13.10. Visualization Development Guide

You are welcome to contribute to the plotting/post-processing commands. This documentation is to explain what is going on inside the plotting library “Get_Rendering” and how you can enhance the functons. As of now, Get_Rendering has functions to plot a structure, mode shapes and shape of a displaced structure and works for 2D (beam-column elements, tri, and quad) and 3D (beam-column, tri, 4-node shell, and 8-node brick) elements.

As of now, all plotting functions should use Matplotlib only and should be able to produce interactive plots. Developers should test the new functions extensively , including on Jupyter Notebook, before submitting a pull request.

Note: A list of test examples will be available soon.

The source code of all the plotting functions is located in OpenSeesPy repository.

As an object oriented approach and to reduce the code repetition, Get_Rendering uses two types of functions.

13.10.1. Internal Database functions

These functions get, write to and read data from output database.

_getNodesandElements() : Gets node and element tags from the active model, stores node tags with coordinates and element tags with connected nodes in numpy arrays.

_saveNodesandElements() : Saves the node and element arrays from _getNodesandElements() to text files with “.out” extension.

_readNodesandElements() : Reads the node and element data into numpy arrays from the saved files.

_getModeShapeData() : Gets node deflection for a particular mode from the active model, stores node tags with modeshape data in numpy arrays.

_saveModeShapeData() : Saves the modeshape data arrays from _getModeShapeData() to text files with “.out” extension.

_readModeShapeData() : Reads the modeshape data into numpy arrays from the saved files.

_readNodeDispData() : Reads the node displacement data into numpy arrays from the saved files (from createODB() command).

_readFiberData2D() : Reads the section fiber output data into numpy arrays from the saved files (from saveFiberData2D() command).

13.10.2. Internal Plotting functions

These functions are helper functions that are called by the user functions once the updated node coordinates are calculated.

_plotBeam2D() : A procedure to plot a 2D beam-column (or any element with 2 nodes) using iNode, jNode and some internal variables as input.

_plotBeam3D() : A procedure to plot a 3D beam-column (or any element with 2 nodes) using iNode, jNode and some internal variables as input.

_plotTri2D() : A procedure to render a 2D, three node shell (Tri) element using iNode, jNode, kNode in counter-clockwise order and some internal variables as input.

_plotTri3D() : A procedure to render a 3D, three node shell (Tri) element using iNode, jNode, kNode in counter-clockwise order and some internal variables as input.

_plotQuad2D() : A procedure to render a 2D, four node shell (Quad, ShellDKGQ etc.) element using iNode, jNode, kNode, lNode in counter-clockwise and some internal variables as input.

_plotQuad3D() : A procedure to render a 3D, four node shell (Quad, ShellDKGQ etc.) element using iNode, jNode, kNode, lNode in counter-clockwise and some internal variables as input.

_plotCubeSurf() : This procedure is called by the plotCubeVol() command to render each surface in a cube using four corner nodes.

_plotCubeVol() : A procedure to render a 8-node brick element using a list of eight element nodes in bottom and top faces and in counter-clockwise order, and internal variables as input.

_plotEle_2D() : A procedure to plot any 2D element by calling other internal plotting commands for 2D elements.

_plotEle_3D() : A procedure to plot any 3D element by calling other internal plotting commands for 3D elements.

_initializeFig() : Initializes a matplotlib.pyplot figure for each of the user plotting commands. This procedure reduced the code repetition.

_setStandardViewport() : Sets a standard viewport for matplotlib.pyplot figure for each of the user plotting commands. This procedure reduced the code repetition.

13.10.3. User functions

These are the functions available to users to call through OpenSeesPy script. The following table describes what they are and how they work.

createODB() : Redords the model and loadcase data to be used by other plotting functions in a user defined output folder.

saveFiberData2D() : Redords the output data from all the fibers in a particular section to plot the distribution.

plot_model() : Gets the number of nodes and elements in lists by calling getNodeTags() and getEleTags(). Then plots the elements in a loop by checking if the model is 2D or 3D, and calling the nodecoord() command for each node to get its original coordinates and internal function _plotBeam3D.

plot_modeshapes() : Gets the number of nodes and elements in lists by calling getNodeTags() and getEleTags(). In a loop, calls nodecoord() and nodeEigenvector() for each node to get original and eigen coordinates respectively. Then plots the mode shape by calling the internal functions.

plot_deformedshape(): Reads the displacement data from the output of createODB() function and original coordinates using nodecoord() function. Then plots the displaced shape of the structure using the internal functions in a manner similar to plot_modeshapes().

plot_fiberResponse2D(): Reads the fiber output data from the output of saveFiberData2D() function and plots the distribution across the section.

animate_deformedshape(): Reads the displacement data from the output of createODB() function and original coordinates using nodecoord() function. Then animates the displaced shape of the structure.

animate_fiberResponse2D(): Reads the fiber output data from the output of saveFiberData2D() function and animates the stress/strain distribution across the section.

13.10.4. Example of an internal function

Here is an example of _plotQuad3D() internal function:

_plotQuad3D(iNode, jNode, kNode, lNode, ax, show_element_tags, element, eleStyle, fillSurface)

This function uses the following inputs:

*Nodes iNode, jNode, kNode, lNode : A list of four nodes in counter-clockwise order.
ax Reference to the Matplotlib fixure axes space. This should not be changed.
show_element_tags
The default is set to “yes” for plotting the model with plot_model() and “no” while plotting mode shapes
or deformed shapes. This should not be changed.
element This is the tag of that particular element as a string which is displayed when show_element_tags=”yes”.
eleStyle
Use eleStyle = “wire” for a wire frame, and “solid” for solid element lines. Wire frame is used when
overlapping the original shape of the structure with the mode shape or displaced shape.
fillSurface Use fillSurface = “yes” for color fill in the elements. fillSurface=”no” for wireframe.

13.10.5. Naming conventions

The names of classes, variables, and functions should be self-explanatory. Look for names that give useful information about the meaning of the variable or function. All the internal functions should start with “_” and be in camelCase (for example _plotCubeSurf) to distinguish them from the user functions. There are two type of user functions, 1) recorder functions and 2) plot functions. All the “recorder” functions should start with “record” and use camelCase (example: recordNodeDisp) and the plotting functions should start with “plot_” use snake_case (example: plot_deformedshape). Try to keep the internal variables (see 3. Example) consistant. New internal variable should be defined only if necessary.

13.10.6. Wish List

Some functions helpful to users might include, but not limited to,

  • plot_stress() : Record stress in all the elements of a function and plot. This will be useful in visualization of shear walls, fluid-structure interaction and soil modeled as brick elements.
  • plot_strain() : Similar functionality as above.
  • plot_sectionfiber() : Visualize stress-strain distribution accross a fiber section in a non-linear beam-column element.
  • plot_elementforces() : Element forces such as moment, shear, axial force.
  • animate_elementstress() : Animation of the stress or strain in a structure.

Feel free to comment and discuss how to streamline your plotting code with the existing Get_Rendering library. Or contact me Anurag Upadhyay.