Diapositive 1

1
Modelling Transport Phenomena during Spreading
and Solidification of Droplets in Plasma
Projection
Dominique GOBIN CNRS France
NGU Seminar
Nova Gorica (November 5, 2009)
2
Contents

1. Motivation
2. Equations
3. Isothermal spreading
4. Spreading with solidification
5. Perspective

3
Building up a coating
The functional properties of the coating depend
on the cohesion and adhesion of the
splats
4
(No Transcript)
5
Characteristic times Spatial scales
Time scales Spatial scales
Splat Formation (spreading solidification) 10 µs 0.5 à 5 µm
Time interval between 2 impacts at the same place 10 à 100 µs
Layer Formation A few ms A few 10 µm
Time interval between two passes of the torch A few s
6
Modelling issues
Modelling the plasma
Spreading and solidification of droplets on a
cold substrate
In-flight melting (vaporization) of particles
Building-up the coating
Define and control the process parameters
7
Droplet spreading and solidification
Impacting Particle
T0 gt Tm V0 ? 100 m/s 20 lt d0 lt 50 µm
Tsplat and dsplat time evolution
Ts
Substrate
8
2. Equations
9
Modelling spreading
Pure fluid dynamics problem. The substrate is a
boundary condition
Mass Conservation
Momentum Conservation
10
Modelling spreading
Non-dimensionalizing variables (choosing
reference values d0, V0, etc) yields the
dimensionless parameters of the problem
Mass Conservation
Momentum Conservation
11
Modelling spreading with solidification
Coupling the equations of fluid dynamics with
the heat transfer equations
Mass Conservation
Momentum Conservation
Energy Conservation - in the splat -
in the substrate
12
Modelling spreading with solidification
During solidification two phases (solide and
fluid) are present. A phase function is defined

1 if liquid 0 if solid

Mass Conservation
Momentum Conservation
13
Modelling spreading with solidification
Heat transfer and enthalpy formulation
Energy Conservation
14
Conservation equations
Mass Conservation
Momentum Conservation
Energy Conservation
1 liquid 0 solid
Liquid Fraction
15
Physical parameters of the problem
Parameters of the particles at impact Nature
Size Velocity
Temperature and state of
melting Parameters of the substrate
Nature Rugosity Initial temperature Surface
chemistry (wettability)
16
Spreading and solidification of a splat
1. Operation parameters

• Dynamic contact angle

2. Adjustable parameters
Splat
Substrate

• Contact thermal resistance

17
Numerical tool
Simulent-Drop a software developed at the
University of Toronto (J. Mostaghimi et al.)
Main hypotheses

• Newtonian fluid
• Constant properties (surface tension, contact
  resistance, conductivities, viscosity, )
• Equilibrium solidification

18
Numerical tool main features

• Finite difference method
• Fixed regular grid (Eulerian formulation)
• Boundary condition using dynamic contact angles
• Interface reconstruction VoF method
• 3-D Geometry (computational domain a quarter
  of the domain)

Typical grid
Symmetry
Full domain
Computational domain
19
Scales
Micrometric droplets (Conditions of plasma
projection)
Millimetric droplets (Free fall conditions)
Similitude ?
Re identiques We 10 à 100 fois plus grand
d
1 mm
gt 10 µm
Vimpact
1 m/s
gt 100 m/s
Characteristic times
ms
µs
- 19 -
20
3. Isothermal spreading
21
Isothermal impact of a water droplet
Simulation F. Loghmari
Water droplet spreading d0 2,75 mm , V0 1.18
m/s on soft wax q (105,95) Rioboo et al.
(2001)
- 21 1 ²
22
Simulation Experiments
Spreading factor d(t)/do
Reduced time t t Vo/ do
23
Wettability effect
?a
Substrat
Forward dynamic contact angle
Backward angle effect (?a 105)
Forward angle effect (?r 95)
- 23 -
24
4. Spreading with solidification
25
mm-size droplet simulation
Simulation Nabil Ferguen
Copper droplet on steel substrate d 3 mm V
4 m/s Ts 25C
26
Impact velocity influence
Vp8 m/s
Vp4 m/s
Vp2 m/s
Time evolution of the spreading factor
- 26 -
- 26 -
27
Impact velocity influence
Vp 8 m/s
Vp8 m/s
Vp4 m/s
Vp2 m/s
Vp 4 m/s
Time evolution of the spreading factor
Vp 2 m/s
- 27 -
- 27 -
28
Contact thermal resistance
Non perfect contact between the drop and a rugous
substrate gt resistance to the heat flux
temperature discontinuity at the interface
CTR Model
- 28 -
- 28 -
29
Influence of the contact thermal resistance


10-5 m²K.W-1
5.10-6 m²K.W-1
2.10-6 m²K.W-1
10-6 m²K.W-1
- 29 -
- 29 -
30
High contact resistance
RTC 10-5 m²K.W-1
Simulation Nabil Ferguen
Copper droplet on steel substrate d 3 mm V
4 m/s Ts 400C
31
Low contact resistance
RTC 10-6 m²K.W-1
Simulation Nabil Ferguen
Copper droplet on steel substrate d 3 mm V
4 m/s Ts 400C
32
Influence of the initial substrate temperature
Ti Cr Cu
To 300 K
To 673 K
From Fukumoto et al. (1995)
33
Splat formation
Morphological transition temperature Tt
Alumina on steel 304L
 Splat  Pre-heated substrate Tsubgt Tt Better
adhesion (? 30 MPa)
 Splash  Cold substrate Tsublt Tt Poor
adhesion of the coating (? 4 MPa)
34
Influence of the substrate temperature
Vp 4 m/s dp 2 mm T0 1100 C, Tf
1080 C
Re 23900 , We 191
Ts 1084 C
Ts 800C
Ts 400C 
Ts 25C
Pre-heating of the substrate higher final splat
diameter
35
Transition Temperature ?

• Desorption of adsorbates and condensates
• Modification of wettability of the substrate
• Modification the thermal resistance
• Possible evolution of the surface state of the
  substrate

36
5. Further developments
37
Non equilibrium Solidification

• Basic hypothesis solidification at equilibrium
• Most models do not take into account
  undercooling, nucleation and growth
  problem of multi-scale
  (micro macro) simulation
• But in plasma projection, the cooling velocity
  measured in the experiments reaches from 106 to
  5.108 K/s
• undercooling about 0,1 to 0,2 Tm.
• ? Include rapid solidification

38
Experiments on mm-size droplets
Film S. Goutier M. Vardelle
Alumina droplet on steel substrate d 5 mm V
10 m/s Ts 400C
39
Thank you for your attention

• Special Thanks to
• Nabil Ferguen SPCTS Laboratory
• Simon Goutier SPCTS Laboratory
• Fahmi Loghmari FAST Laboratory

40
(No Transcript)
41
Isothermal impact of a water droplet
Simulation F. Loghmari
Water droplet spreading d0 2,75mm , V0
1.18m/s on soft wax q (105,95) Rioboo et
al. (2001)
- 41 1 ²