1



FePostProc User Manual

Version 2.2

Robert L. Campbell

May 18, 2004

Table of Contents

1. Introduction 4

2. Installation 4

3. Input/Output Format 4

4. GUI Description 7

4.1. File Menu 8

4.2. Preferences 9

4.3. Data Subset Selection 12

4.4. Model Manipulation 13

5. Examples 14

5.1. Four DOF Spring Mass Example 14

5.2. Nastran Model of Metal Block 19

6. Appendix 21

6.1. Matlab Bulk Data Representation 21

Table of Figures

Figure 3.1. Coordinate system definition 6

Figure 3.2. Quadrilateral element example 6

Figure 4.1. GUI features 8

Figure 4.2. GUI file menu 9

Figure 4.3. Preferences dialog, Display tab 10

Figure 4.4. Preferences dialog, Colors tab 11

Figure 4.5. Color selection dialog 11

Figure 4.6. Preferences dialog, Animation tab 11

Figure 4.7. Preferences dialog, Misc tab 12

Figure 4.8. Data Selection Criteria dialog box for selecting a subset of the total data set 13

Figure 4.9. Shortcut animation controls (from Animation button) 14

Figure 5.1. Four DOF example problem 14

Figure 5.2. Model as shown in fePostProc 17

Figure 5.3. First mode shape (double-click to animate) 18

Figure 5.4. Second mode shape (double-click to animate) 18

Figure 5.5. Third mode shape (double-click to animate) 18

Figure 5.6. Fourth mode shape (double-click to animate) 19

Figure 5.7. First bending mode (double-click to animate) 20

Figure 5.8. First torsional mode (double-click to animate) 20

Introduction

The fePostProc software application is a Matlab graphical user interface (GUI) that enables structural finite element mode shapes or operational deflection shapes to be easily viewed from within the Matlab environment. Working from within the Matlab environment offers numerous advantages over other post processors, including

• the ability to quickly generate and visually evaluate analytical model variations

• offers the user virtually infinite control over the view settings (colors, colormaps, grid symbols, colorbar divisions/precision, post titles, font properties, deflection amplitudes, etc.)

• use as many views at once (with no licensing problems!) to facilitate model comparisons

The viewer has been developed to enable fully dynamic model rotation, zoom, and pan while animating. The file conversion utilities enable Nastran models and results files to be read directly into the Matlab environment. All results file readers (OP2, F06, and Punch) allow the user to select the data sets of interest by data set ID or by an upper and/or lower frequency range. Viewing complex- or real-valued mode shapes (or operating deflection shapes) cannot get much easier! Most importantly, the source code is available for fePostProc as well as all of the file translators!

Installation

The only requirement for installation is that the zipped files be extracted and placed into a folder which is a part of Matlab’s search path. The files can be placed into an existing folder within the search path, or a new folder can be added to the path list. Type “path” at the Matlab command prompt to view a listing of the current Matlab paths. Type “help addpath” at the command prompt to see how to add a path to Matlab’s search path. The next step is to execute the feInstall.exe program (double-click) and then read the license agreement. If the agreement is accepted, the files will be extracted and placed into the chosen folder. The functions are then executed by typing their names at the command prompt. For example, fePostProc is executed by typing “fePostProc” at the command line. To see the required input and available output for any of the functions, type “help functionName” at the command line.

Note: If the source code has been purchased, beware that errors have been known to arise on some platforms when compiling C mex functions with Matlab’s default object code optimization enabled. To disable the optimization, use the mex function “-g” option as follows: mex filename -g. Also note that any executables downloaded are platform dependent and may not execute on your platform.

Input/Output Format

The fePostProc software uses a single data structure to represent the finite element model. The data structure has the following fields (the second column lists the Nastran equivalent name):

|Matlab structure field |Nastran Bulk Data Card |Description |

|grids |grid |Grid point definition |

|cord1Cs |cord1c |Cylindrical coordinate system with grid references |

|cord1Rs |cord1r |Rectangular coordinate system with grid references |

|cord1Ss |cord1s |Spherical coordinate system with grid references |

|cord2Cs |cord2c |Cylindrical coordinate system with point references |

|cord2Rs |cord2r |Rectangular coordinate system with point references |

|cord2Ss |cord2s |Spherical coordinate system with point references |

|quads |cquad |Non-linear plane strain quadrilateral element |

|quad4s |cquad4 |Iso-parametric plane strain quadrilateral element |

|quad8s |cquad8 |Curved quadrilateral shell element |

|tria3s |ctria3 |Iso-parametric plane strain triangular element |

|tria6s |ctria6 |Curved triangular shell element |

|hexas |chexa |Six-sided solid (brick) element |

|pentas |cpenta |Five-sided solid element |

|tetras |ctetra |Four-sided solid element |

|bars |cbar |Bar element |

|beams |cbeam |Beam element |

Note that the quads, quad4s, and quad8s are all identical as far as elements definitions are concerned within fePostProc. Similarly, the tria3s and tria6s are identical, as are the bars and beams.

The grids have the following format:

|Grid ID |Definition Coordinate System ID |x |y |z |Output Coordinate System ID |

The grid ID is a unique integer value that assigns a label to a particular location in space. The location is defined by the x, y, and z values, which can reference any coordinate system as identified by the Definition Coordinate System ID. If the basic (Cartesian) coordinate system is to be used, the value of the Definition Coordinate System ID should be zero (0), and no coordinate card (i.e., cord1Rs or cord2Rs) are required. However, if any other coordinate system is to be used, then this coordinate system must be defined by the appropriate coordinate field of the data structure. The output coordinate system ID indicates the coordinate system used by the results sets. Again, if the basic (Cartesian) coordinate system is used by the results, then this value should be set to zero (0) and no coordinate system field need be defined.

The coordinate systems cord1Cs, cord1Rs, and cord1Ss are based on grid IDs and have the following format:

|ID |G1 |G2 |G3 |

The ID is a unique integer value that differentiates each of the coordinate systems. The general idea is to define the location of the coordinate system origin, specify a point along the z-axis, and also a point that lies within the x-z plane:

[pic]

Figure 3.1. Coordinate system definition

The coordinate systems cord2Cs, cord2Rs, and cord2Ss are based on three points and have the following format:

|ID |Base Coordinate ID |G1 |G2 |G3 |

The ID is a unique integer value that differentiates each of the coordinate systems. The base coordinate ID is the coordinate system used by each of the three defining points, G1, G2, and G3 (see Figure 3.1).

In all cases, the elements have the following general format:

|Element ID |Property ID |Grid 1 |Grid 2 |Grid 3 |… |

where the number of grids depends on the element type. The property ID is not used by fePostProc, and thus any arbitrary number may be used in this column. All entries for the elements are integers and simply define the grid connectivity by specifying the method in which the grids (referenced by their ID) are connected. For example, the two element, six-node structure shown in Figure 3.2 is defined as follows in matlab:

quad4s = [1,1,1,2,5,4; 2,1,2,3,6,5]; % An arbitrary property ID of 1 has been used

[pic]

Figure 3.2. Quadrilateral element example

Additional information can be found in the Nastran reference manuals. Note that additional fields are required for the Nastran structure variable to distinguish between integer and floating point numbers. This information is discussed in the Appendix (Section 6.1).

The results consist of three variables: disp, freq, and gridIDs. The disp variable contains the real- or complex-valued deflections for all grids, frequencies, and degrees of freedom. This variable is a three-dimensional array with the first dimension corresponding to the grids, second dimension corresponding to the frequencies, and the third dimension corresponding to the grid directions (or degree of freedom). So, for a model with 101 grids and 10 frequencies, the disp array will be of size (101x10x3). (Note that the disp array will always have a third dimension of length three, even for 1-D or 2-D models. The third dimension corresponds to the x-, y-, and z-directions, respectively. The array should be populated with zero entries where not used.) The freq variable is a vector containing the resonance (for mode shapes) or response (for operational deflection shapes) frequencies. This vector may be complex-valued, but must have a length that corresponds to dimension 2 of disp. The gridIDs variable is also a vector, but it must consist of real-valued numbers corresponding to the grid IDs in the finite element model.

GUI Description

The fePostProc GUI has several features as shown below in Figure 4.1. The post title displays information about the current data set being displayed, the coordinate axes (basic coordinate system) show the orientation of the model, the colorbar shows the range of the absolute value of the deflection, the status box relays any messages to the user, the animation controls facilitate control of the animation, the dataset list box allows the individual data sets to be selected and animated with a click of the mouse button, and the menubar houses controls for animations, preferences, loading the geometry and results, and so on.

[pic]

Figure 4.1. GUI features

1 File Menu

The components of the file menu are shown in Figure 4.2. This menu allows the entire database to be saved for future evaluation, controls the model and results input, allows the preferences to be modified and saved for future uses of fePostProc.[1] The save mode shapes and save animation functions save the individual mode shapes or an *.avi movie file. The number of frames used by these functions is determined by the number of animation frames specified in the Animation tab of the Preferences dialog (see Section 4.2).

[pic]

Figure 4.2. GUI file menu

2 Preferences

The preference dialog, executed from the file menu, controls the user options for all aspects of the GUI. Figure 4.3 shows the components controlled from the Display sheet. Grids, grid color, grid type, element edges, coordinate axes, colorbar, undeformed shape, etc. are selected in this sheet. There is also an option to display only the surface elements. When selected, only the grids and elements on the surface of a model are used in the animation, and thus large reductions in the required computer overhead can sometimes be obtained.

[pic]

Figure 4.3. Preferences dialog, Display tab

Elements of the Colors sheet of the Preferences dialog are shown below in Figure 4.4. As shown in this figure, the colors are modified by pressing the colored buttons with the mouse. When pressed, a color selection dialog, Figure 4.5, appears and the color can be modified as desired. Various colormaps are available for use; the drop-down box in the colors sheet chooses these. A user-defined colormap can also be used via the Matlab colormapeditor command, which is available in recent Matlab releases.

[pic]

Figure 4.4. Preferences dialog, Colors tab

[pic]

Figure 4.5. Color selection dialog

The Prefences sheet, Figure 4.6, houses the animation renderer, deflection amplitude (which is a percentage of total model size), animation speed, and number of animation frames can be selected. Also available is the use of a movie animation scheme, which sometimes allows a faster animation for very large models. Along with the movie method, the number of animation cycles to complete must be defined using this sheet, which only appears when the movie method is selected.

[pic]

Figure 4.6. Preferences dialog, Animation tab

The miscellaneous (Misc) sheet of the Preference dialog allows the specification of a scratch file directory and the data range prompt. The scratch file directory is the location where all intermediate output files are stored as well as the default fePostProc preference settings file (fePostProc_prefs.mat). If not specified, the scratch directory defined by the Matlab tempdir command is used.

The data range prompt, when selected, prompts the user to specify a subset of the analysis results to select each time results are read into fePostProc. When not selected, the entire data set is read into memory. This option facilitates the evaluation of large data sets. Additional information is provided in Section 4.3.

When the results are from a Nastran analysis and they involve a complex frequency, the Nastran Complex Frequency option should be checked. This option controls the method by which the frequencies are displayed in both the Post Title and the Mode List Box. If this box is checked and the frequencies are complex-valued, then the displayed frequency is the imaginary component of the frequency and the loss factor is computed as: -2*real(frequency)/abs(frequency). Otherwise, the displayed frequency is the real component of the frequency and the loss factor is computed as: 2*imag(frequency)/abs(frequency).

[pic]

Figure 4.7. Preferences dialog, Misc tab

3 Data Subset Selection

When turned on (see the Misc tab within the Preferences dialog), the user can specify a subset of the data set to read into memory through the Data Selection Criteria dialog, Figure 4.8, that automatically opens when the data loading process begins. The set can be chosen based on the data index or by a frequency range. The data indices can be specified using the Matlab format for specifying a vector (e.g., [1:3:100]), or referring to a vector variable in the Matlab workspace can specify them. A single number for the lower frequency limit and a single number for the upper frequency limit specify the frequency range. If all indices are to be read, then the user can simply press the OK button without entering any index or frequency limits.

[pic]

Figure 4.8. Data Selection Criteria dialog box for selecting a subset of the total data set

4 Model Manipulation

Once a model and results are loaded into fePostProc, the view can be modified by several means. First, the mouse buttons facilitate modifying the viewing angle (left mouse button), zooming (middle mouse button or consecutively the left and right buttons), and panning (right mouse button). Alternatively, the default view can be set by selecting Default from the View menu item (or by pressing CNTL-a on the keyboard). Additional preset views can also be selected from the same View menu item. These views can be modified using the mouse or the menu items at any time, even when an animation is taking place. The view can be refreshed/regenerated at any time by choosing Regenerate from the View menu (or by pressing g from the keyboard).

The animations are selected by choosing a mode from the Data List Box (see Figure 4.1) or by pressing the animate/stop, next (mode), and previous (mode) buttons, also shown in Figure 4.1. Shortcut keys have also been defined to view the modes: animate (a), stop animation (s), next mode (n), previous mode (p). The animation button is a shortcut to the most used animation parameters (see Figure 4.9) that are otherwise found in the preferences dialog. Two additional options are available from this button: 1) data animation type (real, imaginary, or complex), which is only available when the data is complex-valued, and 2) a radio button to force the program to use the rotational degrees of freedom as the complex component of the displacement. The utility of this feature is the ability to store complex data results in Nastran F06 files using the translational DOFs for the real component and the rotational DOFs for the imaginary component. This is a feature of fePostProc that will most likely be infrequently used by most analysts.

[pic]

Figure 4.9. Shortcut animation controls (from Animation button)

Note that all Matlab functions can be used to manipulate the view by typing set(0,’showHiddenHandles’,’on’) at the Matlab command line. This command makes all GUI handles visible to the Matlab workspace and the individual settings can be modified as needed (e.g., views can be set using the view command, lighting can be added and oriented by using the lightangle command, etc.)

Examples

1 Four DOF Spring Mass Example

The following example of a four degree-of-freedom spring mass system with complex stiffness coefficients demonstrates the usefulness of the fePostProc software for viewing Matlab results.

Consider the following system of four lumped masses and six springs, with the spring rates arbitrarily chosen to be complex-valued to capture system damping and cause the mode shapes and resonance frequencies to be complex-valued:

[pic]

Figure 5.1. Four DOF example problem

The mass and stiffness values are defined as follows:

|k1 = 1000+i200 N/m |m1 = 30 kg |

|k12 = 800+i50 N/m |m2 = 73 kg |

|k23 = 300+i100 N/m |m3 = 14 kg |

|k3 = 1200+i320 N/m |m4 = 29 kg |

|k24 = 1400+i50 N/m | |

|k4 = 750+i120 N/m | |

The equation of motion (EOM) is derived using Newton’s second law along with the free body diagrams of each of the mass components. The homogeneous EOM is shown below in matrix form:

[pic]

Or, in short hand notation:

[pic].

Assuming harmonic motion, the equation can be cast as follows:

[pic]

and solved using Matlab’s eig command to generate the eigenvalues and eigenvectors. These results can then be viewed in fePostProc. The entire procedure is completed by the following Matlab script:

[pic]

% Define the stiffness components (N/m)

k1 = 1000+i*200;

k12 = 800+i*50;

k23 = 300+i*100;

k3 = 1200+i*320;

k24 = 1400+i*50;

k4 = 750+i*120;

% Define the mass elements (kg)

m1 = 30;

m2 = 73;

m3 = 14;

m4 = 29;

% Assemble the matrices

K = [k1+k12,-k12,0,0;

-k12,k12+k23+k24,-k23,-k24;

0, -k23,k3+k23,0;

0,-k24,0,k4+k24];

M = [m1,0,0,0;

0,m2,0,0;

0,0,m3,0;

0,0,0,m4];

% Label the grids and their direction

gridIDs = [1;

2;

3;

4];

% Solve the system of equations for the eigenvectors and eigenvalues

[phi,om] = eig(K,M);

f = diag(om)/2/pi;

% Put the results in the format for the mode viewer

disp = zeros(length(gridIDs),length(f),3);

disp(:,:,1) = phi; % The results are all in the x-direction

% Create the geometry information

fe.grids = [1 0 1 0 0 0;

2 0 2 0 0 0;

3 0 3 .5 0 0;

4 0 3 -.5 0 0;

5 0 4 .5 0 0; % A grounded grid

6 0 4 -.5 0 0; % A grounded grid

7 0 0 0 0 0]; % A grounded grid

fe.bars = [1 1 7 1;

2 1 1 2;

3 1 2 3;

4 1 2 4;

5 1 3 5;

6 1 4 6];

The model and results are then transferred to the fePostProc program:

>> fePostProc

->File->Load Geometry -> Matlab Workspace (type fe in the input dialog box)

->File->Load Results -> Matlab Workspace (type disp in the first input dialog box, f in the second, and gridIDs in the third).

The model will then appear as shown below in Figure 5.2. (Note that the squares shown in this figure representative of the mass elements are actually the grids of the model, which can be chosen in the preference dialog: ->File->Preferences. To set up the identical view, type set(0,showHiddenHandles,’on’) and then type view(0,90).) The results for the four mode shapes are shown in succeeding four figures.

[pic]

Figure 5.2. Model as shown in fePostProc

[pic]

Figure 5.3. First mode shape (double-click to animate)

[pic]

Figure 5.4. Second mode shape (double-click to animate)

[pic]

Figure 5.5. Third mode shape (double-click to animate)

[pic]

Figure 5.6. Fourth mode shape (double-click to animate)

2 Nastran Model of Metal Block

The following example is for a Nastran model of a block having a constrained layer of viscoelastic material along its mid-plane. The model is read from the following text file (copy and paste this file to disk in order to read):

[pic]

The complex mode shapes, obtained from a Nastran Solution 107 run, are stored in the following Nastran OP2 file:

[pic]

The following two mode shape animations show two sample results:

[pic]

Figure 5.7. First bending mode (double-click to animate)

[pic]

Figure 5.8. First torsional mode (double-click to animate)

Appendix

1 Matlab Bulk Data Representation

Because Matlab programs generally do not distinguish between integer and floating point numbers, additional fields are required in the Nastran bulk data files in order to retain such information. The following paragraphs describe the convention used by the readBulk and writeBulk functions. Note that when generating a bulk data input file and default values are desired, BLANKS must be used for these default entries or errors may result.

The grids field of the data structure has the following columnar format:

ID |CP |X1 |X2 |X3 |CD |PS |SEID | |

ID ~ Grid ID: required

CP ~ Grid point definition coordinate system ID: if blank and not defined by a GRDSET card, a default value of 0 (basic coordinate system) is applied

X1,X2,X3 ~ Grid point location in coordinate system CP: required

CD ~ Output coordinate system ID (the coordinate system used by the displacements): if blank and not defined by a GRDSET card, a default value of 0 (basic coordinate system) is applied

PS ~ Permanent single point constraints: if blank and not defined by a GRDSET card, a default value of -2 is applied

SEID ~ Superelement ID: if blank and not defined by a GRDSET card, a default value of 0 is applied

Note that the shaded region above indicates information not required by fePostProc. Errors will result if the required fields are not specified in the data file. The default values listed are applied if the fields are not specified (i.e., if they are blank) in the file.

There are no auxiliary fields required for any of the coordinate systems: cord1Cs, cord1Rs, cord1Ss, cord2Cs, cord2Rs, or cord2Ss.

The quads field of the data structure has the following format:

EID |PID |G1 |G2 |G3 |G4 |G5 |G6 |G7 |G8 |G9 | |

EID ~ Element ID: required

PID ~ Property ID: required

G1,G2,G3,G4 ~ Corner grid point IDs: required

G5,G6,G7,G8 ~ Edge grid point IDs: default -1 if blank

G9 ~ Center grid point ID: default -1 if blank

Note that the shaded region above indicates information not required by fePostProc.

The psolid field of the data structure has the following format:

PID |MID |CORDM |IN |STRESS |ISOP |FCTN | |

PID ~ Property ID: required

MID ~ Material ID: required

CORDM ~ Material coordinate system ID: default = -20 if blank

IN ~ Integration network: default = -20 if blank, -1 if “BUBBLE”, numerical value otherwise

STRESS ~ Stress output location selection: default = -20 if blank, -1 if “GRID”, 1 if “GAUSS”

ISOP ~ Integration scheme: default = -20 if blank, -19 if “REDUCED”, numerical value otherwise

FCTN ~ Fluid element flag: default 1 if blank or “SMECH”, or -1 if “PFLUID”

Note that the shaded region above indicates information not required by fePostProc.

The quad4s field of the data structure has the following format:

EID |PID |G1 |G2 |G3 |G4 |TH |ZOFFS |T1 |T2 |T3 |T4 |XX | |

EID ~ Element ID: required

PID ~ Property ID: required

G1,G2,G3,G4 ~ Corner grid point IDs: required

TH ~ Material property orientation in degrees (float) or material coordinate system ID (integer): default -1 if blank (see field XX)

ZOFFS ~ Element offset: default -1 if blank

T1,T2,T3,T4 ~ Membrane thickness at grid points G1 through G4: default = -1 if blank

XX ~ Additional control field to indicate the type of value for TH: XX = 1.0 if TH is a float, XX = -1.0 if TH is an integer

Note that the shaded region above indicates information not required by fePostProc.

The quad8s field of the data structure has the following format:

EID |PID |G1 |G2 |G3 |G4 |G5 |G6 |G7 |G8 |T1 |T2 |T3 |T4 |TH |ZOFFS |XX | |

EID ~ Element ID: required

PID ~ Property ID: required

G1,G2,G3,G4 ~ Corner grid point IDs: required

G5,G6,G7,G8 ~ Edge grid point IDs: default 0 if blank

T1,T2,T3,T4 ~ Membrane thickness at grid points G1 through G4: default = -1 if blank

TH ~ Material property orientation in degrees (float) or material coordinate system ID (integer): default 0 if blank (see field XX)

ZOFFS ~ Element offset: default 0 if blank

XX ~ Additional control field to indicate the type of value for TH: XX = 1.0 if TH is a float, XX = -1.0 if TH is an integer

Note that the shaded region above indicates information not required by fePostProc.

The tria3s field of the data structure has the following format:

EID |PID |G1 |G2 |G3 |TH |ZOFFS |T1 |T2 |T3 |XX | |

EID ~ Element ID: required

PID ~ Property ID: required

G1,G2,G3 ~ Corner grid point IDs: required

TH ~ Material property orientation in degrees (float) or material coordinate system ID (integer): default 0 if blank (see field XX)

ZOFFS ~ Element offset: default 0 if blank

T1,T2,T3 ~ Membrane thickness at grid points G1 through G4: default = -1 if blank

XX ~ Additional control field to indicate the type of value for TH: XX = 1.0 if TH is a float, XX = -1.0 if TH is an integer

Note that the shaded region above indicates information not required by fePostProc.

The tria6s field of the data structure has the following format:

EID |PID |G1 |G2 |G3 |G4 |G5 |G6 |TH |ZOFFS |T1 |T2 |T3 |XX | |

EID ~ Element ID: required

PID ~ Property ID: required

G1,G2,G3 ~ Corner grid point IDs: required

G4,G5,G6 ~ Edge grid point IDs: default -1 if blank

TH ~ Material property orientation in degrees (float) or material coordinate system ID (integer): default 0 if blank (see field XX)

ZOFFS ~ Element offset: default 0 if blank

T1,T2,T3,T4 ~ Membrane thickness at grid points G1 through G4: default = -1 if blank

XX ~ Additional control field to indicate the type of value for TH: XX = 1.0 if TH is a float, XX = -1.0 if TH is an integer

Note that the shaded region above indicates information not required by fePostProc.

The pentas field of the data structure has the following format:

EID |PID |G1 |G2 |G3 |G4 |G5 |G6 |G7 |G8 |G9 |G10 |G11 |G12 |G13 |G14 |G15 | |

EID ~ Element ID: required

PID ~ Property ID: required

G1 through G6 ~ Corner grid point IDs: required

G7 through G15 ~ Edge grid point IDs: default 0 if blank

Note that the shaded region above indicates information not required by fePostProc.

The hexas field of the data structure has the following format:

EID |PID |G1 |G2 |G3 |G4 |G5 |G6 |G7 |G8 |G9 |G10 |G11 |G12 |G13 |G14 |G15 |G16 |G17 |G18 |G19 |G20 | |

EID ~ Element ID: required

PID ~ Property ID: required

G1 through G8 ~ Corner grid point IDs: required

G9 through G20 ~ Edge grid point IDs: default 0 if blank

Note that the shaded region above indicates information not required by fePostProc.

The tetras field of the data structure has the following format:

EID |PID |G1 |G2 |G3 |G4 |G5 |G6 |G7 |G8 |G9 |G10 | |

EID ~ Element ID: required

PID ~ Property ID: required

G1 through G4 ~ Corner grid point IDs: required

G5 through G10 ~ Edge grid point IDs: default 0 if blank

Note that the shaded region above indicates information not required by fePostProc.

The bars field of the data structure has the following format:

EID |PID |GA |GB |X1 |X2 |X3 |PA |PB |W1A |W2A |W3A |W1B |W2B |W3B | |

EID ~ Element ID: required

PID ~ Property ID: required

GA,GB ~ Connection grid point IDs: required

X1,X2,X3 ~ Orientation vector components: required unless alternate form is used (G0), in which case X1 = G0, X2 = -1, and X3 = -1

PA,PB ~ Pin flags for bar ends: default -1 if blank

W1A,W2A,W3A, W1B,W2B,W3B ~ Components of offset vectors: default -1 if blank

Note that the shaded region above indicates information not required by fePostProc. Also note that the BEAMOR card is not supported.

The beams field of the data structure has the following format:

EID |PID |GA |GB |X1 |X2 |X3 |BIT |PA |PB |W1A |W2A |W3A |W1B |W2B |W3B |SA |SB | |

EID ~ Element ID: required

PID ~ Property ID: required

GA,GB ~ Connection grid point IDs: required

X1,X2,X3 ~ Orientation vector components: required unless alternate form is used (G0), in which case X1 = G0, X2 = -1, and X3 = -1

BIT ~ Built in twist: default -1 if blank

PA,PB ~ Pin flags for bar ends: default -1 if blank

W1A,W2A,W3A, W1B,W2B,W3B ~ Components of offset vectors: default -1 if blank

SA,SB ~ Scalar or grid point IDs (for warping): default -1 if blank

Note that the shaded region above indicates information not required by fePostProc. Also note that the BEAMOR card is not supported.

The rbars field of the data structure has the following format:

EID |GA |GB |CNA |CNB |CMA |CMB | |

EID ~ Element ID: required

GA,GB ~ Connection grid point IDs: required

CNA,CNB ~ Independent degrees of freedom component numbers: required

CMA,CMB ~ Dependent degrees of freedom: default -1 if blank

Note that the shaded region above indicates information not required by fePostProc.

The elas1s field of the data structure has the following format:

EID |PID |G1 |C1 |G2 |C2 | |

EID ~ Element ID: required

PID ~ Property ID: required

G1,G2 ~ Geometric grid point IDs: required

C1,C2 ~ Component numbers: default -2 if blank

Note that the shaded region above indicates information not required by fePostProc.

The elas2s field of the data structure has the following format:

EID |K |G1 |C1 |G2 |C2 |GE |S | |

EID ~ Element ID: required

K ~ Spring stiffness: required

G1,G2 ~ Geometric grid point IDs: required

C1,C2 ~ Component numbers: default -2 if blank

GE ~ Damping coefficient: default –2 if blank

S ~ Stress coefficient: default –2 if blank

Note that the shaded region above indicates information not required by fePostProc.

-----------------------

[1] When the preferences are saved, a file named fePostProc_prefs.mat is saved in the scratch directory (see Section 4.2). This file can be deleted to reset the default preference settings.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download