FLOWNET

version 5.12

U S E R 'S M A N U A L

CONCEPT March 1992

MicroComputer Modeling of Two-dimensional Steady State

Groundwater Flow in a Rectangular Heterogeneous
Anisotropic Section of the Subsoil

H. van Elburg, G.B. Engelen and C.J. Hemker

Developed by H. van Elburg,
G.B. Engelen and
C.J. Hemker,

in cooperation with the Institute of Earth Sciences

of the Free University of Amsterdam, the Netherlands.

Copyright  1984 .. 92 by H. van Elburg, G.B. Engelen and C.J. Hemker

Elandsgracht 83 - 1016 TR - Amsterdam - The Netherlands

 TABLE OF CONTENTS

page

PROGRAM IDENTIFICATION ..................................................................................... 1

USER LICENSE .............................................................................................................. 2
HARDWARE REQUIREMENTS AND INSTALLATION ........................................... 3
CONTENTS OF FLOWNET DISKETTE ....................................................................... 4
AVAILABILITY, FUTURE UPDATES AND SUPPORT ............................................. 5
USER INSTRUCTIONS .................................................................................................. 6
- Before starting the program .......................................................................................... 6
- Starting FlowNet ........................................................................................................... 7
- Directions for use .......................................................................................................... 8
- Data entry ...................................................................................................................... 9
- Page 3: Define model parameters ............................................................................... 10
- Page 4: Hydraulic head data entry ............................................................................. 12
- Page 5: Hydraulic Conductivity data entry ................................................................ 14
- Page 6: Define drawing parameters ........................................................................... 17
- Page 7: Calculation and drawing of the flownet ......................................................... 23
READING AND WRITING A DATAFILE ................................................................... 32
 1

PROGRAM IDENTIFICATION

Program Name: FlowNet version 5.12

Authors: H. van Elburg, G.B. Engelen and C.J. Hemker

Date: March 1992

Language: Turbo Pascal, version 5.5

Computer: Microcomputer with MS-DOS, a minimum of 256K RAM, a

CGA, EGA, VGA, ATT or Hercules graphic board and optionally
a numerical coprocessor

Availability: Both English and French versions of FlowNet can be ordered

from C.J. Hemker. Reduced versions of FlowNet are available for
demonstration, testing and simple modeling purposes.

Abstract: FlowNet is an interactive program for the modeling and

interpretation of two-dimensional steady state saturated
groundwater flow in a rectangular heterogeneous anisotropic
section of the subsoil. The pattern of flow lines and equipotential
lines is obtained by using the finite-difference method and a
regular grid. The highest possible number of model cells is about
5000, whereas the highest possible number of columns or rows is
400. Furthermore, FlowNet indicates time steps, draws
isochrones, and animates these patterns to a lively flow
simulation. These features, in combination with the foolproof
nature of FLOWNET, make it ideal for educational purposes.
2

USER LICENSE

The FLOWNET program is the property of the authors. The purchaser is licensed to use
his copy of the program.

The conditions for use follow those of the IGWMC software. All files on the
FLOWNET diskette may be copied for backup purposes and use within the purchasing
organization only. The authors do not allow copying for commercial purposes and/or
passing the software on to individuals in other organizations. FLOWNET is not
warranted to be completely coincident with this manual.

Support is restricted to written or verbal communication about installation of the

software, but may be extended to answering questions about other aspects of the
program. The authors will not assist users in solving groundwater modelling problems.
Updates will be provided at cost. The program disk is warranted to be readable for one
month after purchase. Should it be unreadable, please return it and we will send a new
one. Never will the authors be liable for any damages arising out of furnishing, use,
abuse or inability to use, or performance of the FLOWNET program package.

The FLOWNET program has been used and tested for years now. No software is
completely free of errors, however. If you should encounter a bug, the authors will
immediately work on it and supply you with a corrected version at no charge. The
authors disclaim all other warranties. It is the user's responsibility to make sure that the
interpretations derived from the use of this software are correct. The author does not
assume any responsibility for results obtained by using this software.
 3

HARDWARE REQUIREMENTS AND INSTALLATION

The FLOWNET program has been developed for IBM PC, IBM XT, IBM AT, IBM
PS/2 models and all other fully IBM compatible microcomputers with graphics
capability. The computer must have at least 256Kbytes of main memory (RAM) and an
IBM CGA, EGA, VGA, Hercules or ATT400 (Olivetti) graphics adaptor. The Operation
System (DOS) must be version 2.0 or later. An 80(x)87 mathematical coprocessor is
recommended to speed up calculations. Plots of generated flownets can be made on an
HP or compatible plotter.

FLOWNET is multilingual and one of the few programs that respect decimal notation
characteristic of a language. Currently, FNINST can generate Dutch, English, French
and German FLOWNET.TXT files. Just type FNINST at the MS-DOS prompt and
choose one of the possibilities offered. When FNINST has run, type FLOWNET and
you will see that its language is the one you chose. Please feel free to write if you have
any remarks or suggestions on grammar, spelling etc.
4

CONTENTS OF FLOWNET DISKETTE

The complete program is supplied on a single diskette and contains the following files:

READ.ME - Optional textfile containing information not supplied

in this manual
FLOWNET.EXE - The main program
FLOWNET.TXT - FLOWNET text file generated by FNINST
FNINST.EXE - Program to install the FLOWNET conversation language
*.BGI - Driver files for various graphics hardware
FEMDMP.COM - CGA/EGA/VGA screendump program for 9- and 24-pin
matrix printers and for DeskJet and LaserJet printers,
originally developed for Micro-Fem users
FEMDMP.DOC - Documentation on FEMDMP.COM
 5

AVAILABILITY, FUTURE UPDATES AND SUPPORT

FlowNet can be ordered from

C.J. Hemker
Elandsgracht 83
1016 TR Amsterdam
HOLLAND

Price of the FlowNet package (January 1992):

Holland Outside the Netherlands

FlowNet DFl. 400 + VAT US$ 200

The demonstration version of FLOWNET is free to be used and copied, but its
application is limited to 200 cells. Applicants should enclose DFl. 25 (US$ 15) for
disk, manual, mailing and handling.

Major revisions of the program will be distributed as versions 5.3, 5.4, etc. Users of
version 5.1 and older will receive update notices describing the (major) changes and
instructions about how to obtain a new version.

Program support is provided by C.J.Hemker.

Fax: Netherlands (+31) - 20 6234628
6

USER INSTRUCTIONS

Before starting the program

While you are working with the program, model results are presented as graphical
output to the monitor. Graphs and related water balance data can be directed to a matrix
printer by means of a single keystroke if some additional program has been run (loaded)
before FLOWNET is started. If your computer uses a CGA graphics card (a low-
resolution screen with 640*200 pixels) or Olivetti (ATT400) graphics (640*400), the
first thing to be done is to run a program called GRAPHICS.COM (supplied on your
DOS disk).
Type: GRAPHICS <return>.

If, however, you are working with an EGA (640*350) or VGA (640*480) graphics card,
the graphics program will probably not work. In that case you can try the FEMDMP
program, supplied on the FLOWNET disk.
Type: FEMDMP <return>.

If this still does not result in a proper screendump, you can think of selecting another
screentype from the main menu (see next section), look for another screendump
program or try another matrix printer.
 7

Starting FLOWNET

FlowNet is started from DOS by typing (capitals or lower case):

FLOWNET <return>.

Note that when you start FLOWNET, the FLOWNET.TXT file and a suitable BGI file
should be present in the current directory.

FLOWNET detects your graphics hardware and adjusts screen scale etc. to it.
If a VGA screen is detected, the program will run in EGA mode.
If autodetection fails, CGA hires graphics will be default.

In case your PC does have higher resolution graphics than detected, for instance your PC
has a EGA adapter but the program produces CGA resolution graphics, it is possible to
override autodetection. Just add a graphics card number when you start FLOWNET
from the MS-DOS prompt:

flownet 1 for CGA graphics ( 640 x 200 pixels ),

flownet 3 for EGA ( 640 x 350 ),
flownet 7 for Hercules ( 720 x 348 ) and
flownet 8 for ATT400 ( 640 x 400 ).

The aforementioned method enables one to see CGA and EGA graphics on an Olivetti
M240 where ATT400 graphics is autodetected. Unfortunately, Hercules graphics cards
seem to have a problem with mixed text and graphics. Therefore, no texts are shown
when flownet drawings are made on Hercules screens. Also, flow rate calculations
cannot be performed on Hercules screens1.
8

Directions for Use

FLOWNET is designed not to crash on error or nonsense user input that put most
programs back to DOS or worse.
Although the program is not warranted to be completely armed against malevolent users
(they are hereby challenged to try and crash it), FLOWNET has reached a high degree of
security against error input and you need not be worried about hitting a wrong key at all.
If a runtime error should nevertheless occur, please report the current state of the
program (page number, current item or activity), a model description or file and the
runtime number to us in any form such that the error is replicable.

Main program control keys are PG UP and PG DN. The program is, like a booklet,
subdivided into several (7) pages. The PG UP and PG DN keys are used to switch
between these pages.
PG DN moves program control to the next page from 1 to 7.
PG UP moves program control one page back.

If PG UP and PG DN do not seem to work properly, press NUMLOCK.

Also, END moves program control to page 7 if all data are defined, and
HOME takes the program from page 7 to the page where END was pressed.
Press ESC to leave the program, then Q.
PRT SC prints the screen (run the MS-DOS program GRAPHICS first).

Of the function keys, only F1 plays a role during data entry. F1 is used for special
functions like storage and retrieval of model data files, printout, function evaluation etc.
as indicated on the lines at the bottom of the screen.
The bottom line or bottom two lines of each page contain further directions for use.
Often, these change with the position of the cursor on the screen. If you are not familiar
with FLOWNET, read them carefully and soon you will no longer need to read them.
 9

Data entry

Pages 2 through 6 use alphanumeric keys for data entry, while BACKSPACE, INS,
DEL, HOME, END and arrow keys function as in regular text editors.
If a string or number is input, HOME takes the cursor to the beginning of the item. END
takes the cursor to the end of the input string. If END is pressed again, program control
will be moved (see above).
INS toggles between Insert and Overwrite modes. Insert mode is default. Overwrite
mode is indicated in the lower right corner of the screen.
BACKSPACE deletes the character to the left of the cursor.
DEL deletes the character at the location of the cursor.
The left and right arrow keys move the cursor across the item.

In addition, the left and right arrow keys are used as toggle switches to change
statements from, for example, "upper side is open" to "upper side is closed" and vice
versa.
If the input item consists of data entry as well as a switch (e.g. equipotential lines on
page 6 ), the cursor must be at the end of the input string before toggling is effective.
Any time numerical data are input, a formula can be input instead of numbers only. For
instance, type 1/3 for 0.333.
Except for the heads formula evaluation on page 4, where x serves as the column or row
index, x as well as y represent the present value of the input item. See "hydraulic head
data entry" for more details.
10

Page 3 : Define Model Parameters

- printout model data - no printer output

This item appears after all data entry pages have been passed once. Use left or right
arrow keys to toggle printout if you want the model data dumped on the printer. Press
F1 to start printing, once printout is enabled.

- read model from disk - save model to disk

This item can be changed after all data entry pages have been passed once. Default at
program start-up is "read model from disk".

- model path- and file name :

Press F1 to load a model data set or save the current model data as specified at "read"
and "save" respectively, above.

- model name :
- model code :
Any two text strings up to 40 characters each can be input here.

- number of columns is
A maximum of 400, depending on the number of rows.

- number of rows is
A maximum of 400, depending on the number of columns.
The number of model cells equals the number of columns multiplied by the number of
rows.

- model length (m) is

- model height (m) is
These two values determine, together with the number of columns and rows, the model
cell geometry.
A too high contrast in cell length and height is not allowed.
This error will be detected when calculations on page 7 start.
 11

- upper side is open - upper side is closed

- left side is open - left side is closed
- right side is open - right side is closed
- bottom side is open - bottom side is closed

These items can be toggled using the left and right arrow keys. A configuration with all
four sides closed, or with two opposite sides open and the other two closed is not
allowed.
12

Page 4 : Hydraulic Head Data Entry

- upper side from left to right

- left side from top to bottom
- right side from top to bottom
- bottom side from left to right
Hydraulic heads are input only along open boundaries. Each open boundary is treated on
at least one subpage of page 4. A maximum of 100 heads is displayed on each subpage
of page 4. If the number of columns or rows is larger than 100, subpages will be
subdivided. PG UP and PG DN or END take program control along the subpages.

The hydraulic head of columns/rows can be a (function) value, functions dependent on

the column/row number x and/or present head y. Press F1 anytime when the program is
on page 4, and control will move to the function entry page.
As elements of a formula, you can use: abs() sqrt() sqr() sin() cos()
tanh() ln() log() exp() pi
As operators, you can use: + - * / ^
where * is the multiply, / the divide and ^ the power operator.
The last closing parentheses of a formula can be left out.
Also, the multiply operator can often be left out, e.g. 2sin(x) instead of 2*sin(x) but
takes highest precedence then, so cos2x equals cos(2x), whereas cos2*x equals
cos(2)*x.

Function formula examples: y+2-sqr(x),

y+0.1,
5,
-y,
x^(1/3),
0.5x,
tanh(x-7),
-cos((x-1)*pi/8)+x/5.
The last formula is of a type made famous by the work of Toth, in his works on regional
groundwater systems. If the constant 8 or 5 in the formula used here is changed, the
frequency or slope of the function changes. If there is more than oneimum in the
 13

hydraulic head, one can usually distinguish so-called saddle points in the flownet
drawing. These are points where no flow occurs, and recognizable because their
surroundings look empty in an average drawing of the flownet. The degree of
coincidence of the saddle points for the flowlines and the equipotential lines patterns
respectively is a good indication for the consistency of FLOWNET's solution for a
specific problem.

- value or function formula is

Here, you can specify a formula or change the default formula.

- starts at column/row nr. x =

- ends at column/row nr. x =
These two indices determine the formula range. It does not matter whether the lowest
index comes first or not.
Press F1 anytime on the function entry page, and the formula will be evaluated and
allocated for the specified range.
Program control moves to the normal page 4 display.
PG UP or PG DN moves program control without the function being evaluated and
allocated.

If an error should occur, FLOWNET reports:

- * error in formula :
with an arrow indicating the approximate error position.
14

Page 5 : Hydraulic Conductivity Data Entry

Here a schematic map consisting of keyboard symbols represents the model and each
symbol represents a model cell. The default model map consists of the symbol x only. A
blinking cursor can be moved across this map to indicate individual cells. The cursor
position (column 1, row 1 upon entry) is displayed at the top and right side the map:
1
1 xxxxxxxxxxxxx
xxxxxxxxxxxxx etc.

Wherever you move the cursor, it keeps its position so when you enter page 5 again, you
will find the cursor where you left it.
At the left side of the screen, the symbol, conductivities and effective porosity of the
cursor model cell are displayed:
- symbol
- conductivity (m/day)
horizontal =
vertical =
effective porosity =

Below that, the symbol list ("library") is displayed:

- symbol list:
The default model's list consists of an x only.

Then, the directions:

- move cursor with arrow keys

If the model is larger than 40 columns wide or 12 rows high, part of the model is
displayed. The map will move if the cursor is moved to the edge of the map drawing.
 15

- alphanumeric keys extend the symbol list

If your model is inhomogeneous, press any alphanumeric key (but x) and the screen will
change to conductivity entry mode. During conductivity input, the cursor can be moved
across the input items as on page 3 and 4. Hit RETURN or up/down arrow keys to move
the cursor off the data input list. Then, the map will be restored and the cursor will
reappear in it.
Please note that uppercase and lowercase keys yield different symbols.

- allocate conductivity to cursor by pressing a listed symbol or F1

This option is shown when the number of different symbols is larger than one.
Press the above alphanumeric key again, and the cursor will blink faster indicating
conductivity allocation mode. Wherever the cursor is moved, the model cells will get
the new symbol conductivity and the symbol changes to the cursor symbol.
Directions for use then read:

- * use cursor to change cells

- * RETURN and F1 stop cell allocation
To toggle change mode off, press either RETURN or F1. Then, the cursor can be moved
across the map again without changing cells. F1 toggles change mode on and off,
allocating the present cursor cell conductivity to the cursor when switched on.

- RETURN to edit current conductivity

Press RETURN to switch to conductivity entry mode for the
current cursor symbol:
- symbol
- conductivity (m/day)
horizontal =
vertical =
effective porosity =
Only horizontal and vertical conductivity determine, with the model geometry, the flow
rate between grid points. Effective porosity is used for drawing time steps and
isochrones (see pages 6 and 7). If calculations are made, and effective porosity only is
changed, new matrix calculations need not to be made; press END (or PG DN twice)
and FLOWNET will start drawing immediately.
16

- layered cells: Khor = Σ(kd) / D, Kvert = D / Σ(d/k)

Often, model cells are thicker than the strata they represent. These are the formulae to
calculate the horizontal and vertical conductivities of layered cells where D = Σd is the
height of the cell, d the height of each layer and k the conductivity of each layer. This
feature is called macro-anisotropy.

- topography: use : or ; with low Kh and high Kv (no draw for ; :)

Representation of topography is difficult in an obligatory rectangular model.
This way works quite fine: if some part of the model consists of "air", apply cells with
high vertical conductivity and low horizontal conductivity such that the head values
along the boundary will be transferred deeper into the model.
For cells with symbols ";" and ":" no drawing will be made which makes these symbols
appropriate to be used as "air" symbols.

- cursor row/column copied if rows/columns added (on page 3)

If you want to add rows or columns, you might wish to insert copies of a certain existing
row or column. Just move the cursor there, switch back to page 3 (PG UP) and increase
the number of columns or rows. Switch down to page 5 again and you will see that the
new rows or columns are copies of the row or column where you left the cursor.
If you wish to delete rows or columns, this is also done at the position of the cursor.
FLOWNET will delete the cursor column and the columns to the right of it if the
number of columns is decreased, and the cursor row and the rows below if the number
of rows is decreased.
 17

Page 6 : Define Drawing Parameters

- iteration stop -log(r/r0) =

This number determines to a certain degree the level of precision for the approximate
iteration method. Generally, the higher the better but too high is quite useless and takes
more time. Sensible ranges are 3 to 15. r0 stands for the norm of the boundary vector b
of the matrix equation Ax=b to be solved. The iterative process breaks the boundary
vector b down with r being the norm during iterations. Once r is small enough, -log(r/r0)
becomes larger than the iteration stop. The iterations terminate and the drawing can be
made.

- information and stop after iteration

- iteration graph only
If the first possibility is chosen, the iterative process screen will be illustrated with text
and the final error calculated and displayed. The program will halt and wait for a key to
be pressed before the drawing is made.
If the second possibility is chosen, the iteration graph only will be shown during
calculations, and the drawing will be made immediately thereafter.

Note: the above item appears only if the calculations are not yet made.

- no preview / immediate calculation

- vertical/horizontal scale is
This item offers the choice of viewing the model before calculations start. If you choose
to see a preview, you can vary its scale (see below). When the preview is being shown,
press any key to continue or PG UP to return to page 6.

Note: the above item appears only if the calculations are not yet made.

- vertical/horizontal scale is
- scale is fixed
Often, models are elongated and their images need vertical exaggeration. Use the right
arrow key to fix the scale of the drawing. If the scale is fixed, however, and the number
of model cells on display is increased such that the drawing would become too big for
18

the screen, FLOWNET switches fixed scale off automatically and adjusts display scale
to screen size.

Note: the above item appears only after the model calculations are made.

- flow rate calculations

- no flow rate calculations
Switch flow rate calculations on to calibrate your model (see page 7).
Note: Flow rate calculations area and results cannot be plotted.

Note: the above item appears only after the model calculations are made.

- ( A + B )- number of user screens is

This item controls animations and page 7 image storage. A and B are two numbers.
The sum of A and B, n is the total number of screens available. The number of user
screens cannot exceed n or 10.
The number of screens used for animation equals A + B minus n. A user screen is used
to store a FLOWNET image. If you want to store images, increase the number of user
screens if it is zero. The function keys F1 to Fn indicate the user screens in memory
(next page).

Use CTRL -F1 to CTRL- Fn to store a screen into memory

SHIFT-F1 to SHIFT-Fn to recall a screen from memory
ALT -F1 to ALT- Fn to exchange monitor and memory screen
F1 to Fn to combine monitor and memory screen.
When calculations are made, however, B stored images are lost and only the first A
screens are kept in memory.
The numbers A and B depend on available memory, graphics hardware and number of
columns and rows. Mostly, at starting up A is 9 to 12 and B is zero. When the number
of model cells increases, A decreases and B increases.
Once the difference between ( A + B ) and number of user screens is 3 or larger,
drawings of time steps and isochrones will be made such that animation is enabled. The
free screens serve to store the various screens shown when animation is in progress.
Again, time steps and isochrones will be animated if the number of user screens is less
than the total ( A + B ) minus two.
 19

Note: If you do not want isochrones or time steps animated, increase the number of user
screens to ( A + B ) or 10 even if you do not want to store images.

- left border of drawing at column

- right border of drawing at column
- upper border of drawing at row
- lower border of drawing at row
These items allow you to display any rectangular part of the model.
- equipotential lines interval (m) is
- equipotential lines are not drawn
The equipotential interval will determine the number of equipotential lines in the
drawing. If for instance 0.2 is specified and the heads range is from 320.1 to 321.02 m,
the 320.2, 320.4, 320.6, 320.8 and 321.0 equipotential lines will be drawn.
The right arrow key toggles equipotential lines drawing off.

- streamlines will be drawn - streamlines will not be drawn

Use left or right arrow keys to toggle.
- number of streamlines is
This item can be changed if streamlines and/or time steps are drawn.
The drawing is made such that the discharge of each stream lane (area between stream
lines) is equal (except for the innermost and outermost stream lanes which have half the
discharge of the other stream lanes). This comes in very handy if flow rates are to be
determined, for it is sufficient to determine the flow rate of one stream lane only to
quantify flow rates in the entire model.

- time steps are drawn - time steps are not drawn

Time steps are represented as square dots along streamlines.
The number of streamlines above also determines the streamlines along which time
steps are drawn. The distance between two dots along a flowline is the distance of travel
during the time step (see below at "- time step (years) is").
Each dot represents an equivalent area ("volume of water") in the model.
However, flow between grid points is controlled by average conductivity, ICCG(0) is an
approximate method (although a good one). Most importantly, the time step flowline
and driving potential difference are determined by linear interpolation between grid
20

points. Therefore, the statement "Each dot represents an equivalent area" only holds
approximately true in regular parts of the model.
For models with a complex "geology" or subtle head differences, another consequence
of this linear interpolation is that FLOWNET might not draw time steps along all flow
lines successfully, and within some cells the number of time steps might be incorrect.

- time step (years) is

Here, the time interval for time steps is specified.
This is also the time step interval for isochrones (see below).

- output to plotter disabled

Use left or right arrow keys to toggle plotter output on:
- plotter (if used) at :
COM1 is the serial port of your PC. It is usually necessary to run the MSDOS program
MODE first to enable communication between your computer and plotter (see your
MS-DOS manual, subject MODE. You might want to put the MODE command in an
AUTOEXEC.BAT file). If a non-port name is used, a file with plotter instructions will
be generated and saved to disk without further notice.
If communication between computer and plotter fails, FLOWNET reports this error and
makes the drawing on the monitor only. If animation is enabled, the first display only
will be plotted. Switching plotter output on or off will change the screen lines below to
normal or plotter menu, respectively.

Normal menu items include:

- isochrones are not drawn

Use left or right arrow keys to toggle isochrone drawing on.

- number of isochrone flowlines is

Isochrones connect points of equal residence time. The number of isochrone flowlines
determines the precision with which isochrones are drawn. A too high number is
useless. As a rule of thumb, use 3 to 15 times the number of flowlines.
Isochrone drawing can be switched off with the right arrow key.
- isochrone distance (screenpoints) is
 21

Often, isochrone points (time steps) along different flowlines move further apart in
time.This number specifies the maximum distance between isochrone points to be
connected with a line in the drawing.

- conductivity borders are drawn - conductivity borders are not drawn

If these are drawn, the boundaries between regions of different conductivity will be
shown as solid lines. Needless to say, the default model has no conductivity boundaries.

- head distribution is drawn - head distribution is not drawn

If the head distribution is drawn, a schematic display of the heads along open boundaries
will be visible along the model boundaries.

- gridpoints are drawn - gridpoints are not drawn

Little dots might indicate the position of the grid points.

Plotter menu items read:

- (horizontal) model:plotter scale 1 :

The scale factor is from meters (model) to pixels (plotter) and depends on the size of
your model and the pixel range of your plotter. The plotter is controlled in its own
coordinate system. See your plotter handbook for the maximum coordinate values, you
need them also to determine the plotter start x and y coordinates (below).
- delay between plotter commands (ms)
To prevent plotter input buffer overflow, a delay is made each time a plot command is
sent.
- plotter time step box size
The size of the square box plotted to indicate a time step.
- plotter start x coordinate
- plotter start y coordinate
These numbers allow the plot not to be made in the ultimate corner of the plotter paper.
22

Page 7 : Calculation and Drawing of the Flownet

First, the equipotential lines equations are built and solved, then the flowlines equations.
These equations each yield a matrix equation Ax=b which is solved with the ICCG(0)
iterative method.
- condition code
Before calculations start, the matrix equation is examined and the condition code
calculated. If this code is high, ICCG(0) will have a tough time. If this code is higher
than a certain level determined by the "iteration stop" (page 6) a warning will be given:
- note: iteration stop too low
- (erroneous output ?)
Output might be not according to the input data. If this code is higher than a certain
level depending on numerical precision, calculations will be aborted and messages
displayed:
- * * CALCULATION ABORTED : * *
- ** CELLS ARE TOO ELONGATED **
or
- ** CONDUCTIVITIES TOO EXTREME **
Then, FLOWNET will switch to page 3 or 5 after some seconds or when a key is
pressed.

If "information and stop after iteration" is chosen, the number of the iteration is output,
and the current error level, and the program will halt after calculations until a key is
pressed. Thereafter, the drawing will be made according to the specifications of page 6.
 23

Flow rate calculations

First, the drawing is made. Wait until the flow rate calculations messages appear:

- add cells with arrow keys

Here you will move a "cursor" (dotted square) across the model while the water balance
for the dotted area is calculated and displayed.

- flow rate into shaded area:

- from above
- from the left
- from the right
- from below
For example, "from the left" means: across all left boundaries of the dotted area, into the
dotted area.
- error =
Here, the imbalance is displayed. Since the heads used for water balance calculations
(cell centered) are derived from the drawing matrix (grid centered) an imbalance always
exists (especially along open boundaries and along conductivity boundaries). This need
not be a handicap, since all stream lanes have equal discharges and it is sufficient to
determine the discharge of a single stream lane. Remember, however, that the innermost
and outermost lanes have only half the discharge that the other stream lanes have.

- RETURN reduces block size to 1

Press RETURN and the dotted area will be reduced to one cell at the current cursor
location. The dots of vanished water balance cells are blackened. This might damage the
drawing of the flownet somewhat. If PG DN is pressed after the ultimate flow rate
calculation area is determined, the pattern is drawn again.
24

Drawing of Isochrones

For isochrones, an array with coordinates needs to be stored in memory by the program.
If the number of isochrone links to be kept is too big (larger than 8000) FLOWNET
might not be capable of finishing the isochrones pattern correctly. A warning is output
on the bottom of the screen if this occurs:
- * isochrones memory overflow *
To prevent this type of error, you might either change the number of isochrone flowlines
or increase the time step.

Animation

If the difference between number of user screens and total number of screens is three or
larger, and time steps and/or isochrones are drawn, FLOWNET will animate flow
behaviour. In this case, wait until this message appears:

- animation fast / slow : press +++ / ---

Press either "+" or "-" and the show will start.
Subsequent pressing of "+" will speed up animation until delay between screen changes
is zero and "-" will slow down animation. If flow rates also are determined, press PG UP
to switch to flow rates calculations. Thereafter, press PG DN or END and the screen
will be refreshed for animation again.
 25

Reading and writing a datafile

When a set of data has been entered or modified, it can be saved as a datafile.
Select page 3 and set line 2 to 'save model to disk'. Enter path and filename at line 3 and
press F1 to start the saving procedure. If the input of data has not been completed it
cannot be saved.

A file can be read by setting line 2 to 'read model from disk', entering the required path
and filename at line 3 and pressing F1. There is no option to produce a list of all
FlowNet files in a directory at the screen.