ACM Modeling Language Reference

Aspen Custom
Modeler 2004.1
Modeling Language
Reference Guide
Who Should Read this Guide
This guide contains detailed reference information about the modeling

language, including syntax details and examples.
Who Should Read this Guide 2

Contents
INTRODUCING ASPEN CUSTOM MODELER ....................................................... 9
1 MODELING LANGUAGE CONCEPTS ............................................................. 11

Statements ........................................................................................................... 11
Type Definitions..................................................................................................... 12
Declaration Statements........................................................................................... 13
Assignments.......................................................................................................... 14
Assignments to Array Elements ................................................................................ 16
Assignment Lists.................................................................................................... 17
Assignments using WITHIN and ENDWITH ................................................................. 17
Naming Conventions............................................................................................... 18
Duplicate Names: BLOCK and STREAM ...................................................................... 19
Parameters ........................................................................................................... 19
Sets ..................................................................................................................... 20
Set Arithmetic ....................................................................................................... 21
Arrays .................................................................................................................. 23
Atomic Type Conversions ........................................................................................ 24
Inheritance ........................................................................................................... 25
Type Relatives ....................................................................................................... 27
Component Lists .................................................................................................... 29
Defining which ComponentList to use ................................................................... 30
Using More Than One Component List in Models .................................................... 30
Built-In Type Definitions.......................................................................................... 31
Built-In Parameter Types ................................................................................... 32
Built-In Variable Type ........................................................................................ 33
Setting Attribute Values ..................................................................................... 33
Getting Attribute Values ..................................................................................... 34
2 MODELING LANGUAGE FOR SIMULATIONS................................................ 35

Using Units of Measurement .................................................................................... 35
Specifying a Physical Quantity for Variables .......................................................... 36
Defining Physical Quantities, Units of Measurement, and Conversions ....................... 37
Contents 3
Defining Tasks ....................................................................................................... 45
Event-Driven Tasks ........................................................................................... 46
Callable Tasks .................................................................................................. 48
Tasks and Opening Simulations ........................................................................... 51
Using Task Statements ........................................................................................... 51
Assigning a Value to a Variable in Tasks ............................................................... 51
UOM Conversions in Tasks.................................................................................. 52
Using the Random Function ................................................................................ 52
The IF Construct ............................................................................................... 53
Ramping the Value of a Simulation Variable in Tasks .............................................. 54
Suspending and Restarting a Task ....................................................................... 55
Creating Snapshots in Tasks ............................................................................... 57
Executing Tasks in Parallel ................................................................................. 57
PRINT Statements in Tasks................................................................................. 58
Pausing a Simulation from a Task ........................................................................ 59
Invoking a Script from a Task ............................................................................. 60
FOR Loops in Tasks ........................................................................................... 61
REPEAT Loops in Tasks ...................................................................................... 62
WHILE Loops in Tasks........................................................................................ 62
Task Language Quick Reference .......................................................................... 63
Defining Flowsheet Constraints ................................................................................ 65
Defining Optimization Constraints............................................................................. 66
Defining Steady State Constraints ............................................................................ 68
Defining Dynamic Constraints .................................................................................. 69
Path Constraints .................................................................................................... 70
3 MODELING LANGUAGE FOR TYPES ............................................................ 72

Defining Variable Types .......................................................................................... 72
Defining Parameter Types .................................................................................. 74
Defining Port Types ................................................................................................ 75
Defining Stream Types............................................................................................ 76
Defining External Procedures ................................................................................... 77
Defining Structure Types......................................................................................... 80
4 MODELING LANGUAGE FOR MODELS ......................................................... 83

MODEL Syntax....................................................................................................... 84
Declaring Variables in Models................................................................................... 84
Declaring Ports in Models ........................................................................................ 88
Using Port Properties ......................................................................................... 89
Using the IsMatched Attribute of Variables ............................................................ 90
Contents 4
Using Multiports ................................................................................................ 91
Writing Equations in Models ..................................................................................... 94
Mathematical Operators in Equations ................................................................... 94
Equations with Array Variables ................................................................................. 96
Conditional Equations ............................................................................................. 97
Using Sets in Conditional Expressions ......................................................................100
FOR Loops ...........................................................................................................101
ForEach Operator ..................................................................................................103
Equations that Call Procedures ................................................................................105
Equations that Refer to Port Variables ......................................................................106
CONNECT Keyword................................................................................................107
CONNECT Restrictions.......................................................................................108
LINK Keyword ......................................................................................................110
LINK Restrictions .............................................................................................110
EXTERNAL Keyword...............................................................................................112
Connectivity Rules.................................................................................................113
Using SIGMA ........................................................................................................114
Using SIGMA for Multidimensional Arrays and Multiports ........................................115
Using the SIZE Function.........................................................................................117
Assigning Specifications in Models ...........................................................................117
Referring to Sub-Models in a Model Definition ...........................................................119
Using External Properties to Write Sub-Models ..........................................................121
Example of Using External Properties to Write Sub-Models .....................................122
Changing the Default Mapping For External Properties ...........................................123
Notes and Restrictions on Writing Sub-Models ......................................................124
Using Virtual Types in Sub-Models ...........................................................................124
Example of Using Virtual Types in Sub-Models ......................................................124
Restrictions on Using Virtual Types .....................................................................126
Using SWITCH ......................................................................................................126
5 MODELING PDE SYSTEMS .........................................................................129

Overview of PDE Modeling ......................................................................................129
About PDE Modeling ..............................................................................................129
Creating a Custom PDE Model .................................................................................130
Example of Custom PDE Modeling.......................................................................130
Using Domains with PDE Modeling ...........................................................................132
Declaring Domains for PDE Modeling ...................................................................133
Using Discretization Methods for PDE Modeling .....................................................136
Contents 5
Specifying Domain Length .................................................................................138
Specifying Element Spacing Preference ...............................................................139
Using Domain Sections .....................................................................................141
Number and Location of Discretization Nodes .......................................................142
Interpolation of Existing PDE Models ...................................................................143
Using Distributions with PDE Modeling......................................................................148
Declaring Distributions for Distributed Variables ...................................................149
Declaring Distributed Variables that do not Require Partial Derivatives .....................152
Switching On Integrals in Distribution Declarations ...............................................152
Referring to Domains in Distributed Variable Declarations ......................................153
Using an Array of Distributions...........................................................................154
Referencing Domains and Distributions in Sub-Models ...........................................155
About IPDAE Models ..............................................................................................155
Partial Differential Equations (PDEs) ...................................................................156
Writing PDEs in Models ..........................................................................................157
Specifying Partial Derivatives in PDEs..................................................................157
Using the Interior Set in PDEs ............................................................................158
Open Domains for PDE Modeling.........................................................................159
Closed Domains for PDE Modeling.......................................................................160
Using Domain Slices for PDE Modeling.................................................................160
Using Boundary and Initial Conditions for PDE Modeling.........................................163
Boundary Conditions Containing Second-Order Derivatives.....................................165
Writing Integrals in PDE Models..........................................................................168
Using Method of Lines for the Solution of PDEs ..........................................................171
Finite Difference Methods: General .....................................................................172
Discretization Methods: Finite Element Methods....................................................174
Mixer-Reactor-Absorber (PDE) Example Description ...................................................176
Running the Mixer-Reactor-Absorber (PDE) Example ..................................................176
Jacketed Reactor (PDE and Integral) Example Description...........................................177
Running the Jacketed Reactor (PDE and Integral) Example .........................................178
Crystallizer (PDE and Integral) Example Description ...................................................178
Running the Crystallizer (PDE and Integral) Example..................................................179
PDE Modeling Glossary ..........................................................................................180
6 MODELING LANGUAGE CONVENTIONS ......................................................181

Modeling Language Conventions..............................................................................181
Adding Comments ............................................................................................181
General Modeling Language Conventions .............................................................182
Text Conventions ..................................................................................................182
Contents 6
GENERAL INFORMATION..............................................................................183
Copyright.............................................................................................................183
Related Documentation..........................................................................................186
TECHNICAL SUPPORT...................................................................................187
Online Technical Support Center .............................................................................187
Phone and E-mail..................................................................................................188
INDEX ..........................................................................................................189
Contents 7
Contents 8
Introducing Aspen Custom
Modeler
Aspen Custom Modeler (ACM) is an easy-to-use tool for creating, editing and
re-using models of process units. You build simulation applications by
combining these models on a graphical flowsheet. Models can use inheritance
and hierarchy and can be re-used directly or built into libraries for distribution
and use. Dynamic, steady-state, parameter estimation and optimization
simulations are solved in an equation-based manner which provides flexibility
and power.
ACM uses an object-oriented modeling language, editors for icons and tasks,
and Microsoft Visual Basic for scripts. ACM is customizable and has extensive
automation features, making it simple to combine with other products such as
Microsoft Excel and Visual Basic. This allows you to build complete
applications for non-experts to use.
Introducing Aspen Custom Modeler 9

Introducing Aspen Custom Modeler 10
1 Modeling Language
Concepts
The Aspen Custom Modeler modeling language uses a number of common

concepts throughout. This chapter describes the following concepts:
• Statements
• Type definitions
• Declaration statements
• Assignments
• Assigning array elements
• Assignment lists
• Assignments using WITHIN and ENDWITH
• Naming conventions
• Duplicate names: BLOCK and STREAM
• Parameters
• Sets
• Set arithmetic
• Arrays
• Conversions of atomic types
• Inheritance
• Type relatives
• Component Lists
• Built-in type definitions
Statements
Use statements to make assignments of values, define equations, declare
items such as variables or models, and make connections.
Statements include one of the syntax elements shown in the following table.
Syntax Element Type of Statement

AS Declaration
: Value assignment
1 Modeling Language Concepts 11

= Equation
CONNECT AND Connection
LINK AND Link
Statements must end with the ; (semi-colon).
Statement Examples
Create an instance of a variable in a model:
T AS Temperature;
Assign a value to a variable and specify that the variable value is known:
Feed.T: 75.0, Fixed;
Equate two variables in a model:
FlowIn = FlowOut;
Make a connection between the input and output of two blocks:
CONNECT Pump1.Output1 AND Valve2.Input1;
Make a connection between the port of a sub-model and the port of the
containing model:
LINK Tank5.Output1 AND MultiOutput1;
Type Definitions
Type definitions introduce and describe new types for models, streams,
variables, structures, procedures, ports and parameters.
For further information on types, see Modeling Language for Types. Chapter
3.
Type Definition Syntax

The syntax for all type definitions is of the form:
TypeKeyword TypeName USES UsedTypeName
DefinitionBody

END
TypeKeyword One of these keywords: MODEL,

STREAM, VARIABLE, STRUCTURE,
PROCEDURE, PORT, PARAMETER
TypeName The name of the type being defined.
This name must be unique within
the scope in which the definition is
made.
UsedTypeName Optional - The name of a type
(which has been previously defined),
from which this type inherits
properties
DefinitionBody The content of the definition. This
varies depending on the object that
is defined. The definition body can
be one of the types listed in Chapter
3. In the simplest case, the
definition body is just a list of the
declarations or assignments of the
type’s properties.
Type Definition Example

This example defines the new variable type Temperature, and assigns default
values for all variables that use this type.
VARIABLE Temperature
Value: 1.0;
Lower: -40.0;
Upper: 300.0;
Scale: 10.0;
Description: "Temperature";
PhysicalQuantity: "Temperature";
END
Declaration Statements
A declaration statement declares a property by giving it a name and defining
its type. The syntax for all declarations has the following form:
Identifier, Identifier ... AS FlagList TypeName
(AssignmentList);
Identifier .................................. The name of a property being

declared. This can be a list if more
than one properties are to be
declared with the same FlagList,

TypeName and AssignmentList.
FlagList ..................................... Optionlal list of qualifiers, separated
by commas, that apply to the
property. These qualifiers affect the
implementation of the property,
and are optional. For example, the
direction of a port is defined in the
FlagList.
TypeName ................................. The name of the type of which the
property is an instance.
AssignmentList ......................... Optional comma-separated list of
assignments of values to properties
of the property. The assignments in
the list are of the form
PropertyName:Value. If the type of
the value unambiguously identifies
the property, you can omit
PropertyName.
Declaration Statement Remarks

Note that not all declarations are valid in all contexts. For example, you can
declare a port within a model definition, but not within a variable type
definition.
Declaration Statement Example

Below are some example declaration statements that could be found within a
model.
P AS Pressure (fixed, Description: ”Inlet Pressure”);
OUT_V as output MaterialPort
P is declared as a variable of type pressure, with a specification of fixed, and
a textual description. OUT_V is declared as a port of type MaterialPort, and
the flag output identifies it as an output port.
Assignments
Assignments are used to change the value of a property of an object, such as
a variable or a parameter.
Assignment Syntax
The modeling language uses a : (colon) as its assignment operator:
Object.Property: ValueExpression;
Object ....................................... A variable or parameter

Property.................................... A property of the variable or
parameter, for example Value,
Spec, Description etc.
Note: If you do not specify a
property, Value is assumed.
ValueExpression ....................... An expression for the property
l

value
Assignment Remarks
Assignments to expressions are treated as constraints by the compiler. This
means that the expression is re-evaluated whenever the value of any term
within it changes. The result is then automatically assigned to the property.
You can assign values to properties of variables or parameters in several
places, for example in models, flowsheet constraints and interactively. From
highest to lowest, the order of precedence in which variable and parameter
properties are changed is:
• Changes made in the following ways are made immediately and overwrite
any previous assignment:
− Interactive changes, for example, using a table, using the automation
interface, running a script, or applying a snapshot with Use
− Values calculated as the result of running a simulation
• Assignments written into the Flowsheet Constraints section take
precedence over assignments in models.
• If an assignment in a model uses an expression, and one of the values in
the expression changes, the expression is re-evaluated and the assigned
value is updated. However if the value has previously been changed by a
higher precedence operation, this updating is disabled to avoid overriding
this change.
• If a block uses a model that uses hierarchy, assignments made in the
parent model take precedence over those in child models in the hierarchy.
• If a block uses a model that uses inheritance, assignments in the derived
model take precedence over those in the inherited model.
• If properties are not assigned anywhere else, they will remain at the
default values defined for their variable type or parameter type.
Assignment Examples
Assign the value 5 to the Value property of the variable x:
x: 5.0;
Assign the value "Cylinder" to the Value property of the string parameter
Geometry:
Geometry: "Cylinder";
Assign the value True to the Value property of the logical parameter
PressureDriven:
PressureDriven: True;

Assign a value to the variable Lag using an expression, and assign a value to
Lag2 that is dependant on the value of Lag:
Lag: 5/60;
Lag2: Lag/2;
Assign values for the Upper and Lower bounds of variable x:
x.Lower: -10;
x.Upper: 20;
For further examples of assignments, see Defining Specifications in Models

Examples and Defining Specifications in Flowsheets Examples.
Assignments to Array Elements

Assignments to an array variable will assign the specified value to all
members of the array. However you can assign values to individual array
elements.
Assigning to Array Elements Syntax

ArrayVariable(Index) : Value ;
ArrayVariable ............................ The name of the array variable

Index ........................................ The index of the array. This must
be a member of the index set for
the array
Value......................................... The value to be assigned to the
array element
Assigning to Array Elements Example

The following example shows the values of an array can be assigned
individually:
Components AS StringSet(["H2O","H2+","Zn++","Ca++"]);
mw(Components) AS MoleWeight;
mw("H2O"): 18;
mw("Ca++"): 40.1;
Note: For convenience Components is defined as the index set

for the array mw. The array variable mw in the above example
could have been defined directly with the statement:
mw(["H2O","H2+","Zn++","Ca++"]) AS MoleWeight;

The [] characters show the start and end of the list of elements in the set.
Assignment Lists
You can use assignment lists in variable assignments to avoid repeating path
names.
Assignment List Syntax

Property : Assignment, Assignment, ... ;
Property.................................... Property, such as a variable within

a block
Assignment ............................... Value assigned to the Property
Assignment List Example

The following example assigns 25 to the value attribute and Fixed to the
specification attribute of the variable Tank.Q.
Tank.Q: VALUE:25.0, SPEC: Fixed, LOWER: -10.0;
If the property name is omitted it is assumed that the assignment is to the

value or spec attribute. This example can be shortened to:
Tank.Q: 25.0, FIXED, LOWER: -10.0;
Assignments using WITHIN and

ENDWITH
If you want to make several assignments within the same scope, you can
reduce the amount of typing by using the WITHIN and ENDWITH keywords.
WITHIN and ENDWITH Syntax

WITHIN Scope
AssignmentList
ENDWITH
Scope ........................................ The block or variable to which

subsequent assignments apply
AssignmentList ......................... The list of assignments

WITHIN / ENDWITH Remarks
You can use WITHIN / ENDWITH statements within a model when making
assignments to variables in a sub-model.
WITHIN / ENDWITH Example

Assign values to multiple variables within a sub-model used within a model.
Rxn1 AS PowerLawRxn;
WITHIN Rxn1
PreExp: 1.45E6;
EAct : 2374.7;
TExp : 1.2;
ENDWITH
Alternatively, you could type each specification explicitly:
Rxn1 AS PowerLawRxn;
Rxn1.PreExp: 1.45E6;
Rxn1.EAct : 2374.7;
Rxn1.TExp : 1.2;
Naming Conventions
You use names when you:
• Define a type, such as a model or a stream type.
• Add a property to a type, such as a variable.
Important Note: You cannot create a type which has the same
name as a built-in type or a keyword.
Make sure that the names you choose:

• Are unique within their scope, for example variable names must be unique
within a model, but two different models can have a variable with the
same name.
• Include only alphanumeric characters and the _ (underscore) character.
• Start with an alphabetic character.
• Do not begin with am_.
Because you may use names to identify objects from external programs such
as Microsoft Visual Basic or Microsoft Excel, you should apply the naming
rules for Visual Basic.

Duplicate Names: BLOCK and
STREAM
Within a given scope, the names you use for all properties must be unique.
The one exception to this is that you can have a block and a stream with the
same name within your flowsheet. This feature is for compatibility with other
AspenTech products such as Aspen Plus. In general it is still best to avoid
using the same name for a block and a stream.
If you use the same name for a block and a stream you must refer to each
using the following syntax:
Block Syntax
BLOCKS("BlockName")
Stream Syntax
STREAMS("StreamName")
BlockName ................................ Name of a block in the flowsheet

StreamName ............................. Name of a stream in the flowsheet
You can use the same syntax in both input files and in Microsoft Visual Basic.
BLOCK Keyword Example

The following is a global equation which equates two variables in a block and
a stream which have the same name:
BLOCKS("B1").Tin = STREAMS("B1").T;
Parameters
Use parameters to define values that are fixed for the whole simulation, such
as acceleration due to gravity, or the gas constant. You can also use
parameters for fixed physical quantities such as the number of trays in a
column. Like variables, parameters are declared within types such as models
or stream types.
Parameters Syntax
ParameterName AS ParameterType (Value);
ParameterName ........................ Name of the parameter

ParameterType ......................... Can be IntegerParameter,
RealParameter, StringParameter,
LogicalParameter, or a user-
defined parameter type

Value......................................... Optional value assigned to the
parameter
Parameters Remarks
The key difference between variables and parameters is that variables can be
solved for during a simulation run, whereas parameters always have a known
value. Also parameters can be used in conditional statements or loops within
models to determine which variables and equations will be active for a
particular instance of a model.
Parameters may be of one of the built in types IntegerParameter,
StringParameter, LogicalParameter or RealParameter., or of a user defined
type defined using the PARAMETER keyword.
Parameters Examples
Define a parameter for the gravitational constant g:
g AS RealParameter (9.81);
In the next example the model Column has a parameter NStages which
defines the number of stages. This parameter is used to define an array of
Tray models called Stages. NSTages is given a default value of 10, but this
can be changed when the Column model is instanced as a block in the
flowsheet.
MODEL Column
NStages AS IntegerParameter(10);
Stages([1:NStages]) AS Tray;
.
.
END
Sets
Sets are most commonly used to index array variables. For information on
arrays, see Arrays.
A set is a collection of objects that has no duplicate values and is not in any
specific order. You can manipulate sets using set functions, such as union,
intersection, and difference.
Sets Syntax
SetName AS SetType (SetContents) ;
SetName ................................... The set name

SetType..................................... Can be IntegerSet or StringSet
SetContents .............................. An optional list of the contents of
th t

the set
Sets Remarks
You can define sets that contain either integers (IntegerSet) or strings
(StringSet).
An IntegerSet is a set whose members are integers, for example [1,2,5].
A StringSet is a set whose members are strings, for example [“Methane”,
“Ethane”, “Propane”].
You can define the set contents at the same time as you define the set, or
you can define the set contents later. The set contents can be defined by a
list of objects enclosed between [ and ] (brackets), or by a set expression
involving previously defined sets.
Sets Examples
The set MySet is defined as containing objects that are of integer values. The
contents of the set are not yet defined.
MySet AS IntegerSet;
MySet is defined as containing integer objects, and the set is immediately

defined as an empty set. This set has no contents:
MySet AS IntegerSet([]);
MySet is defined as containing four integers 1, 2, 3, and 4:

MySet AS IntegerSet([1, 2, 3, 4]);
MySet is defined as containing four integers 1, 2, 3, and 4 using a colon to

define a range of numbers:
MySet AS IntegerSet([1:4]);
MySet is defined as a set containing the strings "Yes" and "No":

MySet AS StringSet(["Yes", "No"]);
Set Arithmetic
The following set arithmetic functions are available:
Function name Symbol Description

UNION + The UNION of two sets is the set of elements
that are in both sets but without duplicates.
DIFFERENCE – The DIFFERENCE of two sets is the set of

elements in the first set but not in the second
set.

INTERSECTION * The INTERSECTION of two sets is the set of
elements that are common to both sets.
SYMDIFF none The SYMDIFF of two sets is the set of the

elements that are in one set but not in the
other. This can be defined as the difference
between the union and the intersection of the
two sets.
Set Arithmetic Function Syntax

You can use the symbol or the function name for set arithmetic.
SetExpression SetFunctionSymbol SetExpression;
SetExpression ........................... A set expression, such as the

name of a set, or a set definition
enclosed by [ ] (brackets)
SetFunctionSymbol ................... One of the set function symbols:
+ –*
or
SetFunctionName (SetExpression1, SetExpression2, ... );
SetFunctionName...................... One of the set function names,

such as UNION, INTERSECTION,
etc.
SetExpression ........................... A set expression, such as the
name of a set, or a set definition
enclosed by [ ] (brackets)
Set Arithmetic Example

Define SetA and SetB:
SetA AS IntegerSet([1, 2, 3, 4]);
SetB AS IntegerSet([3, 4, 5, 6]);
The union of SetA and SetB is [1, 2, 3, 4, 5, 6]:

SetA + SetB
or
UNION (SetA, SetB);
The intersection of SetA and SetB is [ 3, 4 ]:

SetA * SetB
or
INTERSECTION (SetA, SetB);
The difference of SetA and SetB is [ 1, 2 ]:

SetA - SetB
or
DIFFERENCE (SetA, SetB) ;
The symmetric difference of SetA and SetB is [ 1, 2, 5, 6 ]:

SYMDIFF (SetA, SetB)
Arrays
An array is a collection of objects indexed by either one integer set or one
string set for each dimension of the array.
You can use arrays to define a collection of variables, parameters, ports,
model instances, or streams.
Note: Global arrays are not supported.
Arrays Remarks
Because an array is indexed by a set, it is possible for the index of the array
to not take consecutive integer values. You can define an array that is
indexed by an integer set that has some of the integer numbers “missing”.
This is useful for modeling trays in a distillation column, where, for example,
you may want to use a different model for a feed tray.
As arrays can be indexed by a string set, you can define an array indexed by
component names. For more information, see Chapter 3, Using the
ComponentList Property.
Arrays Examples
The following line defines an array variable. The variable SectionTemp is
indexed explicitly over the integer set from 1 to 10:
SectionTemp([1:10]) AS Temperature;
The variable FeedMolFrac is indexed by the set Components, which contains

three component names:
Components AS StringSet(["N2", "O2", "CO2"]);
FeedMolFrac(Components) AS MoleFraction;
The variable SectionMolFrac is defined as a two dimensional array. The first

dimension is defined as the set Components and the second dimension is the
set of integers from 1 to 10:
SectionMolFrac(Components,[1:10]) AS MolFraction;

The following example shows how to define an array of model instances:
MODEL Tray
.
.
END
MODEL FeedTray
.
.
END
MODEL Column
NTrays AS IntegerParameter(10);
NFeedTray AS IntegerParameter(5);
Stage([1:NTrays] - NFeedTray) AS Tray;

FeedStage AS FeedTray;
.
.
END
The model Column uses the previously defined models Tray and FeedTray.
The parameter NTrays is the total number of trays and NFeedTray is the tray
number of the feed tray. NTray is used to define an array of instances of the
Tray model. Set arithmetic is used to create an array of Trays called Stage.
Stage(5) is excluded from this array because this is the feed stage, which is
represented by a different model, FeedTray..
The definition for Stage is:
Stage([1, 2, 3, 4, 6, 7, 8, 9, 10])
Atomic Type Conversions

The CONVERTTO operator enables you to convert a string or the first element
in a set of strings to the integer value that it represents.
You can also use CONVERTTO to convert a REAL, INTEGER, or LOGICAL value
to a string. For example, you could use CONVERTTO in a task, to print out
variable values (because the PRINT statement in a task allows only strings).

CONVERTTO Syntax
Use the following syntax for CONVERTTO:
CONVERTTO '('ResultType', 'Expression')';
ResultType ................................ Can be INTEGER or STRING

Expression ................................ If ResultType is INTEGER,
Expression must be of type
STRING or STRINGSET
If ResultType is STRING,
Expression must be of type REAL,
INTEGER, or LOGICAL
Example of Using CONVERTTO

The following example shows how to use CONVERTTO to convert to integer
values:
x_string as stringparameter ("5");

x_int as IntegerParameter(CONVERTTO(INTEGER, "3"));
//initializes x_int with the value 3
x_int: CONVERTTO(INTEGER, x_string); //assigns the integer
value of x_string (5) to x_int
The following example shows how to use CONVERTTO to convert to a string:
task MyTask
s2 as StringParameter;
s2: CONVERTTO(string, V); // V is a variable accessible
to the task
print "value of V is " + s2;
end
Note that CONVERTTO cannot be used directly in a print statement.
Inheritance
Inheritance enables you to create new types which build upon existing ones.
You describe an inheritance relationship between two type definitions with the
keyword USES.

Inheritance Syntax
Type TypeName USES TypeName
Statements
END
Type.......................................... One of the supported types, such

as Model or Variable
TypeName ................................. The name of the new type being
defined
Statements ............................... Statements which extend the
definition of the inherited type
Inheritance Remarks
If you define a type using inheritance, the type inherits all the properties of
the type from which it inherits. You can then extend the type definition with
statements that add extra properties to the type.
Inheritance also enables you to refine the definition by changing the default
values of existing properties. Other than this you cannot change or remove
properties that come form the inherited type.
Note: Icons are not inherited between models.
Inheritance Example
In the following example, the variable type ColdTemp inherits all the
properties and default values from the variable type Temperature, but
overrides the default value for the property UPPER.
VARIABLE ColdTemp USES Temperature

UPPER: 273;
END
In the following example, the model HeatX inherits from the model Heater. It
extends Heater by adding the variable UA and an equation to calculate Q. It
also changes the default spec of Q to free, as Q is now calculated by an
equation:
MODEL HeatX USES Heater
UA as HTCoeff (1, Fixed);
Q: free;
Q=UA*DT;
END

Type Relatives
The inheritance relationship enables you to define a family of related types
that can be substituted for each other.
For example, in a procedure definition, you list the types of the input and
output variables expected in a call to that procedure. Using inheritance, you
can define new variables that inherit properties from the type you have
named in the procedure call argument list. You can use these new variable
types in a call to the procedure, because they are recognized as descendants
of the variable types you named in the procedure call list.
You can use related types to determine which ports are allowed to be
connected using custom streams. You can only connect two ports if the port
types are the same or are related port types. Note that if you use the built-in
Connection stream, any two ports can be connected. In this special case,
variables in the ports being connected are matched by name.
Type Relatives Example

The first example shows how two ports are created using inheritance. These
two related ports can be connected together in the flowsheet.
PORT Material
Flow AS Flow_Vol_Liq;
END
PORT NewMaterial USES Material

Temp AS Temperature;
END
MODEL Tank1
.
Out1 AS OUTPUT NewMaterial;
.
END
MODEL Tank2
.
In1 AS INPUT Material;
.
END

STREAM MaterialStream;
In1 AS Input Material;
Out AS Output Material;
In1.Flow = Out.Flow;
END
The port Material carries flow rate, and the port NewMaterial carries flow rate
and temperature values. If we create instances of the two models we can
connect their ports together with a stream of type MaterialStream. This is
because the two ports are related by inheritance. However, the temperature
value carried from the instance of Tank1 is not passed to the instance of
Tank2.In the following example, the procedure TotalMassCalc is written to
accept a variable of type Density as an input argument. However, the same
argument can accept a variable of a type inherited from Density. In this case
the variable types VaporDensity and LiquidDensity are related by inheritance
to Density.
VARIABLE Density
Value: 1000.0;
Lower: 1.0e-2;
Upper: 1.0e5;
PhysicalQuantity: "Density";
END
VARIABLE LiquidDensity USES Density
Value: 1000.0;
Lower: 10.0;
Upper: 1.0e5;
END
VARIABLE VaporDensity USES Density

Value: 1.0;
Lower: 1.0e-2;
Upper: 100.0;
END
PROCEDURE TotalMassCalc
.
.
INPUTS : Area, Length, Density;
OUTPUTS: Mass;

END
MODEL Tank
TankArea AS Area;
TankHeight AS Length;
RhoLiq AS LiquidDensity;
RhoVap AS VaporDensity;
TotalMass AS Mass;
Phase AS PhaseParameter("Liquid");
IF Phase == "Liquid" THEN

CALL (TotalMass) = TotalMassCalc(TankArea, TankHeight,
RhoLiq);
ENDIF
IF Phase == "Vapor" THEN

CALL (TotalMass) = TotalMassCalc(TankArea, TankHeight,
RhoVap);
ENDIF
END
Component Lists
A component list contains two types of information:
• A list of component names
• A list of options associated with these components. Typically this is used
to store options for calculating physical properties for mixtures of these
components
A component set is a simplified version of a component list which does not
include a list of physical property options.
In general you should use a full component list when using a physical
properties package such as Aspen Properties, and a component set if you are
not using a physical properties package.

Defining which ComponentList to use
You can define the name of the default flowsheet component list to be used
for you simulation by editing the value of ComponentList in the Flowsheet
LocalVariables table. By default this has the value “Default” which means the
component list called Default is used.
All blocks, streams and ports have a built-in property called ComponentList.
This contains the name of the component list to be used for the block, stream
or port. When a new block or stream is instanced it’s ComponentList property
is set to the flowsheet component list name. You can change this value from
the AllVariables table for the block or stream if you want to use a different
component list for the block or stream.
You can use ComponentList to index arrays over the set of components to be
used in the block, stream or port.
Syntax for Indexing Arrays using ComponentList

VariableName (ComponentList) AS VariableType;
VariableName ........................... Name you give to a variable

VariableType ............................. The type of the variable
Using ComponentList Example

The following is an extract from a model that shows a mole fraction array
being declared, and indexed over the components in the component list.
MODEL FeederPP
:
Z(ComponentList) AS Molefraction;
:
END
Using More Than One Component List

in Models
If you are using a physical properties package such as Aspen Properties to
calculate physical properties, you can use Procedure calls in your Models and
Stream types to perform the property calculations. Each call automatically
passes the components and options defined in the component list used by the
block or stream.
This mechanism enables you to use different component lists in different
blocks and streams to use different physical property options in different parts
of your flowsheet.
If a model or stream needs to use more than one component list, you can
define special parameters of type ComponentListName for storing these

component list names (ComponentListName is the same type that is used for
the built-in ComponentList property). You can then associate the appropriate
component list with each physical property procedure call by ending the call
with the appropriate component list name.
When you use the model or stream you can then change the value of each
component list name to that of the required component list.
Syntax for Defining Component List Names

ComponentListName AS COMPONENTLISTNAME;
ComponentListName ................. The name of the parameter used

to store the component list name
Using More Than One Component List in Models Example

This example shows two component lists used in the same model. These two
component lists are associated with the two sides of a heat exchanger. The
user of the model can then assign values to the HotSide and ColdSide
component list names to reflect the fluids flowing on each side of the heat
exchanger.
MODEL HeatExchanger
HotSide, Coldside AS ComponentListName;
ColdSideInlet AS INPUT Process (ComponentList:
ColdSide);
ColdSideOutlet AS OUTPUT Process (ComponentList:
ColdSide);
HotSideInlet AS INPUT Process (ComponentList: HotSide);
HotSideOutlet AS OUTPUT Process (ComponentList: HotSide);
:
:
CALL (ColdSideOutlet.H) = pEnth_Mol (ColdSideOutlet.T,
ColdSideOutlet.P, ColdSideOutlet.z) ColdSide;
CALL (HotSideOutlet.H) = pEnth_Mol (HotSideOutlet.T,

HotSideOutlet.P, HotSideOutlet.z) HotSide;
END
Built-In Type Definitions

Aspen Custom Modeler contains a number of built-in parameter and variable
types.

Built-In Parameter Types
The four built-in parameter types are:
• StringParameter
• IntegerParameter
• RealParameter
• LogicalParameter
You can use these parameter types directly within type definitions, such as
models. You can also inherit from them when you define your own parameter
types.
The attributes of each parameter type are listed below:
StringParameter Attributes
Description................................ A text string that can be used to store

a description of the parameter
Value......................................... A text string that contains the value of
the parameter
IntegerParameter
Description ............................... A text string that can be used to

store a description of the parameter
Value......................................... The value of the integer parameter
Upper ........................................ The upper bound for the value
(defaults to 32767)
Lower........................................ The lower bound for the value
(defaults to –32767)
RealParameter Attributes
Description ............................... A text string that can be used to store

Value ........................................ The value of the real parameter
Upper........................................ The upper bound for the value. Default
is 1.0e37.
Lower........................................ The lower bound for the value. Default
is –1.0e37.
PhysicalQuantity ....................... The name of the physical quantity used
for the parameter. Default is none.
Tag ........................................... Label for use in interfaces to online
applications
LogicalParameter Attributes
Description ............................... A text string that can be used to store


Value ........................................ The value of the logical parameter, can
be TRUE or FALSE.
Built-In Variable Type

The built-in type RealVariable is inherited by all variable types that you
define. This means that all variables have the RealVariable attributes.
RealVariable Attributes
Attribute Description Default Type

Derivative The rate of change of the value with 0.0 Real
respect to time. This attribute is only
active if the variable is differentiated in
a differential equation
Description A string of text that contains summary None Text

or help information for the variable
IsConnected A logical property for testing whether False Boolean

the variable is connected to a control
stream in the flowsheet. Read only
IsMatched When the variable is contained in a False Boolean

port, this is used to test whether
connected ports have equivalent
variable names. Read only.
Lower The lower bound for the variable value. -1.0e37 Real
Physical A string of text that contains the name None Text

Quantity of the physical quantity definition used
by the variable. Read only.
Record Controls recording of history for a False Boolean – True or False

variable. This is automatically set to
True when the variable is displayed on
a plot or history table.
Scale The scaling factor for the variable value 1.0 Real
used by the solver
Spec A string of text that contains the “Free” Text – must be one of:-
specification for the variable
“Free”, ”Fixed”,
”Initial”, ”RateInitial”
Tag Label for use in interfaces to online None Text

applications
Units Current UOM symbol. The available Text

values depend on the physical quantity
used for the variable
Upper The upper bound for the variable value 1.0e37 Real
Value The value of the variable 1.0 Real
Setting Attribute Values

In the modeling language an attribute value can be set using the following
syntax.

For real attributes:
e.g. Temperature.Value: 12.34;
For Boolean attributes:

e.g. Temperature.Record: True:
For text attributes:

e.g. Temperature.Spec: Fixed;
You can assign the value and spec attribute in 1 statement

e.g. Temperature: 2.25, Initial;
or only the value or the spec
e.g. Temperature: Initial;
e.g. Temperature: 2.25;
In a script the syntax is:

e.g. Temperature.Value = 12.34
For Boolean attributes:

e.g. Temperature.Record = False
For text attributes:

e.g. Temperature.Spec = “Free”
Note the full path to the variable should be included depending on the context
of the script.
Getting Attribute Values

In the modeling language an attribute value can be used for example in an if
statement.
e.g. if (Temperature.IsMatched) then …
e.g. if (Temperature.spec == Fixed) then …

2 Modeling Language for
Simulations
This chapter describes the following:

• Using units of measurement.
• Defining tasks.
• Using task statements.
• Defining an estimation simulation.
• Defining an optimization simulation.
• Defining a homotopy simulation.
For information on defining types and models, see Chapters 3 to 4; Modeling
Language for Types and Modeling Language for Models.
Using Units of Measurement

For any given physical quantity, such as temperature, you can define any
number of different units of measurement (UOM), and conversions between
these different units of measurement. You can group together units of
measurement into UOM sets, such as SI or metric, and make these UOM sets
available automatically when you start a simulation session or load a new
simulation.
The units of measurement used in your models are referred to as the base
units of measurement. When you write your own variable types and models,
you are implicitly defining the base units of measurement. These are what are
used at run time to solve your simulation. You must ensure that your model
equations are written consistently in these units of measurement.
The user interface converts between the current display UOM and the base
units of measurement. This means that values shown on plots and tables are
automatically displayed in the current UOM and values you input on forms are
automatically converted to the base units of measurement.
To determine the current units of measurement, you can either select a unit
of measurement set or set the Units attribute of individual variables.
Scripts and automation also have features to support unit of measurement
conversion. Tasks always work in the base units of measurement, and that all
values that you enter in tasks should be in these units.
2 Modeling Language for Simulations 35

The base unit of measurement and unit of measurement conversions are
defined as part of a physical quantity. A variable type is associated with a
physical quantity by assigning the name of the physical quantity to the
Physical Quantity attribute in the variable type definition.
The Modeler library works in Metric base units. Sets of conversions to other
units of measurement are built in ready for you to use. We strongly
recommend that you use this unit of measurement set for any models that
you develop. You can then benefit from all of the built in unit of measurement
capabilities, without having to reconfigure these.
You can edit the built in conversion sets, or add your own completely new
UOM set. To perform these operations, you use automation methods from a
script.
The methods you can use are:
Automation Method Description

AddUomSet Adds a unit of measurement set.
AddPhysicalQuantity Adds a physical quantity to the system and

specifies its base units of measurement.
DefineConversion Defines a conversion between base units and

another unit of measurement.
DefaultDisplayUOM Defines the units of measurement in which a

physical quantity is displayed, for an existing units
of measurement set.
SelectUomSet Enables you to select an existing unit of measure

set.
ConvertToBaseUom Converts a value from its current display unit of

measurement to its base unit of measurement.
ConvertFromBaseUom Converts a value from its base unit of measure to

its current display unit of measure.
Specifying a Physical Quantity for

Variables
You specify the physical quantity to be used as part of a variable type or
parameter type definition. The physical quantity:
• Defines the base unit of measure used for all variables of that type
• The unit of measure conversions available for all variables of that type
Syntax for Defining a Physical Quantity

The syntax for defining a physical quantity for a variable type or parameter
type is:
PHYSICALQUANTITY: "PhysQuanName";
PhysQuanName ........................ Name of a physical quantity

Physical Quantity Examples
The following example shows how the physical quantity energy is assigned to
the variable type holdup_heat:
VARIABLE holdup_heat
value: 1;
lower: -1E5;
upper: 1E5;
PhysicalQuantity: "Energy";
END
Defining Physical Quantities, Units of

Measurement, and Conversions
You can define new physical quantities, unit of measurement conversions and
unit of measurement sets. You do this by writing a Visual Basic script.
Defining the UOM Object

Start your script by defining a reference to the UOM object.
Syntax for Defining UOM Object

DIM UOM
SET UOM = ActiveDocument.uom
UOM .......................................... A Visual Basic variable. You can

choose any valid name for this
variable. The following examples
assume you have chosen the name
UOM.
Defining Physical Quantities

To use units of measurement, you need to define the physical quantities used
in your simulation.
Syntax for Defining Physical Quantities

The syntax for defining a physical quantity is:
bOK = UOM.ADDPHYSICALQUANTITY("PhysicalQuantityName",
"BaseUnitSymbol", AspenPlusQuantity, AspenPlusUnit)
bOK ........................................... Dummy boolean variable to accept

the return code from the function
call
UOM Previously defined reference to the
UOM object
PhysicalQuantityName .............. The name of the physical quantity
BaseUnitSymbol ........................ The symbol for the base units of
t f thi h i l

measurement for this physical
quantity
AspenPlusQuantity.................... The integer number of the
corresponding physical quantity in
Aspen Plus®, as entered in the
Aspen Plus file rcunits.dat.
AspenPlusUnit........................... For the given Aspen Plus physical
quantity, this is the integer number
of the Aspen Plus unit that matches
the base units for the physical
quantity.
Note: AspenPlusQuantity and AspenPlusUnit are optional and are

only used if you export models from Aspen Custom Modeler for
use in Aspen Plus.
You can use the full path for the Uom object instead of defining the UOM
variable. For example:
bOK = ActiveDocument.Uom.AddPhysicalQuantity("Length", "m")
If you define UOM once, you can save typing the full path each time.
Physical Quantity Example

The following example creates physical quantities for length and temperature
with base units of measurement of meters and degrees Celsius
respectively.
bOK = UOM.AddPhysicalQuantity("Length", "m")

bOK = UOM.AddPhysicalQuantity("Temperature", "C")
Defining Conversions between the Units of Measurement

After defining the physical quantities and base units used in your simulation,
you can define additional units of measurement. A unit of measure consists of
a symbol for the unit, and the factors for conversion to base units.
Syntax for Defining Conversions between Units of Measurement

The syntax for defining the conversion is:
bOK = UOM.DEFINECONVERSION("PhysicalQuantityName",
"UomSymbol", Multiplier, Offset)

call
UOM .......................................... Previously defined reference to the
UOM object
PhyscialQuantityName .............. Name of a previously defined
physical quantity
UomSymbol............................... The symbol of the UOM to which
you are making the conversion

from the base UOM
Multiplier .................................. The factor by which you need to
multiply the UOM value to convert
to the base UOM value
Offset ........................................ The difference you need to make to
the UOM value to convert to the
base UOM value
The conversion is made using the Multiplier and Offset in the following
equation:
BaseUnit = Multiplier * UomSymbolUnit + Offset
Units of Measurement Conversion Example

The following example defines conversions for the physical quantities
Temperature and Length. The base units of measurement defined in the
physical quantity definitions are "C" for temperature, and "m" for length.
bOK = UOM.DefineConversion("Temperature", "K", 1.0, -

273.15)
bOK = UOM.DefineConversion("Temperature", "F", 0.5556, -
17.7778)
bOK = UOM.DefineConversion("Length", "ft", 0.3048, 0.0)
Creating Unit of Measurement Sets

Unit of measurement sets enable you to group together related units of
measurement. The automation method AddUomSet enables you to add a new
UOM set name to your simulation. You can create a new UOM set using the
following syntax in a Visual Basic script:
bOK = UOM.ADDUOMSET("UomSetName")
call
UOM object
UomSetName ............................ The name you give to the UOM set
Creating Units of Measurement Sets Example

The following example creates three different units of measurement sets: for
SI units, metric units, and English units.
bOK = UOM.AddUomSet("SI")
bOK = UOM.AddUomSet("Metric")
bOK = UOM.AddUomSet("Eng")

Defining the Displayed Units of Measurement for the Units of
Measurement Set
After you have created a unit of measurement set, you can choose the unit of
measurement to be used for each physical quantity within the unit of
measurement set. These are the units of measurement that quantities will be
displayed when the units of measurement set is selected.
Syntax for Defining the Displayed Units of Measurement

The syntax for defining the displayed units is:
bOK = UOM.DEFAULTDISPLAYUOM( "UOMSet",
"PhysicalQuantityName", "UomSymbol")

call
UOM object
UomSet ..................................... The name of a previously defined
UOM set
PhysicalQuantityName .............. The name of a previously defined
physical quantity
UomSymbol............................... The UOM to be used for the
physical quantity within the UOM
set
Note: If a default display unit of measure is not specified for a

physical quantity within a UOM set, then that physical quantity
will be displayed in the base units when using that UOM set.
Defining the Displayed Units of Measurement Example

The following example defines the displayed units for length in the Eng UOM
set as ft:
bOK = UOM.DefaultDisplayUOM("Eng","Length","ft")
Selecting a Unit of Measurement Set for the Simulation

• Once you have defined a UOM set, you can select it from the Units of
Measurement item on the Tools menu. You can also select it from a Visual
Basic script using the syntax:
ActiveDocument.Uom.SelectUomSet("UomSetName")
Selecting a Units of Measurement Set Example

The following example shows how to switch to the Eng units of measurement
set, using a Visual Basic script:
ActiveDocument.Uom.SelectUomSet("Eng")

Running the Units of Measurement Script Automatically
You can ensure that the Visual Basic® units of measurement script runs
automatically each time you start a simulation session or load a new
simulation. This means you can write a script that defines all the units of
measurement sets and conversions you need and ensure that the conversions
are available each time you run a simulation.
To create a UOM script that runs automatically, you need to save the script
text to a file called:
OnNewDocumentScript.vb
You must place this file in the bin directory of your installation. For example:
C:\Program Files\AspenTech\Aspen Custom Modeler 11.1\bin

A default version of this file already exists, and defines the standard unit of
measurement sets available in Aspen Custom Modeler. You can choose to
replace these, but we would recommend that instead you append your
changes to the end of this file.
In addition, the Open Document event will run a script called
OnLoadSimulation located under the Flowsheet in the simulation being
loaded. This allows you to make changes that are specific to the simulation
being loaded e.g. simulation specific units of measurements and screen
layouts.
Converting Between Units of Measurement

You can convert values in Microsoft Visual Basic scripts between your base
units of measurement and your current units of measurement.
Syntax for Converting Between Units of Measurement

Use the following syntax in your Visual Basic Scripts:
ActiveDocument.Uom.ConvertToBaseUom (“BaseUom”,
CurrentUomValue)
and
ActiveDocument.Uom.ConvertFromBaseUom (“BaseUom”,
BaseUomValue)
BaseUom................................... The base unit of measurement

name
CurrentUomValue...................... The value of the variable in the
currently selected unit of
measurement
BaseUomValue .......................... The value of the variable in the
base unit of measure
Converting Between Units of Measurement Example

In this Microsoft Visual Basic script example, the current UOM set is English
units and the base UOM set is metric. A value of temperature of 60 Celsius is
converted to the current temperature unit of measure, Fahrenheit. A value of

length of 5280 feet is converted to the base unit of measure, meters.
Dim TempInF, LengthInMeters, UOM

Set UOM = Application.Simulation.Uom
TempInF = UOM.ConvertFromBaseUom("C", 60.0)
LengthInMetres = UOM.ConvertToBaseUom("m", 5280.0)
Using the Units of Measurement of a Control Variable

Normally variables acquire units of measurement from the variable type that
they use. However it is sometimes useful to have a variable acquire it’s units
of measurement from another variable in the same model or flowsheet
constraints section.
Syntax for Using the Units of Measurement of another Variable

MODEL
VariableName1 AS VariableType1;
VariableName2 AS VariableType2 (UseUOMOf:
"VariableName1");
END
Example of Using the Units of Measurement of a Control

Variable
This example shows how other variables in a controller model can take the
units of measurement of a variable in a different block that is connected to a
control variable.
MODEL PI_Controller
.
PV AS INPUT Control_Signal;
PVMin AS Control_Signal (Fixed, UseUOMOf: "PV");
PVMax AS Control_Signal (Fixed, UseUOMOf: "PV");
.
END
In this example, if a variable of type Temperature is connected to the input
control variable PV using the ControlSignal stream type, the variables PV,
PVMin and PVMax are all displayed in the same units of measurement as the
connected variable.
Advanced Use of the UseUOMOf Property

The UseUOMOf property is used to declare that one variable uses the same
units of measurement of another variable. For introductory information on the
UseUOMOf property, see Using the Units of Measurement of a Control Variable

on page 2-42. The UseUOMOf property is also capable of more complex
associations.
See Also
Using Variables whose UOM is Produced from the UOMs of Other Variables
Applying only Scaling to a Variable
Using Variables whose UOM is Determined from the UOMs of

Other Variables
You can use the UseUOMOf property for a variable whose UOM is a product or
division of the UOMs of other variables.
Example of Using Variables whose UOM is Determined from the

UOMs of Other Variables
The following example is a model that calculates the ratio and product of two
inputs:
Model RatioBlock
input1 as input RealVariable(fixed);
ratio as output RealVariable(UseUOMOf: "input1/input2");

prod as output RealVariable(UseUOMOf: "input1*input2");
ratio*input2 = input1;
prod = input1*input2;
End
Notes:
• The 'ratio' variable is declared as using the UOM of input1

divided by the UOM of input2. If, for example, input1 is
connected to a temperature variable displaying in Fahrenheit
and input2 to a length variable displaying in feet, the display
units of measurement will be Fahrenheit/ft. Note that in this
case the conversion for display units to base units
(Centigrade/meter) will not include the offset between
Fahrenheit and Centigrade: only the scaling is applied.
• The 'prod' variable is handled in a similar way to the ratio

variable except in this case a "*" is used to reflect the
multiplication operation.

Applying only UOM Scaling to a Variable
You can use UseUOMOf to apply only the unit of measurement scale factor to
a variable.
Example of Applying only UOM Scaling to a Variable

The following example is a model that calculates the difference between two
variables:
Model RatioBlock
diff as output RealVariable(UseUOMOf: "ScaleOf: input1");
diff = input1 - input2;

End
In this example, if the 'diff' variable was simply declared to use the UOM of
input1 then when converting to display units, the conversion offset would be
inappropriately applied (for example, a zero Centigrade temperature
difference would be converted to 32.0 in Fahrenheit). To ensure that only the
scale is applied, this example uses the "ScaleOf:" qualifier before the name of
the variable.
Accessing Variable Values in Alternative UOMSets

You can access the values of variables in alternative UOM Sets from the
current UOMSet through automation.
This means you can access variable values in a Microsoft® Visual Basic®
Script in a different unit of measure and use the alternative unit of measure
for intermediate calculations or results output.
Syntax
The syntax for a script at Flowsheet level is:
BlockName.VariableName.VALUE("UOMName")
The syntax for the script at Model level is:
VariableName.VALUE("UOMName")
BlockName ................................ Name of a block on your flowsheet

VariableName ........................... Name of a variable in the block
UOMName ................................. Can take one of the following
values:
BaseUnits Returns the variable value in

base units

CurrentUnits Returns the value in the
current units of measurement
UnitName The name of the unit of
measurement in which the
value is to be returned
Example of Accessing Variable Values in Alternative UOM Sets

For a Visual Basic Script at Flowsheet level:
Application.Msg "Temperature in base units is " +

B1.Out1.Temp("BaseUnits")
Application.Msg "Temperature in current units is " +
B1.Out1.Temp("CurrentUnits")
Application.Msg "Temperature in Celsius is " +
B1.Out1.Temp("C")
Application.Msg "Temperature in Fahrenheit is " +
B1.Out1.Temp("F")
If the BaseUOMSet is Metric and the CurrentUOMSet is US units, the output
from the script is:
Temperature in base units is 22.2222

Temperature in current units is 72
Temperature in Fahrenheit is 72
Temperature in Celsius is 22.2222
Defining Tasks
A task is a set of instructions that defines a sequence of actions that take
place during a dynamic simulation. For example, you can use tasks to:
• Define disturbances in feed conditions or controller set points at a pre-
defined time
• Trigger events when certain conditions are met. For example, you can use
a task to provide basic level control in a tank by closing an output valve
when the fluid level falls below a certain value.
You can define two kinds of tasks:
Task Type Started By Explorer

Icon
Event-driven A start condition, which is written as a
conditional expression.
Callable Being called from other tasks, either

an event-driven task or another
callable task.
Does not have a start condition.

There are three locations where you can define a task:
Location Can be Event- Can be Callable

Driven
Model Yes Yes
Flowsheet Yes Yes
Library Task folder No Yes
Event-Driven Tasks
Event-driven tasks are associated with a defined event and execute when that
event occurs.The events with which a task can be associated are:
Event Example
A specified simulation time AT 20.0
A logical expression involving variable tank.level >= 2.5

values
Variables can appear on both sides of the conditional expression.

The allowable comparison operators are >=, >, = =, <=, <>.
= = means equal to, and <> means not equal to.
For time expressions you can use only >=, > or = =.
Event-Driven Task Syntax

The syntax for an event-driven task is either:
TASK TaskName RUNS AT Time
TaskStatements
END
or
TASK TaskName RUNS ONCE WHEN Condition
TaskStatements
END
TaskName ................................. Name of the task

Condition .................................. A logical expression testing the
value of simulation variables
(including time)
Time An expression evaluating to a
positive real value representing a
simulation time
ONCE Optional keyword (see following
Remarks)
TaskStatements ........................ The sequence of task statements
that are executed when the task is
triggered

You can also use the following system generated events, in place of WHEN
Condition, to trigger your tasks:-
• BEFORE INITIALIZATION
• AFTER INITIALIZATION
• BEFORE STEP
• AFTER STEP
• AFTER ERROR
The associated task triggers when the system generated event occurs. That
is; before or after simulation initialization, before or after a dynamic run takes
a step, or when any error occurs.
Note: An event-driven task cannot take input parameters.
Event-Driven Tasks Remarks

If you include the ONCE keyword in the task start condition, the task runs
only once during the simulation, when then start condition first becomes true.
This can be used to model irreversible events.
A task without the ONCE keyword starts whenever the start condition
becomes true.
The ONCE keyword is implied by AT Time.
The trigger RUNS ONCE When Time == 20.0 is the same as RUNS AT 20.0.
Note: You can also use a SWITCH statement in a model

definition to model an irreversible process.
Examples of Event-Driven Tasks

In the first example, at time 5.0, the flow to a vessel is set to zero. The tank
empties until the fluid level is below 0.01, when the inlet flow is resumed.
TASK test1 RUNS AT TIME 5.0
Input1.Flow: 0.0;
WAIT FOR Level <= 0.01;
Input1.Flow: 5.0;
END
In the second example, the inlet flow is set to 3.0 whenever the level is below
0.01. This can happen at any time in the simulation.
TASK test2B RUNS WHEN Level <= 0.01

Input1.Flow: 3.0;
END

In the following example, the event-driven task is defined to run once only. In
this case a bursting disk is modeled so that once the burst pressure is
exceeded, the disk remains open:
TASK Burst RUNS ONCE WHEN Tank.Press > 9.0
Valve.CV: 1.0E6;
END
Finally, in this example, the associated task triggers after an initialization
occurs.
Task mytask RUNS AFTER INITIALIZATION
Input.Flow: 5.0;
END
Callable Tasks
Callable tasks are called by other tasks instead of being triggered by an
event. You can optionally define callable tasks with parameter call lists, with
the calling task passing the parameter values.
You can make calls to tasks within models, to tasks in a tasks folder, or other
tasks in your flowsheet. You cannot call a task in a model from a flowsheet
task, and you cannot call tasks in the flowsheet from a task in a model or task
folder.
Callable Task Syntax

The syntax for a callable task is:
TASK TaskName (ParameterList)
TaskStatements
END
TaskName ................................. The name of the task

ParameterList ........................... An optional list of parameters to
be passed to this task when it is
called
TaskStatements ........................ The task statements that are
executed when the task is run
Use a CALL statement to call a callable task.
CALL Statement Syntax

The syntax for the CALL statement is:
CALL TaskName (ParameterList);
TaskName ................................. The name of the task being run

ParameterList ........................... A list of assignments to the
parameters of the called task. Only

required if the called task has on or
more input parameters.
Example of a Callable Task

The following example shows two callable tasks being called from an event-
driven task:
TASK SubTaskA
Input1.Flow: 4.0;
END
TASK SubTaskB
Input1.Flow: 1.0;
END
TASK Test RUNS AT Time 1.0

Input1.Flow: 5.0;
WAIT FOR Level >= 3.0;
CALL SubTaskB;
WAIT FOR Level <= 0.4;
CALL SubTaskA;
END
Callable Task Parameter Lists

You can pass parameters from a task to a callable task. Use this to pass
variable names or calculated values from one task to another. You can use
parameterized tasks to create generic tasks that are not dependent on a
particular block name, and then pass information that is specific to a block to
that generic task.
Parameterized Task Syntax

The ParameterList has the following syntax:
(ArgumentName1 AS ArgumentType1; ArgumentName2 AS
ArgumentType2; ...)
ArgumentName ......................... The name of the variable or

parameter within the callable task
ArgumentType .......................... The type of the argument. This can
be RealVariable, RealParameter,
IntegerParameter or
StringParameter.

Note: Use argument type RealVariable to pass in a simulation
variable. Use RealParameter, IntegerParameter and
StringParameter to pass in parameters defined either in the
simulation flowsheet, or in the calling task.
Callable Task Parameters Example

The example shows how you can write a callable task that uses parameters
passed from the task that calls it. The information passed to the callable task
is the name of a variable to be ramped, the target value of the ramp, the
duration of the ramp and the type of ramp used. The task GenericRamp can
be called from any model task or from a flowsheet task.
TASK GenericRamp (VarName AS RealVariable; Target AS

RealParameter; Period AS RealParameter; RampType AS
StringParameter)
IF RampType == "Linear" THEN
RAMP (VarName, Target, Period);
ENDIF
IF RampType == "Scurve" THEN
SRAMP (VarName, Target, Period);
ENDIF
END
TASK DoTheRamp RUNS WHEN B4.Level >= 10.0

TargetValue AS RealParameter(0.0);
RampDuration AS RealParameter(0.1667);
RampType AS StringParameter("Scurve");
CALL GenericRamp(B1.Inlet.Flow, TargetValue, RampDuration,
RampType);
WAIT FOR B4.Level <= 1.0;
TargetValue: 4.25;
RampDuration: 0.2;
RampType: "Linear";
CALL GenericRamp(B1.Inlet.Flow, TargetValue, RampDuration,
RampType);
END

Tasks and Opening Simulations
The state of any active tasks is not stored when you save a simulation, and so
cannot be restored when you load the simulation. Therefore it is
recommended that you do not save your simulation at a time when any tasks
are currently running, or that if you do you ensure that after loading you run
the simulation from time=0.
Using Task Statements

Task statements can perform the following tasks:
• Assign a value to a variable.
• Define a conditional expression that depends upon the value of a
simulation variable.
• Ramp the value of a simulation variable.
• Suspend or restart the task.
• Create snapshots.
• Execute parallel tasks.
• Print.
• Suspend a simulation.
• Invoke a script from the task.
• FOR loops in tasks.
• Task Language Quick Reference
Assigning a Value to a Variable in

Tasks
You can define parameters and sets that are local to a particular task. The
following types can be used:
• IntegerParameter
• RealParameter
• LogicalParameter
• StringParameter
• IntegerSet
• StringSet
The syntax is the same as for defining parameters and sets in models.
You can assign values to local parameters, model parameters, and model
variables using the : (colon) assignment operator.
Example of Assigning a Value to a Variable

The following example shows how a value is assigned to a variable:
TASK Test3 RUNS WHEN h < 0.001

Fin as RealParameter;
Fin: 21.5;
Tank1.flowin: Fin;
Tank1.tempin: 50.0;
END
UOM Conversions in Tasks

Tasks are usually written in the base units. However, limited UOM conversions
are possible with the following syntax:
VarToAssign: Expression {UOMString};
VarToAssign………………………… The task parameter or model

variable to be assigned, which has
Units of measurement defined.
Expression…………………………… A Real Expression specifying the
value to assign.
UOMString…………………………… A string specifying the Units of
measurement of Expression. This
must be a valid unit of
measurement for VarToAssign.
Example of Assigning a Value to a Variable with Units of Measurement

B1.Temperature: 25 {C};
Using the Random Function

In tasks, you can use the RANDOM function in an assignment, with either a
UNIFORM or NORMAL distribution.
The syntax is as follows:
VarToAssign : RANDOM(MeanValue, RangeValue, Distribution);
VarToAssign………………………… The task parameter or model

variable to be assigned.
Mean Value………………………….. A Real Expression specifying the
mid-point of the possible values to
assign.
RangeValue…………………………. A Real Expression specifying the
range of the values to assign.
Distribution…………………………. UNIFORM or NORMAL

Note
• If the Distribution is NORMAL, the RangeValue is taken to be the
standard deviation of a Gaussian Distribution about the specified
MeanValue.
• If the Distribution is UNIFORM, the returned value will be in the range
MeanValue-RangeValue to MeanValue+RangeValue.
Example of Assigning a Value to a Variable with Units of Measurement

S1.Signal: RANDOM(s1.Signal,0.5,NORMAL);
The RANDOM function may be used in an expression:
HeatIn: Q + RANDOM(10,2.0,UNIFORM);
The IF Construct
Use the IF construct to control which statements are executed dependant
upon whether a conditional expression is true or false.
IF Construct Syntax
The IF construct has the following syntax:
IF Condition THEN
TaskStatements1
ELSE
TaskStatements2
ENDIF
Condition .................................. A conditional expression

TaskStatements ........................ The statements to be executed if
the condition is true.
TaskStatements2 ...................... The statements to be executed if
the condition is false.
IF Construct Remarks
The ELSE and TaskStatements2 are optional.
Example of Defining a Conditional Equation

The following example assigns a value to one variable dependant upon the
value of another variable:
TASK OutFlow RUNS WHEN time > 0.001

IF Tank1.Level > 0.1 THEN
Tank1.Flowout: 1.0;

ELSE
Tank1.Flowout: 0;
ENDIF
END
Ramping the Value of a Simulation

Variable in Tasks
There are two ramp functions that you can use to change the value of a
variable over a period of time.
Ramp Function Syntax

RAMP (Variable, Target, Period, Ramptype);
SRAMP (Variable, Target, Period, Ramptype);
Variable .................................... The name of the variable whose

value is to be ramped. The variable
must have a specification of fixed.
Target ....................................... The value to which the variable’s
value is to be changed
Period ....................................... The period of time over which the
ramp takes place
Ramptype ................................. Can be either DISCRETE or
CONTINUOUS.
CONTINUOUS ramps are re-
calculated whenever the integrator
moves its time. Continuous ramps
incur a small performance penalty,
but can be considerably more
accurate when used with a variable
step integrator. DISCRETE ramps
are re-calculated at communication
points only.
The ramptype may be omitted; in
which case the default ramptype of
CONTINUOUS is used.
RAMP is a simple linear ramp function . SRAMP is a sinusoidal ramp, which

gives a smoother S-shaped curve for the ramped variable. With SRAMP, if the
ramp is decreasing the value of a variable then the ramp follows the shape of
a sinusoidal curve between 0 and pi. If the target is an increase then the
curve follows the shape of a sinusoidal curve between pi and 2 x pi.
Example of Ramping the Value of a Variable

The following example shows the use of ramping:
TASK Task4 RUNS AT TIME 4.0

RAMP (Input1.flow, 5.0, 2.0);

/* Flow changes to 5.0 linearly over a period of 2 time
units */
SRAMP (Input1.temp, 15.0, 3.0);
/* Temperature changes with an S-shaped curve to 15.0 over
a period of 3 time units */
END
Suspending and Restarting a Task

You can suspend the execution of a task until a particular condition is met.
You can use the following statements to suspend a task:
Statement Syntax Use

WAIT RealExpression; Suspends the execution of the task for the
number of time units represented by
RealExpression.
WAIT FOR Condition; Suspends the execution of the task until

Condition is met.
RESTART AFTER Suspends the execution of a task for the

RealExpression; number of time units represented by
RealExpression, and then ends the task. The
task will then restart as soon as its start
condition is true.
RESTART WHEN Condition; Suspends the execution of a task. When

Condition is met, the task ends. The task will
then restart as soon as its start condition is
true.
RESTART; Ends the execution of a task. The task will

then restart. Execution will continue from
the first statement after the start condition.
RETURN; Suspends the execution of a task. The task

will then restart as soon as its start condition
is true.
Note:
• You can use RESTART AFTER, RESTART WHEN, and RESTART

in event-driven tasks only. You cannot use these statements
in callable tasks.
• No statements after RESTART will be executed, therefore

these statements should normally be the last statement in a
task.
Examples of Suspending and Restarting Tasks

The first example shows how to use a WAIT statement to suspend a task:

TASK TurnOffFeeds RUNS AT TIME 10
valve1.position: 0.0;
WAIT 0.2;
valve2.position: 0.0;
END
The following example shows the use of a WAIT FOR statement that causes
the task to pause until a condition becomes true.
The task starts when a fluid volume in a vessel falls below a certain value.
The feed flow rate to the task is immediately increased to 10.0 until the fluid
volume in the tank reaches a certain value, at which point the feed flow rate
is returned to its original value:
TASK WaitForTest RUNS WHEN Tank1.Vol =< 0.5

OriginalFlow AS RealParameter;
OriginalFlow: Tank1.Input1.Flow;
Tank1.Input1.Flow: 10.0;
WAIT FOR Tank1.Vol >= 1.0;
Tank1.Input1.Flow: OriginalFlow
END
The following example shows the use of a RESTART WHEN statement.
The task starts running at a time=10. The feed flow rate ramps up and the
feed component molefractions are changed. Once a certain level of ethanol is
reached in the tank, the feed compositions are changed and the flow rate
ramps down. The task now restarts, based on the level of ethanol in the tank.
The cycle carries on repeating.
TASK RestartWhenTest RUNS AT TIME 10.0

RAMP (Tank1.Input1.Flow, 10.0, 5.0);
Tank1.Input1.x("Water") : 0.1;
Tank1.Input1.x("Glucose"): 0.9;
WAIT FOR Tank1.Output1.x("Ethanol") >= 0.6;
Tank1.Input1.x("Water") : 1.0;
Tank1.Input1.x("Glucose"): 0.0;
RAMP (Tank1.Input1.Flow, 1.0, 3.0);
RESTART WHEN Tank1.Output1.x("Ethanol") <= 0.05;
END

Creating Snapshots in Tasks
You can use a task to create a snapshot of the current values of the variables
in the dynamic simulation.
You can use this to record the simulation values when a certain process
condition is reached.
CREATE SNAPSHOT Syntax

The syntax for creating a snapshot is:
CREATE SNAPSHOT SnapshotName;
SnapshotName.......................... The name of the snapshot to be

created
Creating Snapshots Example

The following example shows how to create a snapshot at the moment a
process condition is reached. You can now rewind or apply the values of all
the variables in the simulation at the point when the process condition is
achieved.
TASK Task2 RUNS AT Time 10.0

Flowin.flow: 5.0;
WAIT FOR Level >= 2.5;
CREATE SNAPSHOT "Tank full";
END
Executing Tasks in Parallel

You can execute a number of callable tasks in parallel using the PARALLEL
construct. Each action is executed until it completes. The parallel section
completes when all of the Calls are complete.
PARALLEL Syntax
The syntax for grouping actions in parallel is:
PARALLEL
CALL Task1;
CALL Task2;
:
CALL Taskn;
ENDPARALLEL;
Task .......................................... The name of a callable task to be

called in parallel

Example of Executing Tasks in Parallel
The following example shows how to use of PARALLEL to call three sub-tasks
simultaneously:
TASK P1
RAMP(Input1.Flow, 2.5, 4.0);
END
TASK P2
END
TASK P3
END
TASK Task1 RUNS AT TIME 1.25

PARALLEL
CALL P1;
CALL P2;
CALL P3;
ENDPARALLEL;
Input4.Flow: 0.0;
END
All three Ramps start at time 1.25. The last task to finish is Task P2, which
completes at 6.25, at which time Input4.Flow is changed to 0.0.
Note: If the event task Task1 is created at flowsheet level, the 3

callable tasks P1, P2 and P3 should be created as separate tasks
at flowsheet level. You cannot put the definitions of P1, P2 and
P3 in with the definition of Task1.
However if these tasks were owned by a model definition, all the
task definitions shown above could be included in the model
definition. Any block that used that model definition would have
its own copy of the tasks. The event task could then be activated
separately for each block.
PRINT Statements in Tasks

You can print a message from a task to the Simulation Messages window.

PRINT Syntax
PRINT text1 + text2 + … ;
text ........................................... A quoted text string, or a

StringParameter
PRINT Statement Examples

The following example shows the use of the print statement within a task:
TASK RampFlow RUNS AT Time 1.0

PRINT "Start Task RampFlow";
RAMP (Input1.Flow, 2.5, 5.0);
PRINT "Task RampFlow Finished";
END
The following example shows how to use CONVERTTO to convert to a string:
TASK MyTask
s2 as StringParameter;
s2: CONVERTTO(string, Input1.Flow);
print "Value of Input1.Flow is " + s2;
END
Note that CONVERTTO cannot be used directly within a PRINT statement.
Pausing a Simulation from a Task

You can use a PAUSE statements to suspend a simulation.
PAUSE Syntax
The syntax for suspending a simulation is:
PAUSE;
When a pause statement executes, the simulation is placed in a paused state,
but remains active, and can trigger again once the run is continued. If the
simulation is then run on from the paused time, the task continues execution
at the next statement.
Example of Pausing a Simulation

The following example makes a step change to a variable, waits for a process
condition to be reached and then pauses the simulation:
TASK MyTest1 RUNS AT TIME 5.0

FlowIn.Flow: 10.0;

WAIT FOR FluidHeight >= 4.5;
PAUSE;
END
Invoking a Script from a Task

You can use the Invoke command within a task to invoke a script.
INVOKE Syntax
INVOKE (Output1, Output2, ...) : ScriptLocation.ScriptName
(Input1, Input2, ...);
ScriptLocation ........................... The optional location of the script

to be run. If omitted the script is
assumed to be in the same
location as the Task.
ScriptLocation can be one of:
LIBRARY - The script is in the
Custom Modeling, Scripts folder.
LibraryName - The script is in
the Scripts folder of the named
library. The library must be open
when the task runs.
ScriptName ............................... The name of the script to be run.
Input......................................... An input argument to the script.
Can be a integer, real or string
value, or the name of a variable or
parameter. Use of input
arguments is optional.
Output ...................................... An output argument from the
script. This is the name of a
variable or parameter. Use of
output arguments is optional.
Input parameters passed to the script will be accessible in the script as

members of the Inputs collection. For example the first input argument will be
accessible in the script as Inputs(1).
Output parameters from the script will be accessible in the script as members
of the Outputs collection. For example the first input argument will be
accessible in the script as Outputs(1). On entry to the script the values of the
Outputs collection are initialized to the current values of the output
variables/parameters.
The task will continue with the next statement once the script has completed
Examples of Invoking a Script from a Task

The following example invokes a script called ShowForm. The script has no
arguments:
INVOKE ShowForm;

The following example calls the CalculateRamp script to determine how far to
ramp a variable:
TASK controlstep RUNS AT Time 1.0

target as realparameter;
interval as realparameter;
INVOKE (target, interval):
CalculateRamp(tank1.flowin.flow);
RAMP(tank1.flowin.flow, target, interval);
END
The CalculateRamp script could take the form:
outputs(1) = inputs(1) * 2
outputs(2) = 3.0
FOR Loops in Tasks

You can use a FOR loop in a task to execute a loop a specified number of
times.
FOR Loop syntax
FOR Index IN SetExpression DO
TaskStatements;
ENDFOR;
Index…………………………………. Loop Index, which will be an

integer or string, depending on
the type of the set expression.
SetExpression ........................... A set of values for the Loop Index
to take.
TaskStatements ........................ The statements to be executed.
FOR Loop Remarks

When the SetExpression is an integer set, Index starts with the lowest value
in the set on the first iteration, and takes successively greater values on each
subsequent iteration.
When the SetExpression is a string set, Index takes a different element from
the set on each iteration, but the order in which the elements are taken is
undefined.
FOR Loop example

The following example uses a task to follow a time profile in a parameter
array. The array DataArray is assumed to be indexed by the integer set
DataPoints, both of which are declared in the enclosing model.
Task FollowData Runs Once Always

For iPoint in DataPoints Do
RAMP(X,DataArray(iPoint),1.0);
EndFor;
End // Task FollowData
REPEAT Loops in Tasks

You can use a REPEAT loop in a task to execute a loop until a condition is
met. The statements in the body of the loop will always execute at least once.
REPEAT Loop syntax
REPEAT
TaskStatements;
UNTIL Conditon;
REPEAT Loop example

The following example adds a random value to the level, then waits for a time
of 1. The loop continues executing until the volume exceeds 10. You could
imagine that this is modeling someone filling a tank using a bucket – but it
might be better if the wait period were also a random value!
Task RepeatExample Runs At 0.0

Repeat
Volume:Volume + Random(1,0.3,NORMAL);
Wait 1;
Until volume>10;
End // Task RepeatExample
WHILE Loops in Tasks

You can use a WHILE loop in a task to execute a loop while a condition is met.
If the condition is initially FALSE, the statements in the body of the loop are
not executed.
WHILE Loop syntax
WHILE Expression DO
TaskStatements;
ENDWHILE;

WHILE Loop example
The following example increments the volume of one component in a mixture
while the mixture is acidic. How many times the loop executes depends on
the initial PH of the mixture, and the ratio of the volumes of the components.
The loop includes a wait to allow time for the components to react.
TASK WhileExample RUNS AT 0.0

while (PH < 7.0) do
vol("A") : vol("A") + 0.05;
wait 0.1;
endwhile;
End // Task WhileExample
Task Language Quick Reference

The following is a quick reference guide to the task language.
PARAMETER ParameterType USES BuiltInType

VALUE: DefaultValue;
VALID AS SetType (ValidList);
END
Defines a parameter type.
SetName AS SetType (SetContents) ;

Defines a set
Variable.Attribute : value;
Assigns a value to variable or parameter.
IF Condition THEN
TaskStatements1
ELSE
TaskStatements2
ENDIF
Controls task statements based on a condition.

CREATE SNAPSHOT SnapshotName;
Creates a snapshot
PARALLEL
CALL Task1;
CALL Task2;
:
ENDPARALLEL;
Runs tasks in parallel.
PAUSE;
Pauses the task and the simulation. The task resumes if the simulation is
continued.
PRINT text1 + text2 + … ;

Outputs message.
RAMP (Variable, Target, Period, Ramptype);

SRAMP (Variable, Target, Period, Ramptype);
Ramps a variable value.
WAIT RealExpression;
Suspends the execution of the task for the number of time units.
WAIT FOR Condition;

Suspends the execution of the task until Condition is met.
RESTART AFTER RealExpression;

Suspends the execution of a task for the number of time units.
RESTART WHEN Condition;

Suspends the execution of a task.
CALL TaskName (ParameterList);

Calls another task.

TASK TaskName RUNS ONCE WHEN Condition
TASK TaskName RUNS WHEN Condition
TASK TaskName RUNS AT time
Runs a task based on a condition.
INVOKE (Output1, Output2, ...) : ScriptLocation.ScriptName

(Input1, Input2, ...);
Invokes a script.
Defining Flowsheet Constraints

You can define the following in a flowsheet constraint definition:
• Variables
• Parameters
• Equations and procedure calls
• Assignments
• Inequalities
You can use the flowsheet constraints section to write any equations that
include variables from different blocks on your flowsheet.
To enter flowsheet constraints, within Simulation Explorer go to the Flowsheet
item within Flowsheet and select Edit.
The syntax used for declaring variables, parameters, equations and procedure
calls and assignments is the same as that used within model definitions, as
described in chapter 4. The syntax for in equalities is described in Defining
Optimization Constraints.
To refer to variables within blocks in the flowsheet you need to include the
block name. For example to refer to the variable Tout in the block HX1 use:
HX1.Tout
You can also refer to variables within streams in a similar way. For example to
refer to the variable T in stream S1 use:
S1.T
If you have a block and stream with the same name, or if you are editing an
Aspen Dynamic simulation, you need to explicitly define whether you are
referring to a variable within a block or a stream. To do this use the syntax:
BLOCKS(”BlockName”).VariableName
STREAMS(”BlockName”).VariableName
For example:
BLOCKS(”HX1”).TOut
STREAMS(”S1”).T

If a block is within a hierarchy block you also need to include the hierarchy
block name. For example if HX1 is in the hierarchy block RefrigPlant, then
refer to Tout as RefrigPlant.HX1.Tout.
Flowsheet Constraint Examples

The following example shows a flowsheet constraint equation which equates
the duty for two heat exchangers in the flowsheet.
CONSTRAINT
HX1.Q = -HX2.Q;
END
The next example shows how to specify values for the ambient temperature
variables within three blocks in the flowsheet.
CONSTRAINT
C1.Tamb = 20, Fixed;
END
The last example shows how to declare a new flowsheet variable which
represents ambient temperature, and equates the ambient temperature to
the local values within blocks in the flowsheet.
CONSTRAINT
Tamb as Temperature;
Tamb: 20, Fixed;
C1.Tamb = Tamb;
C2.Tamb = Tamb;
C3.Tamb = Tamb;
END
Defining Optimization
Constraints
Optimization is used to manipulate decision variables to calculate the
maximum or minimum possible value of an objective function. Typically you

need to apply constraints to the optimization to ensure that the solution
meets product specifications, and is within safe and practical operating limits
of the process.
Constraints can be defined within the Optimization tool graphical interface,
and this approach is recommended for new simulations. The ability to define
constraints in input language has been retained for backwards compatibility,
and is documented in this section.
You can define constraints for your optimization simulation in the Flowsheet
Constraints section. You can define two types of optimization constraints:
Type of Used for

constraint
Inequality Ensuring the optimization solution is within process
limits. For example pressure is below maximum
working pressure, impurity levels are below
maximum allowed, and so on.
Equality Defining values that must be exactly satisfied when

performing an optimization run.
Note: An equality constraint is different from an equation in the

Constraints section because:
• An equality constraint does not reduce the degrees of

freedom.
• An equality constraint is only active during optimization

simulations. In other runs modes it will be ignored.
When using constraints you should take care to ensure that there is a feasible
solution within the constraints you specify. This is particularly important when
using equality constraints.
When performing a dynamic optimization the constraints defined in the
Flowsheets Constraints section must only be satisfied at the end of the
dynamic run. They are not applied at other times during the dynamic run. In
other words, they are end point constraints.
Syntax for Constraints

ConstraintName: Expression1 ComparisonOperator Expression2;
ConstraintName ........................ Optional name for identifying the

constraint
Expression ................................ An expression
ComparisonOperator ................. One of:
<= -less than or equal to
>= - greater than or equal to
== - equals

It is recommended that you name your constraint equations so that you can
easily identify constraints when the results are reported.
Examples of Defining Constraints

The following example shows a Flowsheet Constraints definition that contains
a number of inequality constraints.
MAX_EFFLUENT_CONC as RealVariable(fixed, 1E-3);

MAX_TEMPERATURE as RealVariable(fixed, 124);
MIN_TEMPERATURE as RealVariable(fixed, 20);
MAX_PRESSURE as RealVariable(fixed, 6.3);
MIN_PRESSURE as RealVariable(fixed, 1.0);
Conc: PURGE.X1("Methanol") <= MAX_EFFLUENT_CONC;

MinTemp: MEREACTOR.TEMP >= MIN_TEMPERATURE;
MaxTemp: MEREACTOR.TEMP <= MAX_TEMPERATURE;
MinPres: MEREACTOR.PRESSURE >= MIN_PRESSURE;
MaxPres: MEREACTOR.PRESSURE <= MAX_PRESSURE;
Defining Steady State

Constraints
You can define constraints for your steady state optimization simulation
directly in the Steady State Constraints tab in the Optimization Tool dialog.
To specify a constraint on a free variable:
1 Open the Optimization dialog box make sure Steady State Optimization is
selected in the Setup tab.
2 Click the Steady State Constraints tab.
3 To add variable names to the grid, do one of the following:
Type a variable name in the Variable text box and then click the Add
button.
Drag and drop variable names from forms onto the grid.
Click Find, and drag and drop variable names from Variable Find onto the
grid.
4 Edit the upper and lower bound values to the values you wish the variable
to be constrained between during the steady state optimization
simulation. These bounds are the same as the variable bounds by default.

Notes:
• Constraints in the Steady State Constraints grid are applied

only for steady state optimization simulations. They are not
active in any other simulation.
• They do not change the degrees of freedom of your problem.
• Use the Active switch to conveniently change between

different constraints. Constraints are only active if the Active
box is ticked.
• Use the Remove or Remove All buttons to remove one or all

of your constraints.
Defining Dynamic Constraints

You can define constraints for your dynamic optimization simulation directly in
the Dynamic Constraints tab in the Optimization Tool dialog. There are two
types of dynamic constraint you can specify:
Final Time The Constraint is applied at the final time of the

simulation only.
Path The Constraint can be applied at a finite number of
points over the dynamic optimization time horizon.
To specify a constraint on a free variable:
1 Open the Optimization dialog box ensure Dynamic Optimization is selected
in the Setup tab.
2 Click the Dynamic Constraints tab.
3 To add variable names to the grid, do one of the following:
Type a variable name in the Variable text box and click Add.
Drag and drop variable names from forms onto the grid.
Click Find, and drag and drop variable names from Variable Find onto the
grid.
4 Choose the constraint type (Final time or Path) in the Constraint Type
column.
5 If you choose Final Time, press the Edit button to define the upper and
lower bounds you wish the variable to be constrained to at the final time.
These bounds are the same as the variable bounds by default.
6 If you choose Path, press the Edit button to define the Path Constraints.>
Notes:
• Constraints in the Dynamic Constraints grid are applied only

for dynamic optimization simulations. They are not active in
any other simulation.
• They do not change the degrees of freedom of your problem.
• Use the Active switch to conveniently change between

different active constraints.
• Use the Remove or Remove All buttons to remove one or all

of your constraints.
• The values of the variables at each constraint plus the

corresponding Lagrange Multipliers after a successful dynamic
optimization are available in the Final Time Constraint and
Interior Point grids (available from the Edit button).
• At most you can have one final time and one path constraint
active on the same variable.
Path Constraints
These are constraints on free variables at a set of finite interior points in time
over the time horizon of your dynamic optimization simulation. They can be
used to discretize constraints on variables active over all time into a series of
finite interior point constraints. For an example, refer to the Dynamic
Optimization Example.
To define Path Constraints for a dynamic optimization simulation:
1 Select “Path” in the Constraint Type column of the Dynamic Constraints
grid for the variable you wish to define a path constraint.
2 Click Edit.
3 In the Interior Points dialog, choose how many interior points per element
you want to use. This is the finite number of points per element at which
you wish to constrain the variable. The points are spread evenly through
each element, for example: if you choose 1, the constraint is at the end of
each element; if you choose 2, the constraint is at 0.5 through the
element and at the end of the element, and so on.
4 In the Interior Points dialog, specify the upper and lower bounds you want
to constrain the variable to at each interior point.
5 Use the Active switch to define which elements and interior points you
want to constrain the variable. For example, you may only wish to
constrain variable during the first few elements of the simulation.
Notes
• The default values of the upper and lower bounds are the
variables upper and lower bounds.

• If you define some constraints and then change the number
of interior points per element, the existing interior point
constraints will be re-sampled to the new number of interior
points, thus avoiding re-entering data.

Types
This chapter describes how to define the following types:

• Variables
• Parameters
• Ports
• Streams
• Properties
• Procedures
Definition of Models is covered in Chapter 4.
Defining Variable Types

Aspen Custom Modeler includes a large number of pre-defined variable types.
To see these, in Simulation Explorer look in the Variable types folder of the
Modeler library. If an existing variable type meets your needs we recommend
that you use this. However for some applications you may need to define
additional variable types.
Use the VARIABLE keyword to define variable types. Once a variable type is
defined you can then create variables that use that type, for example within
models. These variables inherit their default property values from the variable
type. Variable types are a convenient way of defining variables with common
properties, and normally correspond to physical quantities such as pressure,
flow rate and so on.
Variable Type Definition Syntax

VARIABLE VariableType USES InheritedVariableType
Property1: PropertyValue1;
Property2: PropertyValue2;
:
END
VariableName ........................... The name of the new variable type

InheritedVariableName............. The optional name of an existing
variable type from which the new
3 Modeling Language for Types 72

variable type will inherit its
properties
Property.................................... A property whose default value
you wish to change from that in
the inherited type
PropertyValue ........................... The default value for the property
If you do not inherit from an existing variable type, the new variable type will
inherit its property values from the built in RealVariable variable type. For
information on RealVariable, see Chapter 1.
To change a property value from that in the inherited type, assign a value to
that property in the variable type definition.
Properties which you can change are:
Value......................................... The value of the variable.

Upper ........................................ The upper bound for the value.
Default is 1.0e37.
Lower........................................ The lower bound for the value.
Default is –1.0e37.
Scale ......................................... The scaling factor used by the solver.
Default is 1.0.
PhysicalQuantity ....................... The name of the physical quantity
definition used by the variable.
Description................................ A text string that can be used to store
a description of the variable.
Units ......................................... Current unit of measurement used to
display the value.
Tag............................................ Label for use in interfaces to online
applications.
Record ...................................... Controls whether the time history of
the variable value is always recorded.
Can be True or False. Default is False.
Spec .......................................... The specification for the variable. Can
be Free, Fixed, Initial, RateInitial.
Default is Free.
You can also add properties of type Integer, Real, Logical or String to your
own variable types. These added properties can only be accessed through the
automation interface e.g. from scripts or an external Visual Basic application.
These can be useful if you need to track some additional variable property in
your external application.
VARIABLE Examples
The first example shows the definition of the variable type Temperature:
VARIABLE Temperature
Value: 25;
Lower: -246;
Upper: 5000;

END
The next example shows the definition of a variable type for criogenic
temperatures. This inherits from temperature, but has a different default
value and upper and lower bounds.
VARIABLE CrioTemperature
Value: -100;
Lower: -270;
Upper: 0; END
Defining Parameter Types

You can define new parameter types. This is most useful for creating your
own StringParameter types with lists of valid values. When a parameter of
this type is displayed in a table the valid values will then automatically appear
on drop down selection lists for the parameter value.
Parameter Type Definition Syntax

PARAMETER ParameterType USES BuiltInType
VALUE: DefaultValue;
VALID AS SetType (ValidList);
END
ParameterType ......................... The name of the new parameter type.

BuiltInType ............................... One of the built-in parameter types:
RealParameter.
IntegerParameter.
LogicalParameter.
StringParameter.
DefaultValue ............................. The default value of the parameter
type.
SetType Can be either INTEGERSET (if
BuiltInType=IntegerParameter) or
STRINGSET (if
BuiltInType=StringParameter).
ValidList .................................... A list of valid values for this parameter
type. Can be either a list of integers (if
BuiltInType=IntegerParameter) or text
strings (if
BuiltInType=StringParameter).
A list of valid values can only be supplied for parameter types that inherit
from IntegerParameter or StringParameter. This list is optional.
Parameter Type Definition Example

The following example shows how to define and use a new parameter type
that is used to define the geometry of a vessel.

PARAMETER GeometryParameter USES StringParameter
Valid AS StringSet(["Cylinder", "Cone",
"Hemisphere", "Rectangular"]);
Value: "Cylinder";
END
MODEL Tank
Geometry AS GeometryParameter("Cone");
:
IF Geometry == "Cone" THEN
Volume = PI/3 * Radius * Radius * Height;
ENDIF
IF Geometry == "Rectangular" THEN
Volume = Length * Length * Height;
ENDIF
END
Once the model is instanced as a block you can change the parameter
Geometry from the AllVariables Table for that block. The table will show a
drop down list of the value values.
Defining Port Types

Use port types to define what variables will be passed in or out of a model in
streams connected to the model. For each variable you define its name and
variable type.
PORT Syntax
PORT PortType USES InheritedPortType
PortVariableName1 AS VariableType1;
PortVariableName2 AS VariableType2;
:
END
PortType ................................... The name of the port type

InheritedPortType..................... Optional. The name of an existing
port type from which the port type
will inherit properties.
PortVariableName ..................... The name of a variable to be passed
through this port

VariableType ............................. The variable type for a variable
passed through the port
PORT Remarks
You can optionally make the port type inherit the properties of a previously
defined port with the keyword USES. You can then add additional variable
types to the existing definition so that more information is passed by the new
port type.
PORT Examples
The following example shows the definition of the port type Material. This
passes a molar flow rate and a temperature:
PORT Material
F AS Flow_mol;
T AS Temperature;
END
The next example shows that the definition of port type NewMaterial inherits
the properties of the previously defined type Material. The port variable P is
added to the list of values communicated by the port:
PORT NewMaterial USES Material

P AS Pressure;
END
Defining Stream Types

To simply connect variables between an output port in one model and an
input port in another, you can use the built in Connection stream type. You
can create your own stream types to create more intelligent stream that can
contains additional variables, equations, forms and so on.
Stream types are introduced by the keyword STREAM. A stream is similar in
syntax to a model.
STREAM Syntax
PRIVATE STREAM StreamType USES InheritedStreamType
StreamStatements
END
PRIVATE Optional keyword that makes the

model text inaccessible to the user
when the model is being used from
within a library.
StreamType............................... The name of the stream type.
InheritedStreamType ................ Optional. The name of a stream
t f hi h thi t t

type from which this stream type
will inherit properties.
StreamStatements A list of statements that define the
stream type.
STREAM Remarks
The statements you can use in a stream type definition are the same as those
in a model definition, except that:
• You must define at least one port, but not more than two ports. Most
streams connect two blocks together, and therefore need two ports. You
can choose to define special stream types for feeds or products that have
only one port.
• If you define two ports, one must be an INPUT port, and the other an
OUTPUT port.
If you use the optional PRIVATE keyword and create a library that includes
the stream type, the user of the library cannot:
• See the stream type in the Simulation Explorer view.
• Export the stream type with a Save.
• See the stream type equations and variable names in a Diagnostic view.
• PRIVATE can be used to protect proprietary information contained within
the stream type definition.
STREAM Example
In this example, a stream is defined as having a potential loss of fluid
through a leak. By default the leak flow is fixed and has a value of zero.
STREAM LeakyStream
FLeak AS Flow_mol (Fixed, 0);
StreamInput AS INPUT Material;
StreamOutput AS OUTPUT Material;
StreamInput.F = StreamOutput.F - FLeak;
END
Defining External Procedures

You can use procedure types to interface your models to external Fortran, C,
or C++ routines. These routines can range from simple utility routines to
physical properties calculation, or complete legacy models.
This section explains how to link your models to procedures. For information
on writing procedures and functions, see the online Help Improving Your
Simulations, Designing Models, Using Procedures and Functions.
Procedure Syntax
PROCEDURE ProcedureName
CALL: "RoutineName";

LIBRARY: "dllname";
INPUTS: InputArgumentType, InputArgumentType, ... ;
OUTPUTS: OutputArgumentType, OutputArgumentType, ... ;
OPTIONS: OptionsList;
WORKSPACE: WorkspaceExpression ;
LANGUAGE: "FORTRAN | C | C++";
COMPATIBILITY: "SPEEDUP";
IMPLEMENTATION: SUBROUTINE | FUNCTION "SourceName" |
TEXT
SourceCode
ENDTEXT;
END
ProcedureName ........................ The name of the procedure

RoutineName ............................ The name of the routine in the external
code. If the RoutineName is the same
as the ProcedureName, the CALL
statement is not required
dllname ..................................... The name of the dll containing the
compiled procedure code. You need to
specify the full path unless the dll is
located:
- in a folder that is on the PATH;
- in the simulation Working Folder, or;
- in the same directory as the Input
language text file.
You do not need to include the
extension .dll. A single dll may contain
routines for more than one procedure.
InputArgumentType.................. The types in the input argument

list. Can be any variable type,
RealParameter, IntegerParameter,
or StringParameter.
OutputArgumentType................ The type of a variable in the output
argument list. Can be any variable
type.
OptionsList................................ Optional. A list of the options used
for the procedure.
WorkspaceExpression ............... Optional. Either a fixed integer
value for workspace, or an integer
expression (literal, that is, numbers
not symbols, constants only, for
example, 10*(10-1)/2).
SourceName.............................. Optional. The full pathname of the
file containing the source code for
the procedure. You do not need to
specify the file type extension.
SourceCode ............................... Optional. The FORTRAN, C or C++
source code for the procedure.

The INPUT and OUTPUT argument lists define the types of parameters and
variables that will be passed to and from the procedure when it is used in a
model or a stream type. This information is used to check that you are
passing the correct information to and from the procedure. If you do not want
to restrict the types of variables that can be used as an argument to or from a
procedure, use the type RealVariable in the argument list. This will allow
variables of any type to be passed as this argument.
In the case where an array is passed to the procedure, follow the variable
type name with dummy subscripts, using one asterisk for each dimension, for
example, (*) or (*,*).
The WORKSPACE statement can be a fixed integer number or integer
expression which allocates that amount as workspace to be used by the
external routine.
The LANGUAGE statement indicates whether the routine is written in Fortran,
C, or C++. If there is no LANGUAGE statement the default of Fortran is
assumed.
The IMPLEMENTATION statement indicates whether the external routine is a
subroutine or a function. If no IMPLEMENTATION statement is present, the
default is subroutine.
You can optionally provide either the full path to the file which contains the
source code, or include the source code for the procedure between the
keywords TEXT and ENDTEXT. If you select the Generate Code option for the
procedure you can have ACM automatically compile this code and build your
code into a dll ready for use in your simulation.
The COMPATIBILITY statement enables you to use Fortran code that was
originally developed for use with SPEEDUP 5.5. This means that you do not
need to alter the call list for SPEEDUP compatible code. However, Aspen
Custom Modeler is stricter than SPEEDUP in enforcing the correct argument
list for an external procedure call. If you have unresolved link symbols, check
that you procedure argument list contains all of the required arguments. A
common mistake is to miss out IFAIL or ITYP at the end of the argument list.
The COMPATIBILITY statement can only be used when Fortran is used to
implement external procedures.
The OPTIONS statement is a list of keywords that indicates extra options
about the procedure call. The default is no extra options. The options
available are:
Keyword Meaning
PRECALL The procedure is called before the run starts, in the
following situations:
• After loading a new simulation.
• After changing the structure of a simulation.
• Re-starting a simulation after an Interrupt.
POSTCALL The procedure is called in the following situations:

• Loading a new simulation.
• Changing the structure of the simulation.

• Using Interrupt during a simulation run.
DERIVATIVES Indicates that you have chosen to calculate and return

analytical derivatives of the procedure outputs with
respect to the inputs.
PRESET Indicates that the current values of the procedure

outputs should be passed into the procedure when it is
called. This is useful if wish to use one of these values in
your procedure, for example as an initial estimate of the
new output value to be calculated.
PROPERTIES Indicates that this is a physical properties procedure,

and that the component list identifier should be passed
as an input argument so that the routine can determine
the component name and property calculation options to
use.
CHANGESINPU Enables you to change the value of an input variable in

TS the external code. You can use this to calculate and
return an initial estimate for a variable. Use this option
only if the procedure is always torn; using this feature
may cause the solver to fail to converge if the procedure
is not torn.
ALWAYSCALL By default the solver detects whether the inputs to a

procedure have changed since it was last called, and if
not it skips the call to the procedure. This option forces a
call to the procedure even when the inputs (and, if
PRESET is also specified, the outputs), have not changed
since the last call.
Note: A procedure with no inputs behaves as if the
ALWAYSCALL option has been specified.
External Procedure Example

The following example shows a call to a Fortran subroutine that calculates the
volume of a vessel, given the height and cross sectional area:
PROCEDURE VolCalc
CALL : "vc";
LIBRARY : "C:\Examples\mlp";
INPUTS : Length, Area;
OUTPUTS : Volume;
OPTIONS : DERIVATIVES;
LANGUAGE: "Fortran";
IMPLEMENTATION : SUBROUTINE "C:\Examples\volcal";
END
Defining Structure Types

Structures are designed to support the definition of, access to and persistence
of global data to be used with a simulation. Typically they will contain
variables and parameters the values of which can then be referenced from
model and stream instances.

STRUCTURE Syntax
PRIVATE STRUCTURE StructureName USES StructureName
StructureDescriptionStatements
END

structure type text inaccessible to the
user when used in a library. This can
be used to protect proprietary
information contained within the
structure type.
StructureName The name you give this structure type
description
STRUCTURE Remarks
You can include statements in the structure description to define different
aspects of the structure. Structures do not need to contain specific elements.
For example, you can define a model with no equations and use inheritance to
define different structures with different sets of global data.
You can add the following statements to your structure definition:
• Parameters.
• Sets.
• Variables.
• Ports.
• Equations.
• Procedure calls.
• Sub-structures (instances of other structures).
• Forms and scripts associated with this structure type.
You can mark the structure as PRIVATE. This means that if the structure is
used in a library, the user of the library cannot:
• Display the structure type in the Simulation Explorer view.
• Export the structure type with a Save.
• See the structure type equations and variable names in a Diagnostic view.
Example
EXTERNAL qualifier.
Using EXTERNAL to reference structure instances.
You can use the EXTERNAL keyword to define a reference to an instance of a
structure type from which global data can be obtained.
e.g. Rxn1 as External Powerlaw;
where Powerlaw is the name of a structure type and Rxn1 is the reference
you will be using in this model. Having defined this reference you can use it to
access parameters and variables in the structure type.
e.g. Rxn1.HeatofReaction.

This could then be used wherever local parameters or variables are used.
However note you cannot assign a value to a parameter or variable in a
structure from a reference. You can only do that in the structure type itself.
Example
Rxn1 as EXTERNAL Powerlaw;

Models
Models describe the behavior of a unit operation or other item that you wish
to include in your simulation flowsheet. This chapter describes how to define
models. The following topics are covered:
• Model Syntax.
• Declaring variables.
• Declaring ports.
• Writing equations.
• Equations with array variables.
• Conditional equations.
• Using sets in conditional expressions.
• FOR loops and the ForEach operator.
• Equations that call procedures.
• Equations that refer to port variables.
• CONNECT keyword.
• LINK keyword.
• Connectivity rules.
• Using SIGMA.
• Using the SIZE function.
• Assigning specifications.
• Referring to sub-models.
• Using SWITCH.
• Using external properties to write sub-models.
• Using virtual types in sub-models.
4 Modeling Language for Models 83

MODEL Syntax
PRIVATE MODEL ModelName USES InheritedModel
ModelStatements
END

model text inaccessible to the user
when the model is being used from
within a library.
ModelName ............................... The name of the model.
InheritedModel ......................... Optional. The name of a model
from which this model will inherit
properties.
ModelStatements ...................... A list of statements that define the
model.
If you use the optional PRIVATE keyword and create a library that includes
the model, the user of the library cannot:
• See the model in the Simulation Explorer view.
• Export the model with a Save.
• See the model equations and variable names in a Diagnostic view.
PRIVATE can be used to protect proprietary information contained within the
model.
Model statements can include some or all of the following:
• Variable, Parameter and Set declarations.
• Port declarations.
• Equations.
• Procedure calls.
• Instances of other models (sub-models).
• Connections between sub-models.
• Links between sub-models and ports.
• Assignments to variable or parameter properties.
These are described in more detail in the following sections.
Declaring Variables in Models

Use variable declaration statements in models to define the variables you
need for your model equations. You can also define array variables with one
or more dimensions.

You can define a variable to be measured or manipulated by defining a
direction for the variable. This means you can connect two variables in
different blocks on the flowsheet using the ControlSignal stream.
You can also define a variable as hidden, so that by default the variable is not
seen by the model user.
Model Variable Syntax

VariableName1(Set1 , Set2, ...), VariableName2(Set, …), ...
AS FlagList VariableType (SpecificationList);
VariableName............................ The name of the variable being declared

S Optional index set definition used to
declare the index set for array variables.
Either a set name or a set definition
between [ and ] (brackets). To define
multi-dimensional arrays, define one set
for each dimension.
FlagList ..................................... Optional. List of optional qualifiers for the
variable. Can include a Direction (INPUT,
OUTPUT), HIDDEN, UseUOMof, or
GLOBAL.
The
INPUT, OUTPUT Direction qualifiers that

specify that a variable
will be available as a
control input
(manipulated variable)
or control output
(measured variable).
You can use both
keywords to make a
variable available as
both an input and an
output. For information
on the rules for
connecting variables
with a stream, see
CONNECT Keyword on
page 4-107.
HIDDEN An optional qualifier
that defines a variable
that by default is not
displayed in Tables.
Use HIDDEN for
variables that are likely
to be of little interest
to the end user of the
model. HIDDEN
variables can still be
accessed through the
automation interface.
EXTERNAL Indicates that the
declared property is
not part of the model
but instead is a
reference to a property
of the containing
model.

GLOBAL Indicates that the
declared variable or
parameter is not part
of the model but
instead is a reference
to a global variable.
For example, a global
parameter could be the
Universal Gas
constant.
The
VariableType ............................. The variable type to be used for the

variables.
SpecificationList........................ Optional list of specifications for the
variables.
Model Variable Remarks

You can declare a list of variables with the same type in one statement. Each
variable inherits all of the property values for it’s variable type.
Model Variable Examples

In the first example, a variable called TFeed is declared as being of the
generic built in variable type RealVariable:
FeedTemp AS RealVariable;
In the second example, the variable TFeed is declared to be of variable type

Temperature, and it’s default spec is defined as Fixed:
TFeed AS Temperature (Fixed);
In the next example, TFeed is declared to be of variable type Temperature.

The default value and the lower bound are specified and override those
inherited from variable type Temperature:
TFeed AS Temperature (Value:373.0, Lower:273.0 );
In the next example, no specification keyword has been used. This means the
number in brackets is assigned to the value property of the variable. The
direction keyword is included in the statement, which means the value of
TFeed can be passed to another block on the flowsheet using the built-in
stream type ControlSignal.
TFeed AS OUTPUT Temperature (373.0);
The following example shows two variables defined in one statement. Both
variables have the default value and the lower bound altered from the values
inherited from variable type Temperature.
TFeed, TProd AS Temperature (373.0, Lower:273.0);

In the following example, the variable FlowIn is defined as a four element,
one-dimensional array, and FlowOut is defined as a three element one-
dimensional array of variable type MolFlow.
Nsect AS IntegerParameter(4);
Index AS IntegerSet([1:Nsect]);
FlowIn(Index), FlowOut([1:Nsect-1]) AS MolFlow;
A two-dimensional array variable is declared in the following example:

NPoints AS IntegerParameter(20);
Points AS IntegerSet([1:NPoints]);
Elements AS IntegerSet([1:30]);
BulkTemperature(Points, Elements) AS Temperature;
The following example shows how to declare hidden, input and output
variables. h is hidden and by default is not shown in tables or plots for the
model. T is a control output, and so can be used as a measured variable in a
control scheme. Duty is both a control input control output, and so can be
used as a measured and/or manipulated variable in a control scheme.
h AS HIDDEN Enth_mol;
T AS OUTPUT Temperature;
Duty AS INPUT, OUTPUT Enthflow;
This example shows how to reference the global variable TAmbient. The
statement below can be included in one or more models. As soon as any one
of these models is instanced TAmbient will be instanced. As further models
are instanced they will all share the same global TAmbient. This means that
you can change the value of TAmbient on the Simulation Globals table, and
this will affect all of the models that reference it. This is useful for quantities
such as ambient temperature, where you may want to use the same value
across all of your simulation.
TAmbient AS GLOBAL Temperature;
The last example shows how to define a variable to use the units of
measurement of another variable. You can also see this feature in use in the
PID controller model in the Modeler library.
If the variable, PV, is connected to a level variable in a tank, using a
ControlSignal stream, then PV will automatically be displayed in the units of
measurement of the level variable. The UseUOMof keyword is used to also
ensure that SPRemote will be displayed in the units of measurement of the
level variable.
PV AS Input Control_Signal (Description:"Process
variable");
SPRemote AS Input Control_Signal (Fixed,
Description:"Remote

setpoint", UseUOMof:"PV");
Declaring Ports in Models

Use ports to define what variables and parameters the model can pass to or
from other models via streams. Ports can be connected to graphically within
the flowsheet, or by using the CONNECT and LINK keywords within a model.
Model Port Syntax

PortName AS DirectionType PortType ;
PortName.................................. The name you give to the port.

DirectionType............................ Can be either INPUT or OUTPUT.
PortType ................................... The type of port. A definition for
this port type must already exist
with Custom Modeling or an open
library.
Model Port Remarks

The port name is the name you provide for the port. This name is used to
refer to the port variables related to the port type. You define the names of
the port variables in the port type declaration statement.
Model Port Example

In this example, the ports Input1 and Output 1 are defined as using the port
type Material. Input1 has direction INPUT, and Output1 has direction OUTPUT.
An equation that equates the inlet and outlet flow rates is shown below. The
port type definition for port type Material is also shown.
PORT Material
Flow AS Flow_Mol_Liq;
x(ComponentList) AS Molefraction;
T as Temperature;
END
Model Heater
:
Input1 AS INPUT Material;
Output1 AS OUTPUT Material;
Input1.Flow = Output1.Flow;
:
End

Using Port Properties
A port has properties that you can use in the model to determine how the
port is used. These are:
Property Description
IsConnected A logical parameter that indicates whether a stream
is connected to the port.
ComponentList The set of components being used in this port.
Using Port Properties Examples

This example shows how the variable Output1.flow is calculated, dependant
upon the value of the IsConnected parameter.
If there is no connection to the port Offtake, Offtake.IsConnected has a value
of False, and Input1.flow is equated to output1.flow. If the port Offtake is
connected, then the offtake flow is included in the equation.
MODEL BLEED
Input1 AS INPUT Material;

Output1 AS OUTPUT Material;
Offtake AS OUTPUT Material;
IF NOT Offtake.IsConnected THEN

Input1.flow = Output1.flow;
ELSE
Input1.flow - Offtake.flow = Output1.flow;
ENDIF
END
The following example shows how to use two different port component lists:
PORT Process
Flow AS Flow_Mol_Liq;
END
MODEL ComponentMixer
Quench AS INPUT Process(ComponentList: AqueousFluid);

Inlet1 AS INPUT Process(ComponentList: ProcessFluid);
Outlet1 AS OUTPUT Process(ComponentList: AllComponents);
:
/* Loop on components only in Inlet1 port */
FOR I IN [Inlet1.ComponentList - Quench.ComponentList] DO
Outlet1.Flow*Outlet1.x(i) = Inlet1.Flow*Inlet1.x(i) +
Quench.Flow*Quench.x(i);
ENDFOR
/* Loop on components only in Quench port */

FOR I IN [Quench.ComponentList - Inlet1.ComponentList] DO
ENDFOR
/* Loop on components only in both ports */

FOR I IN [Quench.ComponentList * Inlet1.ComponentList] DO
ENDFOR
:
END
The three FOR loops are designed to catch every situation for the components
entering the mixing vessel, whether there are duplicate components in the
two ports or all components are unique in both ports.
Using the IsMatched Attribute of

Variables
You can connect a port to a port that does not have completely matching
variables, for example, the component list used in each port may be different.
You can test for this by checking the IsMatched attribute of the variables in
ports.
When a port is connected, each variable that is matched in the connected port
has the IsMatched attribute set to True. If it was not matched, IsMatched will
be False.
Note: If a port is not connected, the value of IsMatched is

undefined for the variables in that port.

undefined for the variables in that port.
Example of Using IsMatched

The following example uses the IsMatched attribute in a Mixer model that
combines streams with different component sets. If a component is present in
the inlet port, but is not matched in the stream connected to this port, the
flow rate of that component is equated to zero.
PORT FlowPort
F(componentlist) AS flow_mass;
END
MODEL ddd
Inp AS INPUT MULTIPORT OF FlowPort;
Out AS OUTPUT FlowPort;
FOR c in Inp.componentlist DO
IF (NOT Inp.F(c).IsMatched) THEN
Inp.F(c) = 0.0;
ENDIF
ENDFOR
END
Using Multiports
You can define a collection of ports, which is known as a multiport. A
multiport allows multiple stream connections to be connected to it. A
multiport can use the same port types as a normal port.
Multiport Syntax
The syntax for defining a multiport in a model is:
PortName AS DirectionType MULTIPORT OF PortType
(MIN_CONNECTIONS: mincon, MAX_CONNECTIONS: maxcon);
PortName.................................. The name you give to the port.

DirectionType............................ Can take the values INPUT,
OUTPUT, or both.
PortType ................................... The type of port. A definition for
this port type must already exist
within Custom Modeling or an open
library.
mincon ...................................... Optional. The minimum number of
ti ll d b thi t

connections allowed by this port.
Must be an integer value of 0 or
larger. Default is 0.
maxcon ..................................... Optional. The maximum integer
number of connections allowed by
this port. Must be an integer value
of 1 or larger. By default there is
no limit on the number of
connections.
Multiport Remarks
A multiport contains additional properties about the current connections to it.
You can use these in your model equations. The additional properties are:
ConnectionSet A set of text strings which

contain the names of the
streams connected to the
multiport.
Connection An array of the connected
ports which is indexed over
ConnectionSet.
For information on accessing the properties of a multiport, see Using

Connection with Multiports on page 4-92.
Using Connection with Multiports

You use Connection to access port variables in a multiport.
Connection Syntax
The variables in the port are accessed using the following syntax:
PortName.CONNECTION(StreamName).PortVariable
As Connection is an array, you can use the SIGMA function to sum the values
of PortVariable for all connections made to that port. If the PortVariable is
also an array, you can use SIGMA in the following three ways.
The first way sums all combinations of Connection(i).ArrayVariable(j).
Sum = SIGMA(PortName.CONNECTION.ArrayVariable);
For a detailed explanation, see:
SIGMA for Multidimensional Arrays Examples.
Using SIGMA a second way, for each element of the array variable, the values
of that ArrayVariable in all connections made to the multiport are summed:
FOR i IN ArrayIndexSet DO
Sum(i) = SIGMA(PortName.CONNECTION.ArrayVariable(i));
ENDFOR
Using SIGMA a third way , for each connection made to the multiport, the
values of all the elements of the ArrayVariable are summed:
FOR i IN PortName.ConnectionSet DO
Sum(i) = SIGMA(PortName.CONNECTION(i).ArrayVariable);
ENDFOR

You can use ConnectionSet to size an array variable to the same size as the
number of connections to the multiport:
Variable (Portname.CONNECTIONSET) AS VariableType;
Multiport Example
The following model represents a stream mixer which can be used to mix up
to 10 streams into a single outlet stream. The first equation sums the values
of Flow across all connections to Input1. The FOR loop then performs a mass
balance for each component to calculate the outlet composition. The sigma
function sums the product of flow and fraction of a component across all
connections.
PORT Main
Flow AS Flowrate;
Z(ComponentList) AS Fraction;
END
MODEL MULTIMIX
Input 1 as INPUT MULTIPORT OF

Main (MIN_CONNECTIONS:1, MAX_CONNECTIONS:10);
Output1 AS OUTPUT Main;
Output1.Flow = SIGMA(Input1.Connection.Flow);
FOR i IN ComponentList DO
Output1.z(i)*Output1.Flow =
SIGMA(Input1.Connection.z(i)*Input1.Connection.Flow);
ENDFOR
END
Note that in this example, if there are three streams S1, S2 and S3 connected
to the port Input, and each of these carries water only, then the equations
produced by this model are:
Output1.Flow = Input1.Connection("S1").Flow +
Input1.Connection("S2").Flow +
Input1.Connection("S3").Flow;

Output1.Flow*Output1.z("Water") =
Input1.Connection("S1").Flow*Input1.Connection("S1").z("Wat
er") +
er") +
er");
Writing Equations in Models

Use equations to define relationships between variables and parameters.
Equation Syntax
EquationName: ExpressionA = ExpressionB;
EquationName.................An optional equation identifier which is displayed in

diagnostic output.
ExpressionA, ...................Mathematical expressions involving variables and
ExpressionB parameters.
Equation Remarks
You can have any number of variables on either side of the equality sign. You
should arrange equations to improve the numerical solution of the simulation
for robustness and speed, for example avoid dividing by a variable whose
value may go to zero during solution.
The equations and variables in a model must completely describe the process.
You must ensure that the simulation can be solved, given the number of
known values in your simulation. The number of equations must equal the
number of unknowns in your simulation.
Mathematical Operators in Equations

You can use the following mathematical operators in equations and
assignments:
Operator Usage Description

+ x+y Addition
- x-y Subtraction
- -x Negation
* x*y Multiplication
/ x/y Division
^ x^n Raised to the power of
SIN SIN(x) Sine
SINH SINH(x) Hyperbolic sine

COS COS(x) Cosine
COSH COSH(x) Hyperbolic cosine
TAN TAN(x) Tangent
TANH TANH(x) Hyperbolic tangent
ASIN ASIN(x) Arc sine
ACOS ACOS(x) Arc cosine
ATAN ATAN(x) Arc tangent
SQRT SQRT(x) Square root
SQR SQR(x) Square of
EXP EXP(x) Exponent
LOGe LOGe(x) Natural logarithm
LOG10 LOG10(x) Logarithm to base 10
ABS ABS(x) Absolute value of
SIGMA SIGMA(x,y,...) Sum of values of variables and

array elements
PRODUCT PRODUCT(x,y,…) Product of values of variables and

array elements
MAX MAX(x,y,…) Maximum value in a list of arrays

and variables
MIN MIN(x,y,…) Minimum value in a list of arrays

and variables
TIME TIME Current simulation time
DELAY DELAY x BY lag Applies a time delay of time lag to

variable x
TRUNCATE TRUNCATE(x) Removes the fractional part of

floating point value x and gives the
result as an integer. For example:
TRUNCATE(1.9) -> 1
TRUNCATE(2.1) -> 2
TRUNCATE(-1.9) -> -1
TRUNCATE(-2.1) -> -2
ROUND ROUND(x) Rounds the floating point value x to

the nearest integer. For example:
ROUND(-1.9) -> -2
ROUND(-1.5) -> -2
ROUND(-1.3) -> -1
ROUND(-1.0) -> -1
ROUND(1.9) -> 2
ROUND(1.5) -> 2
ROUND(1.3) -> 1
ROUND(1.0) -> 1
Note: For trigonometric functions ensure that x has units of

radians.

Equation Examples
The following is a simple flow continuity equation. FlowIn and FlowOut are
variables defined in the model. The equation name MaterialBalance is
optional.
MaterialBalance: FlowIn = FlowOut;
Input1.Flow and Output1.Flow are port variables. AdditionalFlow is a variable
defined in the model. Both types of variable can be freely mixed.
Input1.Flow + AdditionalFlow = Output1.Flow;
Equations with Array Variables

Within equations and assignments you can refer to single array elements, a
subset of the array elements, or to the full array.
To refer to a single array element use:
ArrayName(Index)
Where Index is the index of the required element.
To refer to a subset of the array use one of the following:
ArrayName ( [Index1, Index2, …] )
ArrayName ( IndexLower: IndexUpper] )
ArrayName ( [SetName] )
The first refers to the array elements with the specified index values. The
second refers to the slice of elements with indices in the range from
IndexLower to IndexHigher. This can only be used with arrays indexed over
integer sets. The third refers to all elements whose indices are in the named
set.
To refer to the full array use:
ArrayName
Array Variable Examples

In the following example, the statement FlowIn = FlowOut; creates four
equations, one for each of the four array elements. This statement is
automatically treated as an array equation because both the FlowIn and
FlowOut arrays are indexed over the same set. SetFlow is defined as a scalar
non-array variable.
MODEL Test
NComps AS IntegerParameter(4);
FlowIn([1:NComps]) AS Flow_mol;
FlowOut([1:NComps]) AS Flow_mol;
SetFlow AS Flow_mol;
FlowIn = FlowOut;

END
You can also have equations that involve single array elements, such as:
FlowIn(1) = FlowOut(1);
You can also have equations that operate on parts of an array. Both of the
following are equivalent, and create three equations which equate elements
1, 2 and 3 of each array:
FlowIn([1,2,3]) = FlowOut([1,2,3]);
FlowIn([1:3]) = FlowOut([1:3]);
In the next example, because SetFlow is not an array, four equations are
created which equate each element of FlowIn to SetFlow.
FlowIn = SetFlow;
The final example is legal, even though the number of elements on either side
of the equation is not equal:
FlowIn([1, 3, 4]) = FlowOut;
This expression is interpreted as operating on the intersection of the array
elements. Three equations are produced from this expression:
Conditional Equations
You can use conditionals for two purposes:
To switch between different equations during a simulation dependant upon
the values of certain variables.
To make structural changes to the equations and/or assignments used
depending upon the values of parameters or other aspects of the model
configuration.
Conditionals are defined using the following syntax:
Run Time Conditional Equation Syntax

IF ConditionalExpression THEN
Equation1A;
Equation2A;
:
ELSE
Equation1B;
Equation2B;
:
ENDIF

Structural Conditional Syntax
IF ConditionalExpression THEN
Equation1;
Equation2;
:
Assignment1;
Assignment2;
:
ELSE
Equation3;
Equation4;
:
Assignment3;
Assignment4;
:
ENDIF
ConditionalExpression........ A conditional expression that evaluates to true

or false.
Equationx........................... An equation.
Assignmentx ...................... An assignment.
Conditional Equation Remarks

The conditional expression returns a value of either True or False. If the
expression is true, the first set of equations and/or assignments are used. If
the expression is false, the second set of equations and/or assignments are
used.
For structural conditionals each branch of the conditional can take either a list
of equations, a list of assignments, or both equations and assignments. The
two branches can contain different numbers of equations. Also the Else
branch is optional.
For run-time conditional equations, the two lists of equations must contain
the same number of equations and the same variables, to keep the simulation
square and complete. The else branch is always required.
For situations where there are a large number of branches in a run-time
conditional, consider using a Switch statement instead of nested IFs.
You can use the following logical operators:
Operator Meaning
== (two equals signs) Equal to
> Greater than
< Less than
>= Greater than or equal to
<= Less than or equal to

<> Not equal to
AND Logical AND function
OR Logical OR function
NOT Logical NOT function
Notes:
• The three logical functions are evaluated in the priority order:

NOT, AND, OR.
• Two logical functions may not appear in sequence unless the

second one is the NOT operator.
• To avoid ambiguities in your logical expressions, use

parentheses to define the evaluation priority order.
Conditional Equation Example

The following run-time conditional equation represents a tank where both the
inlet and outlet pipes are submerged below the liquid level. The conditional
expression accounts for flow reversal.
IF Input.Press >= Output.Press THEN
Input.Flow - Output.Flow = $Volume ;
Output.Flow = K * SQRT (LiquidLevel) ;
ELSE
Output.Flow - Input.Flow = $Volume ;
Input.Flow = K * SQRT (LiquidLevel) ;
ENDIF
The next example shows a structural conditional based on a logical parameter
whose value is TRUE:
IF IncludeValve THEN
Output1.Pressure = Input1.Pressure -
Cv*Output1.Flow;
Cv: 10.0, FIXED;
ELSE
Output1.Pressure = Input1.Pressure;
ENDIF
The following example shows a structural conditional that switches between
two equations dependant upon whether a feed is connected:
IF NOT Feed.IsConnected THEN
$Holdup = Inlet.F – Outlet.F;

ELSE
$Holdup = Inlet.F + Feed.F – Outlet.F;
ENDIF
In the next example is a structural conditional in which the not-equal-to
operator, <>, is used to decide how pressure is calculated dependent on the
value of the string parameter, ModeOfOperation.
IF ModeOfOperation <> "PressureDriven" THEN
// ModeOfOperation has the value FlowDriven
// Find minimum inlet pressure
Pin = MIN(Inlet.Connection.P);
ELSE
// ModeOfOperation has the value PressureDriven
// Equate inlet pressures
Inlet.Connection.P = Pin;
ENDIF
The following example shows a run-time conditional equation using the logical
AND function to calculate the heat transfer area that is dependent on the level
of liquid present within an evaporator.
IF level<level_low THEN
HT_Area = 0;
ELSEIF (level>level_low AND level<=level_high) THEN
HT_Area*(level_high–level_low) = MaxHT_Area*(level–
level_low);
ELSE
HT_Area = MaxHT_Area;
ENDIF
Using Sets in Conditional

Expressions
You can use sets in a conditional expression to test whether a parameter
value is equal to one of the elements in a set.
Sets in Conditional Expressions Syntax

IF Expression IN SetExpression THEN
:
ELSE
:

ENDIF
Expression ......................... Integer or String Expression

SetExpression .................... IntegerSet or StringSet
Sets in Conditional Expressions Remarks

The condition in this case can be interpreted as “if the Expression exists
within the SetDefinition, then true, else false”.
The Expression must be the same type as the base type of the set. Take care
with operator precedence when the Expression is more than just a simple
variable or parameter - parentheses may be needed for clarification.
Sets in Conditional Expressions Example

The following example tests whether the value of the string parameter
KeyComponent is in the set of strings Components1.
Components1 AS StringSet(["CH4","C2H6"]);
IF KeyComponent IN Components1 THEN ...
The above statement is equivalent to:
IF (KeyComponent == "CH4") OR (KeyComponent == "C2H6") THEN
...
FOR Loops
FOR loops enable you to simplify repetitive statements. FOR loops are useful
for handling arrays with large numbers of elements, or more than one
dimension.
FOR Loop Syntax

FOR Index IN SetExpression DO
statements;
ENDFOR
Index .............................. Index for this FOR loop.

SetExpression ................. An set expression.
statements...................... A number of statements using the Index variable.
FOR Loop Remarks

The FOR loop is repeated with the index given each value in the set
expression. The statement uses the index value to produce a different
statement at each iteration. Note that there is no implied order for the
iteration of the index, particularly when the set is a set of strings.
Statements in the FOR loop can be either equations or assignments.

FOR Loop Example
In the following example an equation and expression are repeated for all
components in a component list.
MODEL Test
FlowIn(ComponentList) AS Flow_Mol;
FlowOut(ComponentList) AS Flow_Mol;
Split(ComponentList) AS Fraction;
FOR I IN ComponentList DO
FlowIn(I) = Split(I) * FlowOut(I);
Split(I): 0.33, FIXED;
ENDFOR
END
If ComponentList is defined as the set ["N2", "O2" and "CO2", ] the index
counter in the FOR loop, I, takes the value "N2", "O2" and "CO2", and three
equations and three assignments are created.
Note that for this example, it is also possible to directly write a vector
equation without using a FOR loop:
FlowIn = Split * FlowOut;
In this case, the number of equations created is based on the intersection of
the sets defining the variables in the equation. As all three array variables are
indexed by the same set, an equation is generated for each component.
The next example assigns a value of 0 to the flow rate of all components
except water:
for i in Componentlist - ["H2O"] do
F(i): 0;
endfor
The next example shows how to create a number of instances of a tray model
within a column model, and how to simplify the repetitive CONNECT
statements that connect the trays.
The loop repeats for all but the top and bottom stages in the column..
MODEL Column
NStage AS IntegerParameter(20);
StageSet AS IntegerSet([1:NStage]);
Stage(StageSet) AS Tray;

FOR J IN [2:NStage-1] DO
CONNECT Stage(J).LiqOutFlow AND
Stage(J+1).LiqInFlow;
CONNECT Stage(J).VapOutFlow AND
Stage(J-1).VapInFlow;
ENDFOR
:
END
In the previous example the number of stages, Nstage, is assigned a default
value of 20 in the model. This value can be overwritten by changing the value
on the AllVariables form for a block that uses the model.
ForEach Operator
The ForEach operator is the equivalent of the For operator, but is used only
with list operators, that is:
• SIGMA
• PRODUCT
• MIN
• MAX
• UNION
• INTERSECTION
• DIFFERENCE
• SYMDIFF
List Operator Syntax

ListOperator (Expression1, Expression2, …)
ListOperator.................... Can be SIGMA, PRODUCT, MIN, MAX, UNION,

INTERSECTION, DIFFERENCE, or SYMDIFF.
Expressionn .................... Any normal modeling language expression, for
example, a variable name, a number, and so on. Can
also be a ForEach expression, as shown in the
following syntax.
ForEach Expression Syntax

ForEach (Index1 in SetExpression1, Index2 in
SetExpression2, …) Expression
Indexn ..............................An index used in the expression.

SetExpressionn .................The set of values through which the index will be
looped.
Expression ........................Must be valid type for the list operator, for example, if
th t t k li t f t i th

the operator takes a list of set expressions then
Expression must be a set expression.
ForEach Examples
The following is a simple matrix multiplication example:
Index([1:n]) as IntegerSet;
A(Index, Index), B(Index, Index), C(Index, Index) as
RealVariable;
For i in Index DO
For j in Index DO
C(i,j) = Sigma(ForEach (k in Index) A(i,k)*B(k,j));
EndFor
EndFor
The result of the above for n = 2 would be the following equations
C(1,1) = sigma(A(1,1)*B(1,1), A(1,2)*B(2,1));
C(1,2) = sigma(A(1,1)*B(1,2), A(1,2)*B(2,2));
C(2,1) = sigma(A(2,1)*B(1,1), A(2,2)*B(2,1));
C(2,2) = sigma(A(2,1)*B(1,2), A(2,2)*B(2,2));
The following example equates the sum of the first three values in an array to
zero:
sigma(foreach (i in set) x(i)) = 0;
When set = [1:3] this is equivalent to:
sigma(x(1), x(2), x(3)) = 0;
The second example sums array values across multiple sub-models:
sigma(foreach(i in set1, j in submodel(i).set)
model(i).x(j)) = 0;
When set1 = ["a", "b"], submodel("a").set = [1,2] and submodel("b").set =
[3,4] , this is equivalent to::
Sigma(model("a").x(1),model("a").x(2),model("b").x(3),
model("b").x(4)) = 0;
The third example sums values from two sets:
sigma(foreach(i in set1) x(i), foreach(i in set2) y(i)) =
0;
When set1 = ["a", "b"] and set2 = [1,2] the result is:
sigma(model("a").x(1),model("a").x(2),model("b").x(1),model
("b").x
(2)) = 0;

Equations that Call Procedures
You can use procedure calls in models to use an external subroutine to
perform calculations. Before doing this the Procedure must be defined in a
Procedure definition.
Procedure Call Syntax

CALL (OutputArgumentList) = ProcedureName(
InputArgumentList ) ComponentList, TEAR;
ProcedureName ............ Name of the procedure to use.

InputArgumentList ....... List of input variables in the same order as the
procedure definition.
OutputArgumentList ..... List of output variables in the same order as the
procedure definition.
ComponentList .............. Optional name of a component list that applies to
this individual call. The component and
thermodynamic properties associated with this
component list are used in property calculation calls.
A value is required only when you want to override
the default component list. For an example, see
Using More Than One Component List in Models in
Chapter 3.
TEAR ............................. Optional. Specifies that this instance of the
procedure is torn. For information on tearing, see
the Aspen Modeler Reference, Chapter 6.
Procedure Call Remarks

The order of the arguments in the argument lists must be the same as that in
the procedure definition. Also each argument must be of the same type as
declared in the procedure definition, or of a type inherited from the declared
type.
You can declare procedure calls at any point in the model definition after you
have defined the variables used in the procedure call.
It is possible to use a slice of a multidimensional array as an argument to a
procedure which expects a single dimension array. The single dimension array
is build from the list of the elements specified from the multidimension array.
For example, the procedure pTestArray takes one single dimension array as
input arguments:
Procedure pTestArray
library: "testarray.dll";
call: testarray;
implementation: subroutine "testarray.f";
language: "fortran";
inputs: real(*);

outputs: realvariable;
End
The model declares a two dimension array C. The first call will send the vector
{ C(1,1), C(1,2], C(1,3), C(1,4), C(1,5) }. The second call will send the
vector { C(1,1), C(2, 1), C(3, 1) }. In other words, the procedure does not
see the shape of the ACM array:
Model TestArray
C([1:3], [1:5]) as realvariable (fixed);
x as realvariable;
x1 as realvariable;
// a column of C
call (x) = pTestArray (C(1, [1:5]));
// a row of C
call (x1) = pTestArray (C([1:3], 1));
End
Procedure Call Example

In the following example, VolCalc is the name of a procedure definition:
CALL (TankVolume) = VolCalc (TankHeight, TankArea);
The next example shows how to use the component list name after a
procedure call to determine the components and properties used in the
procedure calculation.
CALL (EnthOut) = Enthalpy (MolFracOut, PressOut, TempOut)
Process;
The enthalpy calculation is carried out using the component list Process. The
component list determines the physical property calculation options to be
used by this instance of the Enthalpy procedure.
Equations that Refer to Port

Variables
Use port variables in equations to refer to the values being passed between
blocks. Use the following syntax to refer to a port variable.
Port Variables Syntax

PortName.PortVariableName
PortName.............................Name of the port in the model.

PortVariableName................Name of the port variable in the port definition.

Port Variables Remarks
When you declare a port in a model, you can refer to any of the variables
defined in the port’s type definition.
Port Variable Example

This example defines a port type called Main, how this is instanced in a
model, and how variables in this port are used.
PORT Main
Flow AS Flowrate;
Press AS Pressure;
END
MODEL SimpleValve
// Define internal variable

DeltaP AS Pressure ;
/* Defined one inlet and one outlet port*/

Input1 AS INPUT Main ;
Output1 AS OUTPUT Main;
// Equations - pressure change only

Input1.Flow = Output1.Flow;
Input1.Temp = Output1.Temp;
Input1.Press - DeltaP = Output1.Press;
END
CONNECT Keyword
You use connection statements within a model to connect either ports of sub-
models, or variables of sub-models . Ports are connected through streams.
For information on when to use CONNECT, and how to connect ports and
variables, see Connectivity Rules.

CONNECT Syntax
Name: CONNECT BlockName.PortName AND BlockName.PortName
WITH StreamName;
or
Name: CONNECT BlockName.VariableName AND
BlockName.VariableName WITH StreamName;
Name .............................. Optional name of the connection.

BlockName ...................... Name of a sub model within the model.
PortName........................ Name of the port through which the connection is
made.
VariableName ................. Name of a variable that has been defined as an
Input or Output.
StreamName ................... Optional name of the stream to be used for the
connection. For connecting variables, the stream
must be of built-in type ControlSignal.
CONNECT Remarks
When connecting ports, one of the ports must be an input, and one of the
ports must be an output. If you do not specify a stream, the ports are
connected with a stream of an automatically generated name using the built-
in stream type Connection. The Connection stream equates port variables of
the same name and defines them as having equal values.
Caution: If you connect two ports of different types, there may

be a mismatch in the variable list. The values that are passed
over the connections are the intersection of the variables
between the two ports. That is, the variables with names
common to both ports are transmitted.
This means you can have fewer equations in your simulation than you
expected, because the stream does not create an equivalence equation for all
the variables in the ports. This usually makes the flowsheet under-specified.
If you find your simulation is under-specified when you believe that the
models are correctly defined, check for port mis-matching.
If you use the CONNECT statement to connect variables in a flowsheet, you
must use the built-in stream type ControlSignal.
All input and output variables in sub-models are accessible in the containing
model. You do not need to use LINK to make these accessible from the
containing model.
CONNECT Restrictions
The following restrictions apply when using CONNECT:

• You cannot use a CONNECT statement to directly connect arrays of
variables or ports. The following example is therefore illegal:
a([1:nStage]) as ModelType1;
b([1:nStage]) as ModelType2;
connect a.outlet and b.inlet; // Not valid !
• However, you can use a CONNECT statement to connect arrays of
variables or ports using a FOR loop, as shown in the next example:
FOR i IN [1:nStage] DO
connect a(i).outlet and b(i).inlet; // Valid
ENDFOR
• A CONNECT statement without a defined stream cannot appear in nested

FOR loops where the index for any inner loop is dependent upon the index
variable of an outer loop. This is because you cannot declare a suitable
stream array to connect the ports or variables. The following example is
therefore illegal:
nStage as IntegerParameter;
FOR i IN (1:a.nStage] DO
FOR j IN [1:a(i).nArraySize] DO
connect a(i).x(j) and b(i).y; // Not valid !
ENDFOR
ENDFOR
CONNECT Examples
This example connects the ports Valve.Outlet and Tank.Inlet with a stream of
type Connection, and gives this connection stream the name S1:
S1: CONNECT Valve.Outlet AND Tank.Inlet;
In the next example, the stream Stream1 is declared, and then used to
connect Valve.Outlet and Tank.Inlet:
Stream1 as Connection;
CONNECT Valve.Outlet AND Tank.Inlet WITH Stream1;
The final example shows the output variable Tank1.Temp and the Input
variable C101.PV connected with the built-in stream type ControlSignal:
CONNECT Tank1.Temp AND C101.PV;

LINK Keyword
Use the LINK keyword to define a connection to a containing model's ports
from the ports of a submodel within the containing model. This exposes the
sub-model’s port as if it were a port of the containing model, and enables
them to be connected to other models within a flowsheet.
The ports connected by LINK must always be of the same direction.
For information on when to use LINK, and how to connect ports, see
Connectivity Rules.
LINK Syntax
Name: LINK PortName AND SMName.SMPortName;
Name .............................. Optional name of the link.

PortName........................ Name of the port in the containing model.
SMName.......................... Name of the sub-model instance within the
containing model.
SMPortName ................... Name of the port in the sub-model.
LINK Remarks
Use the LINK keyword to connect ports in a sub-model to those in the
containing model. Use the CONNECT keyword for connections between sub-
models within the containing model.
LINK Restrictions
The following restrictions apply when using LINK:
• You cannot use a LINK statement to connect arrays of variables or ports.
The following example is therefore illegal:
Feed ([1:nStage]) as input PortType1;

Stage([1:nStage]) as ModelType2;
link Feed and Stage.inlet; // illegal !
However, you can use a LINK statement to connect arrays of variables or

ports in a FOR loop, as shown in the next example:
Feed ([1:nStage]) as input PortType1;

Stage([1:nStage]) as ModelType2;
for i in [1:nStage] do
link Feed(i) and Stage(i).inlet;
endfor

• A LINK statement without a defined stream cannot appear in nested FOR
loops where the index for any inner loop is dependent upon the index
variable of an outer loop. The following example is illegal:
Feed ([1:nStage]) as input Multiport of Fluid;

Stage([1:nStage]) as StageModel;
for i in [1:nStage] do
for j in Feed(i).Connectionset do
// j is dependent upon i
link Feed(i).Connection(j) and Stage(i).Outlet;
// nested FOR loop too complex for LINK
ENDFOR
ENDFOR
LINK Examples
In the following example, the LINK keyword is used to connect a port from
the model Tank to a port of the model MultiTank. A number of Tank models
are instanced within the model MultiTank. This method enables you to group
models within one containing model.
MODEL MultiTank
/* Define one inlet and one outlet port */

Input1 AS INPUT FluidStream;
Output1 AS OUTPUT FluidStream;
/* Define the Tank sub-models */

Tank1 AS Tank;
Tank2 AS Tank;
Tank3 AS Tank;
/* Use LINK for external connections outside of the

containing model */
LINK Tank1.Input1 AND Input1;
LINK Tank3.Output1 AND Output1;
/* Use CONNECT for internal connections within MultiTank

*/
CONNECT Tank1.Output1 AND Tank2.Input1;

CONNECT Tank2.Output1 AND Tank3.Input1;
END
You can use LINK to connect to multiports, either where the containing model
has a multiport and/or the sub-model has a multiport.
A common situation where you might link to a multiport is where you have a
distillation column model that contains an array of stage sub-models. You can
define the feed to the column as a multiport. This allows multiple feeds to the
stages within the column.
MODEL Column
.
Feed AS INPUT MULTIPORT OF Material;
StageMap(Feed.ConnectionSet) AS IntegerParameter(1);
NStages AS IntegerParameter(20);
Stage([1:Nstages]) AS TrayModel;
FOR i IN Feed.ConnectionSet DO
FeedStage: LINK Feed.Connection(i) AND
Stage(StageMap(i)).TrayFeedPort;
ENDFOR
.
END
In this example, the key is the array StageMap. You can define the stage
numbers in StageMap for each of the connections made to the column. For
example, if you have three connections to the column's Feed port, you can
define which stage these connect to, so that StageMap is an array of three
elements indexed by the connections to the Column. For example:
StageMap("S1").Value: 5;
EXTERNAL Keyword
The External keyword can be used in a definition within a model. The
behaviour depends on the type being used for the definition.
If the type is a variable type then the variable definition is a reference to a
variable in the parent of this model.
e.g. TotalFlow as EXTERNAL moleflow;

See Using External Properties to Write SubModels.
If the type is a structure type then the parameter defined can contain a
reference to an instance of the structure type.
e.g. Rxn1 as External Powerlaw;

where Powerlaw is the name of a structure type and Rxn1 is the reference
you will be using in this model. Having defined this reference you can use it to
access parameters and variables in the structure type.
e.g. Rxn1.HeatofReaction
where HeatofReaction is a variable or parameter in the structure Powerlaw.
This could then be used wherever local parameters or variables are used.
However note you cannot assign a value to a parameter or variable in a
structure from a reference. You can only do that in the structure type itself.
See How to use structures and Defining Structure Types.
Connectivity Rules
Connect ports in models as shown in the following table:
To connect ports To ports Use

belonging to belonging to
Submodels Submodels CONNECT
Submodels Containing model LINK
Connectivity Rules for CONNECT

When using the CONNECT statement, connect ports and variables according
to the following table:
Connect this port / To this port / variable

variable
INPUT OUTPUT
OUTPUT INPUT
Connectivity Rules for LINK

When using the LINK statement, connect ports according to the following
table:
Link this port To this port

INPUT INPUT
OUTPUT OUTPUT

Using SIGMA
Use the SIGMA keyword in models to sum values in expressions. Typically the
expressions contain array variables.
SIGMA Syntax
The syntax for using SIGMA is:
SIGMA (Expression1, Expression2, …);
Expression ................................ An expression
SIGMA Remarks
The SIGMA function is normally applied to expressions containing arrays. You
can sum the whole array or a defined slice of an array. If you use an array
variable in sigma without any qualification, all the values in the array are
summed.
You can apply SIGMA to multidimensional arrays. To control the way SIGMA
works, you define the extent of the arrays to be summed. You can also use
SIGMA to sum the variables across ports within a multiport.
SIGMA Examples
The most basic use of the SIGMA function is to sum non-array variables or
values:
A1, A2, A3, B1 AS RealVariable;
B1 = SIGMA (A1, A2, A3, 2.0 );
The SIGMA command is interpreted as:
B1 = A1 + A2 + A3 + 2.0
More typically, SIGMA is used to sum elements of arrays:
ComponentList AS StringSet(["N2", "O2", CO2"]);
x(ComponentList) AS Fraction;
SIGMA(x) = 1.0;
In this example, x is an array, so SIGMA(x) calculates the sum of the values
in the array. The SIGMA command is interpreted as:
x("N2") + x("O2") + x("CO2") = 1.0
Another way you could write the expression is to list the array elements
explicitly:
SIGMA(x("N2"),x("O2"),x("CO2")) = 1.0;
The following statement is equivalent to the previous SIGMA expression:
SIGMA(x(ComponentList)) = 1.0;
You can use SIGMA on the whole array, or a selected slice of an array as in
the following example:

Nsect AS IntegerParameter(10);
Temp([1:Nsect]) AS Temperature;
X AS RealVariable;
x = SIGMA(Temp([2:Nsect]));
You can remove an element of a set from the calculation in the SIGMA
expression:
NonWaterFraction AS Fraction;
x(ComponentList) AS Fraction;
SIGMA(x(ComponentList - ["WATER"])) =
NonWaterFraction;
In this example, SIGMA is used to sum the fractions of all the components in
a component list except for water.
Using SIGMA for Multidimensional

Arrays and Multiports
You can use SIGMA to sum arrays that contain more than one dimension.
SIGMA (Variable (Dim1, Dim2, ... ) )
or
SIGMA( Array1.Array2. ... )
Variable .......................... Name of a multi-dimensional array variable

Dimn ............................... Set expression that defines the elements to sum
over in the given array dimension
Arrayn............................. An array element in an object path, such as an
array of models, an array of ports, a multiport or
an array variable
SIGMA for Multidimensional Arrays Examples

You can create an expression that contains more than one dimension that can
be expanded:
PORT Material
F AS Flow;
z(ComponentList) AS Fraction;
END
MODEL Tank
M, Mc(ComponentList) AS Holdup;
Total AS RealVariable;

Input1 AS INPUT MULTIPORT OF Material;
Total = SIGMA(Input1.Connection.z);
END
In this example, there are two index sets:
• The multiport set Input1.ConnectionSet,which indexes Input1.Connection
• The component set Input1.Connection(i).ComponentList, which indexes
Input1.Connection(i).z
In this case SIGMA performs an expansion that generates all possible
expansions so that Total would equal the sum of z for all components in all of
the connections to the multiport.
The expression is expanded to:
Total = Input1.Connection("S1").z("C1")
+ Input1.Connection("S1").z("C2")
+ ...
+ ...
Another example is where you have a submodel flowsheet in which the
submodels have multiports. You get a path:
MODEL MultiTank
Ntank AS IntegerParameter(10);
Tank([1:Ntank]) AS MultiMixTank;
CONNECT ...
Sum1 = SIGMA(Tank.Input1.Connection.z *
Tank.Input1.Connection.Flow)
END
In this case, the sigma expression has to evaluate the expression:
SIGMA(Tank([1:Ntank]).Input1.Connection(
ConnectionSet).z(ComponentList) *
Tank([1:Ntank]).Input1.Connection(ConnectionSet).Flow)
As in the previous example, all combinations of model array, multiport array
and component array, are calculated and summed.
Use the ForEach operator to override the default expansion of arrays within
SIGMA. For more information, see the ForEach Operator on page 4-103.

Using the SIZE Function
You can use the SIZE function to determine the number of elements in a set.
Size can be applied to integer and string sets. It returns an integer value of
the number of elements in the set.
Note: Size may not be used in an equation (or procedure call)

statement, or in the condition part of a run time If statement.
SIZE Syntax
SIZE (SetName)
SetName ............................ Name of an integer or string set
SIZE Example
In the following example SIZE is used to determine the number of
components used in a model, and calculates some default values for the
molefractions of the components.
MODEL Feed
:
x(ComponentList): 1/SIZE(ComponentList);
:
END
The SIZE function calculates the size of the ComponentList Set. A value of
one over the number of components is used as a default value for the
molefraction array variable.
Assigning Specifications in
Models
You can assign a specification to a variable either when you declare it, or after
its declaration. Parameters do not have specifications, as their values are
always known. You can also specify a value for a variable or parameter when
you declare it or afterwards.
Syntax for Assigning Specifications in Models

The syntax for assigning specifications to variables in models is:
VariableName.Spec: Specification;
or you can shorten this to:

VariableName: Specification;
The syntax for assigning values to variables or parameters in models is:
Name.Value: Value;
or you can shorten this to:
Name: Value;
You can also assign both the value and specification to a variable using the
following syntax:
VariableName: Value, Specification;
VariableName .................Name of a variable in the model.

Specification ...................Specification for the variable. If you do not specify a
value, Free is assumed. Valid values are one of:
Fixed A variable whose value is known

and fixed.
Initial A variable whose value is known
and fixed at time zero for an
initialization or dynamic run.
RateInitial A state variable whose time
derivative is known at time zero for
an initialization or dynamic run.
Free A variable whose value is being
solved for.
Name ............................ The name of a variable or parameter in the model.

Value............................. The value assigned to the variable or parameter.
Assigning Specifications in Models Example

The following example shows the declaration of a variable called Temp. Temp
is then given a value and a fixed specification. This specification can be
overwritten on the block AllVariables form after the model has been instanced
as a block.
Temp:100.0, Fixed;
Alternatively the specifications can be supplied as part of the variable
declaration.
Temp AS Temperature (Value:100.0, Spec:Fixed);
Or this can be shortened to:
Temp AS Temperature (100.0, Fixed);

The specifications in the model need not provide the complete information for
the simulation. You may choose to include only the value or the specification,
or both in separate statements.
Temp: 100.0;
This means that the temperature is not fixed in the model. You can define
Temp as fixed later when you build the flowsheet. If the variable remains not
fixed, the value supplied is used as the starting estimate for the solution
routines.
Temp: Fixed ;
This means you are making the variable Temp fixed in the model, but you are
not assigning it a value. The default value will be used unless you assign the
value later in the model, or in the AllVariables table of a block that uses the
model.
In the following example, the variable Level is declared and given a value and
an Initial specification:
Level AS Length (1.5, Initial);
The following example shows how the initial value of a derivative is specified
using the RateInitial specification:
Mass AS Holdup_mass;
Mass.derivative: 0;
Mass.spec: rateinitial;
– or –
Mass AS Holdup_mass;
$Mass: 0;
Mass : rateinitial;
Referring to Sub-Models in a
Model Definition
You can refer to variables and parameters in sub-models directly from a
model. This is convenient when using submodels to perform common
calculations.
Syntax for Referring to Sub-Models in a Model Definition

You then refer to a variable or parameter in a sub-model using the following
syntax:
SubModelName.VariableName
SubModelName.ParameterName

When you declare the sub-model, you can optionally use equalities between
sub-model variable names and the containing model variable names. You can
also assign values of model parameters to parameters in the sub-model. To
do this use the following syntax:
SubModelName AS ModelName (SubModelVariable =
ModelVariable, ... , SubModelParameter: ModelParameter, ...
) ;
A useful technique is to conditionally instance sub-models dependant upon
parameter values. To do this use the following syntax:
SubModelName ( [1:(condition)] ) AS ModelName
(SubModelVariable = ModelVariable, ... , SubModelParameter:
ModelParameter, ... );
This declares an array of sub-models. The array will be empty when the
condition is false (0), because the integer set is then [1:0], which is an empty
set. The array will be one element when the condition is true (1), as the
integer set is then [1:1]. The declaration will thus instance the sub-model
only when the condition is true.
When you want to instance the sub-model only when a parameter X is equal
to a value Y, you would use as the condition above the expression (X==Y).
Referring to Sub-Models Examples

The following example shows a sub-model VolCalc being used to calculate
vessel volume within the model Tank:
MODEL VolCalc
Vol AS Volume;
R AS Length (Fixed);
h AS Length (Initial);
PiValue AS RealParameter;
VolumeCalc: Vol = PiValue * R^2 * h;
END
MODEL Tank
k AS Constant (Fixed);
height AS Length;
Radius AS Length;
PI AS RealParameter (3.14159);

// sub-model instance
Vc AS VolCalc (R=Radius, h=height, PiValue: PI);
FlowIn AS INPUT Fluid;

FlowOut AS OUTPUT Fluid;
Vc.vol = FlowIn.flow - FlowOut.flow;
FlowOut.flow*FlowOut.flow = K*K * Vc.h;
END
In the following example, the sub-model used to calculate density is
dependent on the parameter Material. If Material is given the value Alkali the
model will use the RhoAlkaliCalc submodel, otherwise it will use the
RhoAqueousCalc sub-model:
MODEL Tank
Material AS MaterialType (description: "Material

type used");
:
//Using sub-model to calculate density

RhoAl ([1:(Material=="Alkali")]) AS RhoAlkaliCalc
(RhoAlkali=Rho, X("NaOH”)=Inlet.X("NaOH"));
RhoAq ([1:(Material<>"Alkali")]) AS RhoAqueousCalc

(RhoAqueous=Rho);
END
Using External Properties to

Write Sub-Models
Sub-models are often used to package together standard calculations (for
example, property or reaction rate calculations).
Normally, a sub-model declares all of the variables it uses locally. To use a
sub-model the containing model must then equate its local variables with the
variables in the sub-model. To reduce the duplication of variables and
improve simulation efficiency, you can use the EXTERNAL declaration qualifier
in the sub-model. The EXTERNAL qualifier indicates that variables do not exist
in the sub-model but instead come from the parent of this model.

Note: When you declare the variables of the sub-model as
EXTERNAL, you must not include them in equivalence equations
in the containing model.
Example of Using External Properties

to Write Sub-Models
Normally, a sub-model declares all of the variables it uses locally, for
example:
Model FractionToFlow
TotalFlow as flow_mol;
Flow(componentlist) as flow_mol;
X (componentlist) as fraction;
Flow = TotalFlow * X;
End
To use this sub-model, the containing model must equate its local variables
with the variables in the sub-model, for example:
Model Simple
TotalFlow as flow_mol;
Flow(componentlist) as flow_mol;
X(componentlist) as fraction;
convert as FractionToFlow (TotalFlow=TotalFlow,

X=X, Flow=Flow);
:
End
This has introduced duplicates of each Flow and X variable and the TotalFlow
variable. To reduce the duplication of variables, you can use the EXTERNAL
declaration qualifier in the sub-model, for example:
Model FractionToFlow
TotalFlow as EXTERNAL moleflow;
Flow(componentlist) as EXTERNAL flow;
X (componentlist) as EXTERNAL fraction;
Flow = TotalFlow * X;
End
The EXTERNAL qualifier indicates that the variables TotalFlow, Flow, and X do
not exist in FractionToFlow but instead come from the parent of this model, in
this case the model Simple. For example, where TotalFlow is used in
FractionToFlow, it is TotalFlow in Simple that is used.

When you declare the properties of the sub-model as EXTERNAL, you must
remove the equivalence equations in the containing model:
Model Simple
TotalFlow as moleflow;
Flow(componentlist) as flow;
convert as FractionToFlow;
:
End
Now there is only one copy of the variables TotalFlow, Flow, and X.
Changing the Default Mapping For

External Properties
The containing model can redefine a sub-model's external properties using
the IS operator in the declaration of the sub-model.
Example of Changing the Default Mapping for External

Properties
The following example demonstrates extending the example given in Using
External Properties to Write Sub-Models on page 4-121 so that the sub-model
can be used twice in the same model:
Model Simple
TotalX as flow_mol;
TotalY as flow_mol;
FlowX(componentlist) as flow_mol;
FlowY(componentlist) as flow_mol;
Y(componentlist) as fraction;
convertX as FractionToFlow (TotalFlow IS TotalX,

Flow IS FlowX);
convertY as FractionToFlow (TotalFlow is TotalY,
Flow IS FlowY, X IS Y);
:
End

Here convertX.TotalFlow will map to TotalX and convertY.TotalFlow will map
to TotalY.
Notes and Restrictions on Writing Sub-

Models
Note the following points when writing sub-models:
• An assignment to an external property within a sub-model may conflict
with other assignments made to the same property elsewhere.
Although you can assign external properties in the sub-model, you must
be careful. The sub-model assignment can conflict with an assignment to
the same external property from another sub-model. If assignments
conflict no warning is given and the result may be unpredictable.
• External arrays must be declared with consistent indexing in the model
and sub-model. If the indexes of the external property and the resolved
property do not match then array operations may be expanded
incorrectly. The parser does not check the indexes for consistent use: this
is the responsibility of the model writer.
• When you edit models, if you add or remove external properties, you must
recompile any models that use the edited model.
− Recompiling the models ensures that any changes are completely
propagated.
• Models that use external variables can only be used as sub-models: they
cannot be placed directly onto the flowsheet.
Using Virtual Types in Sub-

Models
Virtual types are used in sub-models where the types of the properties can
only be determined when the model is included in a containing model.
Example of Using Virtual Types in Sub-

Models
The following model is an example usage of VIRTUAL types. The sub-model
calculates the average of a vector of values:
Model AverageValue
VIRTUAL ValueType USES RealVariable;
NValues as external IntegerParameter;
Value([1:NValues]) as external RealVariable;
Average as ValueType;
Average = sigma(Value)/NValues;
End

The first statement declares that the type ValueType is a VIRTUAL type
derived from RealVariable. Properties of type ValueType (that is, Average) are
referred to as virtual properties. By using a virtual type, it is then possible for
the containing model to define the types that will be used for virtual
properties.
For example, the following model uses the AverageValue sub-model to
calculate the average temperature from a temperature profile:
Model Pipe
// Discretize temp profile into N sections

NSections as integerParameter;
T([1:NSections]) as Temperature;
Q as heat_flux;
WallArea as Area;
U as notype;
T_Wall as Temperature;
// Use sub-model to calculate average

T_Av as AverageValue (NValues is NSections,
Value is T);
// Calculate heat transfer based on average temp
Q = U * WallArea * (T_Wall - T_Av.Average);
End
Since ValueType has not been redefined, T_Av.Average is a simple
RealVariable. This means that it will appear on tables and plots with no units
of measurement. The ValueType can be redefined by using the IS operation in
the declaration of the sub-model, for example:
Model Pipe
// Discretize temp profile into N sections

NSections as integerParameter;
T([1:NSections]) as Temperature;
Q as heat_flux;
WallArea as Area;
U as notype;
T_Wall as Temperature;

// Use sub-model to calculate average
T_Av as AverageValue (NValues IS NSections,
Value IS T, ValueType IS Temperature);
// Calculate heat transfer based on average temp

Q = U * WallArea * (T_Wall - T_Av.Average);
End
Here ValueType has been redefined to be Temperature (which is derived from
RealVariable). Consequently, T_Av.Average is created as a Temperature
variable rather than as a simple RealVariable, so it will show temperature
units of measurement on tables and plots.
Restrictions on Using Virtual Types

Note the following restrictions when using virtual types:
• Virtual types are only visible in the scope in which they are
declared: they cannot be used directly within nested model definitions.
However, the right side of the IS operator may be a virtual type. In this
way, the redefinition of a virtual type in a container model can be
propagated down to sub-models nested within sub-models.
• When you edit models, if you add or remove virtual properties, you must
recompile any models that use the edited model.
− Recompiling the models ensures that any changes are completely
propagated through the system.
Using SWITCH
Use switch to define different sets of equations that will be used dependant
upon the state of your model. You define an initial state, and the conditions
under which the state will change.
Switch is active in all run modes, but will probably be most useful in dynamic
simulations.
Syntax for SWITCH

SwitchName: SWITCH
INITIAL STATE StateName

SwitchStatements;
IF Condition1 STATE: NewState1;
:
ENDSTATE

STATE StateName
SwitchStatements;
:
ENDSTATE
ENDSWITCH
Switchname .................... Optional name you can give to the Switch definition.
INITIAL........................... The state that applies at the start of the simulation
run. Use the Initial keyword on only one state
definition.
Note: One of the states you define in a SWITCH
definition must be flagged as the initial state.
StateName ...................... Unique name of a state within the current Switch
definition.
SwitchStatements ........... Equations that apply when in the state you are
defining.
Condition Condition for switching from the current state to the
new state.
NewState ........................ Name of the new state to switch to if the condition is
met.
SWITCH Remarks
SWITCH provides a quick and easy way for you to describe both reversible
and non-reversible changes in your process.
Note that you cannot change the structure of the simulation between states.
Each state definition in a switch must maintain a square and complete
simulation status. This means that you must have the same number of
equations in each state within a SWITCH definition.
SWITCH Examples
The following example shows how to model a non-reversible action using a
switch definition. The switch is non-reversible because there is a condition
that enables the state Normal to switch to Burst, but there is no condition to
reverse this switch.
SWITCH
INITIAL STATE Normal
VentFlow = 0;
IF VesselPres > 8.5 STATE: Burst;
ENDSTATE

STATE Burst
VentFlow = 0.01*(VesselPres - 1.0)/8.0;
ENDSTATE
ENDSWITCH
The following example shows a switch statement used to model a tank with
two different cross-sectional areas. Between fluid height 0.0 and 1.0, the
sectional area is 0.25; between the fluid level 1.0 and 11.0, the sectional area
is 2.0. The switch models overflow in this example, where the fluid level
remains constant.
CalculateFluidLevel:
SWITCH
INITIAL STATE BottomZone
FluidLevel = FluidVolume/0.25;
OverflowFlowrate = 0.0;
IF (FluidVolume >= 1.0) STATE: Topzone;
ENDSTATE
STATE TopZone
FluidLevel = (FluidVolume - 1)/2 + 4;
OverflowFlowrate = 0.0;
IF FluidVolume >= 11.0 STATE: Overflowing;
IF FluidVolume < 1.0 STATE: BottomZone;
ENDSTATE
STATE Overflowing
OverflowFlowrate = MAX(FluidInA.Flow +
FluidInB.Flow - FluidOut.Flow,0);
FluidLevel = 9.0;
IF FluidVolume < 11.0 STATE: TopZone;
ENDSTATE
ENDSWITCH

5 Modeling PDE Systems
Overview of PDE Modeling

Aspen Modeler products provide a systematic way to model and solve
distributed parameter systems in which process variables vary with respect to
spatial position as well as time. Such systems are common in engineering and
science and are described mathematically by partial differential equations
(PDEs).
In process engineering, PDEs are the fundamental basis for describing
complex physical phenomena such as mass, heat, and momentum transport
coupled to chemical reactions in 1D, 2D, and 3D. Process equipment items
that you can model in detail using PDEs include tubular reactors, packed bed
reactors, adsorption columns, fuel cells, crystallizers, heat exchangers, and
pipelines.
With the custom PDE modeling feature in Aspen Modeler products, you can
formulate coupled linear and nonlinear partial differential equations for use in
steady-state and dynamic simulations. The PDE modeling language enables
you to write compact and readable PDEs over one, two, or three spatial
dimensions, using rectangular or polar coordinates. The built-in text editor,
configuration forms, and specification analysis also facilitate the efficient
preparation of your PDE models for solution.
This chapter describes:
• How PDE modeling is implemented.
• Creating a custom PDE model.
• Using domains and distributions with PDE modeling.
• Integral Partial Differential Equation (IPDAE) models.
• Using Method of Lines for the solution of PDEs.
About PDE Modeling

Aspen Modeler products offer a number of powerful features for modeling,
solving and analyzing PDE models. Domains and distributions are built-in sub-
models provided to make it quick and easy to build your own PDE models.
Domains determine how to discretize spatial dimensions based on your
selection of the type and order of discretization method, as well as element
spacing. Distributions represent variables distributed over a single domain or
5 Modeling PDE Systems 129

up to three domains in the case of a three-dimensional distribution. They
contain the logic for calculating partial derivatives with respect to the
independent spatial variables.
The well-known Method of Lines (MOL) is used to solve the time-dependent
PDE systems arising from distributed parameter models. It consists of the
following two-phase solution method:
• In the first phase, the partial differential equations in your models are
discretized automatically using the finite difference or finite element
method you select.
• In the second phase, the resulting DAE systems are integrated over time
using the dynamic solver of your choice.
The discretization methods and dynamic solvers can be selected and
controlled in order to assure convergence for different types of problems.
Profile plotting is available for visually examining the results of PDE
simulations. You can use 2D color profile plots to view the distribution of your
process variables along the length of a discretized domain. With 3D color
profile plots you can analyze a PDE solution over two spatial dimensions or
over a single spatial dimension and time.
Creating a Custom PDE Model

To create a new custom PDE model:
1 Create a model.
2 Declare one or more domains. To customize a domain, you can specify the
following properties:
− Highest order derivative
− Discretization method
− Domain length
− Number and location of domain section(s)
− Element spacing preferences
3 Declare a distribution for each distributed variable. To customize a
distribution, you can specify the following properties:
− Highest order derivative of the variable in each direction if it differs
from the default set for the corresponding domain
− Integral calculations
4 Write your partial differential equations over the domain(s).
5 Write your boundary conditions.
6 For a dynamic simulation, write your initial conditions.
Example of Custom PDE Modeling

The following example shows a complete description of a two-dimensional
heated slab:
MODEL HeatedSlab

// Declare domains for X and Y spatial domains
X as LengthDomain (DiscretizationMethod:"BFD1",
HighestOrderDerivative: 2, Length:1,
SpacingPreference:0.1);
Y as LengthDomain (DiscretizationMethod:"BFD1",
HighestOrderDerivative: 2, Length:1,
// Declare a two-dimensional distribution for temperature

T as Distribution2D (XDomain is X, YDomain is Y) of
Temperature(10);
D as constant (1.2, Fixed);
// Heat transfer equation: Time-dependent second order

partial
// differential equation. PDE is specified over the
interior of
// the domain.
$T(X.Interior,Y.Interior) = D * T(X.Interior,
Y.Interior).d2dx2
+ D * T(X.Interior,
Y.Interior).d2dy2;
// Initial conditions - steady state

T(X.Interior, Y.Interior): RateInitial;
// Dirichlet boundary conditions

// Specify temperature at domain boundaries
T(0, [0:Y.EndNode]): Fixed, 31;
T(X.EndNode, [0:Y.EndNode]): Fixed, 15;
// Neumann Boundary conditions - No heat transfer at sides

T([1:X.EndNode-1], 0).ddy = 0;
T([1:X.EndNode-1], Y.EndNode).ddy = 0;
End

Using Domains with PDE
Modeling
Domains are used to define the space over which a PDE system and its
corresponding variables are distributed. A single domain describes one
dimension of the distribution (for example, the axis of a tubular reactor). In
the modeling language, a built-in type definition called Domain represents
this concept. Two other built-in types, LengthDomain and AngleDomain,
inherit properties from Domain and are provided for your modeling
convenience.
For example, a pipe model that contains equations that are distributed along
the axis may contain the following declaration:
Model Pipe
Axis as LengthDomain (Length:10,
SpacingPreference:1.0, NumSections:1,
DiscretizationMethod:"CFD4");
End
In this example, Axis is an instance of the LengthDomain model with
boundaries at 0.0 (the default) and 10.0. Describing the boundaries of a
domain is only a small part of the functionality of a domain model. It also
calculates how the space is discretized. The discretization is based on your
selection of element spacing, number of domain sections, and type and order
of discretization method.
The following is an illustration of using a domain with two sections. The
element spacing for the second section is half that of the first section.
Notes:
• You can change domain properties in the domain declaration

or by using the domain configuration form.
• The built-in Domain model is dimensionless. The units of

measurement for the LengthDomain and AngleDomain are
meters and radians, respectively. You must ensure that these
base units of measurement are consistent for all your model

equations. The user interface converts between the base units
of measurement and your current UOM so that plots and
tables are displayed in the chosen UOM.
See Using Units of Measurement in Chapter 2 for more

information.
Declaring Domains for PDE Modeling

When modeling a distributed parameter system, you start by declaring one or
more domains. For each domain, you can specify the following properties:
• Discretization method.
• Highest order derivative.
• Domain length.
• Element spacing preference.
• Number of sections.
• Section location and spacing.
The ability to assign these properties separately for each domain instance
gives you considerable modeling flexibility.
You can declare domains in a model using the following syntax.
Domain Syntax
Domain, Domain, … AS
DomainModel OF RealParameterType (Property:
PropertyValue, ... );
Domain ........................ Name you give to this domain.

DomainModel ............... Name of a defined domain model. Three domain models are
built in to the system, namely DOMAIN, LENGTHDOMAIN, and
ANGLEDOMAIN. The Domain model is dimensionless. The units
of measurement for the LengthDomain and AngleDomain
models are meters and radians, respectively.
RealParameterType....... Name RealParameter type to use for domain values and length.
Property....................... Parameter or variable defined in the built-in domain model. The
following are domain parameters that can be used in the
assignment list:
DiscretizationMethod Discretization method used to

calculate partial derivatives and
integrals. The default is "BFD1". Valid
values are "BFD1", “BFD2”, “CFD2,
"CFD4", “FFD1”, “FFD2”, "OCFE2",
“OCFE3”, "OCFE4", and "UBFD4".
HighestOrderDerivative Highest-order partial derivative (with
respect to the independent spatial
variable) to calculate for a given
domain. Valid values are 0, 1 and 2.
The default value is 1. No partial

derivatives are calculated for a value
of 0. For a value of 1, only 1st-order
partial derivatives are calculated. A
value of 2 means that both 1st-order
and 2nd-order partial derivatives are
calculated.
Length Length of domain. The default is 1.0.
NumSections Number of sections in the domain.
Default value is 1.
SpacingPreference Preferred spacing for the domain. The
default value is
Length/NumSections/8. Note that the
actual spacing used (SpacingUsed) by
Aspen Custom Modeler may differ from
your preferred spacing. This happens
when your domain length is not a
multiple of your spacing preference,
and it also depends on the order of the
discretization method. The spacing is
chosen so there is an integer number
of element groups, where a group is
defined as the same number of
elements as the discretization order.
Your preferred spacing must also
provide the minimum number of
elements required for the
discretization method (for example,
four elements for 4th-order methods).
If it does not, the spacing used is
equal to the domain length divided by
the minimum number of elements.
Because the value of the SpacingUsed
property is determined by the built-in
domain model, you must not modify it.
For more information, see Specifying
Element Spacing Preference.
Section(*).Location Absolute location of the left-hand
boundary of a section. The locations of
contiguous sections in a given domain
must be in increasing order. The
location of the right-hand domain
boundary is equal to
Section(1).Location plus Length.
Section(*).Spacing Preferred spacing for the section. The
Preference default value is the value of the
SpacingPreference property for the
domain. Element spacing is uniform in
any given section. Different spacing
can be used in different sections. Note
that the actual spacing used,
Section(*).SpacingUsed, may differ
from your preferred section spacing for
the same reasons discussed for the
SpacingPreference property. Because
the value of the
Section(*).SpacingUsed property is
determined by the built-in domain
model, you must not modify it. For
more information, see Element
Spacing Preference
PropertyValue Value to be assigned to the
property.

Domain Remarks
• You can declare a comma-separated list of domains of the same type in a
single declaration statement.
• You may associate up to three domains with each distributed variable in
your process model. This association is made with a distribution.
• Distributed variables can refer only to domains that you have already
declared explicitly in your model.
• You can change the configuration of a domain using the domain
configuration form. This form is particularly useful if you want to use
different discretization methods or element spacing.
Domain Declaration Examples

All the following declarations appear within a model definition.
In the first example, a domain called X is created in the model. X takes on the
characteristics of the built-in definition of a domain model:
X AS Domain;
In this example, the domain properties such as X.Value and X.Length are
simple RealParameters. To provide units of measurement for a Domain the
value type can be redefined using the 'of' qualifier, for example:
X AS Domain of LengthParameter;
For brevity, LengthDomain is provided as a built-in DomainModel that uses
LengthParameter as its ValueType by default. The following example declares
Axis as a domain of type LengthDomain, and requires 2nd-order partial
derivatives:
Axis AS LengthDomain (HighestOrderDerivative:2);
In the next example, the domain Angle is of type AngleDomain, which is a
built-in domain model. The default discretization method used is the 4th-
order central finite difference method.
Angle AS AngleDomain(DiscretizationMethod:"CFD4");
The following example shows three domains defined in one statement. Each
domain contains 2nd-order partial derivatives that are discretized using 4th-
order orthogonal collocation on finite elements:
X, Y, Z AS LengthDomain (HighestOrderDerivative:2,
DiscretizationMethod:"OCFE4");
By using three separate domain declaration statements, a different
discretization method can be used for each domain:
X AS LengthDomain (HighestOrderDerivative:2,
DiscretizationMethod:"BFD1");
Y AS LengthDomain (HighestOrderDerivative:2,
Z AS LengthDomain (HighestOrderDerivative:2,
DiscretizationMethod:"UBFD4");
In the next example, the domain Y of type LengthDomain is discretized using
the 4th-order upwind biased finite difference method. The values of highest

order derivative and domain length are specified. The domain is divided into
three sections having lengths of 3.0, 1.0, and 6.0, respectively. Note here
that by default Section(1).Location is 0.0. A domain spacing preference of 0.2
is applied to the first and third sections, whereas a finer element spacing of
0.1 is preferred for the second section.
Y AS LengthDomain (DiscretizationMethod:"UBFD4",
HighestOrderDerivative:2, Length:10.0, NumSections:3,
SpacingPreference:0.2, Section(2).Location:3.0,
Section(3).Location:4.0, Section(2).SpacingPreference:0.1)
Using Discretization Methods for PDE

Modeling
Aspen Modeler products provide two well-known classes of discretization
methods for approximating partial derivatives and integrals: finite differences
over uniform grids and orthogonal collocation on finite elements (OCFE). As
shown in the following table, the methods differ in the order and type of
approximation used. For the built-in domain models, the default method is
1st-order backward finite difference (BFD1). You can specify another
discretization method when declaring your domain or by going to the domain
configuration form.
Method Order of Type of Approximation

Approximation
BFD1(default) 1st-order Backward Finite Difference
BFD2 2nd-order Backward Finite Difference
CFD2 2nd-order Central Finite Difference
CFD4 4th-order Central Finite Difference
FFD1 1st-order Forward Finite Difference
FFD2 2nd-order Forward Finite Difference
UBFD4 4th-order Upwind Biased Finite Difference
OCFE2 2nd-order Orthogonal Collocation on Finite

Elements
OCFE3 3rd-order Orthogonal Collocation on Finite

Elements
OCFE4 4th-order Orthogonal Collocation on Finite

Elements
Discretization Method Remarks

• The order of approximation for partial derivatives in finite difference
methods and the degree of polynomials used in finite element methods
have a significant impact on the accuracy of the solution. Generally, a
higher order algorithm of the same type implies higher accuracy. This is
especially true if you use a large element spacing (coarse grid).
• The discretization method specified in the domain for calculating partial
derivatives is also used to compute the integral over each element in any
distribution using that domain.

• By writing your PDEs and initial conditions using the Interior set provided
in the built-in Domain model, you can switch between the various
discretization methods without having to modify your model. Aspen
Modeler products automatically handle the underlying requirement that
finite difference and OCFE methods inherently require derivatives,
integrals and model equations to be evaluated at different discretization
points.
• You can use different discretization methods for each domain in a 2D or
3D distribution.
Specifying a Discretization Method in the Domain Configuration

Form
Use the domain configuration form to choose a PDE discretization method for
your domain:
1 In the All Items pane of the Simulation Explorer, double-click on the
Flowsheet folder.
2 In the Contents pane, double-click Configure for the domain block of
interest. The domain configuration form appears.
3 Click the cell corresponding to the DomainName.DiscretizationMethod
(row) and the Value attribute (column).
4 Select the method you want from the menu.
Examples of Specifying a Discretization Method in the

Declaration of a Domain
When writing a PDE model, the discretization method can be specified in the
declaration of the domain.
In the first example, a domain called X is discretized using the 4th-order
central finite difference method. CFD4 is used to calculate partial derivatives
and integrals (if integral calculations are turned on for any distributions using
domain X).
X AS Domain(DiscretizationMethod:"CFD4");
The following example shows two domains declared in one statement. Each
domain contains 2nd-order partial derivatives that are discretized using 2nd-
order orthogonal collocation on finite elements:
Axial, Radial AS LengthDomain (HighestOrderDerivative:2,
DiscretizationMethod:"OCFE2");
By using two separate domain declaration statements, a different
discretization method can be used for each domain:
Axial AS LengthDomain (HighestOrderDerivative:2,
DiscretizationMethod:"BFD1");
Radial AS LengthDomain (HighestOrderDerivative:2,

Specifying Domain Length
Distributed parameter systems are described as distributions over one or
more spatial domains of a given length.
Taking a tubular reactor as an example, the lengths of the axial and radial
domains are the reactor length and radius, respectively.
The built-in domain models contain a property called Length. The default
value for the domain length is 1.0. This value can be changed in the
declaration of your domain or by using the domain configuration form. Tip:
While the built-in Domain model is dimensionless, the units of measurement
for the LengthDomain and AngleDomain models are meters and radians,
respectively. When setting the Length parameter for the built-in
LengthDomain and AngleDomain models, you must ensure that these base
units of measurement are consistent for all your model equations. The user
interface converts between the base units of measurement and your current
UOM so that plots and tables are displayed in the chosen UOM. See Using
Units of Measurement in Chapter 2 for more information.
Specifying Domain Length in the Domain Configuration Form

Use the domain configuration form to specify the length of your domain. The
default value is 1.0.
Flowsheet folder.
3 Click the cell corresponding to DomainName.Length (row) and the Value
attribute (column).
4 Enter the domain length you want.
Examples of Specifying Domain Length in the Declaration of a

Domain
When writing a PDE model, the length of the discretized domain can be
specified in the declaration of the domain. If it is not specified, the length
defaults to the value of 1.0 assigned in the built-in domain models.
In the first example, a domain X of length 25 is discretized using the 4th-
order central finite difference method.
X AS Domain(DiscretizationMethod:"CFD4", Length: 25.0);
The next example shows the specification of the domain lengths for a two-
dimensional reactor problem.
Axial AS LengthDomain (Length:10.4);
Radial AS LengthDomain (Length:2.1);
You can also specify the domain length as a parameter:
R as lengthparameter;
Radial AS LengthDomain (Length:R);

Specifying Element Spacing Preference
Given a fixed domain length, the spacing of the elements determines the
granularity of the discretized grid and the number of elements. A small
spacing gives a fine grid with a large number of elements, whereas a larger
spacing means a coarse grid with a small number of elements. The element
spacing, along with the domain length, number of sections, and section
locations, determine the number and location of discretization nodes.
Element spacing is uniform within a given domain section, but non-uniform
spacing can be applied using multiple sections, each with a different spacing.
The latter is particularly useful for modeling PDE systems with sharp fronts.
The element spacing you use can have a large influence on the trajectory of
the solution. Generally, a smaller element spacing implies higher accuracy at
the expense of additional computation. As you reduce the element spacing,
the computational requirements increase. This happens because the number
of partial derivatives to be computed is in general proportional to the number
of elements. On the other hand, if you specify too large a value, the
simulation may be far from accurate and even fail to converge, especially for
problems with steep gradients.
Aspen Modeler products enable you to specify an element spacing preference
for your domain. The default value of the SpacingPreference property is
Length/NumSections/8, a spacing that gives 8 points per section. A different
value can be specified in the declaration of your domain or by using the
domain configuration form. Different element spacings can be used in
different sections, but element spacing is uniform in any given section.
Note: The actual spacing used (SpacingUsed) may differ from

your preferred spacing. This happens when your domain length is
not a multiple of your spacing preference times the method
order. In other words, the spacing will be chosen so that the
number of elements is an integer for a first order method, an
even integer for a second order method or an integer multiple of
4 for a fourth order method.
In this case, the domain length is divided by your preferred
spacing multiplied by the method order. The result is rounded to
the nearest integer to determine the number of element groups
of size 1 (first order), 2 (second order), 3 (third order) or 4
(fourth order). The domain length is then divided by the
calculated number of elements to produce the actual spacing
used.
If the discretization method is OCFE or order 3 or higher, this is
the average spacing, as the points are non-equally spaced within
each group of 3 or 4.
Your spacing preference must also give enough elements to
support the discretization method you select. If it gives fewer,
the actual spacing used is set to give the minimum. For example,
all of the 4th-order discretization methods require at least four
elements.
In summary, the value of the SpacingUsed property depends not
only on the preferred spacing you specify, but also on domain

length, as well as the order and type of discretization method
you select. Because the value of the SpacingUsed property is
determined by the built-in domain model, you must not modify
it.
Specifying Spacing Preference in the Domain Configuration

Form
Use the domain configuration form to specify the element spacing preference.
The default value is Length/NumSections/8.
Flowsheet folder.
3 Click the cell corresponding to the DomainName.SpacingPreference (row)
and the Value attribute (column).
4 Enter the element spacing you prefer to be used for this domain.
Examples of Specifying Spacing Preference in the Declaration

of a Domain
When declaring a domain in your PDE model, you can specify a preferred
element spacing in the list of property assignments.
In the first example, the preferred element spacing (SpacingPreference) for
domain X is set to a value of 0.2. Because the length is a multiple of the
preferred element spacing, the actual spacing used (SpacingUsed) is also 0.2.
The total number of elements in this domain is equal to Length/SpacingUsed
or 2.0/0.2 = 10.
X AS Domain(DiscretizationMethod:"BFD1", Length:2.0,
For the higher order discretization methods, the elements are considered to
be arranged into groups of 2,3 or 4 depending on the method order. If the
length is not a multiple of the spacing preference you choose multiplied by
the method order, the SpacingUsed property is based on the nearest number
of element groups.
For example, if your domain length is 1.0 and you specify a preferred spacing
of .35 for BFD1, the actual SpacingUsed property will be set to .3333
(1.0/3). If your domain length is 1.0 and you specify a preferred spacing of
0.0875 and the discretization method is CFD4, then the SpacingUsed property
will be set to 0.08333 (`1.0/12), giving three groups of four elements. If the
discretization method is OCFE4, this is the average spacing, but the spacings
within each group of four elements will vary according to the position of the
collocation points.
As another example, take the case where your domain length is 1.0 and you
specify a spacing preference of 0.5 with a 4th-order discretization method.
The value of the SpacingUsed property in this case is set to .25 (1.0/4),
because at least four discretization points are required for a 4th-order
method.

Axial AS LengthDomain (DiscretizationMethod:"CFD4",
Length:1.0, SpacingPreference:0.5);
The next example shows the domain length and spacing specified as
parameters:
R as lengthparameter;
n as integerparameter(10);
Radial AS LengthDomain (Length:R, SpacingPreference:R/n);
Using Domain Sections

The solution path for some PDEs may exhibit steep gradients in one or more
spatial domains. When the location of a sharp front does not move much in
time and is known a priori, you can divide the domain into multiple sections
to handle this problem. You can use a small element spacing (that is, fine
grid) for any section that contains a steep gradient and a large spacing (that
is, coarse grid) elsewhere.
The built-in Domain model contains the NumSections property for specifying
the number of sections. The default value is 1. You have the flexibility to set
the Location and preferred element spacing (Section(*).SpacingPreference)
for each section in your domain. The default location for the first section (that
is, Section(1).Location) is 0.0. By default, the remaining section locations are
set at equally spaced intervals. The default spacing preference for each
section is taken to be the spacing preference specified for the domain.
Section Remarks
• You can change the number of domain sections and their location and
element spacing in the declaration of your domain or by using the domain
configuration form.
• The locations of contiguous sections in a given domain must be in
increasing order. A section location value must not exceed
Section(1).Location plus Length.
• Uniform element spacing is used in a given section. Element spacing can
vary between sections, therefore, non-uniform spacing can be applied
using multiple sections, each with a different spacing. The latter is
particularly useful for modeling PDE systems with sharp fronts.
Specifying Section Parameters in the Domain Configuration

Form
Use the domain configuration form to specify the number of sections and their
location and element spacing.
1 In the All Items pane of the Simulation Explorer, double-click the
Flowsheet folder.
3 Click the cell corresponding to the Value attribute (column) of the
DomainName.NumSections (row).
4 Enter the number of sections you want.

5 For each of the domains, click the cells corresponding the Value attribute
(column) of the DomainName.Section(*).Location (row) and enter the
location. The value you enter represents the absolute location of the left-
hand boundary of the section. The locations of contiguous sections in a
given domain must be in increasing order and a section location value
must not exceed Section(1).Location plus Length.
6 For each of the domains, click the cells corresponding the Value attribute
(column) of the DomainName.Section(*).SpacingPreference (row) and
enter the element spacing you prefer.
Examples of Specifying Section Parameters in the Declaration

of a Domain
When declaring a domain with multiple sections, you can specify the number
of sections using the NumSections parameter in the declaration of the
domain.
The first example shows that domain X is divided into three sections.
X AS Domain(NumSections:3);
You can set the location of one or more of the sections in the domain using
the parameters called Section(*).Location. In the following example, the
second and third sections are located at 0.3 and 0.5, respectively. The default
location for Section(1) is 0.0.
X AS LengthDomain(NumSections:3, Section(2).Location:0.3,
Section(3).Location:0.5);
Next, a different spacing preference is specified for each of the three sections
in the previous example. A much smaller spacing between elements is used
for the second section because this part of the domain is known to contain a
sharp front. Note here also that the Section(*).Location values must not
exceed 1.0 as the domain Length is 1.0 by default.
X AS LengthDomain(NumSections:3, Section(2).Location:0.3,
Section(3).Location:0.5, Section(1).SpacingPreference:0.2,
Section(2).SpacingPreference:0.01,
Section(3).SpacingPreference:0.2);
In the following example, two sections each having size 8.0 are specified. The
spacing preference for the first section is set at a value of 0.2, whereas the
spacing preference for the second section is 1.0 by default
(Length/NumSections/8).
X AS LengthDomain(DiscretizationMethod:"CFD4", Length:16.0,
NumSections:2, Section(1).Location:-8.0,
Section(2).Location:0.0, Section(1).SpacingPreference:0.2)
Number and Location of Discretization

Nodes
Based on the configuration parameters you select, the Domain model
distributes the discretization nodes. More specifically, it determines the
number of nodes (EndNode+1) and their location, Value([0:EndNode]), where

0 and EndNode correspond to the node indices for the left-hand and right-
hand domain boundaries, respectively. Given this convention for indexing
over the discrete domain, Value([0:EndNode]) specifies a closed domain and
Value([1:EndNode-1]) defines a domain open on both ends. The Value array
and EndNode are displayed in the Configuration Form for each domain.
EndNode and Value Array Remarks

• The EndNode parameter and the Value array are calculated by the domain
and must not be modified.
• If you wish to produce a profile plot showing your distribution, use the
Value array in your domain as your X-axis variable. For example, use
XDomain.Value ([0:XDomain.EndNode]) or XDomain
([0:XDomain.EndNode]) for short.
Example of Use of EndNode and Value Array

The following example of Dirichlet boundary conditions demonstrates the use
of EndNode and the Value array:
// Boundary conditions specified using EndNode
T(0,[0:Y.EndNode] ) = 0.0;
T([1:X.EndNode-1],0) = 0.0;
T([1:X.EndNode-1],Y.EndNode) = 0.0;
// Use the Value array, Y([0:Y.EndNode]), to vary the
boundary condition depending
// on location along the boundary
T(X.EndNode, [0:Y.EndNode]) =
sinh(PI)*sin(PI*Y([0:Y.EndNode]));
Aspen Custom Modeler Reference
→Modeling Language Reference
→Modeling PDE Systems
→Using Domain with PDE Modeling
Interpolation of Existing PDE Models

When building PDE models, it is desirable to examine the effect of changing
the mesh spacing preference used. This is necessary when trying to
determine the optimum number of nodes required whilst maintaining
numerical dispersion at an acceptable level.
Should the model prove difficult to converge, when changing the mesh
spacing preference used it is desirable to have good initial guesses for any
new nodes introduced.
Within the domain model, a method has been provided to allow you to
interpolate the new mesh from a previously solved mesh.

How to Interpolate a PDE Problem
To modify and interpolate an existing domain:
1 Explore the model and locate the domain to be modified.
2 Open the form named Configure:
3 Make required changes to the domain using the

Configure form:
4 When changes are complete press the now active “Interpolate” button:

5 Close the configure form.
What is Supported?
The domain interpolation will support dependent distributions and variable
arrays of 1 to 3 dependent dimensions.
Dependent distributions and variable arrays at the same model level as the
domain and lower will only found and interpolated. If the simulation message
window print level is medium or higher, messages indicating the dependents
found will be displayed.
When creating variable arrays to represent distributed variables the array
dimensions must be explicitly declare using the either the Interior set or
EndNode parameter of the domain. For example:
//Declare domain
x As Domain;
//Valid array declarations
Valid1([0:x.EndNode]) As RealVariable;
Valid2(x.Interior) As RealVariable;
Valid3([0]+x.Interior+x.EndNode) As RealVariable;
The following type of array declaration is not supported by the interpolation
scheme:
//Declare domain
x As Domain;
//Declare set
MySet As IntegerSet([0:x.EndNode]);

//Invalid array declarations
Invalid1(MySet) As RealVariable;
Invalid2(MySet) As MyModel;
If the dependent distribution or array is within an array of sub models and the
sub model array is dependent on the domain to be modified, the sub model
array itself will represent one additional dimension.
For example:
Model MySub
//Reference an external domain
x As External Domain;
//Declare distributions
T1D As Distribution1D(X Is x) Of RealVariable;
T2D As Distribution2D(X Is x, Y Is x) Of RealVariable;
T3D As Distribution3D(X Is x, Y Is x, Z Is x) Of RealVariable;
//Model equations
:
:
End
Model Main
//Declare domain
x As Domain;
//Declare variables and distributions
Tdist As Distribution1D(X Is x);
Tarray([0:x.EndNode]) As RealVariable;
//Declare array of submodels
ModelArray([0:x.EndNode]) As MySub;
//Model equations
:
:
End
In the above example the interpolation scheme will correctly interpolate
distributions and arrays Tdist, Tarray, ModelArray.T1D and ModelArray.T2D.
However ModelArray.T3D will not be interpolated as it contains 4 dependent
dimensions.
The interpolation scheme will interpolate variable values, time derivatives,
spatial derivatives and integrals

If the domain length is changed, the interpolation scheme will not extrapolate
initial values for points beyond the length of the original mesh.
Interpolation of PDE Problems through

Automation
To support automation of flowsheets through VB scripts and external
applications such as Microsoft Excel™ the AMPDEInterp132.Automation object
has been supplied. This object can be used to temporarily store information
about the current problem and to execute the interpolation after manipulation
of the domain of interest.
The properties exposed by this object are:
Name Type Read/Wri Description

te
Diagnostics Boolean Read/Write Switch on diagnostic information that is sent to
simulation messages window. The dependent
distributions and arrays found will be listed.
Domain ACM Read/Write Object reference to domain model of interest
Object
DomainChanged Boolean Read Has the domain changed significantly to require
interpolation
The methods exposed by this are:
Name Returns Description

FindDependents Integer Must be called before any domain manipulation takes place.
The method is used find all dependents on the previously
defined domain and store variables values internally within the
object.
Returns the number of found dependents (active and inactive).
Interpolate Boolean Used to execute the interpolation of the previous domain mesh
with respect to the new domain mesh configuration.
Returns a flag to indicate whether interpolation was successful

or not.
An example VB Script that manipulates the spacing preference of domain “x”

inside the model “LaPlace2Ord2D” is as follows:
Option Explicit
Dim x
‘Create an AMPDEInterp132.Automation object
Set x = CreateObject("AMPDEInterp132.Automation")
‘Pass a reference to the domain of interest
Set x.Domain = LaPlace2Ord2D.x
‘Turn on diagnostic output to simulation message window

x.Diagnostics = True
‘Find all dependents on the domain
Application.Msg "Script: Found " & _
CStr(x.FindDependents) & _
" dependents on LaPlace2Ord2D.x"
‘Reduce the current spacing preference by a third
With LaPlace2Ord2D.x
.SpacingPreference.Value = 2/3*.SpacingPreference.Value
End With
‘Has domain changed sufficiently to require interpolation?
If x.DomainChanged Then x.Interpolate
‘Destroy object reference
Set x = Nothing
Using Distributions with PDE

Modeling
Distributions are used to distribute a variable over a spatial domain. A
distribution can reference one or more domains dependent on its
dimensionality (1D, 2D, or 3D). In the modeling language, three built-in type
definitions represent this concept:
• Distribution1D
• Distribution2D
• Distribution3D
The distribution sub-models that you declare in your process model represent
distributed variables. They provide access to the values of the distributed
variable at each point of the discretized spatial domain(s). To refer to the
value of a particular distributed variable at a specific location (for example,
node i in a 1D domain), use Distribution.value(i) or Distribution(i) for short.
Using profile plotting, you can view these distributed variable values as your
Y-axis variable along the length of the discretized domain. As your X-axis
variable, use the Value array in your domain (for example,
XDomain.Value([0:XDomain.EndNode]) or XDomain([0:XDomain.EndNode])
for short).
The built-in distribution models also contain the logic for calculating partial
derivatives and integrals. More specifically, they calculate the derivatives and
integrals of the distributed variables with respect to the independent spatial
variables. First and/or second derivatives are calculated depending on the
order of the PDE problem (as defined by the highest-order derivatives in the
PDEs). Only first derivatives are considered for problems of order one. Both
first and second derivatives are considered for problems of order two. The

exact form of the derivative calculation is based on the discretization method
and order of approximation. For more information on partial derivative
calculation, see Discretization Methods: Finite Difference and Discretization
Methods: Finite Element Methods.
The built-in distribution models calculate integrals depending on the setting of
the Integrals property. Single, double, and triple integrals can be computed
for 1D, 2D, and 3D distributions, respectively. The exact form of the integral
calculation is based on the discretization method(s) specified in the domain(s)
associated with the distribution. For more information on integral calculation,
see Switching On Integrals in Distribution Declarations.
Declaring Distributions for Distributed

Variables
When modeling distributed parameter systems, you declare one or more
distributed domains and associate each distributed variable in your model
with up to three of these domains. The distributed variable is declared using a
distribution. A distribution can refer only to domains that you have declared
previously in your model.
Within a given process model, all variables do not necessarily have to be
distributed over the same domain(s). For example, some variables may be
distributed over both radial and axial domains, while others may be
distributed over the axial domain only.
You can use an array to define a collection of distributions. For example, you
can define a concentration array indexed by component names and
distributed over a domain. For an example, see Using an Array of
Distributions.
In the declaration of your distribution, you can use the IS operator to change
the default domain sub-model names (that is, XDomain, YDomain, ZDomain)
to the domain names of your choice.
You can also specify the highest order derivative of the distributed variable in
each direction if it differs from the default set for the corresponding domain.
To do this, assign a value to the HighestOrder*Derivative (where * is X, Y, or
Z) property in the declaration of your distribution.
You can switch on integral calculations by using the Integrals property. With
this property you control the calculation of single, double, and triple integrals.
In addition, you can define a distribution as hidden so that the model user
does not see the properties for the distributed variable.
You can declare distributions in a process model using the following syntax.
Syntax for Declaring a Distribution

Distribution (Set , Set, ...), Distribution (Set…), ... AS
DistributionModel (DomainSubModel IS Domain, ... ,
Property : PropertyValue, ...) OF
VariableType;

Distribution ........... Name you give to the distribution that you declare in your process model.
You can instance a list of distributions using the same distribution model
in one statement. Each distribution has all the properties defined in its
distribution model. The value of each property will be the value assigned
to the property in the distribution model type (that is, 1D, 2D, or 3D).
You can modify the values of these properties.
Set ........................ Optional set definition. Either a set name or a definition between square
brackets [ and ]. Use a Set to define an array of distribution variables. To
define multi-dimensional arrays, define a list of sets.
DistributionModel.. Name of a built-in distribution model. Valid built-in model names are
DISTRIBUTION1D, DISTRIBUTION2D, and DISTRIBUTION3D.
DomainSubModel .. Name of a built-in domain sub-model declared in a built-in distribution
model. The valid domain sub-model names are dependent on the
distribution model. For DISTRIBUTION1D, the only valid DomainSubModel
value is XDOMAIN. For DISTRIBUTION2D, the valid values are XDOMAIN
and YDOMAIN. For DISTRIBUTION3D, the valid values are XDOMAIN,
YDOMAIN, and ZDOMAIN.
Domain ................. Name you give to the domain in your process model. You can then refer
to properties (variables and parameters) in the corresponding built-in
domain and distribution model using the following syntax:
Domain.PropertyName
Property................ Parameter or variable defined in the built-in distribution model. The
following are distribution parameters that can be used in the assignment
list:
You can assign a value to HighestOrder*Derivative (where * is X, Y, or Z
) for a specific domain in the distribution. Valid values for these
parameters are 0, 1 and 2. The default value is 1. No partial derivatives
are calculated for a value of 0. Use a value of 0 to declare a variable that
is distributed but that does not require partial derivatives. For a value of
1, only 1st-order partial derivatives are calculated. A value of 2 means
that both 1st-order and 2nd-order partial derivatives are calculated.
HighestOrderXDerivative Highest order partial derivative with
respect to the independent spatial variable
in the XDOMAIN. This parameter can be
assigned for all three built in distribution
models: DISTRIBUTION1D,
DISTRIBUTION2D, and DISTRIBUTION3D.
HighestOrderYDerivative Highest order partial derivative with

in the YDOMAIN. This parameter can be
assigned for DISTRIBUTION2D and
DISTRIBUTION3D.
HighestOrderZDerivative Highest order partial derivative with
in the ZDOMAIN. This parameter can be
assigned for DISTRIBUTION3D only.
Integrals On/off switch for integral calculations. This
property is a StringSet whose members are
strings, for example [“idx”, “idy”]. By
default, the calculation of integrals is
switched off and the Integrals property has
a value of “None”. This parameter can be
assigned for all three built-in distribution
models. For DISTRIBUTION1D, the valid
value is “idx”. For DISTRIBUTION2D, the
valid values are “idx”, “idy”, “idxdy” or any
combination of the three. For
DISTRIBUTION3D, the valid values are

“idx”, “idy”, “idz”, “idxdy”, “idxdz”, “idydz”,
“idxdydz” or any combination thereof.
VariableType ......... Name of a defined variable type.
Examples of Declaring Distributions

All the following declarations appear within a process model definition.
In the first example, a temperature variable, T1, is distributed on a two-
dimensional domain and takes on the characteristics of the built-in
DISTRIBUTION2D model. In this example, the default value and the lower
bound on the variable are altered:
T1 AS Distribution2D of Temperature (Value:373.0);
Two distributed temperature variables can be declared in one statement:
T1, T2 AS Distribution2D of Temperature (Value:373.0);
The next example shows how to define an array of distributed temperature
variables. The variable T is indexed explicitly as having integer elements from
1 to 2:
T([1:2]) AS Distribution2D of Temperature (Value:373.0);
Use the IS operator to change the default domain names, XDomain and
YDomain, to the more meaningful local names, Axial and Radial:
T([1:2]) AS Distribution2D(XDomain IS Axial, YDomain IS
Radial) of Temperature (Value:373.0);
Continuing with the preceding example, the Axial and Radial domains can
provide second-order derivatives by assigning a value of 2 to
HighestOrderXDerivative and HighestOrderYDerivative, respectively. These
parameter values override the default value of 1 assigned in the built-in
domain model. They also override any value assigned to
HighestOrderDerivative in the declaration of the Axial and Radial domains:
T([1:2]) AS Distribution2D(XDomain IS Axial, YDomain IS
Radial, HighestOrderXDerivative:2,
HighestOrderYDerivative:2) of Temperature (Value:373.0);
To calculate both single and double integrals for a pressure variable P
distributed in 2D, use the following StringSet syntax for the Integrals
property:
P AS Distribution2D(Integrals:[“idx”, “idy”, “idxdy”]) of
Pressure;
In the next example, the concentration variable, Conc, is declared as a

distribution over three domains:
Conc AS Distribution3D of Conc_Mole;
The following example shows how to declare an array of concentration
variables distributed over three domains. The variable Conc is indexed to the
set called Components, which contains three string elements:
Components AS StringSet(["N2", "O2", "CO2"]);

Conc(Components) AS Distribution3D of Conc_Mole;
Declaring Distributed Variables that do

not Require Partial Derivatives
The recommended way to declare a variable that is distributed but does not
require partial derivatives is as follows:
Model Pipe
XDomain as Domain (HighestOrderDerivative: 2);
T as Distribution1D of Temperature;
TWall as Distribution1D of Temperature
(HighestOrderXDerivative: 0);
End
In the above example, the domain declaration sets the default order for all
distributed variables to 2. This means that both ddx and d2dx2 are calculated
for the distributed variable T. However, for TWall the default is overridden
and set to zero. In this case the ddx and d2dx2 arrays are empty and no
partial derivatives are calculated.
Alternatively you can simply index TWall over the same range as the domain:
Model Pipe
XDomain as Domain(HighestOrderDerivative: 2);
T as Distribution1D of Temperature;
TWall([0:XDomain.EndNode]) as Temperature;
End
Switching On Integrals in Distribution

Declarations
You can switch on integral calculation for a distributed variable using the
Integrals property in the declaration of a distribution model. The Integrals
property uses the following StringSet syntax:
Syntax for Switching On Integrals in Distribution Declarations

Distribution as DistributionModel (Integrals:[SetContents])
where SetContents is a valid comma-separated list of the contents of the
Integrals property from the following table:
Distribution Default Single Double Triple

Model Integrals Integrals Integrals
Distribution1D “none” “idx” Invalid Invalid
Distribution2D “none” “idx”, “idy” “idxdy” Invalid

Distribution3D “none” “idx”, “idy”, “idxdy”, “idxdydz”
“idz” “idxdz”,
“idydz”
Integrals are switched off by default (“none”).

You can use the following syntax to switch on single, double, and triple
integrals for 1D, 2D, and 3D distributions, respectively:
T as Distribution1D (Integrals:“idx”);
T as Distribution2D (Integrals:“idxdy”);
T as Distribution3D (Integrals:“idxdydz”);
The next example shows you how to switch on single integrals for a
temperature variable T distributed in 1D. Note that CFD2 is used to calculate
both partial derivatives and integrals for variable T distributed over XDomain.
XDomain as LengthDomain (DiscretizationMethod:”CFD2”);
T AS Distribution1D(Integrals:“idx”) of Temperature;
To compute single and double integrals for T distributed in 2D, use the
following StringSet syntax for the Integrals property:
T AS Distribution2D(Integrals:[“idx”, “idy”, “idxdy”]) of
Temperature;
The following example shows how to switch on triple integrals for a 3D

distribution:
T AS Distribution3D(Integrals:“idxdydz”) of Temperature;
Referring to Domains in Distributed

Variable Declarations
Distributed variables can refer only to domains that you have declared
previously in your model. For instance this sequence of declarations is legal:
Model Pipe
// Declare distributed domain
Axis as Domain of LengthParameter;
// Declare array of concentration variables
distributed over a 1D domain
Conc(ComponentList) as Distribution1D(XDomain is
Axis) of Conc_Mole (Value: 0.0);
End
whereas the following declarations are illegal because the Axis domain is used
before it is declared:
Model Pipe

distributed over a 1D domain
End
Using an Array of Distributions

Use an array of distributions to model a property (for example, concentration)
of various components distributed over a domain:
Model Pipe
// distributed over a 1D domain
FeedConc(ComponentList) as Conc_Mole;
V as Velocity;
// PDE written over nodes 1 to EndNode
$Conc(ComponentList) (Axis.Interior+Axis.EndNode)
= -V * Conc(ComponentList)
(Axis.Interior+Axis.EndNode).ddx;
// Dirichlet boundary condition at left-hand
// domain boundary
Conc(ComponentList)(0) = FeedConc(ComponentList);
End
Notes:
• The declaration of the distributed array called Conc illustrates

the convention for separating declaration qualifiers (for
example, XDomain is Axis), which appear before the value
type (for example, Concentration), and assignments (for
example, Value:0.0), which are put after the value type.
• To refer to the concentration of a particular component (for

example, "A") at a specific location (for example, node iX),
use Conc("A").value(iX) or Conc("A")(iX) for short.

Referencing Domains and
Distributions in Sub-Models
You can declare a domain and/or distribution in a model and then use this in
a sub-model of that model. To do this you should declare the domain and/or
distribution as External in the sub-model, and pass the name of the domain
and/or distribution to be used when you instance the sub-model. For example
to use the Domain X and distribution T in a sub-model called Conductivity, the
main model would contain:
X as LengthDomain (Length:1, SpacingPreference:0.1);

T as Distribution1D (XDomain is X) of Temperature;
cond as conductivity (XDomain is X, YDomain is Y);
And the sub-model would define XDomain and T as external as follows:
Model Conductivity
XDomain as external Domain;
T as external Distribution1D;
XDomain and T can then be used in the Conductivity model as normal.
About IPDAE Models

The equation-based simulation environment enables you to perform dynamic
simulations involving both lumped and distributed parameter processes.
Distributed systems have properties that vary with one or more spatial
dimensions as well as time. Both space and time are independent variables.
These systems are described in terms of integral partial differential algebraic
equations (IPDAE systems) expressed over a domain. In addition, these
IPDAE systems usually involve both boundary conditions and initial
conditions.
The general form of the IPDAE system over the bounded domain is:
f (t , x, φ, φt , φx , φxx ) = 0
with boundary conditions of the following form:
g (t , x,φ,φt ,φx ) = 0
Where:
t = Time, an independent variable.
x = Vector of at most three independent spatial variables.
φ = Vector of dependent variables of the form f(t,x).
φt = Time derivatives.

φx = First-order spatial derivatives.
φxx = Second-order spatial derivatives.
The partial derivatives with respect to the spatial variables are often written
as follows:
∂φ ( x , t ) ∂ 2φ ( x , t )
φx = , φ xx =
∂x ∂x 2
The well-known Method of Lines (MOL) is used to solve the time-dependent
PDE systems. It is a two-stage process that discretizes in space and leaves
the time domain continuous, thereby converting the IPDAE system to a DAE
system with respect to time.
Partial Differential Equations (PDEs)

Partial differential equations contain one or more partial derivatives and
involve at least two independent variables.
For distributed parameter modeling of physical and chemical processes, the
most prevalent and useful PDEs are those with first- and/or second-order
partial derivatives. Usually, at most four independent variables are involved,
the three spatial coordinates and time.
A first-order, time-dependent PDE can be put in the form:
f = aφ + bφ t + cφ x
Where:
In general, the coefficients can be functions of all the variables and their
derivatives, for example:
a = a (t , x , φ , φ t , φ x )
The following is an example of a widely-used, linear second-order PDE:
f = aφ + bφ t + cφ x + dφ xx
Where a, b, c, and d are a function of (t,x).
To represent a distributed parameter problem fully, the model also requires
auxiliary conditions. These are of two kinds, boundary conditions and initial
conditions. The latter is used when time is one of the independent variables.

Writing PDEs in Models
Once you have declared one or more domains and distributions in your
model, you can now start writing PDEs to describe complex physical
phenomena such as mass, heat, and momentum transfer coupled to chemical
reactions in 1D, 2D, and 3D.
The Aspen Custom Modeler modeling language for PDEs uses a number of
common concepts throughout. The topics in this section describe the following
concepts:
• Partial Derivatives
• Interior Set
• Open Domains
• Closed Domains
• Domain Slices
Specifying Partial Derivatives in PDEs

Specify partial derivatives in your PDE models using the dd* array for first
derivatives and the d2d*2 array for second derivatives, where * can be x, y,
or z depending on the dimensionality of your distribution.
Distribution 1st-Order 2nd-Order Partial

Model Partial Derivatives
Derivatives
Distribution1D ddx d2dx2
Distribution2D ddx, ddy d2dx2, d2dy2
Distribution3D ddx, ddy, ddz d2dx2, d2dy2, d2dz2
For one-dimensional distributions, access to first-order partial derivatives is

via the ddx array defined in the built-in distribution models. First derivatives
can appear in PDE model equations such as this advection equation:
$Temp(Axis.Interior+Axis.EndNode) = -V *
Temp(Axis.Interior+Axis.EndNode).ddx;
and in Neumann boundary conditions for 2nd-order PDE problems:
T(X.EndNode).ddx = 25.0;
The next example shows a LaPlace equation with second-order partial
derivatives:
T(X.Interior,Y.Interior).d2dx2 +
T(X.Interior,Y.Interior).d2dy2 = 0.0;
Partial Derivative Type

The PartialDerivative type is a built-in variable type for use in custom PDE
modeling. It has a reduced set of attributes (that is, value, description, and
units) compared to a normal RealVariable. This saves memory if you are
dealing with distributed parameter systems with a large number of partial
derivatives.

Also refer to: Specifying Partial Derivatives in PDEs.
Using the Interior Set in PDEs

The Interior set enables you to write PDE models that are concise, correctly
specified, and flexible from the standpoint of solution with different
discretization methods. It is defined automatically in the built-in domain
model and contains the grid points (nodes) required for your problem.
As shown in the table below, the contents of the Interior set are based on the
following domain properties: order of your PDEs (HighestOrderDerivative),
the discretization method you select (that is, finite difference vs. OCFE), and
the number of sections and elements in your domain.
For 1st-order PDE problems the Interior set is always contiguous and contains
all interior nodes, 1 to EndNode-1. For 2nd-order PDE problems however, the
Interior set is contiguous only when using finite difference methods with a
single section and OCFE methods with a single element. For all other 2nd-
order PDE problems, the Interior set is not contiguous. In these cases, the
model equation and 2nd-order derivative equation at the node(s) excluded
from the set are replaced by a first-derivative continuity equation. For
example, the Interior set does not include the section boundaries when using
the finite difference methods for domains specified as having more than one
section. For the OCFE methods, all element boundaries are excluded from the
Interior set for 2nd-order PDE problems.
Discretization Highest Number Number of Interior Set

Methods Order of Elements
Derivative Sections per Section
BFD1, BFD2, 1 --- --- [1:EndNode-1]
CFD2, CFD4,
FFD1, FFD2,
UBFD4
2 =1 --- [1:EndNode-1]
2 >1 --- [1:EndNode-1]

– [Section
Boundaries]
OCFE2, OCFE3, 1 --- --- [1:EndNode-1]

OCFE4
2 --- =1 [1:EndNode-1]
2 --- >1 [1:EndNode-1]

– [Element
Boundaries]
• Use the Interior set for writing (that is, indexing) your PDE model
equations and initial conditions. Failing to do so can lead to
specification problems (that is, non-square systems) for 2nd-order PDE
problems, especially when using finite difference methods with multiple
sections or OCFE methods with multiple elements. Using the Interior set
allows you to switch between finite difference and OCFE methods without
having to modify your PDEs. The underlying requirement that finite
difference and OCFE methods inherently require derivatives and model

equations to be evaluated at different discretization points is automatically
handled.
• Do not index your boundary conditions using the Interior set, but
rather use an explicit set of contiguous nodes. The reason is that any
automatically excluded interior node in the Interior set (see the table
above) leads to an underspecified problem. The following equations are
examples of boundary conditions written using a contiguous set of nodes:
T([1:X.EndNode-1], 0) = 0.0;
T(0, [0:Y.EndNode]) = 0.0;
• Do not index your integral equations using the Interior set, but
rather use a set of contiguous nodes. Since the Interior set is not
always full (see the table above), indexing your integral equations using a
set of contiguous nodes ensures that the integral for each individual
element contributes to the calculation of the overall integral.
Using Interior Sets in Domain Sections

Each section in a domain contains its own Interior set. The union of the
Interior sets for all sections in a domain is equal to the Interior set for the
overall domain.
An example of how to access the value of a distributed variable using the
Interior set for a domain section is shown here:
X As Domain (NumSections:2);
C As Distribution1D (XDomain Is x);
W As Distribution1D (XDomain Is x);
For i In X.Section(1).Interior Do
W(i) = 2*C(i);
EndFor
For i In X.Section(2).Interior Do
W(i) = 3*C(i)^4;
EndFor
Open Domains for PDE Modeling

The proper declaration of models of distributed parameter systems requires
the accurate specification of the domain of applicability of each equation and
boundary condition. You can easily distinguish between the interior of the
domain and the different parts of the boundaries. The built-in domain model
provides an integer set called Interior that can be used to facilitate the
efficient writing of PDEs in your model.
An open domain is a domain without its boundaries. You can define an open
domain in your model equations using the Interior set declared in the domain

model. As its name implies, the Interior set does not include the left-hand
domain boundary [0] and right-hand [EndNode] domain boundary.
2nd-order PDEs are typically written over an open domain:
u(X.Interior,Y.Interior).d2dx2 +
u(X.Interior,Y.Interior).d2dy2 = 0.0;
Closed Domains for PDE Modeling

A closed domain is a domain with its boundaries included. The left-hand and
right-hand domain boundaries are represented by nodes 0 and EndNode,
respectively. You will often need to define a closed domain when writing your
boundary conditions:
// Boundary condition for heat transfer to solid
$TS([0:X.EndNode]) = SCon * (TG([0:X.EndNode]) -
TS([0:X.EndNode]));
Important Note: Do not use the Interior set for writing

boundary conditions, especially when using OCFE discretization
methods and finite difference methods with multiple domain
sections. You should always index boundary conditions explicitly
using integer elements.
Using Domain Slices for PDE Modeling

The boundary condition equations in a model may involve just a sub-domain
of an entire domain distribution. To define such sub-domains, you can make
use of the built-in support for set operations. For example, in the case of the
variable Temp which is distributed over both axial and radial domains, the
notation Temp([0:Axial.EndNode], 0) refers to the temperature along the axis
of the reactor:
Axial AS LengthDomain
(DiscretizationMethod:"CFD4", Length: 10.0);
Radial AS LengthDomain
(DiscretizationMethod:"CFD4", Length: 2.0);
Temp AS Distribution2D (XDomain is Axial,
YDomain is Radial) of Temperature;
Temp([0:Axial.EndNode], 0) = 100.0;
Examples of Writing PDEs in Models

The first example shows the advection equation for flow in a tube where T is
temperature and V is velocity. The PDE is 1st-order in time (t) and space (x).
It is written over a domain that is open at the left-hand boundary. This is
handled by using the Interior set and adding EndNode.

Advection Equation
∂T ∂T
= −V
∂t ∂x
// Modeling Language
$Temp = -V * Temp(Axis.Interior+Axis.EndNode).ddx;
For example, Fourier’s second law for heat conduction in one-dimensional
Cartesian coordinates:
Fourier’s Second Law
can be written in the modeling language as:

$T(X.Interior) = k/(rho*cp) * T(X.Interior).d2dx2;
Next, consider three well-known elliptic PDEs (zero-order in time) written
over the interior of a domain defined in two-dimensional Cartesian
coordinates:
LaPlace’s Equation
∂2u ∂2u
+ =0
∂x 2 ∂y 2
u(X.Interior,Y.Interior).d2dy2 = 0.0;
Poisson’s Equation
∂2u ∂2u
+ 2 = sinh( π 2 + 1 x ) sin(πy )
∂x 2
∂y
for ix in x.interior do
for iy in y.interior do
u(iX, iY).d2dx2 + u(iX, iY).d2dy2 =
sinh(sqrt(sqr(PI)+1.0)*X(iX)) * sin(PI*Y(iY));
endfor
endfor
Helmholtz’s Equation
∂2u ∂2u
+ =u
∂x 2 ∂y 2

u(X.Interior,Y.Interior).d2dy2 = u(X.Interior, Y.Interior);
Tip The range of each domain of every distributed variable in a PDE must be
the same as shown in the following heat transfer equation:
// Heat transfer equation: Time-dependent second order PDE.
// PDE is specified over the domain’s interior.
$T(X.Interior, Y.Interior) = D * T(X.Interior,
Y.Interior).d2dx2 + D * T(X.Interior, Y.Interior).d2dy2;
On the other hand, the following PDE example is not legal because the range
of the X domain for the $T variable is not consistent with the range of X
domain for the distributed variables on the right-hand side of the equation:
// Heat transfer equation: Time-dependent second order
partial differential
// equation. PDE is specified over the interior of the
domain.
$T([0]+X.Interior, Y.Interior) = D * T(X.Interior,
Y.Interior).d2dx2 + D * T(X.Interior, Y.Interior).d2dy2;
Using PDEs with a Conservative Term

As an example, consider the following heat transfer equation within a cylinder
of infinite length:
∂T a ∂  ∂T  q
= r +
∂t r ∂r  ∂r  c p
where T is temperature, r is radial position, a is the thermal diffusivity, q is
the heat duty, and Cp is the heat capacity. Due to the use of the conservative
term:
∂  r∂T 
 
∂r  ∂r 
it is necessary to introduce a distribution for temperature and another one for
the inner radial flux expression (for example, rdTdr). For the radial
distribution of T over r, we need first to create a Domain. So we declare:
r as LengthDomain(DiscretizationMethod:"CFD4");
For the temperature, we need a 1D distribution:
T as Distribution1D(XDomain is r) of temperature;
To handle the conservative term we need to create an additional distribution:

rdTdr as Distribution1D (XDomain is r) of realvariable;
The heat conduction equation with cylindrical coordinates can be written as
follows:
for i in r.Interior do
$T(i) = a/r(i)*rdTdr(i).ddx + q/cp;
endfor
Finally, we need the equation to define the additional distribution for the
conservative term:
rdTdr = r*T.ddx;
Using Boundary and Initial Conditions

for PDE Modeling
The solution of PDEs must satisfy certain auxiliary conditions (initial and
boundary) on the boundary of the specified domain. These auxiliary
conditions of a PDE system determine its unique solution from among an
infinite number of solutions. They serve as the starting point for calculating
the interior solution. The boundary conditions are a fundamental part of the
description of the process system behavior, while the initial conditions define
the initial state of the process system and may often differ from one
simulation to the next.
In general, derivatives with respect to the spatial independent variables must
be at least one order lower than the highest-order derivative in the PDE.
Thus, for a 2nd-order PDE in x, the auxiliary conditions can be no higher than
1st-order:
g (t , x, φ, φt , φx ) = 0
where:
The auxiliary conditions associated with 2nd-order PDEs typically involve

values of the dependent variable and/or its first derivatives at particular
locations and time. In these cases, the preceding general form becomes:
f = aφ + bφ x
where a, b, and f are a function of (t,x). The derivative is normal to the
domain boundary.

In the mathematical literature, three of the most commonly recognized
conditions are:
• Dirichlet
• Neumann
• Cauchy
The boundary conditions arising in your applications may be significantly
more complicated than the Dirichlet, Neumann, and Cauchy classifications
discussed here.
Important:
• When writing initial conditions, use the Interior set so that your
initial specifications are only applied to the Interior points. This avoids
over-specification for second-order PDE problems with time derivatives.
This is particularly important when using OCFE discretization methods or
finite difference methods with multiple domain sections.
• When writing boundary conditions, do not use the Interior set as
you typically want them to apply to contiguous locations along the
boundary. Remember here that for second-order PDE problems solved
using OCFE discretization methods or finite difference methods with
multiple domain sections, the Interior set does not contain the contiguous
set of nodes from 1 to EndNode-1 (See “Using the Interior Set in PDEs”
for more detailed information). In view of this, you should always index
boundary conditions explicitly using a set of contiguous nodes.
• When writing boundary conditions involving second partial
derivatives, you may need to explicitly write the derivative
equations and include them in your model. The built-in domain
models do not include all second derivatives at the domain boundaries.
For more information, see Boundary Conditions Containing Second-Order
Derivatives.
Dirichlet conditions (b=0 in the preceding equation) specify the values of the
dependent variables at the boundary points as a function of time. These
Dirichlet points represent the exact solutions at the boundary points.
Neumann conditions (a=0 in the preceding equation) specify the values of the
first-order spatial derivatives on the boundary. This corresponds to specifying
the flux across the boundary. If f=0, we have a homogeneous Neumann
boundary condition, which leads to property conservation within the domain.
Cauchy
Cauchy or mixed boundary conditions comprise property flux and property
value at the boundary ( a, b ≠ 0 in the preceding equation). This is often used
to specify property transport over the boundary (for example, interface
kinetic, oxidizing surface conditions).
Example of Initial and Boundary Conditions

The following model gives examples of Dirichlet and Neumann boundary
conditions, as well as initial conditions:
Model HeatEquation
// Heat conduction example

// Domain declaration
X as LengthDomain(DiscretizationMethod:"OCFE4",
HighestOrderDerivative: 2);
// Declaration of distributed temperature variable
T as Distribution1D(XDomain is X) of Temperature;
// Declaration of thermal diffusivity for the solid
K as RealParameter(1/32);
// PDE - open on both ends of the domain as boundary

// conditions specified at 0 and EndNode
$T(X.Interior) = k*T(X.Interior).d2dx2;
// Initial conditions
T([1: X.EndNode-1]): 10.0;
// Initial specification must only be applied to 'Interior'
// points to avoid over-specification when using OCFE
methods,
// or finite difference methods with multiple sections
T(X.Interior): initial;
// Dirichlet boundary condition at left-hand domain

boundary
T(0) = 0.0;
// Neumann boundary condition at right- hand domain

boundary
T(X.EndNode).ddx = 25.0;
End
Boundary Conditions Containing

Second-Order Derivatives
If your model requires 2nd-order derivatives at a domain boundary and they
are not provided automatically by the Distribution model, you must explicitly
write the 2nd-order derivative equations and include them in your model. The
built-in distribution models explicitly exclude some 2nd-order partial
derivatives at the domain boundaries for the following reasons:

• The majority of process models do not involve boundary conditions
containing 2nd-order derivatives.
• The OCFE discretization methods do not require 2nd-order derivative
calculations at element boundaries including the domain boundaries.
Excluding the unnecessary 2nd-order derivatives at the boundaries saves
on wasted operations and memory and also allows Aspen Custom Modeler
to provide more useful analysis for specification errors.
The table below presents a list of the 2nd-order derivatives that are
calculated by the distribution models.
Distribution 2nd-Order Partial Derivatives

Model
Distribution1D d2dx2 (XDomain.Interior)
Distribution2D d2dx2 (XDomain.Interior, [0 + YDomain.Interior +

YDomain.EndNode])
d2dy2([0 + XDomain.Interior + XDomain.EndNode], YDomain.Interior)
Distribution3D d2dx2 (XDomain.Interior, [0 + YDomain.Interior + YDomain.EndNode],

[0 + ZDomain.Interior + ZDomain.EndNode])
d2dy2([0 + XDomain.Interior + XDomain.EndNode], YDomain.Interior,
[0 + ZDomain.Interior + ZDomain.EndNode])
d2dz2([0 + XDomain.Interior + XDomain.EndNode],
[0 + YDomain.Interior + YDomain.EndNode], ZDomain.Interior)
Second-Order Derivatives for Use in Full Discretization

The table above shows that while Distribution2D does not calculate d2dx2 for
nodes ([0 + XDomain.EndNode], *), it does calculate them for boundary
nodes (*, [0 + YDomain.EndNode]). The reason for providing the 2nd-order
partial derivatives for d2dx2 at the YDomain boundaries and d2dy2 at the
XDomain boundaries for Distribution2D is to allow you to set up a full
discretization case where one domain is a spatial domain and the other is a
time domain. In this case, you can write PDEs containing 2nd-order partial
derivatives over the entire time domain, including at the initial and final time.
The same capability is provided for Distribution3D where any of the three
coordinates, x, y, or z can be used as a time domain.
Writing Second-Order Derivatives at Domain Boundaries

As discussed above, if your model requires second derivatives at the domain
boundaries, you must explicitly write equations for the second derivatives and
include them in your model. For example, you can write standard second-
order forward and backward finite difference equations for the left-hand
boundary and right-hand boundary, respectively.
This example shows a second-order forward finite difference equation that
can be used as a boundary condition for the left-hand domain boundary. The
second-partial derivative of a temperature variable, T, is set to zero.
(T(0)–2*T(1)+T(2)) / (X.value(1)-X.value(0))^2 = 0.0;
The next example shows a 2nd-order backward finite difference equation that
can be used as a boundary condition for the right-hand domain boundary. The
second-partial derivative of a temperature variable, T, is set to zero.

(T(X.EndNode-2)-2*T (X.EndNode-1)+T (X.EndNode)) /
(X.value(X.EndNode)-X.value(X.EndNode-1))^2 = 0.0;
If your boundary conditions contain second derivatives which are not
automatically computed in the distribution model and not included in the
derivative equations in your model, Aspen Custom Modeler® gives an error
indicating that your model has equations that use variables that do not exist.
The error message is displayed in the Simulation Messages window when
checking the syntax of your model.
For example, consider the following HeatEquation model with a boundary
condition involving a second-order partial derivative:
Model HeatEquation
// Heat conduction example

// Domain declaration
X as LengthDomain(DiscretizationMethod:"OCFE4",
HighestOrderDerivative: 2);
// Declaration of distributed temperature variable
using a Distribution1D model
T as Distribution1D(XDomain is X) of Temperature;
// Declaration of thermal diffusivity for the solid
K as RealParameter(1/32);
// PDE - open on both ends of the domain since boundary

// conditions specified at 0 and EndNode
$T(X.Interior) = k*T(X.Interior).d2dx2;
// Initial conditions
T([1: X.EndNode-1]): 10.0;
// Initial specification must only be applied to 'Interior'
// points to avoid over-specification when using OCFE
methods,
// or finite difference methods with multiple sections
T(X.Interior): initial;
// Dirichlet boundary condition at left- hand domain

boundary
T(0) = 0.0;
// Boundary condition containing second partial derivative

// Applied at right-hand domain boundary
T(X.EndNode).d2dx2 = 0.0;
End
The following syntax error message is displayed in the Simulation Messages
window when Aspen Custom Modeler compiles the PDE model shown above:
Model HeatEquation has equations that use variables which
do not exist:
HeatEquation.T.d2dx2(8) - Failed path is (8)
To avoid this error, you can replace the boundary condition involving d2dx2
with a polynomial approximation as discussed above.
Writing Integrals in PDE Models

Integrals occur frequently in PDE models for process engineering, especially in
application areas that require population balance modeling (e.g.,
crystallization).
The calculation of integrals for a distributed variable is switched on by using
the Integrals property available in the distribution models.
The discretization method specified in the domain model for calculating partial
derivatives is also used to calculate the integral over each element in the
distribution.
Specifying Integrals
You can specify integrals in your PDEs using the id* array for single integrals,
the id*d* array for double integrals, and the idxdydz for triple integrals,
where * can be x, y, or z depending on the dimensionality of your
distribution. The following table shows valid values for integrals.
Distribution Single Double Triple

Model Integrals Integrals Integrals
Distribution1D idx Not Applicable Not Applicable
Distribution2D idx, idy idxdy Not Applicable
Distribution3D idx, idy, idz idxdy, idxdz, idxdydz

idydz
The indices of the integral arrays range from 0 to EndNode. The convention is
that the integral over element i-1 to i is stored in location i in the integral
array. By default, the integral value for node 0 is 0.0. For example, the
integral for the 1D variable T over the element i-1 to i is stored in T(i).idx. By
default, T(0)=0.0.
Since an integral is calculated over each individual element in a distribution,
you can compute the integral over the entire distribution by summing up the
individual contributions using the SIGMA operator as shown here:
Integral = SIGMA(T.idx);

The above equation is shorthand for:
Integral = SIGMA(T([0:X.EndNode]).idx);
Note that the summation above is done from 0:X.EndNode and not over the
Interior set since we need to sum up the integrals over all elements. If the
Interior set is used, the calculated integral is incorrect when using an OCFE
method since the Interior set is not full (i.e., does not contain all nodes).
Examples of Writing Integral Equations

The The following example shows the use of an integral to calculate of the
average temperature over a 1D distribution:
X as LengthDomain(Length:10);
T as Distribution1D(XDomain is X, Integrals: ”idx”) of
Temperature;
Tavg as Temperature;
Tavg = SIGMA(T.idx)/X.Length;
The previous example can be extended to 2D in the following way:
X, Y as LengthDomain(Length:10);
T as Distribution2D(XDomain is X, YDomain is Y, Integrals:
”idxdy”) of Temperature;
Tavg = SIGMA(T.idxdy)/(X.Length*Y.Length);
This example uses an integral to computes the average temperature over a

slice (i.e., left-hand edge) of a 2D distribution:
X, Y as LengthDomain;
T as Distribution2D (XDomain is X, Ydomain is Y,
Integrals:”idy”) of Temperature;
Tavg = SIGMA(T(0,[0:Y.EndNode]).idy) / Y.Length;
In another 2D example, in this case a jacketed tubular reactor model, the
energy balance on the cooling medium looks like:
Axial as LengthDomain;
Radial as LengthDomain;
T as Distribution2D(XDomain is Axial, YDomain is Radial,
Integrals:"idx") of Temperature;
Q = U*S*(SIGMA(T([0:Axial.EndNode], Radial.EndNode).idx) /
Axial.Length - Tcw);

The following crystallization example shows how the magma density is
calculated using an integral over the crystal size domain. The integral involves
two distributed variables, one for the crystal volumetric shape factor and one
for the exponential of log crystal population density:
//Domain for crystal size

CSize as LengthDomain;
// Log crystal number density
LogN as Distribution1D(XDomain is CSize) of notype;
// Exponential of log crystal number density
eLogN as Distribution1D(XDomain is CSize, Integrals: "idx")
of notype;
// Kv is crystal volumetric shape factor
Kv as Distribution1D(XDomain is CSize, Integrals: "idx") of
pos_small;
// Magma density
Mt as conc_mass;
// Solid density
RhoS as RealParameter(2660.0);
eLogN = EXP(LogN);
// Definition of magma density
Mt = RhoS * 1E-12 * SIGMA(Kv.idx * CSize^3 *
eLogN.idx)/.015;
The following example demonstrates the use of integrals for a system of

cylindrical geometry (z, r), whereby the average temperature is evaluated by:
L R
2π ∫ ∫ T ( z , r ).r.dr.dz
Tavg = 0 0
πR 2 L
A sample model is as follows:
Model CylindricalTemp
Radius as lengthparameter(1);
L as lengthparameter(2);
pi as realparameter(4*atan(1));
r as lengthdomain(length:Radius);
z as lengthdomain(length:L);

T as distribution2D(XDomain is r, Ydomain is z);
Tr as distribution2D(XDomain is r, Ydomain is z,
integrals:"idxdy");
Tavg as realvariable;
Tavg2 as realvariable;
for ir in [0:r.EndNode] do
for iz in [0:z.EndNode] do
Tr(ir, iz) = T(ir, iz)*r(ir);
endfor
endfor
for ir in [0:r.EndNode] do
for iz in [0:z.EndNode] do
T(ir, iz) = 3*r(ir);
endfor
endfor
// Tavg = 2*pi*integral integral [T(r,z)*r*dr*dz] /
volume
Tavg = 2*pi*sigma(Tr.idxdy)/ (pi*Radius^2*L);
End
Using Method of Lines for the

Solution of PDEs
The Method of Lines (MOL) solves time-dependent PDE systems by
discretizing in space, but leaving the time variable continuous (rather than
time evaluated at discrete nodes). If you think of the solution as a surface
over the space-time plane, the MOL computes cross-sections of the surface
along series of lines, each parallel to the time axis and corresponding to one
of the discrete spatial nodes.
The MOL solution process has two phases. In the first, the partial derivatives
and integrals with respect to the independent spatial variables are replaced
with algebraic approximations (finite difference or finite element) evaluated at
discrete points. In the second phase, a series of DAE systems are solved
using a dynamic integrator (See the Integrator tab of the Solver Properties
dialog box for a list of integrators available). The accuracy of the approximate
solution depends on the element spacing and the integration step size in time.

Phase1 Phase 2
Discretization of Integration
Spatial Domains Over Time
IPDAEs DAEs Solution
Two Key Steps in MOL Approach

1 Replace continuous domain(s) of equations by a discrete grid of points
(nodes) and replace derivatives and integrals in PDEs by finite difference
or finite element approximations.
2 Obtain numerical solution at each time step that is a table of approximate
values at selected nodes in spatial domain(s).
Finite Difference Methods: General

In the finite difference method, the continuous problem domain is discretized
in terms of the values of the dependent variables at discrete points only.
First-order and second-order partial derivatives are approximated by finite
difference formulas derived using Taylor series expansions. The general forms
of these derivative equations are:
∂φ  1  [ n ]]
=  A φ
∂ x  ah 
∂ 2φ  1  [ n ]]
=  B φ
∂ x 2  bh 
Where:
∂φ
= Derivative vector.
∂x
φ = Dependent variable vector.
x = Independent spatial variable.

h = element spacing.
1/a,1/b = multiplying constant.
A, B = (n+1)x(n+1) differentiation matrix. The elements of the matrix are
the weighting coefficients of nth-order approximations at the n+1Taylor series
points.
n = order of approximation.
Finite difference methods differ in terms of type (for example, forward,
backward, central, upwind-biased) and order of approximation.

Discretization Methods: Finite Difference
You can choose between several finite difference options for use as a domain
discretization method. The default method is BFD1. When writing your PDE
model, you can specify a different DiscretizationMethod property when
declaring a domain. You can also select another method using the Domain
Configuration Form.
Method Order of Type of FD

Approximation Approximation
BFD1 1st-order Backward
BFD2 2nd-order Backward
CFD2 2nd-order Central
CFD4 4th-order Central
FFD1 1st-order Forward
FFD2 2nd-order Forward
UBFD4 4th-order Upwind-Biased
Remarks
The finite difference methods are based on the assumption of uniform grids.
1st-Order Backward Finite Difference (BFD1)
PDE LHB Sectio Section Section Sbound Section Section Sectio RHB
Order Node 0 n1 1 1 Node 4 2 2 n2 EndNo
Node Node 2 Node 3 Node 5 Node 6 Node de
1 7
1-ddx FFD1 BFD1 BFD1 BFD1 BFD1 BFD1 BFD1 BFD1 BFD1
2-d2dx2 None CFD2 CFD2 CFD2 ddx Cont CFD2 CFD2 CFD2 None
2nd-Order Backward Finite Difference (BFD2)
PDE LHB Section Section Section Sbound Section Section Section RHB
Order Node 0 1 1 1 Node 4 2 2 2 EndNode
Node 1 Node 2 Node 3 Node 5 Node 6 Node 7
1-ddx FFD2 CFD2 BFD2 BFD2 BFD2 CFD2 BFD2 BFD2 BFD2
2-d2dx2 None CFD2 BFD2 BFD2 ddx Cont CFD2 BFD2 BFD2 None
2nd-Order Central Finite Difference (CFD2)
Node 1 Node 2 Node 3 Node 5 Node 6 Node 7
1-ddx FFD2 CFD2 CFD2 CFD2 BFD2 CFD2 CFD2 CFD2 BFD2

4th-Order Central Finite Difference (CFD4)
PDE LHB Section Section Section Section Sbound Section Section Section Section RHB
Order Node 1 1 1 1 Node 5 2 2 2 2 EndNode
0 Node Node Node Node Node Node Node Node
1 2 3 4 6 7 8 9
1-ddx FFD4 DBFD4 CFD4 CFD4 UBFD4 BFD4 DBFD4 CFD4 CFD4 UBFD4 BFD4
2-d2dx2 None DBFD4 CFD4 CFD4 UBFD4 ddx Cont DBFD4 CFD4 CFD4 UBFD4 None
1st-Order Forward Finite Difference (FFD1)
Order Node 1 1 1 Node 4 2 2 2 EndNode
0 Node Node Node Node 5 Node 6 Node
1 2 3 7
1-ddx FFD1 FFD1 FFD1 FFD1 BFD1 FFD1 FFD1 FFD1 BFD1
2nd-Order Forward Finite Difference (FFD2)
Node 1 Node Node 3 Node 5 Node 6 Node 7
2
1-ddx FFD2 FFD2 FFD2 CFD2 BFD2 FFD2 FFD2 CFD2 BFD2
2-d2dx2 None FFD2 FFD2 CFD2 ddx Cont FFD2 FFD2 CFD2 None
4th-Order Upwind Biased Finite Difference (UBFD4)
PDE LHB Section Section Section Section Sbound Section Section Section Section RHB
Order Node 1 1 1 1 Node 5 2 2 2 2 EndNode
0 Node Node Node Node Node Node Node 8 Node
1 2 3 4 6 7 9
1-ddx FFD4 DBFD4 CFD4 UBFD4 UBFD4 BFD4 DBFD4 CFD4 UBFD4 UBFD4 BFD4
2- None DBFD4 CFD4 UBFD4 UBFD4 ddx Cont DBFD4 CFD4 UBFD4 UBFD4 None
d2dx2
Discretization Methods: Finite Element

Methods
The finite element family of methods divides the domain into many smaller
elements and applies weighted residual methods within each element. The
weighted residual methods assume that the solution can be represented as a
weighted combination of known polynomial approximations of order n with
unknown weighting coefficients. The latter are determined to satisfy boundary
conditions and the system of equations at a finite set of points, called
collocation or polynomial points. The precise criterion used to choose the
weighting coefficients determines the method.

The weighted residual method that is combined with the finite element
concept is the method of orthogonal collocation (OCFE). Second-, third-, and
fourth-order OCFE methods are provided as shown in the following table. You
can specify that you want to use an OCFE method using the
DiscretizationMethod property assignment when declaring a domain.
Method Order of Type of Approximation Polynomial (Collocation)

Approximation Points
OCFE2 2nd-order Orthogonal Collocation on 0.00000, 0.50000, 1.00000
Finite Elements
OCFE3 3rd-order Orthogonal Collocation on 0.00000, 0.21132, 0.50000,

Finite Elements 0.78868, 1.00000
OCFE4 4th-order Orthogonal Collocation on 0.00000, 0.11270, 0.50000,

Finite Elements 0.88730, 1.00000
As shown in the following tables for OCFE2, OCFE3, and OCFE4, the domain is
divided into a number of elements and orthogonal collocation is applied within
each element. The value of EndNode is equal to the number of elements
times the order of the method, where the order is the number of collocation
points plus one.
By using the Interior set provided, the PDEs in your model are evaluated
automatically at the interior collocation points only. The correct derivatives
are also evaluated automatically.
For 1st-order PDE problems, the first derivatives at the element boundaries
are calculated using the element to the left of the boundary. For 2nd-order
PDEs, the OCFE methods are implemented using the condition that the first
derivatives are continuous between elements. This treatment of the
boundaries between elements gives a solution that is continuous, with
continuous flux, as in the exact solution.
OCFE2
PDE Order LHB Element ElemBound Element2 RHB

Node 0 1 Coll 1 Node 2 Coll 1 EndNode
4
1 ddx ddx ddx Elem1 ddx ddx
2 None d2dx2 ddx Cont d2dx2 None
OCFE3
PDE LHB Element Element ElemBound Element Element RHB

Order Node 1 1 Node 3 2 2 EndNode
0 Coll 1 Coll 2 Coll 1 Coll 2 6
1 ddx ddx ddx ddx Elem1 ddx ddx ddx
2 None d2dx2 d2dx2 ddx Cont d2dx2 d2dx2 None

OCFE4
PDE LHB Element1 Element1 Element1 ElemBound Element2 Element2 Element2 RHB
Order Node Coll 1 Coll 2 Coll 3 Node 4 Coll 1 Coll 2 Coll 3 EndNode
0 8
1 ddx ddx ddx ddx ddx Elem1 ddx ddx ddx ddx
2 None d2dx2 d2dx2 d2dx2 ddx Cont d2dx2 d2dx2 d2dx2 None
Finite Element Remarks

One advantage of finite element methods is their ability to handle problems
with steep gradients by concentrating many small elements in areas with
steep gradients.
Mixer-Reactor-Absorber (PDE)
Example Description
This example demonstrates the use of PDE modeling in the dynamic
simulation of an isothermal process flowsheet consisting of a well-stirred
mixing tank, a tubular reactor, and a gas absorber.
Components A and B at a concentration ratio of 2 to 1 in the fresh feed
stream enter the mixer along with the contents of recycle stream. The mixer
outlet stream is sent to the reactor which carries out the gas-phase reaction:
A + B →2C. The reactor product enters the bottom of the countercurrent
absorber where C is partially absorbed in the liquid phase. The remaining gas
is recycled to the mixing tank.
The dynamic simulation requires the simultaneous solution of the ordinary
differential equations (ODEs) describing the mixer and the PDAEs for the
tubular reactor and gas absorber.
This example is taken from the paper by M. Oh and C.C. Pantelides, A
modeling and simulation language for combined lumped and distributed
parameter systems, Computers & Chemical Engineering, Volume 20, Issues
6/7, pp. 611-633 (1996).
Also, refer to the section: Running the Mixer-Reactor-Absorber (PDE) Example
Running the Mixer-Reactor-

Absorber (PDE) Example
Example simulations are included in your Aspen Custom Modeler installation.
If you have installed in the default location, the files for this example will be
in the folder:
C:\Program Files\AspenTech\Aspen Custom Modeler

12.1\Examples\MixerReactorAbsorber

To prepare to run this example:
• Copy the files in the example directory to a convenient working folder, for
example:
C:\Program Files\AspenTech\Working Folders\Aspen Custom

Modeler 12.1\MixerReactorAbsorber
To run this example, follow these steps:
1 Double-click the file MixerReactorAbsorber.acmf to open it.
2 In the All Items pane of the Simulation Explorer, ensure Flowsheet is
selected.
3 In the Contents pane, double-click on Blocks icon.
4 In the Contents pane, double-click the Absorber icon.
5 In the Contents pane, double-click the plot icon to display it.
6 Repeat steps 4 and 5 for the Mixer and Reactor.
7 Ensure that the run mode is Dynamic, and then run the simulation.
The 3D profile plots provided show the concentration profiles of component A

in the reactor and gas absorber as a function of the spatial and temporal
domains. The 2D plot for the mixer shows the time transient concentration of
component A.
Jacketed Reactor (PDE and

Integral) Example Description
In this example, a fixed-bed, jacketed tubular reactor is used to carry out the
oxidation of o-xylene (component A) to phthalic anhydride (component B).
The top-level cooled reactor model contains a pseudo-homogeneous 2D PDE
sub-model for the tubular reactor and a cooling jacket sub-model that
assumes perfect mixing and constant holdup. The two sub-models are linked
by equating the reactor wall temperature to that of the cooling medium, and
expressing the total heat input to the jacket as an integral of the heat loss
from the tube surface over the entire tube length. The resistance to heat
transfer occurs primarily between the reactor contents and tube wall. The wall
of the tube is essentially at the cooling medium temperature.
This example is taken from the paper by M. Oh and C.C. Pantelides, A
modeling and simulation language for combined lumped and distributed
parameter systems, Computers & Chemical Engineering, Volume 20, Issues
6/7, pp. 611-633 (1996).
Also, refer to the section: Running the Jacketed Reactor (PDE and Integral)
Example.

Running the Jacketed Reactor
(PDE and Integral) Example
in the folder:

12.1\Examples\JacketedReactor
example:

Modeler 12.1\JacketedReactor
1 Double-click the file JacketedReactor.acmf to open it.
selected.
4 In the Contents pane, double-click the CooledReactor icon.
5 In the Contents pane, double-click on the three plot icons to display them.
The two 3D profile plots show the concentration profiles of component B and
the temperature profiles at the center of the tubular reactor over time. The
2D plot presents the cooling load and temperature of the cooling medium
over time.
Crystallizer (PDE and Integral)

Example Description
This example considers a continuous mixed suspension mixed product
removal (CMSMPR) crystallization unit that operates in cooling mode to
produce crystals from aqueous solutions. The crystallizer temperature is
assumed to be constant and any crystal breakage is ignored. The IPDAE
model consists of the mass balance (ODE), crystal population balance (PDE),
integral equation for magma density, and several algebraic equations.
The dynamic simulation considers the start-up of the crystallizer, with the unit
initially full of pure water (i.e., the initial liquid concentration of solute and the
logarithmic number density for all non-zero crystal sizes are set to zero. The
liquid phase concentration remains below the saturation line for the initial

0.77 hours of operation and then rises towards the feed value (0.15 kg/kg)
until the rate of crystallization increases sufficiently to reverse this trend.
This example is taken from the Ph.D. Thesis by M. Oh, Modelling and
Simulation of Combined Lumped and Distributed Processes, Imperial College
of Science, Technology and Medicine, London (1995).
Also refer to; Running the Crystallizer (PDE and Integral) Example.
Running the Crystallizer (PDE

and Integral) Example
in the folder:

12.1\Examples\Crystallizer
example:

Modeler 12.1\Crystallizer
1 Double-click the file Crystallizer.acmf to open it.
selected.
4 In the Contents pane, double-click the Crystallizer icon.
5 In the Contents pane, double-click the three plot icons to display them.
6 Right-click on the profile plot for Crystal_Population_Density and select
Properties.
7 Click on the Contours tab and change the Number of levels to 10.
8 Click OK.
The 3D profile plot shows the crystal population density profiles in the
crystallizer as a function of the crystal size and time. The 2D plots show the
time transients for the magma density and the liquid phase solute
concentration.

PDE Modeling Glossary
Dirichlet
Dirichlet is a type of boundary condition that specifies the values of the
dependent variables at the boundary points as a function of time.
Distributed parameter system

A distributed parameter system contains process variables that vary with
respect to spatial position as well as time.
Distribution
A distribution is a variable distributed over one or more domains.
Domain
A domain is a region over which a distributed parameter system is
discretized.
Method of Lines
The Method of Lines is an approach to solving time-dependent IPDAE systems
by discretizing in space, but leaving the time variable continuous.
Neumann
Neumann is a type of boundary condition that specifies the values of the first-
order spatial derivatives on the boundary.

6 Modeling Language
Conventions
This section describes:

• Conventions used within the modeling language.
• Test conventions used within this manual when describing the modeling
language.
Modeling Language
Conventions
This section describes the conventions you must follow when writing modeling
language.
Adding Comments
You can use two comment markers:
Symbol Effect
// Everything on the same line to the right of the // characters
is commented out. A second // on the same line has no
effect.
/* ... */ All text between /* and */ is commented out. /* and */ pairs

can be nested.
In the model editor, all commented out text appears highlighted in green.
However, note that the text editor does not highlight commented information
after a nested comment.
Comments Examples
The following examples show the use of the comment markers:
// This is a comment
// This is a comment // This is a comment also
6 Modeling Language Conventions 181

/*All text in this multi-line sentence is
commented out */
General Modeling Language

Conventions
The modeling language is free format. This means that you can write syntax
over several lines and insert spaces for clarity.
The modeling language is not case sensitive.
Text Conventions
The conventions used in the documentation that describes the modeling
language are:
Convention Example Meaning

Text in courier font FlowIn = FlowOut ; Examples of modeling
language.
... Item , Item ... A list of arguments.
: MODEL HeatedTray USES More text is required but not

Heater shown.
:
END
Text in bold, upper MODEL, CONNECT, AND Aspen Custom Modeler

case keywords.
Text in lower case VariableName, Value You must supply a name,

with initial capital quantity, etc.
letters
Italic text VALUE:Value , This input is optional.

HIBOUND:HiBound ;
6 Modeling Language Conventions 182

General Information
Copyright
Version Number: 2004.1
April 2005
Copyright © 1982-2005 Aspen Technology, Inc, and its applicable

subsidiaries, affiliates, and suppliers. All rights reserved. This Software is a
proprietary product of Aspen Technology, Inc., its applicable subsidiaries,
affiliates and suppliers and may be used only under agreement with
AspenTech.
Aspen ACOL™, Aspen Adsim®, Aspen Advisor™, Aspen Aerotran®, Aspen
Alarm & Event™, Aspen APLE™, Aspen Apollo Desktop™, Aspen Apollo
Online™, Aspen AssetBuilder™, Aspen ATOMS™, Aspen Automated Stock
Replenishment™, Aspen Batch Plus®, Aspen Batch.21™, Aspen BatchCAD™,
Aspen BatchSep™, Aspen Calc™, Aspen Capable-to-Promise®, Aspen
CatRef®, Aspen Chromatography®, Aspen Cim-IO for ACS™, Aspen Cim-IO
for Csi VXL™, Aspen Cim-IO for Dow MIF™, Aspen Cim-IO for G2™, Aspen
Cim-IO for GSE D/3™, Aspen Cim-IO for Hewlett-Packard RTAP™, Aspen Cim-
IO for Hitachi PLC (H04E)™, Aspen Cim-IO for Intellution Fix™, Aspen Cim-IO
for Melsec™, Aspen Cim-IO for WonderWare InTouch™, Aspen Cim-IO for
Yokogawa Centum CS™, Aspen Cim-IO for Yokogawa Centum XL™, Aspen
Cim-IO for Yokogawa EW3™, Aspen Cim-IO Interfaces™, Aspen Cim-IO
Monitor™, Aspen Cim-IO™, Aspen Collaborative Demand Management™,
Aspen Collaborative Forecasting™, Aspen Compliance.21™, Aspen
COMThermo TRC Database™, Aspen COMThermo®, Aspen Cost Factor
Manual™, Aspen Crude Manager™, Aspen Crude Margin Evaluation™, Aspen
Custom Modeler®, Aspen Data Source Architecture™, Aspen Decision
Analyzer™, Aspen Demand Manager™, Aspen DISTIL™, Aspen Distribution
Scheduler™, Aspen DMCplus® Composite, Aspen DMCplus® Desktop, Aspen
DMCplus® Online, Aspen DPO™, Aspen Dynamics®, Aspen eBRS™, Aspen
Enterprise Model™, Aspen ERP Connect™, Aspen FCC®, Aspen FIHR™, Aspen
FLARENET™, Aspen Fleet Operations Management™, Aspen Framework™,
Aspen FRAN™, Aspen Fuel Gas Optimizer Desktop™, Aspen Fuel Gas
Optimizer Online™, Aspen General Construction Standards™, Aspen Hetran®,
Aspen HX-Net®, Aspen Hydrocracker®, Aspen Hydrotreater™, Aspen HYSYS
Amines™, Aspen HYSYS Crude™, Aspen HYSYS Dynamics™, Aspen HYSYS
OLGAS 3-Phase™, Aspen HYSYS OLGAS™, Aspen HYSYS OLI Interface™,
General Information 183

Aspen HYSYS Tacite™, Aspen HYSYS Upstream Dynamics™, Aspen HYSYS
Upstream™, Aspen HYSYS®, Aspen Icarus Process Evaluator®, Aspen Icarus
Project Manager®, Aspen InfoPlus.21®, Aspen Inventory Balancing™, Aspen
IQ Desktop™, Aspen IQ Online™, Aspen IQmodel Powertools™, Aspen
Kbase®, Aspen LIMS Interface™, Aspen Local Security™, Aspen LPIMS™,
Aspen MBO™, Aspen MIMI®, Aspen MPIMS™, Aspen Multivariate Server™,
Aspen MUSE™, Aspen NPIMS™, Aspen OnLine®, Aspen Operations Manager -
Event Management™, Aspen Operations Manager - Integration
Infrastructure™, Aspen Operations Manager - Performance Scorecarding™,
Aspen Operations Manager - Role Based Visualization™, Aspen Order Credit
Management™, Aspen Orion Planning™, Aspen Orion™, Aspen PEP Process
Library™, Aspen PIMS Blend Model Library™, Aspen PIMS Distributed
Processing™, Aspen PIMS Enterprise Edition™, Aspen PIMS Mixed Integer
Programming™, Aspen PIMS Simulator Interface™, Aspen PIMS Solution
Ranging™, Aspen PIMS Submodel Calculator™, Aspen PIMS XNLP
Optimizer™, Aspen PIMS™, Aspen PIPESYS™, Aspen PIPE™, Aspen Planning
Accuracy™, Aspen Plant Planner & Scheduler™, Aspen Plant Scheduler Lite™,
Aspen Plant Scheduler™, Aspen Plus OLI Interface™, Aspen Plus®, Aspen
Polymers Plus®, Aspen PPIMS™, Aspen Process Data™, Aspen Process
Explorer™, Aspen Process Manual™, Aspen Process Order™, Aspen Process
Plant Construction Standards™, Aspen Process Recipe®, Aspen Process
Tools™, Aspen Product Margin & Blending Evaluation™, Aspen Production
Control Web Server™, Aspen ProFES® 2P Tran, Aspen ProFES® 2P Wax,
Aspen ProFES® 3P Tran, Aspen ProFES® Tranflo, Aspen Properties®, Aspen
Pumper Log™, Aspen Q Server™, Aspen RateSep™, Aspen RefSYS
CatCracker™, Aspen RefSYS Spiral™, Aspen RefSYS™, Aspen Report Writer™,
Aspen Resource Scheduling Optimization™, Aspen RTO Watch Cim-IO
Server™, Aspen RTO Watch Server™, Aspen Scheduling & Inventory
Management™, Aspen SmartStep Desktop™, Aspen SmartStep Online™,
Aspen SQLplus™, Aspen Supply Chain Analytics™, Aspen Supply Chain
Connect™, Aspen Supply Planner™, Aspen Tank Management™, Aspen TASC-
Mechanical™, Aspen TASC™, Aspen Teams®, Aspen Terminals
Management™, Aspen TICP™, Aspen Transition Manager™, Aspen Turbo
PIMS™, Aspen Utilities™, Aspen Voice Fulfillment Management™, Aspen
Watch Desktop™, Aspen Watch Server™, Aspen Water™, Aspen Web
Fulfillment Management™, Aspen WinRace Database™, Aspen XPIMS™,
Aspen Zyqad Development Version™, Aspen Zyqad™, SLM™, SLM
Commute™, SLM Config Wizard™, the aspen leaf logo, and Plantelligence are
trademarks or registered trademarks of Aspen Technology, Inc., Cambridge,
MA.
All other brand and product names are trademarks or registered trademarks
of their respective companies.
This document is intended as a guide to using AspenTech's software. This
documentation contains AspenTech proprietary and confidential information
and may not be disclosed, used, or copied without the prior consent of
AspenTech or as set forth in the applicable license.
Corporate
Aspen Technology, Inc. Phone: (1) (617) 949-1000
Ten Canal Park Toll Free: (1) (888) 996-7001
Cambridge, MA 02141-2201 Fax: (1) (617) 949-1030

USA URL: http://www.aspentech.com

Related Documentation
In addition to this document, the following documents are provided to help
users learn and use the Aspen Utilities applications.
Title Content
Aspen Custom Modeler 2004.1 Getting Started Contains basic hands-on tutorials to
Guide help you become familiar with Aspen
Custom Modeler.
Aspen Custom Modeler 2004.1 User Guide Contains a general overview of ACM
functionality and more complex and
extensive examples of using Aspen
Custom Modeler.
Aspen Custom Modeler 2004.1 Library Contains reference information on

Reference control models, property procedure
types, utility routines, port types, and
variable types.
Aspen Custom Modeler 2004.1 Aspen Modeler Contains information on using

Reference automation, solver options, physical
properties, the Control Design
Interface (CDI)), Simulation Access
eXtensions, online links, and using
external nonlinear algebraic solvers.
Aspen Custom Modeler 2004.1 DMCplus® Contains information on using

Controllers Interface DMCplus with Aspen Custom Modeler
or Aspen Dynamics™.
Aspen Custom Modeler 2004.1 Polymer Polymers Plus is a layered product of

Simulations with Polymers Plus Aspen Custom Modeler. It provides
additional functionality to the
properties package, Properties Plus,
enabling polymers to be fully
characterized in Aspen Custom
Modeler models.

Technical Support
Online Technical Support

Center
AspenTech customers with a valid license and software maintenance
agreement can register to access the Online Technical Support Center at:
http://support.aspentech.com
You use the Online Technical Support Center to:
• Access current product documentation.
• Search for technical tips, solutions, and frequently asked questions
(FAQs).
• Search for and download application examples.
• Search for and download service packs and product updates.
• Submit and track technical issues.
• Search for and review known limitations.
• Send suggestions.
Registered users can also subscribe to our Technical Support
e-Bulletins. These e-Bulletins proactively alert you to important technical
support information such as:
• Technical advisories.
• Product updates.
• Service Pack announcements.
• Product release announcements.
Technical Support 187

Phone and E-mail
Customer support is also available by phone, fax, and e-mail for customers
who have a current support contract for their product(s). Toll-free charges are
listed where available; otherwise local and international rates apply.
For the most up-to-date phone listings, please see the Online Technical
Support Center at:
http://support.aspentech.com
Support Centers Operating Hours

North America 8:00 – 20:00 Eastern time
South America 9:00 – 17:00 Local time
Europe 8:30 – 18:00 Central European time
Asia and Pacific Region 9:00 – 17:30 Local time
Technical Support 188

Index
parameter types 34
/ solved type 36
/ used as comment symbol 190
C
A C language
ALWAYSCALL keyword 83 modeling language for external procedures
81
Arithmetic operators 99
CHANGESINPUTS keyword 83
Arrays
Comments in modeling language 190
distinguished from sets 22
Comparison operators 102
in equations 100
Compatibility with SPEEDUP 5.5 83
modeling language for 25
ComponentList property 92
values assigned to elements 17
Conditional equations
with many elements or dimensions 106
Assignments
sets in 105
array elements 17
CONNECT keyword 114
lists 18
Connection built-in stream type 114
Connection port type, built-in 36
statements for 11
Connections
values to variables in tasks 55
modeling language for streams 79
WITHIN and ENDWITH 18
ConnectionSet property 96–97, 119
available 197
ControlSignal built-in stream type 114
B Conventions
documentation 191
Base types 34
modeling language 190
BLOCK keyword 20
Converting
Blocks
strings and integers 26
passing values between 78
units of measurement 38
Built-in types
CONVERTTO operator 27
connection stream type 114
Crystallizer (PDE) Example Description 187
control signal stream type 114
Index 189
D Fortran
modeling language for external procedures
Declaration statements 14 80
Definitions, modeling language for re-using 28 Free specification, defined 125
DERIVATIVES keyword 83 Functions
DIFFERENCE operator 23, 108 in sets 23
Discontinuities modeling language for 81
Disturbances G
modeling language for 49 GLOBAL qualifier 88
documentation 195
Duplicate names, BLOCK and STREAM 20 H
HIDDEN qualifier 88
E
ENDPARALLEL keyword 61 I
ENDWITH keyword 18
IF keyword
Equations
defining conditionals in tasks 56
calling procedures 110
in conditional equations 102
conditional 102
Indexed assignments 17
logical operators 103
Inheritance
mathematical operators 98
defined 28
related types 29
referring to port variables 113
Initial specification, defined 125
switching between 134
INPUT qualifier 88
with array variables 100
IntegerParameter built-in parameter type 35
External procedures
Integers, converting to strings 26
IntegerSet 22
External programs
INTERSECTION operator 23, 108
identifying objects from 19
IS keyword 131, 134
EXTERNAL qualifier 129
IsConnected property 92–93
IsMatched attribute 94
F
Fixed specification, defined 125 J

Fixed values 21
Jacketed Reactor (PDE and Integral) Example
Flow, direction of 91 Description 186
FOR loops
L
ForEach loops LINK keyword 117
modeling language for 108 Logical operators 99, 102
Index 190
LogicalParameter built-in parameter type 35 P
M PARALLEL keyword 61
Parameters
Matching port variables 94
built-in types 34
Mathematical operators 98
defined 21
MAX operator 99, 108
Method of Lines 180
relating to variables 98
MIN operator 98, 108
Path names, assignment lists for 18
Mixer-Reactor-Absorber (PDE) Example
Description 185 Pausing a simulation 63
MODEL keyword 87 PDAE models 165
Modeling language PDE models 138
conventions for writing 190 Physical properties
Models modeling language for 32
connecting 114, 117 Physical quantities 40
modeling language for model types 87 PhysicalQuantity property 40
specifications in 125 PORT keyword 78
variables in 88 Ports
Modes of operation 134 collections of 95
MOL for PDEs 180 connecting 92
Multidimensional arrays, SIGMA for 122 modeling language for 78, 91, 114, 117
Multiports 95 properties for 92–93

restrictions in streams 79
N variables in equations 113
POSTCALL keyword 83
Names
PRECALL keyword 83
duplicate 20
Predefined types 34
rules for choosing 19
PRESET keyword 83
O Printing
task messages to Simulation Messages
Objective specification, defined 125
window 62
Operators
Procedures
ForEach expressions 109
called from equations 110
logical 102
mathematical 98
PRODUCT operator 98, 108
Optimization simulations
Properties
adding to variables 34, 76, 80
OUTPUT qualifier 88
for ports 92–93
naming conventions 19
Index 191
writing sub-models 129 in models 125
PROPERTIES keyword 83 SPEEDUP
COMPATIBILITY statement 83
Q SRAMP keyword 57
Qualifiers for variables 88 Statements
repeating 106, 109
R syntax for 11
Ramp functions in tasks 57 Step changes during simulation 102
RAMP keyword 57 STREAM keyword 20, 79
RateInitial specification, defined 125 Streams
RealParameter built-in parameter type 35 types for 79
RealVariable built-in solved type 36 StringParameter built-in parameter type 34
Relational operators 102 Strings, converting to integers 26
Repeating statements 106, 109 StringSet 22
RESTART statement 58 Stringsets 26
Running the Crystallizer (PDE) Example 188 Sub-models 127, 129, 132
Running the Jacketed Reactor (PDE and Switch definitions 134

Integral) Example 186 Symbols
Running the Mixer-Reactor-Absorber (PDE) for sets 23
Example 185
SYMDIFF operator 23, 108
S SystemLibrary library 34
Sets T
arithmetic for 23
Task statements
defined 22, 25
assigning values to variables 55
in conditional expressions 105
IF construct 56
SIZE function 124
modeling language for creating snapshots 60
SIGMA
overview 55
for multidimensional arrays 122
parallel 61
ForEach operator 108
pausing a simulation 63
syntax for 121
printing to Simulation Messages window 62
Simulation Messages window
ramping value of variables 57
task messages in 62
suspending 58
Simulations
Tasks
optimization 69
callable 52
SIZE function 124
Snapshots
opening simulations 54
task statements 55
Specifications
Testing
Index 192
parameters modeling language 21 W
Text conventions in documentation 191
WAIT statement 58
Types
WITHIN keyword 18
built-in 28, 34
model 87
modeling language for 13, 14
naming conventions 19
parameters 21
physical properties 32
port 78
related 29
stream 79
variable 74
virtual types in sub-models 132
UNION operator 23, 108

Units of measurement
automation 40
UOM
USES keyword 28, 78
UseUOMof qualifier 88
Values
array elements 17
fixed 21
VARIABLE keyword 74
Variables
adding properties to 34, 76, 80
in models 88
relating to parameters 98
solved for 36
Vary specification, defined 125
Index 193

ACM Modeling Language Reference

Uploaded by

Copyright:

Available Formats

ACM Modeling Language Reference

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ACM Modeling Language Reference

Uploaded by

Copyright:

Available Formats

What are the main sections covered in the document?

What are the main sections covered in the document?

What modeling concepts are discussed?

What modeling concepts are discussed?

Aspen Custom

This guide contains detailed reference information about the modeling

Who Should Read this Guide 2

1 MODELING LANGUAGE CONCEPTS ............................................................. 11

2 MODELING LANGUAGE FOR SIMULATIONS................................................ 35

3 MODELING LANGUAGE FOR TYPES ............................................................ 72

4 MODELING LANGUAGE FOR MODELS ......................................................... 83

5 MODELING PDE SYSTEMS .........................................................................129

6 MODELING LANGUAGE CONVENTIONS ......................................................181

Introducing Aspen Custom Modeler 9

The Aspen Custom Modeler modeling language uses a number of common

Syntax Element Type of Statement

1 Modeling Language Concepts 11

CONNECT AND Connection

LINK AND Link

Statements must end with the ; (semi-colon).

Feed.T: 75.0, Fixed;

Equate two variables in a model:

Make a connection between the input and output of two blocks:

CONNECT Pump1.Output1 AND Valve2.Input1;

LINK Tank5.Output1 AND MultiOutput1;

Type Definition Syntax

1 Modeling Language Concepts 12

TypeKeyword One of these keywords: MODEL,

Type Definition Example

Identifier .................................. The name of a property being

1 Modeling Language Concepts 13

Declaration Statement Remarks

Declaration Statement Example

Object ....................................... A variable or parameter

1 Modeling Language Concepts 14

1 Modeling Language Concepts 15

For further examples of assignments, see Defining Specifications in Models

Assignments to Array Elements

Assigning to Array Elements Syntax

ArrayVariable ............................ The name of the array variable

Assigning to Array Elements Example

Note: For convenience Components is defined as the index set

1 Modeling Language Concepts 16

Assignment List Syntax

Property.................................... Property, such as a variable within

Assignment List Example

Tank.Q: VALUE:25.0, SPEC: Fixed, LOWER: -10.0;

If the property name is omitted it is assumed that the assignment is to the

Tank.Q: 25.0, FIXED, LOWER: -10.0;

Assignments using WITHIN and

WITHIN and ENDWITH Syntax

Scope ........................................ The block or variable to which

1 Modeling Language Concepts 17

WITHIN / ENDWITH Example

Make sure that the names you choose:

1 Modeling Language Concepts 18

BlockName ................................ Name of a block in the flowsheet

BLOCK Keyword Example

ParameterName ........................ Name of the parameter

1 Modeling Language Concepts 19