Modeling lamps using commercial packages

Modeling lamps using commercial packages

ROUFFET Jean-Baptiste

CPAT – Université de Toulouse, France

COST – Model Inventory Workshop, Funchal, April 2005

CONTENTS

Our modeling of HID lamps

Use of FEMLAB

Why have we choose commercials packages?

Conclusion

COST – Model Inventory Workshop, Funchal, April 2005

OUR MODELING OF HID LAMPS

COST – Model Inventory Workshop, Funchal, April 2005

FLUID MODEL

We have to calculate transport coefficients versus T and p

LTE

We have to calculate discharge composition versus T and p

ASUMPTIONS:•LTE•No demixion•No erosion•Ra=cte•DC

OUR MODELING OF HID LAMPS

COST – Model Inventory Workshop, Funchal, April 2005

We solve equations:

•Energy balance

•Momentum balance

•Current conservation

We calculate and tabulate transport coefficients versus T and p

We calculate discharge composition versus T and p(Tw)

MTDaTa

FEMLAB

We calculate the pressure that we have in the burner versus wall temperature

for a given mixture

OUR MODELING OF HID LAMPS

COST – Model Inventory Workshop, Funchal, April 2005

DISCHARGE

PCA

4PCArad SBT

Anode, we choose V=0

( )( ) ( )

( )( )

2

. 0

.

. .

. 0

m

m m

m p net

u

u u u p g

c u T E U T

V

ìï Ñ =ïïï Ñ =Ñ Ñ - Ñ +ïïïíï Ñ = - +Ñ Ñïïïï Ñ Ñ =ïïî

( )( )

2. .

. 0

m p netc u T E U T

V

ìï Ñ = - +Ñ Ñïïíï Ñ Ñ =ïïî

g

Boundaries conditions given

by the M. Benilov’s team

model

OUR MODELING OF HID LAMPS

COST – Model Inventory Workshop, Funchal, April 2005

Current density Temperature

FEMLAB

COST – Model Inventory Workshop, Funchal, April 2005

PLAN

COST – Model Inventory Workshop, Funchal, April 2005

FEMLAB

COST – Model Inventory Workshop, Funchal, April 2005

FEMLAB

COST – Model Inventory Workshop, Funchal, April 2005

FEMLAB

COST – Model Inventory Workshop, Funchal, April 2005

FEMLAB

COST – Model Inventory Workshop, Funchal, April 2005

FEMLAB

COST – Model Inventory Workshop, Funchal, April 2005

FEMLAB

COST – Model Inventory Workshop, Funchal, April 2005

FEMLAB

COST – Model Inventory Workshop, Funchal, April 2005

TEMPERATURE VELOCITY CURENT (Log)

Why have we choose commercials packages?

COST – Model Inventory Workshop, Funchal, April 2005

Lamps can be considered like a converter of electric energy in radiative energy

Why have we choose commercials packages?

COST – Model Inventory Workshop, Funchal, April 2005

LAMPPOWER SUPPLY

Electric network

CANDELABRAELECTRODEST

radE

T

j

V (or the lamps conductivity)

I

T

OUT PUT OF THE SYSTEM

Cathodique fall

Why have we choose commercials packages?

COST – Model Inventory Workshop, Funchal, April 2005

Simulink is a Matlab toolbox that enable to make model with module

•LAMP MODULE

•BALLAST MODULE

•ELECTRODE MODULE

FEMLAB CAN GENERATE SIMULINK BLOCK AND IS MULTIPHYSICS

LAMPPOWER SUPPLY

Electric network

CANDELABRAELECTRODEST

radE

T

j

V (or the lamps conductivity)

I

T

OUT PUT OF THE SYSTEM

Cathodique fall

Algorithms development are made optimization is made (rapidity)

Conviviality (user interfaces)

User community

Support

Cost

no access to all parameters

+

COST – Model Inventory Workshop, Funchal, April 2005

Why have we choose commercials packages?

-

CONCLUSION

COST – Model Inventory Workshop, Funchal, April 2005

The use of FEMLAB enable us to have quickly results without to develop algorithms and programs

The possibility of convert a FEMLAB model in Simulink is a good things when we want to work with other peoples (collaboration)

We have to add demixion and go toward the 3D

LIMITATIONS We can’t model very high compressible flow (shock tube) 3D (memory problems)

COST – Model Inventory Workshop, Funchal, April 2005

THANK YOU

COST – Model Inventory Workshop, Funchal, April 2005

COST – Model Inventory Workshop, Funchal, April 2005

COST – Model Inventory Workshop, Funchal, April 2005

COST – Model Inventory Workshop, Funchal, April 2005

COST – Model Inventory Workshop, Funchal, April 2005

m

t . mu 0

m

u

t u. u

p mg .u

mcp

T

t u.T

E2 U net . T

. 0V

Les équations utilisées pour modéliser le plasma

Boundary Condition on temperature

Condition on velocity

Condition on potential

q.n=0 with

TuCTq p n.u=0 n.σE=0

T=TB

TB is given by M.Benilov

u=0

n.σE=jB

where jB is given by

M.Benilov

T=TB

TB is given by M.Benilov

u=0 V=0

T=TB

TB is given by M.Benilov

n.σE=0

q.n=0 u=0 n.σE=0

T=Tw n.σE=0

Les conditions aux limites

Le projet NumeLiTe

Evolution of the relative proportion of elements in gas phase versus wall temperature.

Evolution of system pressure versus wall temperature.

For Tw>1090K, we have in gas phase : 62,5% of Hg 18.75% of Tl 18,75% of I

Mercury vaporisation

Thallium iodine vaporisation

Only gas phase

Wall temperature changes system pressure and modifies proportions of each element in gas phase

Calcul des densités des espèces

Tw Psys %Hg700 K 1,83 atm 99,90800 K 2,10 atm 98,88900 K 2,43 atm 93,891000 K 2,93 atm 80,051100 K 3,73 atm 62,501400 K 4,76 atm 62,50

Les coefficients de transport

Le coefficient d’émission net

1.

4N RF

Pour un plasma à l’équilibre thermodynamique local, on peut relier le coefficient d’émission nette à la divergence du flux radiatif:

N T ,Rp L0 T K T exp K T Rp d

0

K étant le coefficient d’absorption monochromatique corrigé de l’émission stimulée

En considérant un plasma homogène et isotherme de rayon Rp, nous pouvons écrire:

Le coefficient d’émission nette

2

.( ) 0

.

. .

. 0

mm

m m

m p net

t

pt

Tc T U T

t

V

u

uu. u g u

u E

.

. 0

PCA PCA PCAm p

PCA

Tc T

t

V

Les équations du modèle

r (m)

r

z

Temperature profile versus radius for different z

Résultats