2015 Fall

“Phase Transformation in Materials”

11. 11. 2015

Eun Soo Park

Office: 33-313
Telephone: 880-7221
Email: espark@snu.ac.kr
Office hours: by an appointment 1
Contents for previous class Solidification: Liquid Solid
< Nucleation > & < Growth >
• Equilibrium Shape and Interface Structure on an Atomic Scale

Thermal Roughening

• Growth of a pure solid 1) Continuous growth

: Atomically rough or diffuse interface
2) Lateral growth
: Atomically flat of sharply defined interface
a) Surface (2-D) nucleation b) Spiral growth

Kinetic roughening
When the growth rate of the singular
Interface is high enough, it follows the
2
ideal growth rate like a rough interface.
 “Removal of latent heat” → Heat Flow and Interface Stability
1) Superheated liquid 2) Supercooled liquid
: Extraction of latent heat by conduction in the crystal : conduction of latent heat into the liquid

TS’↓ & TL’↑→v ↓ TS’0 & TL’↓ →v ↑

3
 Q: Alloy solidification?
1. Solidification of single-phase alloys

• Three limiting cases

1) Equilibrium Solidification: perfect mixing in solid and liquid

2) No Diffusion in Solid, Perfect Mixing in Liquid

3) No Diffusion on Solid, Diffusional Mixing in the Liquid

- Planar S/L interface → unidirectional solidification

- Superheated liquid

- Cellular and Dendritic Solidification 4

- Constitutional supercooling
 1. Solidification of single-phase alloys

Fig. 4.19 A hypothetical phase diagram.

k = XS/XL is constant.
XS
k= <1
XL

k : partition coefficient
X : mole fraction of solute

In this phase diagram of

straight solidus and liquidus,
k is const. (independent of T).

Planar S/L interface

→ unidirectional solidification

5
 1. Solidification of single-phase alloys
• Three limiting cases
1) Equilibrium Solidification (perfect mixing in solid & liquid)
2) No Diffusion in Solid, Perfect Mixing in Liquid
3) No Diffusion on Solid, Diffusional Mixing in the Liquid

1) Equilibrium Solidification Fig. 4.19 A hypothetical phase diagram.

(perfect mixing in solid & liquid) k = XS/XL is constant.
→ low cooling rate
: infinitely slow solidification

XS
k=
XL
partition coefficient

- Sufficient time for diffusion in solid & liquid

- Relative amount of solid and liquid : lever rule
6
- Solidification starts at T1 (Xs=kX0) and ends at T3 (XL=X0/k).
Composition vs x at T2

AL
AS

AS = AL

Fig. 4.20 Unidirectional solidification of alloy X0 in Fig. 4.19. (a) A planar S/L
interface and axial heat flow. (b) Corresponding composition profile at T2
assuming complete equilibrium. Conservation of solute requires the two
shaded areas to be equal. AS = AL
7
 1) Equilibrium Solidification : perfect mixing in solid and liquid

T1-ΔT
Xs = kXL
X0+α

kX0+α

T2
AL
AS
Conservation of solute requires
the two shaded areas to be equal.
AS = AL
* Equilibrium solute concentration
kX0 ≤ Xs ≤ X0
X0 ≤ XL ≤ X0/k < XE T3+ΔT X0/k-α

X0-α

Liquid α
2) Non-equilibrium Solidification: No Diffusion in Solid, Perfect Mixing in Liquid
: high cooling rate, efficient stirring
- Separate layers of solid retain their original compositions
mean comp. of the solid ( X S) < Xs
- Liquid become richer than X0/K → XE at the last part of solidification.
- Variation of Xs: solute rejected to the liquid → solute increase in the liquid
(∵ X S < Xs) local equil. at S/L interface

solid → xs < xs
T1-ΔT
liquid > X 0 k → X E

T2

Fig. 4.21 Planar front solidification of alloy X0 in fig. 4.19 TE

assuming no diffusion in the solid, but complete mixing in the liquid.
(a) As Fig. 4.19, but including the mean composition of the solid. (b)
Composition profile just under T1. (c) Composition profile at T2
(compare with the profile and fraction solidified in Fig.4.20b) (d) 9
Composition profile at the eutectic temperature and below.
2) No Diffusion in Solid, Perfect Mixing in Liquid
: high cooling rate, efficient stirring
- Separate layers of solid retain their original
compositions
- mean comp. of the solid ( X S ) < Xs

T1-ΔT

T2 Liquid Primary α + Eutectic

solid → xs < xs liquid > X 0 k → X E T3 TE

10
Mass balance: non-equilibrium lever rule (coring structure)

When cooled by dT from any arbitrary T,

determine the followings.

: solute ejected into the liquid = ?

→ solute increase in the liquid

Ignore the difference in molar volume

between the solid and liquid.
fs: volume fraction solidified

solute ejected into the liquid=? → proportional to what? dfs (XL – XS)
solute increase in the liquid=? → proportional to what? (1-fs) dXL

( X L − X S )dfS =
(1 − fS )dX L Solve this equation.

when fS = 0 → XS, XL? XS = kX0 and XL = X0

11
Initial conditions
 dfS
fS XL dX L XL dX L XL dX L
∫0 1 − fS
= ∫=
X0 X − X
L S
∫X=
0 X
L − kX L
∫ X0 X L (1 − k )
fS XL
∫0
(1 − k )(−1)d ln(1 − f S ) = ∫ d ln X L
XO

XL
ln = (k − 1) ln(1 − f S )
XO

( k −1)
∴ L
X = X f
O L Xs = kXL

( k −1)
X S kXO (1 − fS )
=
: non-equilibrium lever rule
(Scheil equation)
→ quite generally applicable even for nonplanar solid/liquid interfaces provided
here, the liquid composition is uniform and that the Gibbs-Thomson effect is negligible.
12
“If k<1: predicts that if no diff. in solid, some eutectic always exist to solidify.”
(Xs < XL)
3) No Diffusion on Solid, Diffusional Mixing in the Liquid
: high cooling rate, no stirring→ diffusion

- Solute rejected from solid

→ diffuse into liquid with limitation
- rapid build up solute in front of the solid
→ rapid increase in the comp. of solid forming
(initial transient)
- if it solidifies at a const. rate, v, then
a steady state is finally obtained at T3
- liquid : X0/k, solid: X0

local equil. at S/L interface

Composition profile
at T2 < TS/L < T3?

Steady-state profile at T3?

at TE or below ? 13
Interface temperature

* Steady-state at T3. The composition

solidifying equals the composition of
liquid far ahead of the solid (X0).
 No Diffusion on Solid, Diffusional Mixing in the Liquid
During steady-state growth,
(Interface →liquid: Diffusion rate)
Rate at which solute diffuses down the concentration gradient away from the interface
= Rate at which solute is rejected from the solidifying liquid
(Liquid→Solid: solute rejecting rate )

Set up the equation. Solve this equation.

J = DC L ' = v(C L − CS ) X S = X 0 for all x ≥ 0steady-state
∂X L
J=
−D v ( XL XS )
=− dX L v
= − dx
∂x XL − X0 D
( Solidification rate of alloy: excess solute control)
v
ln( X L − X 0 ) = − x + c
D
K STS′ K LTL′ +vLV
= x = 0; X L = X 0 / k
steady-state
( Solidification rate of pure metal: latent heat control, X0
104 times faster than that of alloy) c = ln( − X0)
k
15
 XL − X0 v
ln =− x
1 D
X 0 ( − 1)
k
1− k D
vx
−
XL − X0 = X0( )e
k
1− k x
X L = X O [1 + exp(− )]
k ( D /ν )

( XL decreases exponentially from X0/k at x=0, the interface, to X0 at large distances

from the interface. The concentration profile has a characteristic width of D/v. )

- The concentration gradient in liquid in contact with the solid :

XL − XS
v ( XL − XS )
−DX L′ =
J= X L′ = −
Dv
16
“Alloy solidification” - Solidification of single-phase alloys
* No Diffusion on Solid, Diffusional Mixing in the Liquid

When the solid/liquid interface is within ~D/v of the end of the bar the bow-wave
of solute is compressed into a very small volume and the interface composition 17
rises rapidly leading to a final transient and eutectic formation.
No Diffusion on Solid, Diffusional Mixing in the Liquid

Fig. 4.22 Planar front solidification of alloy X0 in

Fig. 4.19 assuming no diffusion in solid and
no stirring in the liquid.
(a) Composition profile when S/L temperature is
between T2 and T3 in Fig. 4.19.
(b) Steady-state at T3. The composition solidifying
equals the composition of liquid far ahead of
the solid (X0).
(c) Composition profile at TE and below, showing
the final transient.
D/v 18
Concentration profiles
in practice
: exhibit features
between two cases

Zone Refining
19
 Q: Cellular and Dendritic Solidification
by “constitutional supercooling” in alloy

20
2. Cellular and Dendritic Solidification
Fast Solute diffusion similar to the conduction of latent heat in pure metal,
possible to break up the planar front into dendrites.
→ complicated, however, by the possibility of temp. gradients in the liquid.

What would be Te along the

steady-state solidification
at a planar interface concentration profile ahead
of the growth front during
steady-state solidification?

TL’
Te
temp. gradients in the liquid

21
* Constitutional Supercooling No Diffusion on Solid, Steady State
Diffusional Mixing in the Liquid

* Actual temperature gradient in Liquid

TL’
* equilibrium solidification temp. change At the interface,
TL = Tequil. (not TE) = T3
Tequil.

TL' > (T1-T3)/(D/v) : the protrusion melts back - Planar interface: stable
TL' /v < (T1-T3)/D : Constitutional supercooling→ cellular/ dendritic growth 22
Q: Planer → Cell structure → Dendrite?
by constitutional supercooling in superheated liquid

23
Cellular Solidification: formation by constitutional supercooling in “superheated liquid”
If temperature gradient ahead of an initially planar interface is gradually reduced
below the critical value, (constitutional supercooling at solid/liquid interface)

Solute Lower Te Formation of

pile up : Local melting
other protrusions

First protrusion

Break down of the Protrusions develop into long

interface: formation arms or cells growing parallel
of cellular structure Heat flow to the direction of heat flow

Fig. 4.24 The breakdown of an initially planar solidification front into cells
24
Cellular Solidification: formation by constitutional supercooling in “superheated liquid”
If temperature gradient ahead of an initially planar interface is gradually reduced
below the critical value, (constitutional supercooling at solid/liquid interface)

S L

Fig. 5.30. Supercooling ahead of planar interface

Heat flow

<The breakdown of an initially planar solidification front into cells>

Cellular Solidification: formation by constitutional supercooling in “superheated liquid”
If temperature gradient ahead of an initially planar interface is gradually reduced
below the critical value, (constitutional supercooling at solid/liquid interface)

Convexity

First protrusion

Fig. 5.14. Solute diffusion ahead of a convex interface

Break down of the
interface: formation
of cellular structure

<The breakdown of an initially planar solidification front into cells>

Cellular Solidification: formation by constitutional supercooling in “superheated liquid”
If temperature gradient ahead of an initially planar interface is gradually reduced
below the critical value, (constitutional supercooling at solid/liquid interface)

Solute pile up→ Te ↓→ TL’ ↑ → v↓

Solute Lower Tequil
Convexity pile up : Local melting

First protrusion

Break down of the

interface: formation Heat Balance Equation K STS′ K LTL′ +vLV
=
of cellular structure
K: thermal conductivity

<The breakdown of an initially planar solidification front into cells>

Cellular Solidification: formation by constitutional supercooling in “superheated liquid”
If temperature gradient ahead of an initially planar interface is gradually reduced
below the critical value, (constitutional supercooling at solid/liquid interface)

Formation of other
Solute Lower Te protrusions
Convexity pile up : Local melting → “an array of cells”
: most of cells
having 6 neighbers
First protrusion

Break down of the Protrusions develop into long

interface: formation arms or cells growing parallel
of cellular structure Heat flow to the direction of heat flow
28

<The breakdown of an initially planar solidification front into cells>

Tips of the cells grow into the hottest liquid and therefore contain the least solute.

Liquid Primary α + Eutectic

Even if X0 << Xmax

Solute file up → eutectic solidification → formation of 2nd phases
at the cell wall

Fig. 4.25 Temperature and solute

distributions associated with cellular
solidification. Note that solute enrich-
ment in the liquid between the cells, and
coring in the cells with eutectic in the
cell walls. 29
* Cellular microstructures
Note that each cell has virtually the same orientation as its neighbors and
together they form a single grain.

(a) A decanted interface of a cellularly solidified Pb-Sn alloy (x 120)

(after J.W. Rutter in Liquid Metals and Solidification, American Society for Metals, 1958, p. 243).
(b) Longitudinal view of cells in carbon tetrabromide (x 100)
(after K.A. Jackson and J.D. Hunt, Acta Metallurgica 13 (1965) 1212). 30
* Temp. and solute distributions associated with cellular solidification.
1) Note that solute enrichment in the liquid between the cells, and
coring in the cells with eutectic in the cell walls.

T3 < T1

2) Tips of the cells grow into the

hottest liquid and therefore
contain the least solute.

3) Even if X0 << Xmax

Solute file up → eutectic solidification
→ formation of 2nd phases at the cell wall

31
The change in morphology from cells to dendrites
* Cellular microstructures are only stable for a certain range of temp. gradients.
→ Sufficiently low temp. gradients → Creation of constitutional supercooling in the liquid between the cells
causing interface instabilities in the transverse direction (although, No temp. gradient perpendicular to the growth direction)

→ Develop arms, i.e. dendrites form & Change in the direction of the primary arms
away from the direction of heat flow into the crystallographically preferred directions
i.e. (100) for cubic metals.

Fig. 4.27 Cellular dendrites in carbon tetrabromide. Fig. 4.28 Columnar dendrites in a transparent organic alloy.
( After L.R. Morris and W.C. Winegard, Journal of Crystal Growth (After K.A. Jackson in Solidification, American Society for Metals,
6 (1969) 61.) 1971, p. 121.)
32
Cellular and Dendritic Solidification
At the interface, TL = Te (not TE) = T3 → TL, liquid = T1 : T’= T1-T3 (superheating)
• Criterion for the stable planar interface:
TL' > (T1-T3)/(D/v) : the protrusion melts back_steeper than the critical gradient
( T1-T3 : Equilibrium freezing range of alloy)
TL' /v > (T1-T3)/D
Large solidification range of T1-T3 or high v promotes protrusions.
need to well-controlled experimental conditions (temp. gradient & growth rate)

• Constitutional supercooling: TL' /v < (T1-T3)/D

Formation of Cell and Dendrites Structures

Solute effect : addition of a very small fraction of a percent solute with
very small k ( k=
XS
) → (T1-T3 ) ↑ promotes dendrites.
XL

Cooling rate effect : Higher cooling rate allow less time for lateral diffusion
of the rejected solute and therefore require smaller cell or dendrite arm
spacings to avoid constitutional supercooling.
33
Solidification of Pure Metal : Thermal gradient dominant

Solidification of single phase alloy: Solute redistribution dominant

a) Constitutional supercooling

Planar → Cellular growth → cellular dendritic growth → Free dendritic growth

응고계면에 조성적 과냉의 T↓ → 조성적 과냉영역 증가 성장하는 crystal로 부터 발생한
thin zone 형성에 의함 Cell 선단의 피라미드형상/ 가지 잠열을 과냉각 액상쪽으로 방출함
Dome 형태 선단 / 주변에 들의 square array/ Dendrite 에 의해 형성
hexagonal array 성장방향쪽으로 성장방향 변화 Dendrite 성장 방향/ Branched
rod-type dendrite
→ “Nucleation of new crystal in liquid”
성장이 일어나는 interface 보다 높은 온도

b) Segregation
: normal segregation, grain boundary segregation, cellular segregation,
dendritic segregation, inversegregation, coring and intercrystalline
segregation, gravity segregation
34