High-PressureThermalPlasmasandSources(PlasmaSourcesII)TonyMurphy20thEuropeanSummerSchool,October2016“LowTemperaturePlasmaPhysics:BasicsandApplicaJons”MANUFACTURING

WhatisaThermalPlasma?Atorclosetoatmosphericpressure(atleast0.1atm)Temperature~1to2eV(10000to20000K)forelectronsandheavyspeciesHighlyionised(typically100%,atleast5%)HighelectrondensiDes(~1023m-3)Formedbydcoracelectricarcs,radio-frequencyormicrowaveelectromagneDcfieldsDominatedbycollisionsVerywidelyusedinmanufacturingandotherindustries•  Highpowerfluxes•  HighfluxesofreacDvespecies•  StrongradiaDveemission

ThermalPlasmaApplicaJons:ArcWelding

ThermalPlasmaApplicaJons:PlasmaCuTng

ThermalPlasmaApplicaJons:PlasmaSpraying

ThermalPlasmaApplicaJons:WasteTreatment

ThermalPlasmaApplicaJons:ArcLighJng

Carbon arc lamp Xenon arc lamp

ThermalPlasmaApplicaJons:MineralProcessing

Electric arc furnaces

Aluminium dross recovery

Plasma remelting

ThermalPlasmaApplicaJons:CircuitBreakers

ThermalPlasmaApplicaJons:ICP-OES,ICP-MS

Inductively-coupled plasma – optical emission spectroscopy

Inductively-coupled plasma – mass spectroscopy

ThermalPlasmaApplicaJons:NanoparJcleSynthesisandParJcleSpheroidizaJon

ThermalPlasmasinNature:Lightning

TheFirstThermalPlasmaApplicaJon?

The Miller-Urey experiment

Outline1.  ThermalplasmaproperDes

•  Localthermodynamicequilibrium(LTE)•  ComposiDon•  ThermodynamicandtransportproperDes•  RadiaDveemission

2.  GeneraDonofthermalplasmas•  Transferredarcs•  Non-transferredarcs•  RFinducDvely-coupledplasmas•  Microwaveplasmas•  Typesofplasmaflow

3.  Modellingofthermalplasmas•  EquaDons•  Turbulence,electrodesheaths,gasmixtures

4.  ThermalplasmadiagnosDcs•  Enthalpyprobes•  Emissionspectroscopy•  LaserscaXering

5.  ThermalplasmaapplicaDons•  PlasmawastedestrucDon

CollisionsbetweenElectronsandIons

Ar+

Anode

Cathode

e

E Ar+

Anode

Cathode

e

E = =F qE

am m

CollisionsbetweenElectronsandHeavySpeciesTransferLiYleEnergy

Ar+

Anode

Cathode

e

E

Δ ≈kin 2 e hE m mAr+

Anode

Cathode

E e

LotsofCollisionsbetweenElectronsandIonsTransferALotofEnergy

Ar+

Anode

Cathode

e

E Ar+

Ar+

Ar+

Ar+

Ar+ e e

e e

e

e

e

e e

Ar+

Anode

Cathode

E

Ar+

Ar+

Ar+

e

e

e 6

5

For 15 000 K,1 atm:

3 10 m ~7 10 m/s

~ 5 ps

e e el vτ−

=

×

×

EffectofPressureonEquilibrium

Increasing pressure increases the number density, and therefore the collision rate and the transfer of energy from electrons to heavy species

LocalThermodynamicEquilibrium(LTE)LTEexistsifallspeciessaDsfy• MaxwelliandistribuDonfortranslaDonaltemperatures• BoltzmanndistribuDonforexcitaDontemperatures• ChemicalequilibriumequaDonsforreacDons,e.g.,ionisaDonandallthesetemperaturesarethesameLTEexistsifthecollisionrateismuchgreaterthantheratesofdiffusionandconvecDon

Itisgenerallyvalidinthebulkoftheplasma(awayfromtheedgesandelectrodes)

IfLTEexists,thenifweknow(atagivenpointintheplasma)

•  thepressure•  thetemperature(orinsteadthedensityofanyspecies)

•  ifthereisamixtureofgases,theproporDonsofeachgases

thenwecanfullydescribethecomposiDonoftheplasmaatthatpoint

ThisgreatlysimplifiesmodellinganddiagnosDcs

0 5000 10000 15000 20000 25000 300001020

1021

1022

1023

1024

1025

1026

e-,Ar+

Ar3+

Ar2+N

umbe

r den

sity

(m-3)

Temperature (K)

Ar

Ar+

e-

WhyWeLoveLTE

Example: argon arc at 1 atm

WhyCalculatePlasmaProperJes?1.TheyareneededforcomputaJonalmodelling

( )

2

( ) 0

( ) ( )

where 2 ,etc

5( ) ( )2

where isafunctionof

0

rr r

p

B

pc c

h

t

Pt

v r

kh h h ht e

T

U κ

ρρ

ρρ ρ

ρ

τ

φ

η

σ

ρσ

∂+∇⋅ =

∂∂

+∇⋅ = −∇ −∇⋅ + × +∂

= − ∂ ∂

⎛ ⎞∂+∇⋅ = − −∇⋅ ∇ + ⋅∇⎜ ⎟⎜ ⎟∂ ⎝ ⎠

∇ ⋅ ∇ =

v

v vv τ

v

j B g

j j

WhyCalculatePlasmaProperJes?2.Measurementsareinaccurateandinadequate

0 5000 10000 15000 20000 25000 300000

1

2

3

4

5

6

Nitrogen, 1 atm

Temperature (K)

Ther

mal

con

duct

ivity

(W m

-2K

-1)

Calculated (Murphy) Measured (Hermann & Schade) Measured (Plantikow)

0 5000 10000 15000 20000 25000 300000.00000

0.00005

0.00010

0.00015

0.00020

0.00025

Nitrogen, 1 atm

Temperature (K)

Vis

cosi

ty (k

g m

-1s-

1 )

Calculated (Murphy) Measured (Guevara et al.)

ClassificaJonofProperJesThermodynamicproperDes•  Massdensity(kg/m3)•  Specificheatatconstantpressurecp(J/kg/K)•  Specificenthalpy(J/kg)

TransportproperDes(transportcoefficients)•  Viscosity(kg/m/s)•  ThermalconducDvity(W/m/K)•  ElectricalconducDvity(S/m)•  Ordinarydiffusioncoefficient(m2/s)•  Thermaldiffusioncoefficient(kg/m/s)RadiaDonproperDes•  Netemissioncoefficient(W/m3/sr)

RelaJonbetweenBasicDataandMaterialProperJes

Species t hermo - d ynamic d ata

Binary collision integrals

Composition

Thermodynamic properties

Transport properties

Fluid dynamic equations

Basic data

Properties

PlasmaComposiJon

Useeither:•  SahaequaDonsforionisaDonreacDons&Guldberg-WaageequaDonsfordissociaDonreacDons–  Solvesimultaneously(oneequaDonforeachreacDon)

•  MinimisaDonofGibbsfreeenergyG=H–TS(H=enthalpy,S=entropy)–  RestricDons:chemicalelementsconserved,chargeneutrality–  Inputdata(specificheatofeachspecies/parDDonfuncDonsforeachspecies)canbecalculatedfromspectraldata,ortakenfromtables

Examples

0 5000 10000 15000 20000 25000 300001020

1021

1022

1023

1024

1025

1026

e-,Ar+

Ar3+

Ar2+

Num

ber d

ensi

ty (m

-3)

Temperature (K)

Ar

Ar+

e-

0 5000 10000 15000 20000 25000 300001020

1021

1022

1023

1024

1025

Ar3+

N2+Ar2+

Num

ber d

ensi

ty (m

-3)

Temperature (K)

e-

Ar, N2N

N2Ar+

N+

Argon, 1 atm 50% Argon, 50% Nitrogen, 1 atm

ThermodynamicProperJes

TheseareeasilycalculatedoncetheplasmacomposiDonisknown

3

1

1

1

1 1

Density (kg m )

1Enthalpy (J kg ) (+ minor correction term)

Specific heat (J kg K )

N

i ii

N

i i ii

pP

n m

h nm h

hCT

ρ

ρ

−

=

−

=

− −

=

=

∂=∂

∑

∑

SpecificHeatforVariousPlasmas

0 5000 10000 15000 20000 25000 300000

50

100

150

200

250

300 Argon Nitrogen Oxygen Helium Hydrogen

Spe

cific

Hea

t (kJ

kg-

1 K-1)

Temperature (K)

EnthalpyforVariousPlasmas

0 5000 10000 15000 20000 25000 300000

200

400

600

800

1000

1200

1400

1600

1800

2000 Argon Nitrogen Oxygen Helium Hydrogen

Ent

halp

y (M

J kg

-1)

Temperature (K)

TransportProperJes

For plasmas, have many species present – need information about collision cross-sections between all pairs of species: •  neutral-neutral •  neutral-ion •  neutral-electron •  charged-charged (Coulomb)

•  Viscosity: transport of momentum perpendicular to flow •  Thermal conductivity: transport of heat •  Electrical conductivity: transport of charge •  Diffusion coefficients: transport of mass Determined by collision cross-sections (mean free paths) of all species

Viscosity

( )

( ) ( )

η

δ δ δ δ

=

=

= ∑

=∑

= ∑+

⎡ ⎤× − Ω + + Ω⎣ ⎦

102

000

1

0021

(1,1) (2,2)32

Tofirstorder:

where 5

163

5

B j jj

qij j i

j

qi i k k

ijkj i k

j ij jk k ij jkik ik

k T n b

Q b n

n m n mQ

m m m

m m

CollisionIntegralsareCalculatedfromtheIntermolecularPotenJals(using3integraJons)

( ) ( ) ( )( , ) ( )2 2 30

exp where 2 2

ijl s lsij ij ij ij ijij ij

ij

kTT Q g d gkTµ

γ γ γ γπµ

⌠⎮⌡

∞+Ω = − =

( )0

( ) 2 1 cosl lijQ g bdbπ

π χ⌠⎮⌡

= −

are averages over a Maxwellian distribution of the gas kinetic cross-section

02

2 2 212

( , ) 2(1 )

mrr

dr rb g bg b r

χ πµϕ

⌠⎮⎮⎮⌡

= −− −

intermolecular potential ϕ(r) reduced mass µ initial relative speed g impact parameter b

where the angle of deflection is

ViscosityforVariousPlasmas

0 5000 10000 15000 20000 25000 300000.0000

0.0001

0.0002

0.0003

0.0004

Argon Nitrogen Oxygen Helium Hydrogen

Vis

cosi

ty (k

g m

-1s-

1 )

Temperature (K)

Viscosity describes relationship between shear stress and velocity gradient

, is collis~ ion cross e n s ctio

x

i

dvFdy

mT

η η

η

= −

ΩΩ

ThermalConducJvityforVariousPlasmas

0 5000 10000 15000 20000 25000 300000

2

4

6

8

10

12

14 Argon Nitrogen Oxygen Helium Hydrogen

Ther

mal

Con

duct

ivity

(W m

-1K

-1)

Temperature (K)

Thermal conductivity describes relationship between heat flux density an

~

d temperature gradient

, is collision cross-secti onp ii

dTk q kdx

C Tkm

= −

ΩΩ

ComponentsofThermalConducJvityforNitrogen

0 5000 10000 15000 20000 25000 300000

1

2

3

4

5

6 Total Heavy species Reaction Electron Internal

Ther

mal

Con

duct

ivity

(W m

-1K

-1)

Temperature (K)

ElectricalConducJvityforVariousPlasmas

0 5000 10000 15000 20000 25000 300000

2000

4000

6000

8000

10000

12000

14000

16000

Argon Nitrogen Oxygen Helium Hydrogen

Ele

ctric

al C

ondu

ctiv

ity (S

m-1)

Temperature (K)

Electrical conductivity describes relationship between current density and voltage gradient

, is electron density, is density of other species, is collision cross sectio n~ e e

dVjdx

n n nTn

σ σ

σ

=

ΩΩ

RadiaJveEmissionImportantinthermalplasmasAfulltreatmentisdifficult:•  DetermineemissionandabsorpDonateveryvolumeelement

•  AlargenumberofwavelengthintervalsrequiredtocoverlineandconDnuumradiaDon

NetemissioncoefficientsareaverysuccessfulapproximaDon•  Foragiventemperature,integrateemissionoverallwavelengths

•  TakeintoaccountabsorpDoninasphereofagivenradius

ForapplicaDonsinwhichradiaDonfluxatboundaryisimportant,needmorecomplextreatments

J. J. Lowke J Quant Spectrosc. Radiat. Transf. 16 (1974) 111 A. Gleizes et al. J. Phys. D: Appl. Phys. 26 (1993) 1921

NetEmissionCoefficientsforArgonat1atm

L is the radius of the absorbing sphere J. A. Menart PhD Thesis Uni. of Minnesota (1996)

2.GeneraJonofThermalPlasmasa)ElectricArcs:TransferredArcs

•  Arcbetweenoneelectrode(usuallycathode)andmetalorconducDngworkpiece

•  Highenergytransferefficiencytoworkpiece•  Lowgasflow•  RelaDvelyinexpensive•  Highpeaktemperature,narrowdistribuDon(highgradients)•  Usedinwelding,plasmacuing,electricarcfurnaces,arclamps,

wastedestrucDon,etc.

b)ElectricArcs:Non-transferredArcs

•  Arcisconfinedwithinplasmatorch•  HighefficiencyheaDngofbulkgas(“archeater”)•  Highgasflow•  BroaderheatfluxdistribuDon,lowerpeaktemperature•  Usedinplasmaspraying,wastedestrucDon,etc.

c)RFInducJvely-CoupledPlasmas

•  Noelectrodes,solowcontaminaDon•  Generatesalarger,moreuniform,plasmavolume•  Moreexpensive;lowerefficiency•  MoresensiDvetoprocessvariaDons•  Usedforpowderprocessing(densificaDon,spheroidisaDon),

nanoparDcleproducDon,etc.

d)MicrowavePlasmas

•  No electrodes •  Several different designs •  Relatively small and expensive •  Strongly non-equilibrium (gas temp. < 5000 K, electron temp. > 10 000 K •  Applications include MP-AES (microwave plasma - atomic emission spectroscopy),

gas treatment

TypesofPlasmaFlowGravity-drivenflow•  Lowcurrent(<20A)arcs•  Flowdrivenbybuoyancy•  Arclamps

FlowdrivenbyjxBforces•  Highercurrent(>30A)arcs•  Pressuregradient=jxB(themagneDcpincheffect)

•  MaximumvelociDesfrom100to1000m/s

•  Arcsforwelding,plasmacuing,etc

j

MagneJcPinch(LorentzorjxB)Force

B F

TypesofPlasmaFlowFlowsdrivenbythermalexpansion•  Inplasmatorches,thearcoccursinaconfinedregion,causingthepressuretorise

•  SupersonicvelociDes(over2000m/s)typicallyachieved•  Arcstabilisedbyhighgasflowandopenswirlofgas•  Typicalinplasmaspraying

Shock diamonds indicating supersonic flow

3.ModellingofThermalPlasmas•  UsecomputaDonalfluiddynamicequaDonsforviscous

incompressibleflow•  AconservaDonofenergyequaDonisrequiredbecausethe

temperatureisnotconstant•  Maxwell’sequaDonsareusedtodescribecurrentconDnuity

(chargeconservaDon)andmagneDcfields•  AddiDonaltermsdescribing‘plasma’effectsareincluded(ohmic

heaDng,radiaDveemission,magneDcpincheffect,electrondiffusion)

•  TheequaDonofstateisimplicitinthedependenceofthermophysicalproperDesontemperature

•  ModificaDonsrequiredforelectroderegions,turbulentflows,departuresfromLTE,highMachnumberflow,etc.

ConservaJonEquaJons:SingleGas

( )

2

02

( )

( )

( )

0

52

0

( )

(

)

B

p

κ

P

kU

t

t

ch h

eht

T

ρ

σ

ρ

ρ

ρ

ρ

ρ

µσ φ

ρ

∇⋅

⎛ ⎞∇ ⋅ ∇ +⎜ ⎟⎜ ⎟⎝ ⎠

+ =

+ = − − + +

∇⋅ ∇

+ = − − +

=

∂∇⋅

∇ ⋅

∇ ⋅

∇ ×

⋅

∂∂

∂

∂

∂

= −∇

∇

v

vv

v

j B g

j

v

j

τ

A j

Time-dependent term Convective term Diffusive term Source term

ConservaJonofMass(MassConJnuity)

0

mass density time velocity

Note: if there is a source of mass (e.g., evaporation of a surface) then a source term is re u ed

(

q ir

)t

t

ρρ

ρ∂

∇⋅+∂

=

v

v

ConservaJonofMomentum

where

mass density, time, velocity viscous stress tensor, visco

2

sity pressure current de

( )

2 ,

(

nsity, a

,

)

3

m

ji iii iji j i

P

vv v i jx x

t

t

P

x

ρ

ρ

η

η

ρ

τ τ

ρ

η

∇⋅

⎛ ⎞∂⎛ ⎞∂ ∂= − ∇⋅ = + ≠⎜ ⎟⎜ ⎟ ⎜ ⎟∂ ∂ ∂⎝ ⎠

∇ ×+ = − − +∇

⎝ ⎠

+∂

⋅∂ τ

v

vv

τ

g

j

v

B

B

v

j

gnetic field strength acceleration due to gravityg

Magnetic pinch force

Gravity

Pressure gradient

ConservaJonofEnergy2

mass density, time, velocity, temperature enthalpy, thermal conductivity, speci

(

fic heat

current density, electr

5

ical conductivity

(2

))

p

B

p

kht

U Te

t Th c

h κc

U

h

ρ

κ

ρσ

σ

ρ∇ ⋅∇⎛ ⎞

∇ ⋅ ∇ +⎜ ⎟⎜+ = − − +

∂⋅

⎟⎝∂ ⎠

j j

v

j

v

radiative emission coefficient Boltzmann constant, elementary chargeBk e

( )

For compressible flow (Mach number > 0.5), need to add

. , where , to right-hand sideDp Dp p pDt Dt t

∂− ⋅∇ + ≡ + ⋅∇

∂τ v v

%

Ohmic heating

Radiative cooling

Electron diffusion

ChargeConJnuity,MagneJcPotenJal,Maxwell’sEquaJons

( )2

0

0

0

electrical conductivity, electric potential magnetic potential, current density permeability of free space

0

σ φ

µ

σ φ

µ

σ

µ

φ∇ ⋅ ∇

∇

=

= −

= − ∇

∇× =

jB j

A

A

j

j

ComplicaJons:TurbulenceManythermalplasmaflowsareturbulent•  Plasmajets(e.g.,forplasmaspraying)•  PlasmacuingarcsDifferentapproaches•  DirectnumericalsimulaDon(DNS)

–  Solvesallthescalesofflow–  Veryexpensive,andunfeasibleforindustrialproblems

•  LargeeddysimulaDon(LES)–  Solvesforlargescalesandapproximatelymodelssmallscalesoftheflow–  Turbulencemodelneededtoapproximatethesmallscales

•  Reynolds-averagedNavier-Stokes(RANS)–  Approximatelymodelsallscalesoftheflow–  Themostcommonapproachforindustrialproblems–  Include:

–  0equaDons:Prandtlmixinglength–  1equaDon:Spalart-Allmaras–  2equaDons:K-ε,K-εRNG,K-ω

TheK-εModelTurbulenceismodelledasanaddiDonaldiffusionmechanismSolveaddiDonaltransportequaDonsforK(turbulencekineDcenergy)andε(turbulenceenergydissipaDon),whichthengiveviscosityandthermalconducDvity

2

1 2

2

2

1

( ) ( ) 2

( ) ( ) 2

2

is turbulent viscosity, is turbulent thermal conductivityPr

, , ,

t

K

t

Tt

t pt t

t

K

K K K Gt S

C G Ct S K K

G

CKC

S S C

ε

η

ε

ηρρ η ρε

ηρε ε ερ ε η ε ρ

η

ηη ρ κ

ε

⎡ ⎤⎛ ⎞∂+∇⋅ =∇⋅ + ∇ + −⎢ ⎥⎜ ⎟

∂ ⎝ ⎠⎣ ⎦

⎡ ⎤⎛ ⎞∂+∇⋅ =∇⋅ + ∇ + −⎢ ⎥⎜ ⎟

∂ ⎢ ⎥⎝ ⎠⎣ ⎦

= ∇ +∇

= =

v

v

v v

2, Pr are constantstC

EffectofTurbulence

Mass fraction of air for argon plasma jet discharging into air A. B. Murphy and P. Kovitya, J. Appl. Phys. 73 (1993) 4759

Laminar

Turbulent

Argon

Air

Argon

Air

EffectofTurbulence

Temperature (1000 K) of argon plasma jet discharging into air A. B. Murphy and P. Kovitya, J. Appl. Phys. 73 (1993) 4759

Laminar

Turbulent

Ar plasma jet into Air T (K): 300 12,000

ω (s-1): -50,000 50,000

Temperature

Vorticity

(mm)

captures small~large eddies.

Pfender et al. (1991)

(2D-axisym)

Eddy-breakup

Eddy-breakup

Direct numerical simulation

(mm)

Courtesy Dr Masaya Shigeta, Osaka University

ComplicaJons:ElectrodeSheathRegions

10-1 mm 10-5 mm

Te > Th ne ≠ ni

AnodeTreatments

( ) ( )*2 * 30 0

the temperature adjacent to the anode is low, so if assume LTE, then 0 0

Add electron diffusion term (Sansonnens et al.)

0

and ar

Problem:

Approach

el

1:

e

e

a e e e e

e

e a

e

n

D

e

n n n n n n

D

D n

D

σ

σ φ

α

≈ ⇒ ≈

= − ∇ +

∇⋅ ∇ + − =

∇j

ectron and ambipolar diffusion coefficients is three-body recombination coefficient, * indicates LTE value

"LTE-diffusion approximation" (Lowke and Tanaka)Set mesh size near anoApp

de roach

to 2:

width

α

of region in which electron diffusion dominates ( 0.1 mm).Then use LTE everywhere.

Bk T eEΛ = ≈

L. Sansonnens, J. J. Lowke and J. Haidar J. Phys. D: Appl. Phys. 33 (2000) 148 J. J. Lowke and M. Tanaka J. Phys. D: Appl. Phys. 39 (2006) 3634

ComparisonofApproaches1and2

(Sansonnens et al)

(Lowke & Tanaka)

Anode

ThermionicCathodeTreatment

• Applies to high melting point materials e.g., W, C, Mo, Zr • Electrons emitted from cathode • Cathode heated through ion bombardment • Current densities around 100 A/mm2

( )25 2 2

Richardson equation

6 10 A mm K for most metals (work function) 4.5 V for W

exp

2.5 V for W(Th)

w B

w

j AT k

A

e Tφ

φ

− −≈ ×

≈

−

≈

=

ThermionicCurrentDensityvsTemperature

Decreasing work function by 2 V reduces the cathode temperature by 1700 K (for 100 A mm-2) Metals that don’t have high melting points (Cu, Al, Fe etc) cannot be thermionic emitters

0 1000 2000 3000 4000 500010-6

10-4

10-2

100

102

104

106

Cur

rent

den

sity

(A m

m-2

)

Temperature (K)

φw = 4.5 V φw = 2.5 V

ComplicaJons:PlasmasinGasMixturesIfmorethanonegaspresent,thesegasescanmixordemix

EveninLTE,needtoknowtheirrelaDvefracDons

Foreachspeciesi

( ) ( )

2

1

: mass fraction : reaction source term : diffusion mass flux

: ordinary diffusion coefficient

: thermal diffusion coefficient: mole

l

fr

n

ρ ρ

ρ =

∂ +∇⋅ = −∇⋅ +

= ∇ − ∇∑

i i i i

q Tii j ij j

i i

i

ijTi

i

i

j

Y dt Y r

n

Y

m m D x D

r

D

Dx

T

v J

J

J

action

A. B. Murphy J. Phys. D: Appl. Phys. 34 (2001) R151

CombinedDiffusionCoefficients

Groupspeciesintogases

E.g.,forAr-Hearc

•  argongas=Ar,Ar+,Ar2+,Ar3+,e-

•  heliumgas=He,He+,He2+,e-

Modeldiffusionofheliumgasthroughargongas

Typically50ordinaryandthermaldiffusioncoefficientsreplacedbytwocombineddiffusioncoefficients

ExactforhomonucleargasesthatdonotreactforplasmasinLTE

GasConservaJonEquaJon

SpeciesconservaDonequaDonsforeachspecies(Ar,Ar+,

Ar2+,Ar3+,He,He+,He2+,e-)

( )i i iY rρ∇⋅ = −∇⋅ +v Jreplaced by a single gas conservation equation

( ) ( )( )Ar Ar Ar

2Ar Ar He Ar,He He Ar/ ln

x T

Y t Y

n m m D x D T

ρ ρ

ρ

∂ ∂ +∇⋅ = −∇⋅

= ∇ − ∇

v J

J

The overbarred quantities are averaged over all species A. B. Murphy Phys. Rev. E 48 (1993) 3594

Combined diffusion coefficients

Example:DemixinginanArgon-HeliumArc

A. B. Murphy Phys. Rev. E 55 (1997) 7473

Example2:MetalVapourinWeldingArc

All emission Atomic iron line Atomic argon line

Argon arc with steel wire

GoeckeSF,MetzkeE,Spille-KohoffAandLangulaM2005ChopArc.MSG-LichtbogenschweißenfürdenUltraleichtbau(StuXgart:FraunhoferIRB)

MeasurementsShowaRelaJvelyCoolArc,andaLocalTemperatureMinimumonAxis

ZielińskaS,MusiołK,DzierżęgaK,PellerinS,ValensiF,deIzarraC,BriandFPlasmaSourcesSci.Technol.16832–8(2007)

Current = 326 A Steel wire 6 mm above workpiece 4.5 mm above workpiece 3 mm above workpiece

ModellingIndicatesCoolingCanOccurduetoIncreasedRadiaJveEmissionfromMetalVapour

SchnickM,FüsselU,HertelM,Spille-KohoffAandMurphyABJ.Phys.D:Appl.Phys.43022001(2010)

5000 10000 15000 20000 25000100101102103104105106107108109

1010

Net

em

issi

on c

oeffi

cien

t (W

m-3

sr-1

)

Temperature (K)

1% Fe, 99% Ar 1% Al, 99% Ar 100% Ar

4.ThermalPlasmaDiagnosJcsProbes•  Langmuirprobes•  EnthalpyprobesSpectroscopy•  Emissionspectroscopy•  AbsorpDonspectroscopyLaserscaXering•  ElasDcscaXering

–  RayleighscaXering–  ThomsonscaXering–  RamanscaXering

•  InelasDc–  LIF(laser-inducedfluorescence)–  Two-photonLIF–  CARS(coherentanD-StokesRamanscaXering)

EnthalpyProbesShut-off valve

Coolant in

Coolant out

( ) ( )( )stagnation atm

water gas flow no gas flow

2

Enthalpy Flow velocity

is cooling water mass flow rate is gas sampl

ing w

w

g

pg

P Pmh

m

T v

m

C Tm Tρ

−= Δ −Δ =

&&

&&

mass flow rate

Can also measure composition using gas analyser

EnthalpyProbeResults

Entrained air %, temperature (K) and velocity (m/s) for enthalpy probe (●) and laser scattering (Δ) Plasma torch, 900 A, argon, 2 mm downstream from orifice J. C. Fincke, S. C. Snyder, W. D. Swank, Rev. Sci. Instrum. 64 (1993) 711

OpJcalEmissionSpectroscopy

x

Arc

EmissionSpectroscopy:DeterminingtheTemperature

( ) ( )( )

Intensity exp

is energy of upper levelA is transition probability

is statistical weight of upper level is transition frequency

is number density of species (e.g.

mmn m mn

B

m

mn

m

mn

n T ES T A g hZ T k T

E

g

n

ν

ν

⎛ ⎞= −⎜ ⎟

⎝ ⎠

atoms) is partition function of speciesZ

The measurement system has to be calibrated Three standard methods: 1.  Use a calibrated radiation source 2.  Use ratio of intensity of two or more lines 3.  Fowler-Milne method – use maximum in S(T)

m

n

hνmn

LaserScaYeringLightisscaYeredfromelectrons–bothbound(toatomsorions)andfreeTheintersecJonofthelaserbeamandtheobservaJonaxisdefinesapoint

scattered laserλ λ=laser transitionElastic scattering: λ λ≠

Thomson scattering: elastic scattering from free electrons Rayleigh scattering: elastic scattering from bound electrons Raman scattering: elastic scattering from bound electrons

scattered laserλ λ≠Resonance scattering: inelastic scattering from bound electrons

Laser-induced fluorescence: inelastic scattering from bound electrons

laser transitionInelastic scattering: λ λ=

scattered laserλ λ≠

scattered laserλ λ=

+

_ Dipole oscillates with E field; emits perpendicular to oscillation axis

ThomsonScaYeringSpectrum of scattered light depends on Te and ne Approach 1. Wavelength-resolved scattering: Measure scattered light as a function of wavelength – need to be careful of heating the plasma! Approach 2. Integrated scattering: No wavelength resolution Accept all light in a given range of wavelengths around laser wavelength

LaserScaYering:Wavelength-ResolvedScaYering

M. Kuhn-Kauffeldt, J.-L. Marques, J. Schein, J. Phys. D: Appl. Phys. 48 (2015) 012001

Sca

ttere

d si

gnal

Wavelength (rel. to laser wavelength) (nm)

T (1

04 K

)

n e (1

023 m

-3)

r (mm)

LaserScaYering:IntegratedScaYeringRayleighscaXering(elasDcscaXeringfromatoms/ions)ThomsonscaXering(elasDcscaXeringfromfreeelectrons)ResonancescaXering(inelasDcscaXeringfromatoms/ions)

100 A argon arc: T vs r

Spectroscopy

Laser scatt.

A. B. Murphy and A. J. D. Farmer, J. Phys. D: Appl. Phys. 25 (1992) 634

Ar ion laser – 514.5 nm

5.ThermalPlasmaApplicaJonsArcweldingPlasmasprayingPlasmacuingMineralprocessing•  Arcfurnaces•  PlasmaremelDngPlasmaspheroidisaDonArclighDngCircuitinterrupDonChemicalsynthesisProducDonofnanostructuredmaterials•  NanoparDcles•  Bulknanostructuredmaterials•  Carbonnanomaterials(nanotubes,graphene,etc.)Plasmawastetreatment

PlasmaWasteTreatmentComparedtohigh-temperatureincineraDon,plasmagives:

HighertemperatureandenergydensiDes•  ShorterresidenceDmes•  Morecompactsystems

Heatsourceisindependentofwaste•  Exhaustgasflowdecreased•  Rapidshut-offpossible

Nolowerlimitonsize•  EasyintegraDonintoprocesses•  On-sitedestrucDonpossible

BUT:

Highcost

ComponentlifeDmescanbelimited

Highpowerplasmasourcedifficulttomake

J. Heberlein and A. B. Murphy J. Phys. D: Appl. Phys. 41 (2008) 053001

TrendfromNichetoMainstreamApplicaJons

Highly-concentratedwastesorwastesrequiringhightemperatures:Asbestos-containingwastes

Incineratorfly-ashandgrate-ash

Ozone-depleDngsubstances

Otherconcentratedchemicals

Generalwastes:Municipalsolidwastes

Medicalwastes

Auto-shredderresidues

Tyres

Sewagesludge

WesJnghousePlasma:WasteTreatment&FuelProducJon

DCplasmatorch,upto2.5MW

TreatMSW,auto-shredderresidue,industrialliquidandsolidwastesPlantscapaciDesfrom25to300tonnesperday

E.g.,EcoValleyUtashinai,Japan:165tauto-shredderwasteperday,8MWelectricalpowerfromburningsyngas(CO+H2)produced

PLASCON

Treatsliquidandgaseouswastestreams

150kWnon-transferredarc10plantsinAustralia,Japan,USA,Mexico

Treats•  Ozone-depleDngsubstances•  PCB-contaminatedoils•  Greenhousegases•  Liquidorganicwastes

A. B. Murphy, T. McAllister, Appl. Phys. Lett. 73 (1998) 459

FluorocarbonGreenhouseGasesareFormedasaBy-productofHCFCProducJonHFC-23=CHF3=trifluoromethane

GlobalwarmingpotenDalof11700

ByproductofCHClF2(HCFC-22)producDon(2.4%bymass)

185tonne/yearproducedatQuimobasicosSAdCV,Monterrey,Mexico+ → + + +3 2 2 36HF 2CHCl CHCl F CHClF CHF 6HCl

185tonnesperyearofHFC23 is equivalent to 2.2 million tonnes per year CO2

i.e., the annual emissions from a 300 MW coal-fired power station

a 600 MW gas-fired power station

or 500 000 cars

TrifluoromethaneisSuccessfullyDestroyedByPLASCON

+ + → +13 2 2 22CHF H O O CO 3HF

PLASCONDestrucJonofHCF23isVeryProfitable

Economics (under Kyoto Protocol Clean

Development Mechanism): Destruction of 1 t HCF 23

•  Equivalent to destruction of 11 663 t CO2 •  Carbon credits (@ $10/t CO2) have value of $120

000 •  Cost of destruction < $10 000

185 t/year gives a profit of ~$20M/year

Carbon accounting: Destroying 1 t HFC 23

Destroys equivalent of 11 700 t CO2 Produces direct emissions of 0.6 t and indirect emissions of 6 t CO2

ThankyouCSIROManufacturingTonyMurphyChiefResearchScienDstt +61294137150e tony.murphy@csiro.auw www.csiro.au

MANUFACTURING