Towards a Metadata Scheme for the Description of Materials –...

G.J.Schmitz , 2nd International Workshop on Software Solutions for ICME, Barcelona, Spain April 2016

The European Materials Modeling Council EMMC

Towards a Metadata Scheme for the Description of Materials

– the Description of Microstructures – Georg J. Schmitz, Access e.V., Aachen, Germany Coordinator of ICMEg CSA

The talk touches also: requirements on metadata and interoperability from a continuum software owner perspective

Information flow between materials models

boundary conditions T(ti), ci(ti),…

temperature T , concentrations ci, phase fractions Фi

local equilibrium fractions interfacial properties (T,ci) per phase properties(T,ci)

crystallography kinetic properties

Material Structure (ti)

effective properties functionality (ti)

e/a/m/c models

c models (mesoscale) Material Structure (ti+1)

c models (ti) c models (ti+1)

Perfomance tools (process(t);functionality(t))

per phase properties

effective properties h-tools

HDF5-type-file

getting the initial conditions

vtk-type-file

restart-file preceeding simulation

digital experimental data e.g. CT scans, SEM/EDX,LOM,

EBSD stacks

synthetic microstructures e.g. voronoi tesselation

designed structures, assumptions

microstructure database

experimental data e.g. images

ASCII text + pgm2txt

or .dri

System Description • outer geometry of the system • material composition • initial state, (time lapse)* • boundary conditions/process condition *not addressed yet

This presentation only discusses metadata keywords to describe a

static microstructure Adding keywords for any dynamics will happen in future steps/activities

Das Bild kann zurzeit nicht angezeigt werden.

Huge variety of existing codes

only common denominator: MATERIAL

What is a material ?

Any material is „a number of atoms

arranged in a volume“

The number of atoms may be large and they typically will belong to several chemical elements

What is a material in terms of keywords?

Any material is „a NumberAtoms

arranged in a Volume“

The NumberAtoms may be large (NumberMoles) and they typically will belong to a NumberChemicalElements

Volume

• must be large enough to host the number of atoms • should be large enough to be representative for the „material“

Atoms are described by electronic and atomistic models

Position/Volume/Orientation of RVE Keyword Example ReferenceFrame (component frame CF)

RVEID Identifier

Position = (x0 y0 z0) (in ComponentFrame)

Centroid = (x1 y1 z1) (in RVE Frame)

Origin Position-Centroid

Orientation Orientation e.g Euler Angles or Quaternions (w.r.t CF)

Volume NumberCellsX*CellSizeX NumberCellsY*CellSizeY* NumberCellsZ*CellSizeZ

or lenghtX*lenghtY*lenghtZ

Surface separate slides…

Note: RVE may have arbitrary shape Discretisation needs NOT to be voxel type

Generic microstructure: Atoms in an RVE

Some further keywords: NumberConstituents, NumberIsotopes, NumberAtoms

ConstituentID, ConstituentName(ConstituentID), IsotopeID, IsotopeName(IsotopeID)

Keyword Example

NumberChemicalElements 2 (Al and Cu)

ChemicalElementID „CEID“ =1 (for Al) (=2 (for Cu))

ChemicalElementName(CEID) = Al, CU

NumberMoles(CEID) = see „composition“

NumberMoles all Atoms

Mol(CEID)_Fraction

(NumberMoles(CEID) /NumberMoles)

Composition Vector of all Mol_Fractions for all CEID

Note: „Properties“ are introduced via CEName e.g. MolecularMass Al = 27g/mol

Generic microstructure: Phases in an RVE

Keyword Example

PhaseType Multiphase solid liquid

NumberPhases in the RVE = 3

PhaseID = 3 , 1 and 2

PhaseName(PhaseID) = Liquid = α- Al = Al2Cu

Volume(PhaseID)_Fraction Volume(PhaseID) /Volume(RVE)

Note: Equilibrium volume fractions of phases can be calculated for given composition and conditions, but NO information about spatial distribution of the phases in the RVE

Generic microstructure: Ensembles/Phases I

Keyword Example

PhaseID 1

PhaseName α- Al

Volume Volume of this phase

Centroid Centroid of this phase

Orientation …

NumberChemicalElements 2 (Al and Cu)

ChemicalElementID „CEID“ =1 (for Al) (=2 (for Cu))

ChemicalElementName(CEID) = Al, CU

NumberMoles(CEID) = see „composition“

NumberMoles all Atoms in PhaseID

Mol(CEID)_Fraction

(NumberMoles(CEID) /NumberMoles)

Generic microstructure: Ensembles/Phases II

Keyword Example

CrystalStructure solid-solution FCC

CrystalSymmetry cubic

Crystalline_Fraction 1

LatticeConstants e.g. numbers, xyz or CIF file information

An RVE from the engineering perspective

homogeneous and isotropic piece of material „effective“ (i.e. homogenized/averaged) properties Different RVEs at different positions in component „Properties“ typically taken from tables/databases Microstructure determines properties: statistical

information about microstructure features thus is highly desirable and sometimes available e.g.

– Average grain size (Hall-Petch relation) – Texture (anisotropy) – Secondary dendirte arm spacing (permeability, strength,..) – Phase fractions (precipitation hardening) – Phase composition…(solution hardening) – Dislocation densities..(work hardening)

Keywords for spatially resolved information

How to describe

„arranged“

Atoms may „self arrange“ (see also e/a/m models) and form

objects/features and ensembles of objects/features e.g. – molecules, particles, precipitates, crystals/grains – different phases/crystal structures – defect structures

….and eventually – multiphase, polycrystalline materials

A technical Ni-base Superalloy component

"A scale-bridging journey into the microcosmos of a Ni-base superalloy"

by Ruhr University Bochum

Full Video on YouTube:

https://www.youtube.com/watch?v=wYHch5QIWTQ

Feature Ensemble

of Features Field

Contineous Fields (x,y,z,t0): structuring space

Note: RVE may have arbitrary shape , Cells may have arbitrary shape and may be defined by vortices, edges and faces

Continous Fields (x,y,z,t0)

Keyword Example

CellID defines position of Cell

Mol_Fraction(CEID) e.g. at% Al

Composition

Orientation ….

Strain ….

Defect_Density(type) ….

FlowVelocity …..

FeatureID next slide

Further: Fields corresponding to any other keyword in case spatial variations are relevant

Describing features by fields

1D space

x [mm]

Finite Volume dx

FeatureID

: feature indicator function ( ….a FIELD) = 0 if feature is absent = FeatureID if feature is present

x-r x+r

Example Feature = grain with Volume (r) and Centroid at x0

Generic microstructure: Features in RVE

NumberFeatures in RVE : 4 FeatureIDs = 1-4

NumberFeatures belonging to Phase 1 = 2 FeatureIDs = 1 & 2

Position/Volume/Orientation of Features

Keyword Example FeatureID Identifier

PhaseID to which phase the feature belongs

Centroid Centroid of feature

Orientation Orientation e.g Euler Angles or Quaternions (w.r.t RVE Frame)

Volume Volume of feature/grain

Composition Average composition in feature

Surface See separate slides

Note: Feature may have arbitrary shape which is described by FeatureId Field

Further Features/Fields: Defects

NumberDefectTypes (e.g. point defects, line defects, planar defects, 3D defects, dislocations, porosity,…) NumberDefects(Type) Defect(Type)_Density = NumberDefects(Type)/Volume(RVE) (alt.: Volume(Ensemble, Feature, Field) DefectVolume_Distribution

These keywords apply to RVE (shown here), Ensemble, Feature, or Field

Interfaces

The whole is more than the

sum of its parts

…..when considering interfaces as a part of the story….

Dream3D

Source: Dream3D manual

Feature1 „below“

Interfaces/Surfaces/Boundaries: SurfaceElement

SurfaceElement/Face: FaceID FaceLabel (FaceID,Feature1,Feature2) Area Orientation (of face normal) Centroid Curvature

Feature2 „above“

FaceID

FaceFeature

FaceFeatureLabel (FeatureID1,FeatureID2) Ensemble of Faces shared between two Features e.g 3 and Feature1

Area (FaceFeatureLabel) InterfaceArea consisting of all faces shared between FeatureID1 and FeatureID2

Centroid

Orientation Curvature

FaceFeatureEnsembles

FaceFeatureEnsemble: e.g:

Surface(FeatureID)

SurfaceRVE InterfacesRVE PhaseBoundary(PhaseID1,PhaseID2)

GrainBoundary(PhaseID1,PhaseID1)

Feature Label (4,3)

Face Feature Label (1,3) Feature 3

Canonical set of keywords?

• Is there a hierarchical and canonical base of keywords being sufficient to describe any type of material/microstructure?

• Can all further required keywords be derived by suitable schemes?

Canonical keywords (Volumetric)

RVE Phase Feature Field at Cell NumberElements NumberElements NumberElements NumberElements NumberAtoms NumberAtoms NumberAtoms NumberAtoms NumberPhases

PhaseID PhaseID NumberFeatures NumberFeatures

FeatureID FeatureID NumberDefects NumberDefects NumberDefects NumberDefects Composition Composition Composition Composition Volume Volume Volume Volume Centroid Centroid Centroid Centroid Orientation Orientation Orientation Orientation Crystallinity Crystallinity Crystallinity Crystallinity

Summary: Canonical keywords (2Dimensional)

Surface Interfaces/ surfaces

FaceFeature Face

Centroid Centroid Centroid Centroid Area Area Area Area Composition Composition Composition Composition

Curvature Curvature Curvature Curvature Orientation Orientation Orientation Orientation

Area/Volume Area/Volume Area/Volume Area/Volume

Scheme to derive further keywords

Derivation Scheme

Keyword appendix

New keyword _example

time derivative _dot Centroid_dot smallest/largest value _min/ _max Composition_min

Calculate gradient (s) _grad Composition_grad Calculate fractions _fraction Volume_fraction normalizing by volume _density NumberMoles_density,

Mass_density Distributions _distribution

_distribution_mean _distribution_sigma

Volume_distribution Mean value of distribution Standard deviation of distribution

Differences _diff Centroid_diff

Products _product_ Mass_times_Centroid_dot Averaging (time/space) _mean (time)

_average (space) Centroid_mean Volume_average

other types of „operators“ (to be refined..)

_integral , _sum _ratio , _fit, _total

Entire equations….

Microstructure description: What‘s next?

SD Microstructure at (t0)

Keywords for time

dependantBC

MR for thermodynamic

and kinetic properties of pure phases

and interfaces

Microstructure (t0 +∆t)

possibility to describe „evolution“ of materials functionality, performance

post-processing/

anaylsis/ visualisation

SD (t0 +∆t) input for

subsequent simulation

200x200x200 Cells 50 µm x 50 µm x 50 µm

Al – 1.564 at% Cu

green = Al2Cu violet = α-Al

Coarsening of Al-Cu eutectic

Microstructure description: What‘s next?

Microstructure (t0)

Keywords for „effective“ material

properties „post processing“

Keywords for properties of

phases „Materials Relations“

possibility to predict materials properties: „functionality, performance

A. Hashibon. Joint ICMEg EU-MMC EMMC Workshop, 24th November 2015, Covent Garden, Brussels

SimPhoNy SD: Example for ContinuumModels SD

• name, • description, • uuid • dataContainer

Internal Regions (walls, moving parts) • name • uuid • description • options? • type (CUBA)

Boundaries

DS:ABC: 2DMesh name, description,

CUBAtype

Face UUID,

DataContainer

FaceFeature UUID,

DataContainer

Surface/Boundary UUID

DataContainer

DS: ABCMesh name, description,

CUBAtype

Cell UUID,

DataContainer

Feature UUID,

DataContainer

Phase properties

• Name • Description • UUID • kind:CUBA • parameters: Shared

material properties

Geometry/Arrangement

RVE UUID

DataContainer

Phase UUID

DataContainer

interface UUID

DataContainer

Atom properties

RVE properties

Interatomic properties

interface properties

…….

Metadata for properties to complement

An HDF5 file generated based on keywords for mechanical properties from Abaqus-Python, showing Material Elasticity, Plasticity and Expansion, tables.

Elasticity table Expansion table Plasticity table

first keywords for properties

Work performed by Anshuman Singh UPC BarcelonaTech

Summary: metadata description of a microstructur

The proposed metadata keyword structure describes – a volume – distinct features in the volume – continous fields – interfaces – surfaces/boundaries

The proposed metadata keyword structure accounts for – the hierarchical structure of materials – descritpions of phase-transitions – future amendments

The proposed metadata keyword structure allows to assign – attributes to an ensemble of objects („phases“) – attributes to features – attributes to interfaces – conditions on boundaries

The proposed metadata keyword structure is independent of models and numerics – in view of interoperabilty with experimental data – in view of possibly available analytic solutions

The proposed metadata keyword structure may even be applicable to e/a/m - models – to be verified….

Publication has been submitted by as invited regular article to a special issue on "Materials Genome and Informatics“ of

„Science and Technology of Advanced Materials”

Towards a Metadata Keyword Scheme for the Description of Materials

-Keywords for the Description of Microstructures- G.J.Schmitz, B.Böttger, M.Apel, J.Eiken, G.Laschet, R.Altenfeld, R.Berger, G.Boussinot, A.Viardin

Article is scheduled to appear in spring 2016 on-line as open access HDF5 file comprising the defined keywords will be provided as

template along with publication Keyword list repreents a first version and will be/has to be regularly

amended/revised according to future findings/discussions e.g. within the present meeting

Next steps II

Towards a Metadata Scheme for the Description of Materials –...

Documents

Transcript of Towards a Metadata Scheme for the Description of Materials –...

Skyway:ConnectingManagedHeaps inDistributedBigDataSystemspeople.cs.uchicago.edu/~shanlu/paper/p56-nguyen.pdf · Skyway provides SkywayObjectOutputStream and SkywayObjectInputStream

RTXXX~N688~P56-58 - USAL

P56 · 2019. 12. 5. · Title: P56.pdf Created Date: 1/13/2004 7:09:40 PM

(19) 대한민국특허청(KR) (12) 등록특허공보(B1)diamond.kist.re.kr/DLC/publication/patent/p56.pdf · 2020. 7. 28. · (19) 대한민국특허청(KR) (12) 등록특허공보(B1)

Validation of a structure of a multiscale model with OWL and ...congress.cimne.com/icme2016/admin/files/filepaper/p28.pdfAdaptive Multiscale Modelling Methodology (AM3) Classical approach

Y2 P56 6009 the Song at the Sea Spanish

OPERATING SYSTEMS (O.S.) PILO TNG DSR C ( .) · p20 p56 p48 p56 p31 p56 p14 p20 p63 p25 p12 p34 p08 ... sen eca bristol clyman harr is fay ette oxford halsey norway ridgeway clinton

RAPID PRELIMINARY SEISMIC ASSESSMENT METHODOLOGY …congress.cimne.com/eccomas/proceedings/compdyn2011/compdyn2011_full/394.pdfRecent earthquakes have repeatedly illustrated the deficiencies

Neutrino-nucleus (nucleon) Reaction Measurement by J-PARC MLF sterile neutrino search experiment (J-PARC P56) Takasumi Maruyama (KEK) for J-PARC P56 working.

P56 Saber vs Conocer

Thermodynamics and HDF5: Towards a Metadata Keyword …congress.cimne.com/icme2016/admin/files/filepaper/p88.pdf · Thermodynamics and HDF5: Towards a Metadata Keyword Scheme for

Situation Recognition from Multimodal Data Tutorial (ICME2016)

contributed articles - University of Washingtoncourses.washington.edu/info445/docs/p56-hellerstein.pdfa GrouP oF database researchers, architects, users, and pundits met in May 2008

P56 Pharma Presales Activities

828,S30 Abraha~ S6 P56 S20,P82 P28Mattos, Flor R. McClain, Jon V4T

Accelerating multi-scale microstructure simulations by exploiting …congress.cimne.com/icme2016/admin/files/filepaper/p29.pdf · 2016-05-02 · rm 1 Accelerating multi-scale microstructure

Best practice in information exchange in process scale modellingcongress.cimne.com/icme2016/admin/files/filepaper/p84.pdf · 2016-04-29 · Best practice in information exchange in

CAPSUL - CIMNEcongress.cimne.com/icme2016/admin/files/filepaper/p104.pdfCAPSUL: a tool for computational homogenization of polycrystals A. Cruzado1, B. Bellón1, J. Segurado1,2, J.

IJCST Mechanicalpropertiesoffabric materials for draping ...repository.ust.hk/ir/bitstream/1783.1-436/1/p56.pdfMechanicalpropertiesoffabric materials for draping simulation Z.Wu Department

Takasumi Maruyama (KEK) for J-PARC P56 working group