User Guide

From TexGen
Revision as of 21:45, 11 December 2006 by WikiSysop (talk | contribs)
Jump to navigationJump to search

There are several ways to interface with TexGen. A little bit of technical information is necessary to understand how this all fits together. The core TexGen functionality has been written in C++, on top of this a Python interface exists that has access to all the C++ functions of TexGen. There are many advantages of a Python interface that may not be immediatly obvious to the end user.

A user friendly Graphical User Interface has been built on top of the Python interface to provide a quick and easy method to interface with TexGen. But the real power comes with writing small Python scripts which will be discussed later.

Graphical User Interface

The GUI is the easiest way to use TexGen, although a lot of TexGen features are not accessible via the graphical user interface. It is a good starting point to explore the capabilites of TexGen, but for the serious user the Python interface is recommended.

Layout

The GUI layout is similar to that of ABAQUS CAE for those familiar with it.

The left side of the screen contains the controls to interact with TexGen. These controls are identical to the ones shown in the drop down menu at the top. Use which ever is most convenient.

The right of the screen contains the 3d rendering of the model(s). Several models can be opened at once, all operations apply to the current model in view.

And finally at the bottom of the screen contains an interactive Python Console along with two log windows. The interactive console allows the user to input Python commands directly. This can be usefull for small operations and learning how to use the python scripting interface. For inputting a large number of Python commands it is more convenient to save it in a Python script and execute it all at once.

The Python Output log window displays all the Python code executed by the GUI. Essentially every button pressed in the user interface will create a Python command. This is done by design, making it possible to record Python scripts while interacting with the GUI. If you know how to do something with the GUI, you then automatically know how to do it with a Python script because the code is displayed.

The TexGen Output log window displays log and error messages from TexGen. These generally include information on what TexGen is doing, if something goes wrong it is usefull to be able read the log to find out what happened. Error messages will be displayed in red so that they can easily be located.

If new messages are posted one of the log windows while it is inactive, the icon above the window will change. In the case of receiving an error message the icon will change to an error icon.

Python Interface

Python is a high level interpreted programming language. In this case it is used as a scripting language to allow easy access to TexGen functionality. Python has been embedded in the TexGen GUI which allows the GUI to run python scripts. The goal of this tutorial is not to teach Python programming, there are plenty of resources online for that. A few of them are listed below:

An advanced knowledge of Python is not needed in order to write simple TexGen scripts. However, bear in mind that Python is a very powerful language with many libraries included for performing common tasks. I suggest you try to follow through the examples below and if you get stuck refer to one of the documents above.

Example 1

First of all, a Python script is just a plain ASCII text file with a .py extension. To create one simply open up your favorite text editor. Notepad will do if you have nothing else but I suggest a more powerfull freeware text editor like PSPad. Code that goes into the python script will be displayed like so:

# All Python code will be shown in a box like this one.
# Lines begining with a '#' character are comments

In this example, we will create a simple 2x2 plain weave model with a Python script. Its time to dive in and see some actual Python code. At this point if you are completely lost please refer back to the resources on Python programming. The first step is to create an instance of the CTextileWeave2D class and give it a name, weave:

# Create a 2D weave textile
weave = CTextileWeave2D(2, 2, 1, 0.2, True)

The CTextileWeave2D class takes 5 parameters in the order shown below:

  • Number of weft yarns in the unit cell
  • Number of warp yarns in the unit cell
  • Spacing between the yarns
  • Thickness of the fabrics
  • Refine model (True/False)

The only parameter that really needs an explanation here is the Refine model option, this controls whether or not the interference correction algorithm will be applied to keep the yarn volumes from intersecting which is a common problem. The interference correction algorithm generally does a good job but is not instantaneous so sometimes it can be beneficial to switch it off.

This is the bare minimum information necessary to create a 2D woven fabric. In the following steps we will add more information to create the model the way we want it.

Let's set the weave pattern, this determines whether a warp or weft yarn will appear on top at a particular cross over. Cross overs are arranged in 2d grid where the size of the grid depends on the number of warp and weft yarns created. To swap the positions of the weft and warp yarns, we call the function SwapPosition with the grid coordinates of the cross over. In order to create a plain weave we need to do this on two diagonally opposite cross overs like so:

# Set the weave pattern
weave.SwapPosition(0, 0)
weave.SwapPosition(1, 1)

Now let's add our weave model to the database of models so that TexGen can render it:

# Add the textile
AddTextile(weave)

Thats it! This is a very simple script but it will generate a plain weave model. To recap your Python script should look like this:

# Create a 2D weave textile
weave = CTextileWeave2D(2, 2, 1, 0.2, True)

# Set the weave pattern
weave.SwapPosition(0, 0)
weave.SwapPosition(1, 1)

# Add the textile
AddTextile(weave)

Save it to disk as PlainWeave.py or give it whatever name you want as long as it has a .py extension.

In order to test the Script we need to load it into TexGen. So start up the TexGen GUI and from the Python menu select Run Script. Select the file you have just saved and click ok. With a bit of luck you should see a 2d plain weave on screen.

Congratulations you have created your first TexGen Python script!

Example 2

In the first example we made use of the CTextileWeave2D which is the reason the script is so compact. This was great for creating a 2D Weave, but what if we want to create something other than a 2D Weave? In this example we will create yarns directly rather than through one of the textile sub-classes.

First we will define a general textile and call it Textile:

# Create a textile
Textile = CTextile()

The CTextile class doesn't take any parameters, instead yarns will be created and added to it. Next we will create 4 yarns to go inside our textile, it is convenient to make use of the Python lists to do this. Create a list called Yarns with 4 elements, each one an instance of the class CYarn:

# Create a python list containing 4 yarns
Yarns = [CYarn(), CYarn(), CYarn(), CYarn()]

Yarns now contains 4 empty yarns. Note that the yarns are not currently associated with the Textile, first we must fill in the details then make them a part of the textile. The first thing to do is define their paths with nodes. This is done like so:

# Add nodes to the yarns to describe their paths
Yarns[0].AddNode(CNode(XYZ(0, 0, 0)))
Yarns[0].AddNode(CNode(XYZ(0.22, 0, 0.05)))
Yarns[0].AddNode(CNode(XYZ(0.44, 0, 0)))

Yarns[1].AddNode(CNode(XYZ(0, 0.22, 0.05)))
Yarns[1].AddNode(CNode(XYZ(0.22, 0.22, 0)))
Yarns[1].AddNode(CNode(XYZ(0.44, 0.22, 0.05)))

Yarns[2].AddNode(CNode(XYZ(0, 0, 0.05)))
Yarns[2].AddNode(CNode(XYZ(0, 0.22, 0)))
Yarns[2].AddNode(CNode(XYZ(0, 0.44, 0.05)))

Yarns[3].AddNode(CNode(XYZ(0.22, 0, 0)))
Yarns[3].AddNode(CNode(XYZ(0.22, 0.22, 0.05)))
Yarns[3].AddNode(CNode(XYZ(0.22, 0.44, 0)))

We have added 3 nodes to each yarn creating a woven pattern. The number in the square brackets after Yarns refers to a particular yarn in the list. The function AddNode is then called for the yarn. The function takes an instance of the class CNode which in turn is constructed with an instance of the class XYZ representing the position of the node. The class XYZ is constructed with 3 numbers representing the X, Y and Z coordinates of the point.

We still need to define a few more things to obtain a model. For the sake of simplicity, we will give all the yarns the same cross sectional shape, width, height, etc... The easiest way to do this is to loop over the yarns and assign the same properties to each yarn one by one. The syntax for looping over Yarns is shown here:

# Loop over all the yarns in the list
for Yarn in Yarns:

Inside the loop, the variable Yarn refers to the current yarn in the list. So within the loop we just assign properties to Yarn and they will be applied to all the yarns in the list. In python the begining and end of a loop are defined by indentation, hence the following lines are indented to represent the contents of the loop.

Let us define periodic cubic spline interpolation between the nodes:

    # Set the interpolation function
    Yarn.AssignInterpolation(CInterpolationCubic())

This is done with the function AssignInterpolation which takes an instance of a class derived from CInterpolation. In this case we used CInterpolationCubic to specify periodic cubic spline interpolation.

Next we need to define a cross section for the yarn, to keep things simple a constant cross section all along the length of the yarn will be defined:

    # Assign a constant cross-section all along the yarn of elliptical shape
    Yarn.AssignSection(CYarnSectionConstant(CSectionEllipse(0.18, 0.04)))

The function AssignSection takes an instance of a class derived from CYarnSection. In this case we used CYarnSectionConstant which in turns take an instance of a class derived from CSection as a constructor. CSectionEllipse defines an elliptical cross section with given width and height defined by the first and second parameters respectively.

    # Set the resolution of the surface mesh created
    Yarn.SetResolution(20)

This line sets the resolution of the surface mesh created. The first parameter to the function SetResolution represents the number of points around a cross section. The number of points along the length of the yarns is automatically calculated such that the distance between points is similar to around the cross section.

Next we must add repeat vectors to define how the yarns are repeated.

    # Add repeat vectors to the yarn
    Yarn.AddRepeat(XYZ(0.44, 0, 0))
    Yarn.AddRepeat(XYZ(0, 0.44, 0))

The function AddRepeat takes an instance of the class XYZ. As we saw earlier the class XYZ is constructed with 3 numbers defining the X, Y and Z components.

Finally the yarn is fully defined, it can now be added to Textile:

    # Add the yarn to our textile
    Textile.AddYarn(Yarn)

This is the end of the loop, hence the following lines return to the original indentation. Now let's define a Domain within which the textile will be placed:

Textile.AssignDomain(CDomainPlanes(XYZ(0, 0, -0.02), XYZ(0.44, 0.44, 0.07)))

The function AssignDomain takes an instance of a class derived from CDomain. In this case we will create a box shaped domain, this can be accomplished by using the class CDomainPlanes. A box can be created by passing in opposite corners of the box (the minimum and maximum coordinates). This is done again with the class XYZ.

And finally Textile must be added to so that it can be rendered. Each Textile added must be assigned a name to identify it, this is the name that will appear at the top in the TexGen GUI. This name can either be generated automatically or we can specify it explicitly like so:

# Add the textile with the name "polyester"
AddTextile("polyester", Textile)

That's it! You now have the knowledge to create any type of fabric imaginable. Well, almost. The model created here is pretty simple, there is plenty of oppertunity to create more complex shapes. For example non-constant cross sections can be defined, complex domain shapes defined by arbritrary planes, different interpolation functions.

Here is the completed script:

# Create a textile
Textile = CTextile()

# Create a python list containing 4 yarns
Yarns = [CYarn(), CYarn(), CYarn(), CYarn()]

# Add nodes to the yarns to describe their paths
Yarns[0].AddNode(CNode(XYZ(0, 0, 0)))
Yarns[0].AddNode(CNode(XYZ(0.22, 0, 0.05)))
Yarns[0].AddNode(CNode(XYZ(0.44, 0, 0)))

Yarns[1].AddNode(CNode(XYZ(0, 0.22, 0.05)))
Yarns[1].AddNode(CNode(XYZ(0.22, 0.22, 0)))
Yarns[1].AddNode(CNode(XYZ(0.44, 0.22, 0.05)))

Yarns[2].AddNode(CNode(XYZ(0, 0, 0.05)))
Yarns[2].AddNode(CNode(XYZ(0, 0.22, 0)))
Yarns[2].AddNode(CNode(XYZ(0, 0.44, 0.05)))

Yarns[3].AddNode(CNode(XYZ(0.22, 0, 0)))
Yarns[3].AddNode(CNode(XYZ(0.22, 0.22, 0.05)))
Yarns[3].AddNode(CNode(XYZ(0.22, 0.44, 0)))

# Loop over all the yarns in the list
for Yarn in Yarns:

    # Set the interpolation function
    Yarn.AssignInterpolation(CInterpolationCubic())

    # Assign a constant cross-section all along the yarn of elliptical shape
    Yarn.AssignSection(CYarnSectionConstant(CSectionEllipse(0.18, 0.04)))

    # Set the resolution of the surface mesh created
    Yarn.SetResolution(20)

    # Add repeat vectors to the yarn
    Yarn.AddRepeat(XYZ(0.44, 0, 0))
    Yarn.AddRepeat(XYZ(0, 0.44, 0))

    # Add the yarn to our textile
    Textile.AddYarn(Yarn)

# Create a domain and assign it to the textile
Textile.AssignDomain(CDomainPlanes(XYZ(0, 0, -0.02), XYZ(0.44, 0.44, 0.07)))

# Add the textile with the name "polyester"
AddTextile("polyester", Textile)