Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
76 views

Lecture 05: Modelling Advection (Coupled Water and Heat Flow)

1. The document discusses numerical modeling of coupled water and heat flow through porous media, known as advection. 2. It presents the heat balance and water balance equations, and describes how saturated hydraulic conductivity depends on temperature via fluid viscosity. 3. Upwind differencing is used to model advection, but this introduces numerical diffusion that increases with smaller time steps. Alternative techniques like marker-in-cell are proposed to better capture advection.

Uploaded by

Maria Tabares
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
76 views

Lecture 05: Modelling Advection (Coupled Water and Heat Flow)

1. The document discusses numerical modeling of coupled water and heat flow through porous media, known as advection. 2. It presents the heat balance and water balance equations, and describes how saturated hydraulic conductivity depends on temperature via fluid viscosity. 3. Upwind differencing is used to model advection, but this introduces numerical diffusion that increases with smaller time steps. Alternative techniques like marker-in-cell are proposed to better capture advection.

Uploaded by

Maria Tabares
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 14

Lecture 05: Modelling Advection (Coupled water

and heat flow)


Timo Heimovaara

Section Geo-Engineering, Delft University of Technology

CIE4365 Period 4, 2017

1 / 14
Outline

2 / 14
Heat Balance equation

bw T
+ qh = 0
t
with
qh = qDiff + qconv = T + vT
where v is the velocity of the water in the pores (v = qw /w ).
Rewriting in 1D gives:

bw (w )T
= (qh )
t z
T bw
bw (w ) +T = (qh )
t t z
so:
T bw
bw (w ) = (qh ) T
t z t

3 / 14
Water Balance equation

w
+ qw = 0
t
with
qw == Ksat krel (w ) (hw + 1)
The saturated hydraulic conductivity is a function of temperature:
sat
K sat =
w g(T )

4 / 14
Temperature and Viscosity
Table: Dynamic viscosity as a function of temperature

T(o C) (mPa/s)
0 1.787
5 1.519
10 1.307
20 1.002
30 0.798
40 0.653
50 0.547
60 0.467
70 0.404
80 0.355
90 0.315
100 0.282

5 / 14
Advection and diffusion

1 Advection: Heat moves because phase (water) is flowing;


2 Diffusion: Heat moves because of a gradient in temperature;
Similar for solute transport. Why?

6 / 14
Eulerian advection methods


= vx
t x
upwind
discretization
t
it i1
it+t = it vx t
x
stability
criterium:
x
t
vx

7 / 14
Numerical Diffusion increases with smaller t

8 / 14
Exposing the diffusion term in upwind differences

it i1
t
it+t = it vx x t can be rewritten as

vx t
it+t = it (2it 2i1
t t
+ i+1 t
i+1 ),
2x
or

!
it+t it t i1
t
vx x t
i1 2it + i+1
t
= vx i+1 + .
t 2x 2 x 2
Numerical diffusion term in upwind differences

!
it+t it t
t i1 vx x t
i1 t
2it + i+1
= vx i+1 +
t 2x 2 x 2

simplifies to :
   central  2 

= vx +D
t i x i x2
 central t t  2  t 2 t + t
where x = i+12xi1 and x2 = i1 xi2 i+1
i i
and D = vx x/2
other types of differencing no solution
Alternative Techniques
Marker in Cell

Mixed Eulerian and Lagrangian algorithm


Advection is solved by allowing Lagrangian particles to move
along flow field;
Diffusion and reactions are solved using Eulerian approach
This a so-called operator splitting technique ...
Cannot make use of ODE solvers of MATLAB ... :-(

13 / 14
Marker in Cell Algorithm

1 Distribute a (large) number of markers over across your Eulerian Grid;


2 Use an interpolation technique to initialize the temperatures of each individual
marker;
3 Use the information from the velocity field to calculate the velocity of each
individual marker;
4 Move the markers using the time step xAt+t = xAt + vxA t (but you can also
use a Runge Kutta method!), recycle markers that have left the model
domain;
5 Calculate the averaged temperature at the nodes using the advected
markers;
6 Use the averaged temperature to calculate the change in nodal temperatures
due to diffusion
7 Update temperatures of the markers by interpolating the temperature
difference
8 repeat from step 3.
9 A detailed explanation can be found in Chapters 8 and 10 of Gerya, 2010.

You might also like