                                ELCUT
                    Field Analysis Software for PC

                             Version 3.0C
                             User's Guide


Copyright TOR Cooperative Enterprise, 1989-1994
All rights reserved.

TOR Coop
6 Uglovoj per., St.Petersburg, Russia
Internet: ELCUT@tor.spb.su
Voice:    +7 812 292 1965
Fax:      +7 812 110 1334

ELCUT is a trademark of TOR Cooperative Enterprise.
286, 386 and 486 are trademarks of Intel Corporation.
IBM PC/AT and PC-DOS are trademarks of International Business Machines 
Corporation.
Microsoft, Microsoft Word, MS-DOS and Windows are registered trademarks 
of Microsoft Corporation.



1.	Introduction


Welcome to ELCUT FEA System.

In this chapter, you will find:

*	A description of ELCUT system.
*	ELCUT hardware requirements.
*	An explanation of how to use this manual.


1.1.	What Is ELCUT Software?

ELCUT is PC-oriented interactive environment for electromagnetic, 
thermal and stress analysis. Standard analysis types include:

*	Electrostatics.
*	Linear and nonlinear magnetostatics.
*	Linear and nonlinear heat transfer and diffusion.
*	Linear stress analysis.

During a 15-minute session, you can describe the problem (geometry, 
material properties, sources and other conditions), obtain solution with 
high accuracy and analyze field details looking through full color picture. 
With ELCUT, complicated field problems can be solved at your PC 
instead of large mainframes or workstations. At usual PC with standard 
configuration, ELCUT is able to solve problems of electrostatics, 
magnetostatics and heat transfer with up to 500 nodes (shareware version)
or 100,000 nodes (professional version), and for stress analysis up to
250 nodes (shareware version) or 50,000 nodes (professional version.)


1.2.	Required Hardware Configuration

Computer:
A 286, 386 or 486 based IBM PC/AT compatible.

Coprocessor:
Intel 80287 or 80387 math coprocessor.

Memory:
640K minimum, 2MB recommended (1MB extended), 
4-8MB recommended for very huge problems.

Display:
EGA, VGA color or LCD monochrome.

Mouse:
Microsoft mouse or 100% compatible.

Operating System:
PC-DOS or MS-DOS (Rev. 3.3 or above).


1.3.	Keyboard Formats

In this manual we use CAPITALS to specify the names of keys on 
your keyboard. For example, ENTER, ESC, or ALT. Four arrows on the 
keyboard, collectively named the DIRECTION keys, are named for the 
direction the key points: UP ARROW, DOWN ARROW, RIGHT ARROW, and 
LEFT ARROW.

A plus sign (+) between key names means to hold down the first key 
while you press the second key. A comma (,) between key names means 
to press the keys one after the other.


1.4.	How to Use this Manual

This manual is divided into seven chapters:

Chapter 1, "Introduction", briefly discusses ELCUT capabilities.

Chapter 2, "Theoretical Description" contains mathematical formulations 
for all problem types that can be solved with ELCUT. Read this chapter 
to learn if ELCUT can solve your particular problem.

Chapter 3, "Getting Started", describes first steps of using ELCUT. In this 
chapter, you will learn how to install and start the package.

Chapter 4, "Basic Skills", tells you about working patterns with ELCUT.

Chapter 5, "Model Geometry Definition", explains how to describe 
geometry of the model, build the mesh, and define material properties 
and boundary conditions.

Chapter 6, "Problem Parameters Description", introduces non-geometric 
data file organization, and the way to attach this file to the model.

Chapter 7, "Obtaining and Analyzing the Results", describes how to start 
the solver and introduces ELCUT Postprocessor, its features and 
capabilities.

Appendix A, "Example Problem from Start to End", contains full 
step-by-step instructions how to solve the real problem from scratch.


2.	Getting Started


2.1.	Installing ELCUT

After unzipping (don't forget -d switch to restore directory
structure) ELCUT directory should contain:
eight ASCII files:
FILE_ID.DIZ  - descriptor;
README.TXT   - primary information;
MANUAL.TXT   - system documentation;
REG_FORM.TXT - registration form;
FREE_REG.TXT - alternative way of FREE registration;
VENDOR.TXT   - information for shareware distributors;
FAQ.TXT      - frequently asked questions about ELCUT.
PACKING.LST  - packing list
and two subdirectories:
BIN          - contains all the files needed to run ELCUT;
EXAMPLES     - contains example problems.

The BIN directory contains all the necessary to run the package. These 
files are to be copied to the special directory on the hard disk, for 
example, C:\ELCUT\BIN. For your convenience, we recommend to 
add the name of this directory to DOS PATH environment variable.

The EXAMPLES directory contains a number of problems solved with 
ELCUT. We suggest you to look through the examples in the area of 
your interest, before you start with your own problems.
ELCUT.BAT starts the system from the EXAMPLES directory.

To operate with large amount of data, ELCUT creates temporary file in 
the current working directory. You can specify alternate place for this 
file by assigning path of the preferred directory to ELCUTTMP 
environment variable, e.g., to store temporary files in the C:\TEMP 
directory, you could include SET ELCUTTMP=C:\TEMP line in 
your AUTOEXEC.BAT. We highly recommend to have enough free 
space for the ELCUT working files. About 1.3 MB is required for every 
ten thousand nodes.

ELCUT can use the extended memory of PC. This feature is available 
through the XMS driver, e.g., HIMEM.SYS distributed with 
MS-DOS 5.0 and Microsoft Windows 3.x.

        Note. ELCUT uses part of the extended memory that is 
        not occupied by RAM disks, disk caching systems or 
        any other resident program.

The BIN directory of ELCUT distribution kit contains four files with 
.CFG extension. These files contain optional color tables:

SCREENC.CFG	-	color table for VGA and EGA color monitors;
SCREENL.CFG	-	color table for LCD monitor of Laptop;
SCREENG.CFG	-	gray scale color table, can be used when 
                        obtaining screen hard copy on a monochrome 
                        printer;
SCREEN.CFG	-	default color table, the same as 
                        SCREENC.CFG.

You can change the default by replacing SCREEN.CFG on your hard 
disk with a copy of the color table you prefer.


2.2.	Starting ELCUT

To start ELCUT, go to the directory you reserved for this work and 
enter ELCUT at your system prompt. The command line may include 
the problem file name. If it does not, the name is taken from 
ELCUT.INI file of the previous session. If ELCUT.INI is absent in the 
current directory or does not contain the problem file name, ELCUT 
will ask for the name of the problem to work with.


2.3.	Quitting ELCUT

To exit from ELCUT to operating system environment, choose Exit in 
the Files menu (ALT+F, X) or press ALT+F4.



3.	Basic Skills


This chapter explains common working patterns you will use with ELCUT.
ELCUT is a menu driven system. The meaning of the selected menu 
item is explained by the prompt message occupying the bottom line of 
the screen. The same line is used for other messages.

In most context the ESC key may be used to cancel or to interrupt the 
current action. The right mouse button is completely equivalent to the 
ESC key. The left mouse button is used to click objects.


3.1.	Terminology

The following terms are used to describe your actions when working 
with ELCUT.

Choose	-	To use a mouse or key combination to pick an item that 
                begins an action. For example, choosing a menu item 
                usually causes the execution of ELCUT command.

Click	-	To press the mouse button while the tip of the mouse 
                pointer rests on the item of choice.

Select	-	To mark an item by highlighting it with key combinations 
                or by clicking it with a mouse. Selecting does not initiate an 
                action.


3.2.	Working with Menus

To choose a menu item click it with a mouse. You can also use the UP 
and DOWN ARROW keys to select the item you want; then press ENTER. 
If the item name has an underlined letter, you can type it to choose the 
menu item with one step. To select an item on the horizontal menu bar 
press its underlined letter while holding down the ALT key.

Pressing the ESC key or clicking the right mouse button returns to the 
previous menu level. If you press ESC while in main menu, you will be 
asked about exiting to DOS.


3.3.	Working with Dialog Boxes

ELCUT uses dialog boxes to request information from you and provide 
information to you. For example, when ELCUT needs additional 
information to carry out a command you have chosen, a dialog box 
requests the information. You complete the dialog box by providing the 
missing information. Whenever you see an ellipsis (...) after a menu 
command, another menu or a dialog box follows.

For example, when you choose Open from the File menu, ELCUT 
displays a dialog box asking for the filename of the file you want to open.
 
Most dialog boxes contain options, each asking for a different kind of 
information. After you supply all the requested information, you choose 
a command button to carry out the command.

Often you need to move around within a dialog box to make several 
selections. The current option is marked by a highlight or dotted 
rectangle (or both) around the name of the option or button. To move 
within a dialog box:

*	Click with a mouse the option you want to move to.
*	Press TAB to move forward (generally from left to right and top 
        to bottom) or SHIFT+TAB to move in opposite direction.
*	You can use the DIRECTION keys to move in desired direction.
*	Or, while you hold ALT, you can type the underlined letter in the 
        option name or group name.

The options that are unavailable for some reason are dimmed.

The next few sections describe each kind of dialog box option and how 
to move and select options.


3.3.1.	Command Buttons

Command buttons initiate an immediate action. One command button 
in each dialog box carries out the command you choose, using the 
information supplied in the dialog box. This button is usually named 
OK. Other command buttons let you cancel the command or choose 
from additional options.

Command buttons marked with an ellipsis (...) open another dialog box 
so you can provide more information. The currently selected, or default, 
button has a highlighted green name or, in a monochrome mode, a 
darker border than the other buttons. You can choose the selected 
button by pressing ENTER.

You can close the dialog box without completing a command by 
choosing Cancel button.

To choose a command button:

*	Click it with a mouse.
*	Move to the command button you want. A dotted rectangle 
        around the button text marks the selected button. Press the 
        SPACEBAR (or ENTER) to choose the button.
*	Or, while you hold ALT, you can type the underlined letter in the 
        button name.

One command button in each dialog box carries out the command you 
choose, using the information supplied in the dialog box. This button is 
usually named OK. Other command buttons let you cancel the 
command or choose from additional options.

Some dialog boxes are so small that does not contain any command 
button. Such dialog boxes are usually arise at the right-hand side of the 
screen and have gray background. In spite of missing the OK command 
button, it is still possible to use a mouse to carry out the command you 
choose. Click the dialog box background anywhere outside options. The 
effect will be the same as if using the OK command button.


3.3.2.	Text Boxes

A text box is a rectangle into which you type information.
When you move to an empty text box, a text cursor appears at the left 
side of the box. The text you type starts at the cursor position.

If the box already contains text when you move to it, all the text in the 
box is automatically selected and any text you type replaces it. Or, you 
can erase the existing text by pressing DEL. To discard the selection 
simply move the cursor to the point where you want to enter or erase 
text. Use LEFT and RIGHT ARROW, HOME or END keys to move the cursor.

The text exceeding the length of the text box is scrolled automatically.


3.3.3.	List Boxes

The list box shows a column of available choices. If there are more 
choices than can fit in the list box, a scroll bar is provided so that you can 
use your mouse to move up and down quickly through the list.

To scroll one line click one of the scroll arrows. To scroll one window up 
or down click the gray background of the scroll bar above or below the 
white rectangle.

When the required item is already visible in the list box, you can select it 
by clicking it with a mouse. Clicking the previously selected item is 
equivalent to pressing the ENTER key.

To select an item using a keyboard press UP or DOWN ARROW key until 
you reach your choice. You also can use PAGE UP and PAGE DOWN to 
move one window a time, and HOME or END to move to the first or to 
the last item of the list. Or, type the first letter of the item you want, the 
highlight will be moved to the first item that starts from that letter.


3.3.4.	Drop-Down List Boxes

A drop-down list box appears initially as a rectangular box with the 
current choice (default) displayed in the box. The arrow in a square box 
at the right opens into a list of available choices when you select it. If 
there are more choices than can fit in the drop-down list box, a scroll bar 
is provided.
 
A selected drop-down list box can be opened without a mouse by 
pressing ALT+DOWN ARROW. An opened drop-down list box is closed 
when you select an item in it, select other option in the dialog box, or 
press ALT+DOWN ARROW.


3.3.5.	Option Buttons

Option buttons appear in dialog boxes as a list of mutually exclusive 
items. You can select only one option from the list at a time. You can 
change a selection by selecting a different button.

The selected option button contains a black dot.
 
To select an option button:

*	Click it with a mouse.
*	Press TAB to move into the option group you want; then use the 
        DIRECTION keys to select the option button you want.
*	If the option name contains an underlined letter, you can hold 
        down ALT and press the underlined letter from anywhere in the 
        dialog box to select an option button.


3.3.6.	Check Boxes

Check boxes offer a list of options you can switch on and off. You can 
select as many or as few check box options as are appropriate. When an 
option in a check box is selected, it contains X. Otherwise, the box is 
empty.
 
To select or clear check box options:

*	Click with a mouse each empty check box you want to select. 
        Click a selected box again to clear the selection.
*	Press TAB to move to the empty check box you want to select. 
        Press the SPACEBAR to enter an X. Press the SPACEBAR again if 
        you want to clear the selection.
*	If the check-box name has an underlined letter, hold down ALT 
        and press the underlined letter for each check box you want to 
        select or clear.


3.4.	Selecting the Geometric Objects

When editing the model geometry or analyzing the results you may need 
to enter the coordinates of a point. The plus sign cursor arises to 
indicate the point locating mode. This cursor can be moved by mouse or 
using the DIRECTION keys. The HOME, END, PAGE DOWN and PAGE UP 
keys move the cursor in four diagonal directions. You can control the 
keyboard cursor step by the MINUS and PLUS keys. The MINUS key 
approximately halves the cursor step, the PLUS key increases it back. 
You also can get very fine cursor movement by holding down CTRL 
while pressing the DIRECTION keys.

To select a point of the model move the cursor to the position of choice 
and click left mouse button or press the ENTER key. ESC or the right 
mouse button cancels the operation. If you prefer numerical form of the 
coordinates input press TAB and you will get a dialog box with two text 
boxes for coordinates typing. Press ENTER or click the dialog box 
background to complete the dialog and carry out the command.

When working with the model you often need to select some of its 
constituent geometric objects. The picking mode is indicated by the X-
shaped cursor. You can move this cursor the same way as while locating 
a point. To pick a massive object like a block place the center of the 
cursor on that object and click the left mouse button or press the ENTER 
key. To pick a vertex or an edge it is not necessary to point cursor 
exactly on the object. The selected object is always the closest to the 
cursor.


3.5.	Using the Rubber Band Rectangle

The rubber band rectangle is used to specify the rectangular part of the 
model or of the X-Y plot, which is to be enlarged to occupy all available 
screen area in the result of zoom-in operation. The rubber band 
rectangle is controlled using a mouse or the DIRECTION keys. First, posit 
the lower left corner, then the upper right one.



4.	Problem Description

4.1.	Structure of Problem Database

A special database is built for each problem solved with ELCUT. The 
kernel of the database is the problem description, which is stored in file 
with the extension .PBM. The problem description contains the basics of 
the problem: its subject, plane, precision class, etc., and also references 
to all other files, which compose the problem database. These files are 
the model file, with standard extension .MOD, and physical data files, 
with extension .DES, .DMS, .DHT, or .DSA, with respect to the subject 
of the problem.

The problem description may refer to one or two physical data files. 
Both files have the same format, and differ only in purpose. Usually, the 
first data file contains specific data concerning the problem, as the 
second file is a library of standard material properties and boundary 
conditions, which are common for a whole class of problems.

Owing to the problem is distributed among several files, you may use 
single model file or single data file in several similar problems at once.

While solving the problem, ELCUT creates one more file-the file of 
results with the extension .RES. This file always has the same name with 
the problem description file, and is stored in the same directory.


4.2.	Creating New Problem Description

To create new problem description, choose New in the Files menu 
(ALT+F, N), and then enter name for new problem. While created, new 
problem inherits settings of the preceding problem. To change these 
settings, choose Problem in the Edit menu (ALT+E, P). The dialog box 
appears, containing problem description parameters. The problems of 
time-dependent heat transfer have additional parameter set. To edit 
these values, choose the Time Params button. To exit from problem 
description editing, choose OK. You can cancel editing by choosing 
Cancel button, or pressing ESC, or clicking right mouse button.
 
Choosing the Browse button allows you to select the file from the list of 
files and directories when defining the model or data filename. The 
button acts on that type of file, which box is currently selected.

When the file chosen, you can immediately edit that file by choosing the 
Open button. It acts upon currently selected file.


4.2.1.	Choosing the Length Units

ELCUT allows to use different metric units for coordinates when editing 
model's geometry. You can use millimeters, centimeters, meters and 
even kilometers. To set the units, choose Length Units in the Options 
menu (ALT+O, U).

Chosen unit is stored with the problem description, you can choose it for 
each problem independently. The choice of length units does not affect 
units for other physical parameters, which always use SI units. E.g., the 
current density is measured in   independently of length units. The 
only physical value that is measured in chosen units of length, is 
displacement vector in stress analysis problems.



5.	Model Geometry Definition


This chapter describes how to define the region geometry and build the 
mesh using ELCUT preprocessing utility-the Model Editor.


5.1.	Terminology

Vertex, edge, and block are three basic types of geometric objects, which 
the Model Editor operates with.

Vertex is a point on the plane with coordinates defined by the user or 
calculated automatically as intersection of the edges. For each vertex you 
can define mesh spacing value and the label. The mesh spacing value 
defines approximate distance between mesh nodes in the neighborhood 
of the vertex. The label is used, for example, to describe a singular field 
source.

Edge is a line segment or a circular arc connecting two vertices. It can't 
intersect any other edge of the region. If an edge being created contains 
an existing vertex, two adjacent edges are created. New vertices are 
automatically created in all points where new edge intersects the existing 
ones and all intersected edges are split by these vertices. While 
discretizing the region, edges are divided into elementary fragments 
according to the spacing values defined in the vertices. The edge can be 
labelled, for example, to specify the boundary condition.

Block is a contiguous subregion with the boundary consisting of edges 
and possibly isolated vertices. A block may contain holes which can be 
formed by chains of edges or by isolated vertices. Each block has to be 
labelled to describe material property. Labels of the blocks are also used 
to define distributed field sources. Unlabelled block is not included in 
calculation of field even it is covered by the mesh. The Mesh is created 
block by block according to the mesh spacing value defined in particular 
vertices.

The Label is a string literal up to 16 characters length, which establishes 
the correspondence between geometrical objects of the model and 
physical values assigned to those. Any printable characters including 
letters, digits, punctuation marks, space character are permitted, except 
for asterisk (*) and question mark (?) characters. The label cannot begin 
with space character; trailing spaces are ignored. Labels are case-
sensitive.

The Mesh Spacing value is the parameter of the units of length associated 
with the vertex. The spacing defines the density of the mesh around the 
vertex. Changing these values, you can control the accuracy of the 
solution.


5.2.	How to Create the Model

Model development consists of three stages:

*	region geometry definition;
*	region discretization;
*	definition of properties, field sources and boundary conditions.

To describe model geometry you define vertices and edges which form 
boundaries of all subregions having different physical properties. You 
can create or delete a vertex; create, delete or copy an edge.

To get the finite element mesh you have to define the spacing values at 
few specific vertices. Spacing values for other vertices are calculated 
automatically to make the division smooth. Once the spacing value is 
defined at any vertex, the edges are divided automatically. To build the 
mesh it is necessary to define the spacing value at least at one vertex.


5.2.1.	Starting and Quitting the Model Editor

To start the Model Editor, choose Geometry from Edit menu (ALT+E, 
G) or while editing the problem description select the model filename 
and choose Edit button.

The Model Editor uses interactive graphics. Two graphic windows are 
displayed on screen permanently. The small window presents general 
view of the problem region, while the large one provides more detailed 
view. Below the graphic windows is a prompt line. Upper right screen 
area is normally occupied by the menu or a temporary window used for 
editing values and labels.

The Model Editor has hierarchical structure of menus. Pressing ESC or 
right mouse button causes returning to preceding menu level or quitting 
the Model Editor in the main menu.

To quit the Model Editor select Exit from the main menu or press ESC 
while in main menu. You will be prompted to save the model.


5.2.2.	Region Geometry Definition

With ELCUT Model Editor, you can avoid many routine operations 
while describing the geometry. We recommend to start with look throw 
the draft to escape from inserting vertices, which could be obtained with 
intersection of edges. Also you might detect the objects which could be 
built by copying some other objects, think over the auxiliary elements, 
which would simplify the procedure and be removed afterwards. We 
recommend not to avoid creating edges over existing ones, although you 
should take care especially with arcs, because sometimes it takes more 
time to remove incorrect edge thrown over the whole region than to 
rebuild the model from scratch. We proclaim saving the model to file 
regularly to be able to continue with that state if such an error occurs.

It happens that improper elements rest out of window. Zoom Natural 
arranges window limits to contain all the elements of the model.

When new model is created, the default window is set to correspond the 
unit square. It is convenient to assign the window with region's 
dimensions and create the outward boundary of the model at once, and 
than describe the details. To change default window dimensions choose 
Keyboard from Zoom submenu. The default window dimensions are 
saved with the model to be restored in later editing sessions.

Use the Model submenu to change the geometry. To create new vertex 
choose Add Vertex. Then locate required position with the cursor or 
press TAB to enter coordinates from the keyboard. New vertex appears 
in the window. Then you can continue creating vertices or return to 
Model menu by pressing ESC or right mouse button.

If at least two vertices are defined in the region you can connect them 
with an edge. To do this, choose Add Edge and type in angle size in 
degrees for new edge. Zero value corresponds to the line segment. 
Positive value defines an arc directed from the first vertex to the second 
counter-clockwise, or clockwise if negative. After that pick the vertices 
to be connected by the edge in corresponding order. New edge appears 
in the window. Picking vertex by vertex you can create several edges of 
equal angle size. To break the chain and start new one, press ESC. To 
return to menu press ESC twice.

If you want to copy some edges using parallel displacement, choose 
Copy Edge. Then edit values of shift vector components and pick the 
edges to copy. To return to Model menu press ESC twice. After the first 
ESC you can change shift vector and continue copying.

You can remove auxiliary or incorrectly defined vertices and edges by 
choosing Remove Vertex or Remove Edge. When some vertex is 
removed, all adjacent edges are deleted.


5.2.3.	Building the Mesh

After you get ready with geometry of the model or its part, you can 
proceed with building the finite element mesh. The Model Editor allows 
to build meshes of uniform or essentially non-uniform density of 
elements, e.g. the mesh can be densed to some vertices or subregions 
and rarefied to the outer boundary. The mesh have to be finest where 
the field changes most rapidly, and also where you need the most 
precision.

The mesh density is controlled by spacing values in vertices. The spacing 
value defines approximate distance between mesh nodes around that 
vertex. You never need to define the spacing in all model's vertices. To 
obtain uniform mesh you can set the spacing in any one vertex. This 
value is spread among all other vertices automatically. If you need the 
non-uniform mesh, define spacing values only in those vertices where 
you need finest and roughest mesh. The spacing values are automatically 
interpolated to other vertices to smooth the mesh density distribution. 
The group selection mechanism allows to assign the value to several 
vertices at once.

If the spacing value is defined at least in one vertex, you can proceed 
with the mesh building. The mesh is built block by block. You may 
order to build the mesh in one block or in selected blocks or in all region 
at once.

Changing the density of pre-built mesh (e.g. if solution results show that 
you need more precision somewhere in the region) obey some rules:

*	when you change the spacing value in some vertex, the mesh is 
        removed automatically in those blocks which contact that vertex;
*	the mesh that is not removed, freezes spacing values along its 
        boundary from recalculation as if those values were defined 
        manually; so whether you need radical changes to the mesh 
        density, first remove the mesh in the whole region.

All operation with spacing values and the mesh is performed in the 
Mesh submenu. To define the spacing, choose Set Spacing, then pick the 
vertex and type in the spacing value. Then you can alter spacing at any 
other vertex or return to menu by pressing ESC. If there are some 
previously selected vertices, the spacing values are set for all of them at 
once. If some edges are selected, the spacing values are set for all 
vertices situated at them. The same way, if there are some selected 
blocks, the spacing values are set for all vertices situated at their 
boundaries or inside them.

To build the mesh in separate blocks, choose Build Mesh and pick the 
blocks by turn. Use Build All to build the mesh in whole region except 
already meshed blocks.

Remove Mesh removes meshes in chosen blocks, while Remove All 
removes all the mesh at once.

If the spacing visibility switch is on (Show Spacing in the Options 
menu), the explicitly set spacing values are shown as small circles 
around the vertices. You can check out mesh building process when 
Show Breaking or Show Mesh switches in the Options menu are on.


5.2.4.	Labelling Vertices, Edges and Blocks

The correspondence between geometrical objects and physical 
properties, boundary conditions or field sources is established by use of 
labels. All operation with assigning, editing and checking the labels is 
done in the Label menu.

To assign a label, choose one of Label Block, Label Edge or 
Label Vertex, and then pick the object which you need to label. The 
dialog box will appear, which allows to type in the label from keyboard 
or to select one from already defined in the model or in data files 
assigned to the problem. After you define the label, you can continue 
picking the objects or return to the menu by pressing ESC key.

You may assign the equal labels to several objects of similar type at 
once. To do so, select those objects and choose Label Selected. How to 
select objects by picking them is described in following division.

To check labels assignment or to select objects possessing the same 
label, use Find Label. The dialog box will appear, which allows to select 
the type of objects, and contains the full list of labels as appears in the 
model for the given object type. Picking some label in the list causes 
selection of all objects carrying that label.


5.2.5.	Objects Selection

The Model Editor provides possibility to make some group operation 
upon several objects at once if those have been previously selected. To 
enter select mode, choose Select and then one of  Select Blocks, 
Select Edges or Select Vertices in submenu that will appear.

While in select mode pick an object to select or unselect it. All selected 
objects are highlighted on the screen. To leave the select mode press ESC 
or right mouse button. Objects of different types cannot be selected 
simultaneously.

Unselect All cancels all previous selection.


5.2.6.	Graphics Window Management

There are two graphic windows on the screen while editing the model. 
The small window always displays the general view of the model. The 
large window is used to present more delayed picture of the whole 
model or its selected part. Zoom menu provides several options to 
control which part of the model will be seen in each graphic window.

Keyboard allows you to type in dimensions of the visible region. 
Specified region is displayed in both windows and becomes the default. 
This is a normal way to extend the default visible region. Limits of the 
visible region can be automatically adjusted to preserve equal X and Y 
scales.

Natural sets minimum visible region large enough to contain the whole 
model. Resulting region is displayed in both windows and becomes the 
default.

Default sets visible region dimensions to the default values. The region 
visible in the large window becomes the same as in small one.

Large Window specifies graphic input of visible region dimensions using 
rubber rectangle in the large window. The rubber rectangle is controlled 
using mouse or cursor keys. First, posit lower left corner, then upper 
right one.

Small Window does the same, but rubber rectangle appears in the small 
window. This mode is recommended when whole or part of the required 
region lies outside the large window.


5.2.7.	Obtaining Model Information

You can get the detailed information on the current model state entering 
Info submenu. After you pull it down, choose the object type to deal 
with (vertices, edges or blocks). The complete information about that 
type will appear (amount of objects, presence of labels, mesh 
parameters, etc.). Then pick an object you are interested in to get local 
information. Pressing ESC returns to menu. This mode is used to get 
total number of mesh nodes, to check label values, to examine the mesh 
spacing values and so on.


5.3.	Additional Options

5.3.1.	Saving Model

The Save item of File menu saves model to disk. The Save As does the 
same but lets you change the name of output file. It is wise to save mesh 
not only when editing is over, but regularly during the session to avoid 
troubles due to errors or supply failure.

                Note Changing output file name with Save As does not 
                cause changes in problem description.


5.3.2.	Opening Model File for Edit

To create new empty model choose New in the File menu. To open 
existing model file choose Open in the same menu, and then type or 
select the file name. If the active model was changed you will be 
prompted to save it before transition to the new one.

With Open, you can import old geometry model files created with the 
first or second versions of ELCUT and having .TRI extension. To do so, 
you need only to specify .TRI extension when selecting the file.

Note While loading old-formatted model, all digital labels 
are transferred to literal form automatically.


5.3.3.	DXF File Import

You can import model geometry or its fragments from the DXF file 
produced with any major CAD system. To do so, choose Import DXF in 
the File menu and then type or select required file name. The visible 
region is automatically extended if needed to assure visibility of all 
imported geometric objects. If the model is not empty when reading the 
DXF file, it is recommended to save current model state before the 
operation. This will give you a chance to return back if the imported 
objects incidentally overlap the existing part of model.


5.3.4.	Discretization Visibility Options

There are three switches Show Spacing, Show Breaking and Show Mesh, 
which affect discretization visibility level. Those are accessible in 
Options menu. By default all the switches are on (ticked off in the 
menu). When all these switches are off, region is displayed without 
discretization. This mode is useful for region geometry description and 
label setting. If the Show Spacing mode is switched on, all explicitly set 
spacing values are shown as circles with the appropriate radii.

When Show Breaking switch is on, you can see the result of automatic 
edges fragmentation. It is convenient to use both Show Spacing and 
Show Breaking when specifying the mesh spacing values. Show Mesh 
lets you see triangular mesh where it exists. Turn it on to check mesh 
building.


5.3.5.	Attraction Distance Parameter

To avoid very fine unrecognizable inaccuracies in geometry definition 
new vertices or edges cannot be created very close to existing ones. The 
creation of new geometric objects is controlled by the E parameter also 
called the attraction distance.

The following rules concern creating new vertices and edges.

*	Creating a new vertex is prohibited within 2E-neighborhood of 
        the existing one.
*	A new edge cannot be added if it joins the same vertices as an 
        existing edge and the maximum gap between them does not 
        exceed E.
*	If the distance between a vertex to add and some edge is less 
        than or equal E, the vertex is attracted by the edge and the edge 
        is automatically split into pair of new edges to incorporate the 
        vertex. The same is true when new edge is added, but in this case 
        the new edge may be attracted by existing vertex.

The default value of E is 0.5 per cent of the visible region size. You can 
set different value by choosing Epsilon position in Options menu. 
Decreasing E would allow you to describe very fine details of the model. 
But the most convenient way to get the same result is to zoom in the 
window.



6.	Problem Parameters Description


To solve the problem it is needed to describe the material properties, 
field sources and boundary conditions. These parameters are stored in 
the property description file. The correspondence between records of 
these files and subdomains or boundaries of the region is established by 
the labels assigned to geometrical objects during editing the model. 
Labelling vertices, edges and blocks is described in "Model Geometry 
Definition".

Physical values for a problem may be defined in one or two data files 
attached to the problem. Both files have the same format and distinguish 
only in purpose. The first file is the basic data file of the problem and is 
assumed to contain the data specific for the problem. The second is the 
library file, that contains common material properties and standard 
boundary conditions for a class of problems.

To edit the basic data file, choose Data in the Edit menu (ALT+E, D) to 
edit the library file, choose  Library (ALT+E, L.) The alternate method to 
start editing data is to select the name of one of the data files and to 
choose the Open button  while editing the problem description.

When entering data file editing mode, the dialog box appears, containing 
three lists of labels, which corresponds to blocks, edges and vertices.
 
Option buttons above the list boxes allow to choose which group of 
labels is currently dealt with. One label in the list may be highlighted. In 
this case it appears also in the text box below the list boxes. The label in 
the text box is the current label, so all the immediate actions are done to 
this label.

When editing the basic data file of the problem (not library), label lists 
contain not only labels defined in the data file, but also those labels the 
model file refers to although not defined in neither basic nor library data 
files. Such labels are marked with asterisk in the list.

Labels of blocks, excluded from consideration (i.e., having empty 
material properties,) are marked with double exclamation mark.

Options for editing the data file could be divided into three groups. The 
first group considers actions on creating new labels and editing data for 
existing ones. Operations on copying, renaming and removing some data 
set, which corresponds to some label, belong to the second group. The 
third group assumes the actions on saving the current state of data file 
under another file name and merging two data files.

To exit from data file editing, choose the Close command button or 
press the ESC key. You will be prompted to save the changes to file.


6.1.	Creating New Label

To create new label:

*	choose appropriate type of geometrical object-block, edge or 
        vertex;
*	type the name of the label in the text box titled Label. You may 
        use the INS key in addition to usual methods to move to text box. 
        If the label's name already exists in the list, but is marked with 
        asterisk (which means that the data for the label is not defined,) 
        you need not type the name, but simply select it in the list using 
        keyboard or mouse;
*	choose Add button or press ENTER to start editing the data.

After you define the data, new label appears in the list of existing labels. 
If data editing was canceled, new label is not created.


6.2.	Editing Label Data

To edit the data assigned to some label, select that label and choose Edit 
button or press ENTER. The dialog box appears, its view depends on the 
class of current problem and on the type of geometrical object which the 
label corresponds to.

To finish label data editing, choose OK button. Choosing Cancel button 
will end the editing with discarding all changes to the values.


6.2.1.	Editing Data in Electrostatics
 
Block label data for electrostatics problem contain two components of 
electric permittivity and possibly distributed charge density.

When creating new label, the text boxes for electric permittivity 
components contain None instead of numbers. The word None in these 
boxes or absence of value means that the block containing such label is 
excluded. If you want to define the material properties (and therefore 
include the block into consideration,) simply type in the value of electric 
permittivity, which will replace highlighted None.

If you need to define two components different from each other, first 
check the Anisotropic box.
 
The data for the edge label allow to assign one of possible boundary 
conditions. Select the type of condition and then type in the values.
 
The vertex in the problem of electrostatics may have known potential or 
concentrated charge. Check one of these options and then enter a value.


6.2.2.	Editing Data in Magnetostatics

With problems of magnetostatics, block label data contain two 
components of magnetic permeability tensor, the current density and 
two components of coercive force, if the subregion is the permanent 
magnet.

With nonlinear materials, you need to define the magnetization curve, 
instead of magnetic permeability. In this case the dialog box contains the 
B-H Curve button. Choose the button to edit the curve. Check the 
Nonlinear box to switch between linear or nonlinear material properties. 
Editing the magnetization curve is discussed in "Editing the Curves".

When creating new label, the text boxes for magnetic permeability 
components contain None instead of numbers. The word None in these 
boxes or absence of value means that the block containing such label is 
excluded. If you want to define the material properties (and therefore 
include the block into consideration,) simply type in the value of 
magnetic permeability, which will replace highlighted None.

If you need to define two components different from each other, first 
check the Anisotropic box.
 
The data for the edge label allow to assign one of possible boundary 
conditions. Select the type of condition and then type in the values.
 
The vertex in the problem of magnetostatics may have known potential 
or the concentrated current may flow through the vertex. Check one of 
the options and then enter a value.


6.2.3.	Editing Data with Heat Transfer Problems
 
The data for block label contain two components of thermal conductivity 
tensor and, possibly, the volume power of heat source. With time-
dependent problems, the value of specific heat per unit volume is 
significant.

To describe the thermal conductivity as a function of temperature, check 
the Nonlinear box. The  l = l(T) button will appear, choose it to edit 
the dependency.

Also the volume power of heat source could be described as a function 
of temperature. To do so, check the Function of Temperature box 
related to the heat source field. To edit the dependency, choose the 
q = q(T) button which will appear. The templates for editing the 
dependencies are described in "Editing the Curves".

When creating new label, the text boxes for thermal conductivity 
components contain None instead of numbers. The word None in these 
boxes or absence of value means that the block containing such label is 
excluded. If you want to define the material properties (and therefore 
include the block into consideration,) simply type in the value of thermal 
conductivity, which will replace highlighted None.

If you need to define two components different from each other, first 
check the Anisotropic box.
 
The data for edge label allow to describe boundary conditions. Check the 
condition which you need, and then type in the parameters. The 
condition of the second kind and the convection and radiation condition 
could be combined together which means that the heat flow through the 
surface is compounded from several components.

The vertex in heat transfer problem may have known temperature or 
string source. Check one of these possibilities, and then enter the 
numeric parameter.


6.2.4.	Editing Data with Stress Analysis Problems

When editing the data for the block label with stress analysis problem, 
there are two sets of properties to be edited simultaneously. To switch 
from one set to another, use option buttons at the top of dialog box or 
press PAGE UP or PAGE DOWN keys.

When creating new label, the text boxes for Young's moduli contain 
None instead of numbers. The word None in these boxes or absence of 
value means that the block containing such label is excluded. If you want 
to define the material properties (and therefore include the block into 
consideration,) simply type in the value of the Young's modulus, which 
will replace highlighted None.

The Anisotropic boxes, which applied to elastic moduli or coefficients of 
thermal expansion, allow to describe anisotropic properties in each set 
independently.

The values for allowable stresses do not affect the solution. Those are 
only used in postprocessing stage to calculate the Mohr-Coulomb and 
Drucker-Prager criteria. You need not define allowable stresses, if you 
are not interested in these criteria.
 
The data defined for some edge label may include restraints along one or 
both coordinate axes and the surface force described as surface pressure 
or Cartesian components. To apply fixed displacement along the axis, 
check the appropriate box and then enter a value of displacement.
 
The node label data may define rigid or elastic support along one or both 
coordinate axes, or concentrated external force. To describe rigid 
constraint along some axis, check the appropriate box, and then enter 
the value of fixed displacement.


6.2.5.	Editing the Curves

Curveous functions, which describe some physical dependencies, are 
implemented as tables containing two columns: an argument and a 
function, e.g., magnetic field intensity and flux density or temperature 
and thermal conductivity. Editing the table is supported with graphical 
presentation of the dependency, which is interpolated with cubic spline 
among the entered points. The solver uses just the same curve as you 
see on your screen.
 
To add the new point to the dependency, type in two values (B and H in 
shown example,) and press ENTER key or choose Add button. If the 
argument of a new point coincides with the argument of existing one, 
new point replaces the old one.

To remove the point, select it in the table and choose the Delete button 
or press the DEL key.
You may control the scaling of the graph with use of the Zoom In or 
Zoom Out buttons.
To exit from editing the curve, choose the Close button or press ESC. 
Note that subsequent canceling label data editing with ESC key or the 
Cancel button will discard all changes including the curve editing.


6.3.	Copying, Renaming and Deleting the Label

In order to copy or rename the label (to be precise, the data set related 
to the label,) select the label in the list and choose the Copy or Rename 
button. The new name is entered in a dialog box, which appears when 
you choose the button.

To remove the label, select it and choose Delete or press DEL.


6.4.	Merging and Copying Data Files

The Merge command button allows you to expand the contents of the 
data file being edited with the labels contained in some other data file for 
the same type of problems. The labels with coinciding names will not be 
replaced.

To copy current state of edited data under some name, use the Save As 
button. This function does not affect the name of the file being edited.


7.	Obtaining and Analyzing the Results

This chapter tells how to solve the problem and how to use the 
postprocessor to analyze  the results.


7.1.	Solving the Problem

Several conditions have to be met to solve the problem. The problem 
type, plane, required precision and other parameters have to be specified 
in the problem description file. The model geometry file must contain 
complete model with mesh and labels. Each label referred by the model 
file is to be defined in the problem's private or library data file.

To obtain the problem solution choose Solve Problem from the Results 
menu (ALT+R, S). You may skip this action and directly proceed to the 
results analysis by choosing Analyze from the Results menu (ALT+R, A). 
If the problem has not been solved yet or its results are out of date, the 
solver will be invoked automatically.

Special bar indicator lets you see the progress of the solution process. 
Linear problems are solved by using the preconditioned conjugate 
gradients method. The preconditioning based on the domain 
decomposition technique guaranties very high speed and very weak 
dependence between number of nodes and the required number of the 
CG iterations. Nonlinear problems are solved using the Newton-
Raphson method. The Jacobian matrix arising at the each step of the 
Newton-Raphson method is inverted the same way as it is done for 
linear problems.

 The solver writes the results to the file with the extension .RES, and 
with name and directory matching the problem description file.


7.2.	The Postprocessor

This section explains how to arrange the results for detailed examination 
using the ELCUT postprocessing utility.

The Postprocessor provides various ways of results presentation:

*	field pictures,
*	local field values,
*	integral quantities,
*	X-Y plots,
*	tables.


7.2.1.	Interpreted Quantities and Forms of Presentation

The set of the physical quantities which can be displayed by the 
Postprocessor depends on the problem type.

For the electrostatic problem these quantities are:

*	scalar electric potential;
*	vector of electric field intensity;
*	vector of electrostatic induction;
*	electric permittivity

For the magnetostatic problem:

*	vector magnetic potential;
*	vector of magnetic flux density;
*	vector of magnetic field intensity;
*	magnetic permeability.

For heat transfer problem:

*	temperature;
*	vector of heat flux density;
*	thermal conductivity.

For stress analysis problems:

*	displacement vector;
*	stress tensor and its principal values;
*	Von Mises, Tresca, Mohr-Coulomb and Drucker-Prager criteria.

Several methods possess drawing of 2D field picture:
*	color map of the distribution of a chosen scalar quantity. The 
        color map is accompanied by the legend showing the 
        correspondence between colors and numerical values.
*	field lines. Those are isotherms for temperature fields, lines of 
        equal potential in electrostatics and flux lines for magnetostatic 
        problems;
*	family of line segments showing magnitude and direction of the 
        vector quantity. The base point of each vector is marked by dot;

The following methods are used only for stress analysis problems:

*	stress tensor display as a pair of eigenvectors reflecting the 
        direction of principal axes, magnitudes and signs of principal 
        stresses (blue color denotes tension, red color-compression);
*	deformed boundary drawing;
*	deformed shape indicated by means of deformed and original 
        rectangular grid.

The Postprocessor draws vector and tensor symbols suited in the nodes 
of equidistant rectangular grid with user-controlled cell size. The same 
grid is used for deformed shape display.

It is possible to combine several visualization methods in the same 
picture to obtain the most expressive one.


7.2.2.	Starting and Quitting the Postprocessor

To examine the results choose Analyze from the Results menu (ALT+R, A).

Screen layout when working with the Postprocessor is very similar to 
one of the Model Editor. There are two graphic windows displayed on 
the screen. The small one presents general view of the model, while the 
large one shows detailed field picture or an X-Y plot. The message bar is 
at the screen bottom. Upper right screen area is normally occupied by 
the menu, legend box, or a temporary window used to display text 
information.

When you decided to leave the Postprocessor and exit to the ELCUT 
main menu, choose Exit from the main Postprocessor menu or press 
ESC key in that menu.


7.2.3.	Field Picture Constructing

When entering the Postprocessor, the default form of the field picture 
appears on the screen. You may use View menu to select other methods 
and quantities to display by checking menu positions. If the color map is 
checked, the quantity to map may be chosen from the lower part of the 
menu. You may select several methods of presentation, but only one 
quantity at once. If the list of available quantities is too long to fit in the 
menu, some quantities are accessible through Others submenu.

                Note. Selecting the Shape position (deformed shape 
                display) for stress analysis problems will turn on the 
                Boundary position automatically.

The picture will be re drawn on exit from the View menu. If you select 
none of the visualization methods, drawn will be the contours only.


7.2.4.	Editing Scales and Ranges

To adjust the scale of the field picture choose Zoom from the menu. This 
command is very similar to the analogous command of the Model 
Editor, but the default window dimensions cannot be altered, they retain 
values set in the Model Editor.

The default values of color scales, ranges of X-Y plots, field lines 
spacing, scales of vectors, tensors and displacements are calculated 
automatically when the results are loaded by the Postprocessor. For 
dynamic problems all the defaults are suited to the final field picture, i.e., 
the last time step. By choosing the Scaling command you can adjust 
those parameters that affect the current field picture.

The density of the field lines is controlled by the distance between two 
neighboring lines. This distance is measured in potential units. The color 
scale is determined by two parameters-the lower and upper limits of 
the range of the shown quantity. The specified range is mapped to the 
available set of colors.

Vector and tensor symbols are drawn in the nodes of the regular 
rectangular grid. The same grid is used with deformed shape display. 
You can change the grid cell size and the scaling factor for desired 
feature.

Sizes of the vector symbols of all vector quantities except the 
displacement vector are determined by the corresponding physical value 
multiplied by the scaling factor and by the cell size. Similar method is 
used for stress tensor components. Unlike other interpreted physical 
quantities, the size of the displacement vector on the screen does not 
depend on the cell size. It is determined by the dimensionless scaling 
factor, the unit value of which means that the displacement is shown in 
its natural scale.


7.2.5.	Local Data Display

The Postprocessor displays local field data in Values mode. Click the 
point where you need to know the values of the field quantities, or press 
TAB and then enter the coordinates of the point with the keyboard. Once 
you choose the point, the values at this point are displayed on the screen. 
To leave the Values mode use ESC key, or press right mouse button.

The values of the local physical quantities obtained in the Values mode 
can be logged to the table file. This file has self explaining ASCII format, 
its default extension is .TBL. The table file can be used for printing 
numerical results, or to pass them to other application program, e.g., a 
spreadsheet program to produce the report.

To open the table file choose Table File from the Options menu. You 
will be asked about name of the table file, and when you provide it the 
file will be opened. Existing files may be appended or overwritten. Every 
point you click in the Values mode causes writing a line to the table file. 
To close the table file choose Table File from the Options menu again.


7.2.6.	Editing Contours for X-Y Plots and Integrals

The drawing of the X-Y plot or a calculation of the integral quantity is 
based on the to be evaluated.

The contour is needed when the X-Y plot is to be drawn or the integral 
quantity is to be calculated. The contour is the line consisting of the 
edges of the model and/or other line segments and arcs. Closed 
contours are used for integrals only, while the open ones have sense for 
X-Y plots and for some integrals. The contour can be specified in the 
Edit Contour mode and preserved until next modification. The following 
operations change the current contour state:

Add Line	 -	attaches a line segment or an arc to the 
                        contour. The arc is specified by its degree 
                        measure (zero means line segment) and two 
                        end points. The contour may be initiated by an 
                        arbitrary line, but only adjacent lines are 
                        accepted later. The line cannot be added to the 
                        closed contour. Adding lines is terminated by 
                        pressing ESC or when the contour becomes 
                        closed.

Add Edge	 -	append the contour with an edge of the model. 
                        The contour may be initiated by an arbitrary 
                        edge, but only adjacent edges are accepted 
                        later. The edge cannot be added to the closed 
                        contour. Adding edges is terminated by 
                        pressing ESC or when the contour becomes 
                        closed.

Add Block	 -	considers the current closed contour as a 
                        frontier of the plane region and updates this 
                        region by adding (or subtracting) a block of the 
                        model in the sense of set theory. Adding blocks 
                        is terminated by pressing ESC.

Change Direction -	alters the contour direction. The direction is 
                        shown by the arrows at the contour elements.

Undo	         -	reverses the last action done with the contour.

Clear	         -	deletes the contour.

The contour may not intersect itself. Open and closed contours are 
discerned. Depending on the current contour state some editing 
operations may be prohibited.

The direction of the contour is significant in the following cases:

*	for volume integrals the domain of integration lies to the left 
        from the contour. It corresponds to the contour's interior if the 
        contour is directed counter-clockwise, or to the contour's 
        exterior otherwise.

*	for surface integrals the positive normal vector points to the right 
        relatively to the contour direction.

*	the starting point of the contour corresponds to zero point at the 
        x-axis of the X-Y plot.

*	if the plotted or the integrated function has different values to the 
        left and to the right of the contour, the right-hand value is used.


7.2.7.	X-Y Plots

In X-Y Plot mode you can examine the distribution of some field 
quantities along the selected contour. The X-Y Plot mode menu allows 
you to edit the contour, to select the shown quantity, to adjust the scale 
of the current plot, or to tabulate the field quantities along the current 
contour.

To select a quantity to display choose View from the menu. Selecting the 
quantity for the X-Y plot is similar to selecting the quantity for color 
map. See "Field Picture Constructing" earlier in this chapter.

Zoom option allows you to change the y-scale of the plot with the 
procedure similar to zooming the field picture. The default displayed 
range of the shown quantity can be set by choosing the Scaling option.

                Note. When entering plotting mode, ELCUT draws the 
                quantity last selected for color map. Altering the quantity 
                shown on the X-Y plot causes modification of the quantity 
                for color map. Default ranges of the quantities are also 
                common for X-Y plots and for color maps.

To tabulate the field quantities along the current contour choose 
Tabulate from menu. Each element of the contour is divided into 20 
equal pieces and data for all these points are written to the file. Existing 
files may be appended or overwritten. Output format of the Tabulate 
command is the same as for the table file output described earlier in 
"Local Data Display." This command does not affect state of the table file 
if it was opened to log the values in the selected points.


7.2.8.	Integral Quantities

The set of calculated integral quantities depends on problem type.
The quantities available for electrostatic problems are:

*	electric charge;
*	electrostatic force;
*	torque of electrostatic forces;
*	electric field energy.

For magnetostatic problems:

*	magnetostatic force;
*	torque of magnetostatic forces;
*	magnetic field energy;
*	flux linkage.

For heat transfer problems:

*	heat flux.

No integral quantities have sense for stress analysis.

Equations for above integral quantities are provided in chapter 
"Theoretical Description". For plane-parallel problems all integral 
quantities are considered per unit length in z direction.

Torque is calculated relative to the origin of the coordinate system and 
only for plane problems.

To get the integral quantities choose Integrals from menu. In the 
Integrals mode you can edit the contour and select the required quantity 
from menu. Some integrals require closed counter-clockwise oriented 
contour, otherwise they have no physical sense.

Apart from the main list of integral quantities some additional integrals 
are available in Line/Surf. Int. menu (line and surface integrals) and in 
Volume Int. menu (volume integrals). All these integrals are defined by 
equations provided in the menu.

When the electrostatic or magnetic force, torque, electric charge, or heat 
flux are to be calculated the domain of integration may be chosen by 
many different ways. The only requirement for the surface of integration 
is to contain all the necessary bodies, but to avoid any extra bodies or 
field sources. It is significant to understand that the precision will be the 
best if you choose the integration surface as far as possible from the 
places with strong inhomogeneity of field, e.g., field sources or 
boundaries of conducting or ferromagnetic bodies.

When calculating the flux linkage the domain of integration must exactly 
fit the cross section of the coil.


7.2.9.	Saving the Postprocessor State

Current state of the Postprocessor can be saved to the special .SST file 
and restored from it later. The current state includes: chosen method of 
presentation, selected quantity, scales, ranges, current contour state, 
color table, etc. If you analyze several similar problems or the results of 
the same problem several times, you can save a lot of work by reusing 
the same Postprocessor parameters once saved in the .SST file.

Choose Save Setup from the Options menu to save current 
Postprocessor state and Load Setup from the same menu to restore this 
state from the file. In both cases you will be inquired to supply the file 
name. The default file name is constructed from the current problem 
name and .SST extension.


7.2.10.	Printing the Results

ELCUT supports special color scheme suited for creating monochrome 
screen hardcopies. You can select this color scheme  with Options 
Colors command, which is available in main ELCUT and Postprocessor 
menus.

If you are running MS-DOS 5.0 and the GRAPHICS program is resident 
you can obtain the screen hardcopy by pressing SHIFT+PRINT SCREEN on 
the keyboard.

Many word processing programs, e.g., Microsoft Word, provide way to 
obtain screen image in the file, and then include it into the text 
document. This is the most convenient and common way of producing 
high quality reports.


Appendix A. Example Problem from Start to End


Title:

Microstrip Transmission Line Capacitance.

A shielded microstrip transmission line consists of a substrate, a 
microstrip, and a shield.


Problem Type:

Plane-parallel problem of electrostatics.


Geometry:

The transmission line is directed along z-axis, its cross section is 
shown at the sketch. The rectangle ABCD is a section of the shield, 
the line EF represents a strip.

                       a
   D  |<------------------------------>| C
      |________________________________|_________
      I                                I        |
      I                                I        |
      I                                I        |
      I               Air              I        |
      I                                I        |
      I                                I        | a
      I         E    Strip    F        I        |
      I___________=====w=====__________I______  |
      I            Substrate           I     |b |
      I________________________________I_____|__|
   A                                     B


Given:

Relative permittivity of air  e = 1;
relative permittivity of substrate  e = 10;
dimensions: a = 10 cm, b = 1 cm, w = 1cm.


Problem:

Determine the capacitance of a transmission line.


Solution:

There are two approaches to calculate the capacitance of the line:

*	to apply some distinct potentials to the shield and the strip 
        and to calculate the charge that arises on the strip;
*	or, to apply zero potential to the shield and to describe the 
        strip as having constant but unknown potential and carrying 
        the charge, and then to measure the potential that arises 
        on the strip.

Both these approaches make use of the equation for capacitance C = q / U.
The first approach needs to get the charge as a value of integral 
along some contour, and the second one uses a local value of 
potential. Therefore, we recommend the second way of calculating 
the capacitance (or partial capacitances, if the number of 
conductors exceeds two.)

The first approach is illustrated in the STRIP1.PBM problem on 
your EXAMPLES disk, and the STRIP2.PBM explains the second 
approach.

The method of calculating the capacitance via the energy of electric 
field could not be applied to this problem, since the energy integral 
does not converge in theory due to infinite singularity on the ends 
of strip.


Results:

Theoretical result:	C = 178.1 pF/m.
Approach 1:		C = 186.4 pF/m.
Approach 2:		C = 186.5 pF/m.


Step-by-step Description

Let us learn, how to solve this problem from scratch, using the 
second approach. We'll forget the solution made in STRIP2.PBM, 
and start a new problem, STRIP3.PBM.

To create new problem:

*	Choose New in the Files menu (ALT+F, N); the dialog box 
	appears, asking for the filename for new problem.
*	Change, if needed, the drive and directory in the Directories 
	list box.
*	Type strip3 in the Filename box.
*	Choose OK.

The extension .PBM will be added automatically.

To select convenient length measurement units (centimeters):

*	Choose Length Units in the Options menu. A dialog box 
	appears.
*	Select Centimeters.
*	Choose OK.

To assign the problem with appropriate features:

*	Choose Problem in the Edit menu (ALT+E, P). The 
	Problem Description dialog box appears.
*	Select Electrostatics in the Problem Type drop-down list 
	box.
*	Select XY Plane.

We'll agree with suggested model and data file names 
(STRIP3.MOD and STRIP3.DES.) If the Library Data filename 
box is not empty, clear it, since we'll define all the labels in the local 
data file (STRIP3.DES.)

We can start editing the model or the data directly in the Problem 
Description dialog box. To edit the model:

*	select the Geometry text box (click anywhere in the box 
	with a mouse, or press ALT+G, or press TAB until the 
	selection reaches the box.)
*	choose the Open button.

The Model Editor starts.

The first step with the new model in Model Editor is to adjust 
window dimensions fitting the problem's region. As the problem 
has vertical axis of symmetry, it is convenient to set zero of x-axis at 
the axis of symmetry. So the region fits the square (-5 <= x <= 5, 
0 <= y <= 10). To assign these values to the window limits:

*	Choose Keyboard in the Zoom menu.
*	Type the values in appropriate text boxes.
*	Press ENTER or click the dialog box background anywhere 
	outside options.

Now we can proceed with defining the geometry itself. To define 
the vertices which correspond to points labeled with letters A 
through F in the sketch:

*	Enter Add Vertex mode in the Model menu. The plus sign 
	cursor (*) arises in the large window indicating the point 
	locating mode.
*	Use DIRECTION keys or a mouse to move cursor from point 
	to point and press ENTER or click left mouse button where 
	you want the vertex to arise. New vertices immediately 
	appear in the window.
	Or, press TAB, type coordinates, and press ENTER for each 
	new vertex. Don't worry of making mistakes--you can 
	remove erroneous vertices later.
*	Press ESC to return to the Model menu.
	If you have created excess points, you can remove it now:
*	Choose Remove Vertex. The X-shaped cursor appears to 
	indicate the picking mode. Consequently pick excess 
	vertices, which immediately disappear.
*	Press ESC to return to the Model menu.

Now we can create edges connecting the vertices:

*	Choose Add Edge. A dialog box appears asking the arc 
	angle for new edges.
*	Press ENTER (or click gray background of the dialog box) to 
	agree with suggested zero value, which means creating the 
	straight lines. The picking mode (X-shaped) cursor appears 
	in the window.
*	Pick points A, B, C, D, A consequently to create edges, 
	which constitute the rectangle ABCD. The edges 
	immediately appear on the screen.
*	Press ESC to break the chain of edges and start a new one.
*	Pick point E and then F to create the corresponding edge.
*	Press ESC twice to return to the Model menu.

You can remove erroneous edges, using the Remove Edge 
command.

We need some additional constructing to create edges which 
separate substrate from air. The easy way to create them is to copy 
the edge AB shifting it 1 cm upper. No doubt that new edge will 
partly coincide with already created edge EF--while creating, the 
coincidence is checked out, and only non-existent parts of new edge 
are really created. To make a copy:

*	Choose Copy Edge. A dialog box appears asking for the 
	components of the shift vector.
*	Press TAB or DOWN ARROW to move to the text box 
	corresponding to y-component.
*	Type 1.
*	Press ENTER. The picking mode (X-shaped) cursor appears 
	in the window.
*	Pick the edge AB to make its copy. New vertices and edges 
	arise on the screen.
*	Press ESC to cancel copying mode and return to the Model 
	menu.

We've done with the model's geometry. Now we can assign labels 
to geometrical objects to describe material properties, sources and 
boundary conditions.

The model contains two blocks having different material properties: 
the air and the substrate. To be clear, we can use the word Air to 
label the upper block and Substrate for the lower one. To assign 
these labels to blocks:

*	Press ESC to close the Model menu and return to the main 
	menu of the Model Editor.
*	Choose Label Blocks in the Label menu. The picking mode 
	cursor arises.
*	Pick the upper block. It becomes highlighted and the dialog 
	box appears asking for the label value.
*	Type Air and press ENTER. The X-shaped cursor appears 
	again.
*	Pick the lower block, type Substrate and press ENTER.
*	Press ESC to return to the Label menu.

Edge labels are used to define specific boundary conditions on 
inner and outward boundaries of the region. In our case, we need 
to specify boundary conditions for the shield (rectangle ABCD) and 
for the strip (line EF). To assign labels to edges:

*	Choose Select Edges from the Select menu (we'll select 
	several edges to assign a label to them at once.) The picking 
	mode cursor arises.
*	Pick consequently six edges, which constitute the rectangle 
	ABCD. Those edges become highlighted that indicate the 
	selection. If you have selected excess edge, pick it once 
	more to unselect it.
*	Press ESC to cancel selection mode.
*	Choose Label Selected. A dialog box appears asking for the 
	label value.
*	Type Shield and press ENTER to assign the label to selected 
	edges.
*	Choose Label Edges. The picking mode cursor appears. 
	This mode is convenient for assigning labels edge-by-edge.
*	Pick the edge EF. A dialog box appears.
*	Type Strip and press ENTER to assign the label.
*	Press ESC to return to the Label menu.

We also need to assign vertex label to any vertex contacting the 
strip, to specify that the strip is charged. No matter which vertex 
you choose, the charge will be distributed through all the 
conductor. To assign the label to a vertex:

*	Choose Label Vertices. The picking mode cursor arises.
*	Pick any of vertices E or F. A dialog box appears asking for 
	the label value.
*	Type Charge and press ENTER to assign the label to vertex.
*	Press ESC to return to the Label menu.

Now we have finished with assigning labels to geometrical objects. 
You can check their values in the Find Label mode.

We can proceed with building a mesh of finite elements. To define 
the mesh density, we need to define spacing parameters in several 
vertices of the model. We suppose that the electric field is most 
non-homogeneous near the ends of the strip, so the mesh there 
must be maximum dense. Therefore, we'll assign the spacing value 
of 0.2 cm to the vertices E and F and the value of 0.8 cm to the 
vertices A, B, C and D to build the mesh of approximately 400 
nodes. To define spacing values:

*	Press ESC to return from the Label menu to the main menu.
*	Choose Mesh to open the mesh building menu.
*	Choose Select Vertices from the Select menu, which allows 
	to assign one spacing value to several vertices at once.
*	Select vertices A, B, C and D and press ESC to return to the 
	Mesh menu.
*	Choose Set Spacing. A dialog box appears asking for the 
	spacing value.
*	Type 0.8 and press ENTER. This value is now assigned to 
	selected vertices.
*	Choose Unselect All from the Select menu to unselect all 
	previously selected objects.
*	Select vertices E and F.
*	Choose Set Spacing, type 0.2 and press ENTER. The model 
	is now ready to build the mesh.
*	Choose Build All to build the meshes for both blocks at 
	once.

Now the model is ready. To exit the Model Editor with saving the 
file:

*	Press ESC twice to close the Mesh menu and quit. The box 
	appears prompting you to save the model file.
*	Choose Yes to confirm saving operation.

You have returned to the Problem Description dialog box. Let us 
continue with defining data for material properties and boundary 
conditions. To start editing data file:

*	Select the Data text box.
*	Choose the Open button. A dialog box appears warning you 
	that the file STRIP3.DES does not exist.
*	Choose OK to create new data file.

The Properties Description File dialog box appears. It contains 
labels, which you have just defined in the model. The label names 
are marked with asterisks to outline the fact, that the data for these 
labels are not yet defined. Now we need to select the labels one-by-
one and to define the data for them.
To define the data for block label Air:

*	Select its name in the list box (click it with a mouse, or 
	press ALT+B and use DOWN key to highlight the name's field.)
*	Choose Open button (or press ENTER as Open is the default 
	button, or click a mouse once more on the name's field.)

A dialog box appears, prompting to enter material properties and 
distributed source for block label Air. To assign values:

*	Type 1 in any text box for components of electric 
	permittivity tensor.
*	Choose OK.

Repeat last actions for the label Substrate. The value of relative 
permittivity of substrate is 10.

Now we'll continue with edge labels' data. We'll define the label 
Shield as homogeneous Dirichlet boundary condition (U = 0) and 
the label Strip as conductor (U = const) condition.

To define the data for edge label Shield:

*	Select its name and choose Open. A dialog box appears, 
	allowing to assign to edge label any of possible boundary 
	conditions.
*	Select the Dirichlet Condition box. Zero value of pre-
	defined potential will be suggested.
*	Choose OK.

To define the data for edge label Strip:

*	Select its name and choose Open.
*	Select the Conductor box.
*	Choose OK.

We need to define the vertex label Charge to assign the charge to 
the strip. While determining the capacitance, no matter what exact 
value of the charge to specify. We'll issue the value of 1.
To define the data for vertex label Charge:

*	Select its name and choose Open. A dialog box appears, 
	allowing to specify the charge, or to assign the Dirichlet 
	boundary condition.
*	Select the Electric Charge box.
*	Type 1 in the text box for charge value.
*	Choose OK.

All the data needed to solve the problem is now defined. To exit 
from data editing mode:

*	Choose Close button in the Properties Description File 
	dialog box. A dialog box appears, prompting you to save 
	changes to data file.
*	Choose Yes to save changes. Now you return to the 
	Problem Description dialog box again.
*	Choose OK.

At last, we can solve the problem and analyze the solution. To do 
this in one step:

*	Choose Analyze from the Results menu. You will be 
	suggested to solve the problem first, as the results are 
	absent.
*	Choose OK.

The solver starts, showing you the process in dynamics. After 
finishing, the Postprocessor starts automatically. There are many 
possibilities to analyze the field in the Postprocessor. We will show 
only those steps needed to determine the capacitance:

*	Choose Values from the menu. Cross-shaped cursor 
	appears allowing you to pick points to determine the local 
	field data.
*	Move the cursor to the point (0.0, 1.0) (exactly; the 
	convenient way is to use the DIRECTION keys, or to press 
	TAB and type coordinates from keyboard.)
*	Press ENTER.

The box appears, showing you the local field data. The potential of 
the strip is 4.98E9 Volts. The capacitance is

	C = q / U = 1 / 4.98E9 = 2.01E-10 F/m = 201 pF/m

Note that, with plane-parallel problems, we specify the sources as 
specific values per unit depth (e.g., the charge of the strip), and the 
result is specific capacitance per unit depth, measured in F/m.

If you'll now look through the model provided with STRIP2.PBM, 
STRIP.MOD, you find some hints that allow to get more fine mesh 
near the vertices E and F--the points of singularity.

As the Model Editor cannot squeeze the mesh spacing inside the 
edge, but only from one end to another, two extra vertices have 
been added in central points of edges AB and EF. The first of these 
vertices allows the Model Editor to adjust automatically the spacing 
around it, since the strip is very close to this point. The second 
vertex is aimed to specify manual spacing value for it, which allows 
to decrease the total number of nodes in the mesh, without loss of 
precision.
