virtual experiments: numerical computations as a powerful

Virtual Experiments: Numerical Computations Virtual Experiments: Numerical Computations as a Powerful Tool for Engineers

P Schmitz* 1 4 A Cockx 2 4 S Geoffroy 3 4 and J Gunther 1 4P. Schmitz* 1 4, A. Cockx 2 4, S. Geoffroy 3 4 and J. Gunther 1 4

1 Biochemical Engineering Dpt., 2 Chemical Engineering Dpt.,

3 Mechanical Engineering Dpt. 4 Université de Toulouse;

INSA,UPS,INP; 135 Avenue de Rangueil, F-31077 Toulouse, France

*Corresponding author: [email protected]

135 avenue de Rangueil – 31077 Toulouse cedex 4, France – Tel : 05.61.55.94.35 - www.insa-toulouse.fr

Presented at the COMSOL Conference 2009 Milan

INSA: National Institute of Applied Sciences

In Toulouse• 2100 Students • 500 Graduates each year 500 Graduates each year • More than 500 Foreign Students • 223 Permanent Academic Staff • 10 Engineering Specialities 10 Engineering Specialities • 9 Research Laboratories

RouenStrasbourg

RouenStrasbourg

Toulouse

Lyon

Rennes

Toulouse

Lyon

Rennes

INSA Network


INSA Network

Virtual Experiments Course

INSA Toulouse Teaching

Secondary Education INSA

Maths,2 3 4 51Year

Physics,Chemistry…

Stage 1 Stage 3Stage 2

BAC

Stage 1Basic

Stage 3Expert

Stage 2Pre-Expert


Bachelor Master

Virtual Experiments

INSA Toulouse Teaching: Course details

Course

1st Year 2nd year Pre-orientation

4th year Specialization

5th year Engineer

3rd year Pre-orientation

Chemical & Env.

Biochem.Scientific Basis

ICBE Chemic, Bio, Environ

Chemical

Biochem.

Civil Eng.

Mechanical Eng.

Systems Eng.

Basis

Humanities

IC Construction Engineering

IMACS

Civil Eng.

Mechanic Eng

Control Elec

TRAI

TRAI

TRAI

TRAI

Control, Electrical

Physics

Computer Sciences

Humanities

Personal

IMACSMaterial, Compon., Syst.

MIC

Control, Elec

Physics

Computers

NING

NING

NING

NING


Math, models

Network, Telec.

Personal Options Modeling, Information

Communication.Math, models

Network, Telec.

Virtual Experiments Course

Objective: To initiate student to multiphysics numerical simulation by approaching concrete cases

Outline: - Quick description of Finite Element Method- 4 projects performed by students (about 10h each)4 projects performed by students (about 10h each)

1 N l

2. Application of buoyancy to airship (Aerospace

Adour Technologie)1. Natural convection in a pan(www.nanoscience.info)

Adour Technologie)

3. Hollow fiber cartridge fil

4. Heat exchanger(www.directindustry.fr)


to filter tap water (Dom source)

2. Application of buoyancy to airshipFA

• Buoyancy force

FG

Buoyancy force • Gravity force due to natural convection

Static of fluid t idoutside

Combined incompressible NS equation and convection diffusion heat transfer equation


inside

Application of buoyancy to airshipz

Quiescent fluid with gravitational

1st step : buoyancy analysis Pressure around a sphere

• Quiescent fluid with gravitational force

• Pressure map in the fluid obtained with N-S equations

npwith N S equations

• Integrate the pressure to obtain the buoyancy force

• Recover the Archimede law by yvarying properties

dSenpF


dSe.npF zA

T t d t li

Application of buoyancy to airship

Temperature and streamlines 2nd step : thermal coupling Flow induced by thermal convection

•Coupling N-S and heat transfer with the Boussinesq approximation •Buoyancy force calculated by surface

Re < 40 pressure integration•Weight calculated by volume integration

Discussion on the available force of

dVg)T(F

Discussion on the available force of the airship


dVg)T(FG

Real Filter Simplified Model

3. Hollow fiber cartridge to filter tap water

• Regular arrangement of cylinders

• The fluid flow is

Hollow fiber module

p

• The fluid flow is calculated on a unit cell of this arrangement

H l’ f f S b di i i iCartridge

Happel’s free surface model *:

• The inner cylinder consists of one single hollow fiber

Sub-divising into sub cells

hollow fiber

• The outer cylinder is a fluid envelope with a free surface used to account for packing

Hollow fiber


2D modelaccount for packing effect

* Happel, Viscous flow relatives to array of cylinders, AIChe, 1959

Hollow fibers

Hollow fiber cartridge to filter tap water

1st step : Fluid flow without particles•Coupling Navier Stokes and Darcy Brinkman model

•Good agreement with experimental results

750

p

•Discussion on flow rate values

DRextz DRextz

h-1.m

-2] 450

600

ow

rate

In / Out filtration

II MM EE LII MM EE L

J [l

.

150

300

Numerical data

Flo/ Out t at o

Out / In filtration


rr

[%]

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.90

Experimental data

Packing density


2nd step : Fluid flow with particles2nd step : Fluid flow with particles

•Coupling Navier Stokes and Darcy Brinkman model

DRextz DRextz

•Particles follow the streamlines

•Using moving mesh ALE to In / Out filtration

II MM EE LII MM EE L

model cake formation/ Out t at o

Out / In filtration


rr


2nd step : Fluid flow with particles2nd step : Fluid flow with particles

At what time should we replace the filter cartridge?

0,12t=10 II

0,12

len

ght

[m]

0,06

0,08

0,10t=20t=30

t len

ght

[m]

0,06

0,08

0,10t=0t=10t=20t=30

t

Fib

re l

0,00

0,02

0,04

(a)

50 40 30 20 10 0

Fib

re l

0,00

0,02

0,04

(c)


Cake height [µm]

50 40 30 20 10 0OO Cake height [µm]

50 40 30 20 10 0OO

4. Heat Exchanger

• heated cylinder cooled by a laminar fluid flow heated cylinder cooled by a laminar fluid flow • external flow normal to the axis of the circular

cylinder

• fluid-thermal coupling :combined incompressible NS equation and convection diffusion heat transfer equation


1st step : circular cylinder in a cross flow,

Heat Exchanger

Re < 40 Re > 40

p y ,isothermal condition

Stagnation point,Rise in pressure

St d fl U t d flSteady flow• Two symmetric vortex

Unsteady flow• Karman vortex street observed with a

repeating pattern of swirling vortices. • Period of the vortex shedding


M. Schäfer & S. Turek, ‘Benchmark computations of laminar flow around cylinder’, E.H. Hirschel (editor), Flow Simulation with High-Performance Computers II, 547–566, 1996.

Comparison with literature results

2nd step : introduction of thermal coupling fl d h t d li d

Heat Exchanger

1.5Temperature and streamlines

flow around a heated cylinder

0.5

1

ln(N

u)

Numerical Data

Hilpert Correlation

01 1.5 2 2.5 3 3.5

ln(Re)

Hilpert Correlation

Re < 40

• Maximum of the local heat flux at the stagnation point

• Total heat flux calculated by integration of the normal local

• Several fluids studied Correlations of the overall average Nusselt

number obtained• Very good agreement between numerical


integration of the normal local heat flux on the whole cylinder Nusselt number

Very good agreement between numerical results and Hilpert Correlation

Conclusions

COMSOL Multiphysics: a useful tool for teachersCOMSOL Multiphysics: a useful tool for teachers

-To illustrate physical phenomenaex: buoyancyex: buoyancy

fluid flow / porous media flow convection heat transfer

To approach techniques of numerical simulation-To approach techniques of numerical simulationgeometry,mesh, boundary conditions, solver, postprocessing

In order to initiate future engineers to


gmultiphysics numerical simulation

Thank youy


Natural Convection in a pan

virtual experiments: numerical computations as a powerful

Documents