next up previous contents index
Next: Nodal Results Local Systems Up: Format of the neutral Previous: Loading Data Records

Subsections


User Defined Results

 

Overview of User Defined Results

  FEMVIEW is able to display any type of finite element results; e.g. heat flow, or fluid dynamics, as well as structurally related results. This is achieved by allowing the format of results data sets to be variable. A FEMVIEW results data set will thus contain the results data plus some additional data set headers which describe the form in which the results are given. This section explains how User Defined Results data sets are formatted.

Entities

  Finite element results may be attributed to various types of geometric object in a model, henceforth refered to as entities. FEMVIEW is able to display results attributed to the five types of entity illustrated below. Note that multi-surface results may be entered for all entity types except `Materially Independant Nodal'.

Incomplete sets of results are acceptable for both nodal and element-wise results, ie: it is not necessary to give results data for each node or element in the model. However, nodal results data must be input in the same order as the nodes are given in the nodal point coordinate data, and element-wise results must be input in the same order as the elements are defined in the element topology. In addition, for element nodal results, the order in which the nodes are defined in the element topology must be maintained.


  
Figure B.2: Nodal Results, Materially independant
\begin{figure}
\centerline{
\psfig {figure=app_a.fig2,height=1.5in}
}\end{figure}


  
Figure B.3: Nodal Results, Materially dependant
\begin{figure}
\centerline{
\psfig {figure=app_a.fig3,height=1.5in}
}\end{figure}


  
Figure B.4: Element-wise Results, at Element Nodes
\begin{figure}
\centerline{
\psfig {figure=app_a.fig4,height=1.5in}
}\end{figure}


  
Figure B.5: Element-wise Results, at Element Gauss Points
\begin{figure}
\centerline{
\psfig {figure=app_a.fig5,height=1.5in}
}\end{figure}


  
Figure B.6: Element-wise Results, Invariant over the Element
\begin{figure}
\centerline{
\psfig {figure=app_a.fig6,height=1.5in}
}\end{figure}

Attributes

  An attribute is a type of result which occurs at an entity. Stress, displacement, velocity, heat flow are examples of attributes.

Components

  An attribute can be composed of one or more components. These components might be the X,Y, and Z components of a vector quantity, or the matrix components of a tensor field for example.

Attribute Variants

  In the most simple case, the components of an attribute will be the same for all entities in a model at which that attribute occurs. Sometimes, this will not be the case, taking for example nodal stresses in a model comprised of three-dimensional brick elements and two-dimensional flat plate elements. The finite element analysis will produce a 3-D stress field for the brick elements and a 2-D stress field for the plate elements. The stress components will thus vary between entities. This variation in the number of components for an attribute can be defined.

User Defined Results Header Records

  A User Defined Results data set is denoted by a `KEY' of 100 in the data set header record. The data set header record will then be followed by an Attribute Header Record, up to twenty Component Definition Records, and optionally up to five Attribute Variant Records. Materially Independant Nodal results will have no Attribute Variants.

The Attribute Header Record is used to assign a name to the results attribute that is to follow (e.g. `STRESS', `DISPLACE'). It also specifies the total number of results components, the entity type to which these components are attributed, and the number of attribute variants if any. It is not possible to have more than one entity type for a single results data set.

The Component Definition Records are used to give information about each of the components of the results attribute. A Component Definition Record assigns a name to each component (e.g. `SXX', `FZ' etc) and describes the type of component (e.g. scalar, vector X-component etc). Furthermore, it is possible to specify whether the component is to be used as it is, or combined with other components using a pre-defined FEMVIEW calculation.

The Attribute Variant Records are optional. FEMVIEW will assume that all of the results components are present at each entity (and in the order in which they are defined by the Component Definition Records), if no Attribute Variant Records are used. Where there is a variation in the number of components at each entity, an Attribute Variant Record must be used to define which components are present and the order in which they appear in the results records which follow. Up to five such mappings may be specified, each being identified by a Record Variant Identification number. These Record Variant Identification numbers are then referenced by the entity header records (see later) to determine which result components are present at a particular entity.

Attribute Header Record

 

This record contains the characteristics of the attribute of the form:

KEY, Name, NCOMPS, IRTYPE, NORCTY, Name_orig

to be written with the format:

FORMAT (1X, I2, 2X, 8A1, 3I5, 10X, 8A1)

Where:


KEY = -4
Name = Attribute name to be used in menu
NCOMPS = Total number of components (up to 20)
IRTYPE = Entity type to which components are attributed :
1 = Nodal : material independant
2 = Nodal : material dependant
3 = Element : at nodes
4 = Element : at Gauss points
5 = Element : constant for element
NORCTY = Number of data record variants :
0 = components are as defined
1-5 = number of mappings
Name_orig = Name of attribute to be used as a base attribute.
(Used only when incrementally updating results. ref: B.2)

Component Definition Record

  For each component of the attribute details are given in the form:

KEY, Name, MENU, ICTYPE, ICIND1, ICIND2, IEXIST, ICNAME, ICDATA

To be written with the format:

FORMAT (1X, I2, 2X, 8A1, 5I5, 8A1, 8A1)
Where:


KEY = -5
Name = Component name
MENU = Menu inclusion indicator
0 = not in menu
1 = name in menu
ICTYPE = Type of component
0 = enumerative
1 = scalar
2 = vector : three cartesian components and optionally
three rotations
3 = vector : magnitude and direction cosines
4 = matrix
5 = Homogeneous Transformation Matrix
11 = complex scalar
12 = complex vector : three cartesian components and
optionally three rotations
13 = complex vector : magnitude and direction cosines
14 = complex matrix


ICIND1 = Sub-component index
for ICTYPE = 0, position of enumerative data item in current list of enumerative data items.
For ICTYPE = 2
1 = X-component
2 = Y-component
3 = Z-component
and optionally
4 = rotation about X-axis
5 = rotation about Y-axis
6 = rotation about Z-axis
For ICTYPE = 3
1 = magnitude
2 = direction cosine for X-axis
3 = direction cosine for Y-axis
4 = direction cosine for Z-axis
For ICTYPE = 4, row number
ICIND2 = Column number for ICTYPE = 4
For ICTYPE=3, in order to associate up to three such vectors together as for Principal Stresses. then ICIND2 must be set = 1, 2 and 3.
Note that if ICTYPE=3 and ICIND2=0 then the data will be assumed to be as for Principal Stresses and the ICIND2=1, 2,and 3 association will be applied internally.
In other cases where ICTYPE=3, then set ICIND2=1.
IEXIST = Existence indicator
0 = exists as data in results record.
1 = to be calculated by FEMVIEW calculation.
For ICTYPE = 0, set = -n, where `n' is the length of the current enumerative list (max list length = 5).
ICNAME = Name of FEMVIEW calculation.
ICDATA = Attribute to be used in calculation (leave blank if other components in this attribute are to be used).

Notes:

1.
The menu inclusion indicator is used to define whether the component is to appear in the FEMVIEW results menus at run time.
2.
For ICTYPE = 4 or 14, the matrix is symmetric of size 3 by 3, only the upper or lower triangular members are input.

3.
Matrix component row and column indices refer to the position of the component in the matrix definition of a tensor field.

4.
The integer IEXIST indicates whether the component is to be displayed as it is, or operated on by one of the pre-defined FEMVIEW calculations. If IEXIST = 1, the name of the FEMVIEW calculation must be supplied (see section B.11). If the FEMVIEW calculation operates on more than one component, the location of the other components must be specified. If the other components are defined in the same data set, the field ICDATA should be left blank. If the other components are defined in a different data set with the same attribute, the name of the attribute should be supplied in the field ICDATA.

5.
Where a data set is going to be used as source data for a FEMVIEW pre-defined calculation such as Principal Stress or Von Mises Stress, then the first given component in the data set must be the SXX component of the data.

6.
The enumerative data type (ICTYPE=0), enables up to five additional values to be associated with a component. The value given in the component data for an enumerative data item will be taken as an integer value. When presenting a component of results that has associated enumerative data then, for numerical presentation modes, the enumerative data items will be displayed adjacent to the component value.

In multi-component attributes it is possible to associate different enumerated data with different components. Enumerated data will be associated with all those subsequently defined components of the attribute up to, and until, further enumerated data is encountered which will then become the current enumerated data.

Attribute Variant Record

  These records define the mapping between the data supplied by a particular instance of an entity and the data order as defined in the attribute component definition data record. A maximum of five Attribute Variant Records is allowed for any single data set.

Of the form:

KEY, IRECTY, NUMCPS, (LSTCPS(I), I=1, NUMCPS)

to be written in the format:

FORMAT (1X, I2, 2I5, 20I3)

Where:


KEY = -6
IRECTY = Record variant identification number
NUMCPS = Number of components in this variant (up to 20)
LSTCPS = For each variant component, the position of the corresponding component in the attribute definition.

Entity Header Records

  All results data for a particular entity is defined using two record types; an Entity Header Record followed by a number of Entity Data records. The exception to this occurs with materially independant nodal results for which it is only necessary to include the Entity Data Records.

Each Entity Header Record defines the entity identification number, the number of repetitions of results at that entity, and the particular variation of components that are valid for this entity.

Note that incomplete sets of results are acceptable for both nodal and element-wise results, ie: it is not necessary to give results data for each node or element in the model. However, nodal results data must be input in the same order as the nodes are given in the nodal point co-ordinate data, and element-wise results must be input in the same order as the elements are defined in the element topology. In addition, for element nodal results, the order in which the nodes are defined in the element topology must be maintained.

Material Dependant Nodal

  Material dependent nodal entity data records are preceded by a header of the form:

KEY, NUMB, NMATS, NSRF, IRECTY, ISYSTM

to be written with the format:

FORMAT (1X, I2, 4I5, 15X, I5)

or: FORMAT (1X, I2, I10, 3I5, 15X, I5) for 10 digit node number format

Where:


KEY = -1
NUMB = node number
NMATS = number of materials (inner repeat)
NSRF = number of surfaces (outer repeat)
IRECTY = record variant identification number


ISYSTM = local system indicator
= 0, results not in local system
= -1, results in local system

Element at Nodes

  Element at nodes entity data records are preceded by a header of the form:

KEY, NUMB, ITYPE, IGROUP, IRECTY, NODAL, NSRF, ISYSTM

to be written with the format:

FORMAT (1X, I2, 6I5, 5X, I5)
or: FORMAT (1X, I2, I10, 5I5, 5X, I5) for 10 digit element number format

Where:


KEY = -1
NUMB = element identification number
ITYPE = FEMVIEW element type (optional). If this is zero, `NODAL' must be set to the number of nodes for this element type.
IGROUP = element group number
IRECTY = record variant identification number
NODAL = $\gt$0 : nodal result indicator (inner repeat). Set to number of nodes for this element. If set=1 then the number of nodes for this element will be determined according to the FEMVIEW element type given in `ITYPE'.
NSRF = number of surfaces (outer repeat).


ISYSTM = local system indicator
= 0, results not in local system
= -1, results in local system

Note that results at the element nodes must be entered in the same order as the nodes are defined in the element topology.

Element at Gauss Points

  Element at Gauss points entity data records are preceded by a header of the form:

KEY, NUMB, ITYPE, IGROUP, IRECTY, NODAL, NSRF, INTEG, ISYSTM

to be written with the format :

FORMAT (1X, I2, 8I5)

or: FORMAT (1X, I2, I10, 7I5) for 10 digit element number format

Where:


KEY = -1
NUMB = element number
ITYPE = FEMVIEW element type (optional)
IGROUP = element group number
IRECTY = record variant identification number
NODAL = 0
NSRF = number of surfaces (outer repeat)
INTEG = number of Gauss points (inner repeat)


ISYSTM = local system indicator
= 0, results not in local system
= -1, results in local system

Element : Constant for element

  Element results that are constant for an element are preceded by a header of the form:

KEY, NUMB, ITYPE, IGROUP, IRECTY, NODAL, NSRF, INTEG, ISYSTM

to be written with the format:

FORMAT (1X, I2, 8I5)
or: FORMAT (1X, I2, I10, 7I5) for 10 digit element number format

Where:


KEY = -1
NUMB = element number
ITYPE = FEMVIEW element type (optional)
IGROUP = element group number
IRECTY = record variant identification number
NODAL = 0
NSRF = number of surfaces (outer repeat)
INTEG = 0


ISYSTM = local system indicator (not currently used).
= 0, results not in local system
= -1, results in local system

Gauss Point Coordinate Record

  For results at element Gauss points, each occurrence of the data records for a Gauss point should be immediately preceded by a record of the form:

KEY, NP, X, Y, Z

to be written with the format:

FORMAT (1X, I2, I5, 3E12.5)

Where:


KEY = -2
NP = identification number of the Gauss point
X = global X co-ordinate
Y = global Y co-ordinate
Z = global Z co-ordinate

Entity Data Record

  The data values of the components for the particular repeat for an entity are of the form:

KEY, NUMB, COMP1, COMP2, COMP3, COMP4, COMP5, COMP6

to be written with the format:

FORMAT (1X, I2, I5, 6E12.5)
or: FORMAT (1X, I2, I10, 6E12.5) for 10 digit node/element number format

Where:


KEY = -2
NUMB = inner repeat identification number
= node number if material independent
= material number if nodal material dependent
= node number if element at nodes
= Gauss point number if element at Gauss points
= element number if constant for element
COMP1-6 = values of the components in order specified
in the record variant definition.

This record is repeated with six components per record until all the required components have been entered. The final card may contain less than six components.


next up previous contents index
Next: Nodal Results Local Systems Up: Format of the neutral Previous: Loading Data Records

Femsys Limited
17th August 1999