APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient...

21
COMPARISON OF THE RADIATIVE TWO-FLUX AND DIFFUSION APPROXIMATIONS Charles M. Spuckler NASA Glenn Research Center 21000 Brookpark Rd. Cleveland Ohio 44135 Abstract Approximate solutions are sometimes used to determine the heat transfer and temperatures in a semitransparent material in which conduction and thermal radiation are acting. A comparison of the Milne-Eddington two-flux approximation and the diffusion approximation for combined conduction and radiation heat transfer in a ceramic material was preformed to determine the accuracy of the diffusion solution. A plane gray semitransparent layer without a substrate and a non-gray semitransparent plane layer on an opaque substrate were considered. For the plane gray layer the material is semitransparent for all wavelengths and the scattering and absorption coefficients do not vary with wavelength. For the non-gray plane layer the material is semitransparent with constant absorption and scattering coefficients up to a specified wavelength. At higher wavelengths the non-gray plane layer is assumed to be opaque. The layers are heated on one side and cooled on the other by diffuse radiation and convection. The scattering and absorption coefficients were varied. The error in the diffusion approximation compared to the Milne-Eddington two flux approximation was obtained as a function of scattering coefficient and absorption coefficient. The percent difference in interface temperatures and heat flux through the layer obtained using the Milne-Eddington two-flux and diffusion approximations are presented as a function of scattering coefficient and absorption coefficient. The largest errors occur for high scattering and low absorption except for the back surface temperature of the plane gray layer where the error is also larger at low scattering and low absorption. It is shown that the accuracy of the diffusion approximation can be improved for some scattering and absorption conditions if a reflectance obtained from a Kubelka-Munk type two flux theory is used instead of a reflection obtained from the Fresnel equation. The Kubelka-Munk reflectance accounts for surface reflection and radiation scattered back by internal scattering sites while the Fresnel reflection only accounts for surface reflections. https://ntrs.nasa.gov/search.jsp?R=20060051747 2020-07-11T21:56:09+00:00Z

Transcript of APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient...

Page 1: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

COMPARISON OF THE RADIATIVE TWO-FLUX AND DIFFUSION APPROXIMATIONS Charles M. Spuckler NASA Glenn Research Center 21000 Brookpark Rd. Cleveland Ohio 44135 Abstract

Approximate solutions are sometimes used to determine the heat transfer and temperatures in a semitransparent material in which conduction and thermal radiation are acting. A comparison of the Milne-Eddington two-flux approximation and the diffusion approximation for combined conduction and radiation heat transfer in a ceramic material was preformed to determine the accuracy of the diffusion solution. A plane gray semitransparent layer without a substrate and a non-gray semitransparent plane layer on an opaque substrate were considered. For the plane gray layer the material is semitransparent for all wavelengths and the scattering and absorption coefficients do not vary with wavelength. For the non-gray plane layer the material is semitransparent with constant absorption and scattering coefficients up to a specified wavelength. At higher wavelengths the non-gray plane layer is assumed to be opaque. The layers are heated on one side and cooled on the other by diffuse radiation and convection. The scattering and absorption coefficients were varied. The error in the diffusion approximation compared to the Milne-Eddington two flux approximation was obtained as a function of scattering coefficient and absorption coefficient. The percent difference in interface temperatures and heat flux through the layer obtained using the Milne-Eddington two-flux and diffusion approximations are presented as a function of scattering coefficient and absorption coefficient. The largest errors occur for high scattering and low absorption except for the back surface temperature of the plane gray layer where the error is also larger at low scattering and low absorption. It is shown that the accuracy of the diffusion approximation can be improved for some scattering and absorption conditions if a reflectance obtained from a Kubelka-Munk type two flux theory is used instead of a reflection obtained from the Fresnel equation. The Kubelka-Munk reflectance accounts for surface reflection and radiation scattered back by internal scattering sites while the Fresnel reflection only accounts for surface reflections.

https://ntrs.nasa.gov/search.jsp?R=20060051747 2020-07-11T21:56:09+00:00Z

Page 2: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

1G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Com

pari

son

of th

e R

adia

tive

Tw

o-fl

uxan

d D

iffu

sion

App

roxi

mat

ions

Cha

rles

M. S

puck

ler

NA

SA G

lenn

Res

earc

h C

ente

rC

leve

land

Ohi

o

Page 3: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

2G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Obj

ecti

ve

Det

erm

ine

the

erro

r in

usi

ng th

e di

ffus

ion

solu

tion

com

pare

d to

the

Miln

e-E

ddin

gton

two-

flux

sol

utio

n

Det

erm

ine

if u

sing

a r

efle

ctan

ce f

rom

the

Kub

elka

-M

unk

2-fl

ux th

eory

cou

ld r

educ

e th

e di

ffus

ion

solu

tion

erro

r

Page 4: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

3G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

App

roac

h

A p

aram

etri

c st

udy

of a

one

dim

ensi

onal

sem

itran

spar

ent

gray

pla

ne la

yer

and

a no

n-gr

ay la

yer

on a

sub

stra

te w

as

perf

orm

ed

Fres

nela

nd K

ubel

ka-M

unk

refl

ecta

nce

wer

e us

ed

Bas

elin

e a

= 0

.13

cm-1

and σ s

= 94

.38

cm-1

Bon

d co

at e

mis

sivi

ty c

hang

ed f

rom

0.7

to 0

.3C

utof

f w

avel

engt

h 5 μm

for

the

laye

r on

a s

ubst

rate

Page 5: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

4G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

App

roxi

mat

ions

Dif

fusi

on a

ppro

xim

atio

n●

sim

ples

t●

radi

atio

n tr

eate

d as

a d

iffu

sion

pro

cess

●ra

diat

ion

is a

bsor

bed

em

itted

and

ref

lect

ed a

t the

su

rfac

e

Tw

o-fl

ux a

ppro

xim

atio

n●

mor

e co

mpl

icat

ed r

equi

res

com

pute

r so

lutio

n●

ther

e is

a r

adia

tive

flux

trav

elin

g in

pos

itive

and

nega

tive

x di

rect

ions

and

act

at d

ista

nce

●ra

diat

ion

is a

bsor

bed

and

emitt

ed in

side

the

mat

eria

l●

surf

ace

refl

ectio

n st

ill o

ccur

Page 6: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

5G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Ref

lect

ions

Fres

nelr

efle

ctio

ns ρ

oan

d ρi

are

surfa

ce re

flect

ions

and

are

func

tions

of r

efra

ctiv

e in

dex.

Kub

elka

-Mun

kre

flect

ivity

take

s in

to a

ccou

nt F

resn

elsu

rface

refle

ctio

ns a

nd th

e ra

diat

ive

ener

gy s

catte

red

out o

f the

laye

r. It

is a

func

tion

of ρ

o , ρi

, a, σ

s, an

d D

Fre

snel

Kub

elka

-Mun

k

Page 7: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

6G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

a σ

s

T s1

= 20

00K

q r1o

h 1=

250

w/m

2 K

T g1

= 20

00K

n =

2.1

q r2o

T s2

= 80

0K

h 2=

110

w/m

2 K

T g2

= 80

0K

1 m

m

ρo

ρik =

0.8

w/m

K

Hea

t Tra

nsfe

r M

odel

ρiρo

Page 8: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

7G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Com

paris

on o

f Exa

ct R

adia

tive

Tran

sfer

and

Miln

e-E

ddin

gton

Two-

Flux

App

roxi

mat

ion

Pla

ne L

ayer

Fron

t sur

face

tem

pera

ture

s 0

.075

–0.

28%

Bac

k su

rface

tem

pera

ture

-0

.44

–0.

21%

Hea

t flu

x

0

.14

–3.

6%

100

TT

Tdifference

Percent

solution

exact

solution

exact

solution

flux

two

⋅ ⎟⎟ ⎠⎞⎜⎜ ⎝⎛

−=

Page 9: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

8G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Com

paris

on o

f the

Diff

usio

n an

dM

ilne-

Edd

ingt

onTw

o-Fl

ux A

ppro

xim

atio

n

100

TT

Tdifference

Percent

solution

flux

two

solution

flux

two

diffusion

⋅ ⎟⎟ ⎠⎞⎜⎜ ⎝⎛

−=

Page 10: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

9G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Per

cent

Dif

fere

nce

betw

een

Dif

fusi

on a

nd T

wo-

flux

So

luti

ons

for

Fro

nt S

urfa

ce o

f Pla

ne L

ayer

Fre

snel

Kub

elka

–M

unk

100

101

102

103

104

23

456

72

34

567

23

456

72

34

567

Scat

terin

g co

effic

ient

cm

-1

-2-1012345678910 Percent difference in temperatureso

lid li

ne F

resn

el re

flect

ivity

brok

en li

ne K

ubel

ka-M

unk

refle

ctiv

ity

a =

0.00

13 c

m-1

a =

0.01

35 c

m-1

a =

0.13

46 c

m-1

a =

1.34

6 cm

-1

a =

13.4

6 cm

-1

Page 11: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

10G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Per

cent

Dif

fere

nce

betw

een

Tw

o-fl

ux a

nd D

iffu

sion

Solu

tion

s fo

r B

ack

Surf

ace

of P

lane

Lay

er

Fre

snel

Kub

elka

–M

unk

100

101

102

103

104

23

456

72

34

567

23

456

72

34

567

Scat

terin

g co

effic

ient

cm

-1

-6-4-202468

Percent difference in temperature

solid

line

Fre

snel

refle

ctiv

itybr

oken

line

Kub

elka

-Mun

k re

flect

ivity

a =

0.00

13 c

m-1

a =

0.01

35 c

m-1

a =

0.13

46 c

m-1

a =

1.34

6 cm

-1

a =

13.4

6 cm

-1

Page 12: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

11G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Per

cent

Dif

fere

nce

in H

eat F

lux

betw

een

the

Dif

fusi

onan

d T

wo-

flux

Sol

utio

ns fo

r a

Pla

ne L

ayer

Kub

elka

–M

unk

Fre

snel

100

101

102

103

104

23

456

72

34

567

23

456

72

34

567

Scat

terin

g co

effic

ient

cm

-1

-50

-250255075100

125

150

175

200

225

Percent difference in heat fluxso

lid li

ne F

resn

el re

flect

ivity

brok

en li

ne K

ubel

ka-M

unk

refle

ctiv

ity

a =

0.00

13 c

m-1

a =

0.01

3 cm

-1

a =

0.13

46 c

m-1

a =1

.346

cm

-1

a =1

3.46

cm

-1

Page 13: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

12G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

a σ

s

T s1

= 20

00K

q r1o

h 1=

250

w/m

2 K

T g1

= 20

00K

n =

2.1

q r2o

T s2

= 80

0K

h 2=

110

w/m

2 K

T g2

= 80

0K

1 m

m

ρoρik =

0.8

w/m

K

0.79

4 m

m

ks= 33 w/mKε b

c

Hea

t Tra

nsfe

r M

odel

ε m=0

.6

Page 14: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

13G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Per

cent

Dif

fere

nce

in T

empe

ratu

res

betw

een

Tw

o-fl

ux

and

Dif

fusi

on S

olut

ions

for

Fro

nt S

urfa

ce o

f a L

ayer

λ c=

5.0 μm

and

εbc

= 0.

7

100

101

102

103

23

45

67

23

45

67

23

45

67

Scat

terin

g co

effic

ient

cm

-1

-505101520

Percent difference in temperature

a =

0.00

13 c

m-1

a =

0.01

35 c

m-1

a =

0.13

46 c

m-1

a =

1.34

6 cm

-1

a =

13.4

6 cm

-1

a =

67.2

8 cm

-1

solid

line

Fre

snel

refle

ctiv

itybr

oken

line

Kub

elka

-Mun

k re

flect

ivity

Fre

snel

Kub

elka

–M

unk

Page 15: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

14G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

100

101

102

103

23

45

67

23

45

67

23

45

67

Scat

terin

g co

effic

ient

cm

-1

-6-4-2024681012

Percent difference in temperature

a =

0.00

13 c

m-1

a =0

.013

5 cm

-1

a =

0.13

46 c

m-1

a =

1.34

6 cm

-1

a =

13.4

6 cm

-1

a =

67.2

8 cm

-1

solid

line

Fre

snel

refle

ctiv

itybr

oken

line

Kub

elka

-Mun

k re

flect

ivity

Per

cent

Dif

fere

nce

in T

empe

ratu

res

betw

een

Tw

o-fl

ux

and

Dif

fusi

on S

olut

ions

for

Fro

nt o

f Sub

stra

te

λ c=

5.0μ

m a

nd ε

bc=

0.7

Kub

elka

–M

unk

Fre

snel

Page 16: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

15G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Per

cent

Dif

fere

nce

in H

eat F

lux

betw

een

the

Dif

fusi

onan

d T

wo-

flux

Sol

utio

ns fo

r a

Lay

er w

ith

a Su

bstr

ate

λ c=

5.0μ

m a

nd ε

bc=

0.7

100

101

102

103

23

45

67

23

45

67

23

45

67

Scat

terin

g co

effic

ient

cm

-1

a =

0.00

13 c

m-1

a =

0.01

35 c

m-1

a =

0.13

46 c

m-1

a =

1.34

6 cm

-1

a =

13.4

6 cm

-1

a =

67.2

8 cm

-1

-20

-1001020304050

Percent difference in heat flux

solid

line

Fre

snel

refle

ctiv

itybr

oken

line

Kub

elka

Mun

k re

flect

ivity

Page 17: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

16G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Eff

ect o

f Bon

d C

oat E

mis

sivi

ty o

n th

e P

erce

nt D

iffe

renc

e in

T

empe

ratu

re b

etw

een

Tw

o-F

lux

and

Dif

fusi

on S

olut

ions

F

ront

Sur

face

of L

ayer

F

resn

elR

efle

ctiv

ity

λ c

= 5.

0μm

100

101

102

103

23

45

67

23

45

67

23

45

67

Scat

terin

g co

effic

ient

cm

-1

-505101520 Percent difference in temperature

solid

line

ε

bc =

0.7

brok

en li

ne ε

bc =

0.3

a =

0.00

13 c

m-1

a =

0.01

35 c

m-1

a =

0.13

46 c

m-1

a =

1.34

6 cm

-1

a =

13.4

6 cm

-1

a =

67.2

8 cm

-1

Page 18: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

17G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Eff

ect o

f Bon

d C

oatE

mis

sivi

ty o

n th

e P

erce

nt D

iffe

renc

e in

T

empe

ratu

re b

etw

een

Tw

o-F

lux

and

Dif

fusi

on S

olut

ions

F

ront

Sur

face

of L

ayer

K-M

Ref

lect

ivit

y

λ

c=

5.0μ

m

100

101

102

103

23

45

67

23

45

67

23

45

67

Scat

terin

g co

effic

ient

cm

-1

-5-4-3-2-10

Percent difference in temperature

solid

line

ε

bc =

0.7

brok

en li

ne ε

bc =

0.3

a =

0.00

13 c

m-1

a =

0.01

35 c

m-1

a =

0.13

46 c

m-1

a =

1.34

6 cm

-1

a =

13.4

6 cm

-1

a =

67.2

8 cm

-1

Kub

elka

–M

unk

Page 19: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

18G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Max

imum

Cha

nge

in P

erce

nt D

iffer

ence

fo

r a C

hang

e in

Bon

d C

oat E

mis

sivi

ty

from

0.7

to 0

.3

-4.0

4-1

.90

-1.1

10.

86K

-Mre

flect

ion

-11.

33-2

.94

-2.9

8-1

.96

Fres

nel

refle

ctio

n

Hea

t flu

xB

ack

ofsu

bstra

teFr

ont o

fsu

bstra

teFr

ont o

fla

yer layer

substrate

layer

substrate

layer

substrate

Page 20: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

19G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Con

clus

ions

The

per

cent

err

or u

sing

the

diff

usio

n so

lutio

n is

low

est

for

all s

catte

ring

con

side

red

whe

n th

e ab

sorp

tion

is h

igh

The

larg

est e

rror

s oc

cur

mos

tlyfo

r hi

gh s

catte

ring

and

lo

w a

bsor

ptio

n w

hen

the

Fres

nels

urfa

ce r

efle

ctio

n is

use

d

The

err

or a

t hig

h sc

atte

ring

can

be

redu

ced

by u

sing

a K

ulbe

lka-

Mun

kre

flec

tanc

e

For

a la

yer

with

a s

ubst

rate

the

Kub

elka

-Mun

kre

flec

tanc

e is

abl

e to

han

dle

a ch

ange

in b

ond

coat

emis

sivi

ty b

ette

r

Page 21: APPROXIMATIONS Charles M. Spuckler€¦ · 1 10 2 10 234567 2 34567 234567 3 Scattering coefficient cm-1 a = 0.0013 cm-1 a = 0.0135 cm-1 a = 0.1346 cm-1 a = 1.346 cm-1 a = 13.46 cm-1

20G

lenn

Res

earc

h C

ente

r at L

ewis

Fie

ld

Gen

eral

ized

Con

clus

ions

For

the

heat

ing

cond

ition

s pr

esen

ted

here

the

abso

rptio

n th

ickn

ess

(a∙D

) no

t the

opt

ical

thic

knes

s [(

a + σ s

)∙D

] is

the

impo

rtan

t fac

tor

in d

eter

min

ing

the

use

of th

e di

ffus

ion

appr

oxim

atio

n. N

eed

a la

rge

abso

rptio

n th

ickn

ess.

The

wor

st c

ondi

tions

to u

se th

e di

ffus

ion

appr

oxim

atio

nar

e fo

r hi

gh s

catte

ring

thic

knes

ses

(sσ∙

D)

with

no

abso

rptio

n or

low

abs

orpt

ion

thic

knes

s an

d on

ly s

urfa

cere

flec

tions

acc

ount

ed f

or.

Kub

elka

-Mun

kre

flec

tivity

can

red

uce

the

diff

usio

n so

lutio

nap

prox

imat

ion

erro

r fo

r la

rge

scat

teri

ng th

ickn

esse

s