the materials world · computational models in the materials world - we are...

Documents