ArtiSynth User Interface Guide

When the grid is enabled, a box labeled Grid: appears in the toolbar on top of the main ArtiSynth frame which gives the current resolution of the grid, displayed as S/N, where S is the size of each major grid cell and N is the number of subdivisions per cell. If there are no subdivisions, then the /N is omitted. For example, in Figure 3, this appears as Grid: 1/10, which means that the major grid cells have a size of 1.0 and are each divided into 10 subdivisions. The numeric value of the ratio S/N gives the minor cell size.

By default, the grid automatically resizes itself to the current viewer zoom level, choosing well-rounded numbers for the grid cell size. Auto-sizing can be enabled or disabled by right clicking on the Grid: label and choosing Turn auto-sizing on or Turn auto-sizing off, as appropriate. The user can also specify an explicit value for the grid resolution by entering the desired S/N value (or just an S value) into the Grid: box. Specifying an explicit value will disable auto-sizing, unless S is specified as 0 or the special value * is entered, both of which will re-enable auto-sizing.

Hitting the ‘l’ key within the viewer enables or disables labeling of the major divisions along the horizontal and vertical axis (Figure 4). The division lines along which these labels appear are automatically adjusted so as to ensure proper label visibility, and do not necessarily correspond to the x, y, or z axes.

It is possible to control various properties associated with axis labeling, such as which axes are labeled, and the label size and color. See the next section on Grid properties.

The grid has a number of properties that can be set by right clicking in the viewer and choosing Set viewer grid properties (or by right clicking on the Grid: label and choosing Set properties). This will bring up a property dialog, such as that shown in Figure 5.

Properties that can be set include:

resolution

Grid resolution, as described above.

autoSized

If true, causes the grid resolution to be recomputed as the user adjusts the view position, orientation, and zoom.

minCellPixels

Minimum number of pixels that should appear in a minor cell division when autosizing.

majorColor

Color to use for the major axis lines.

minorColor

Color to use for the minor axis lines.

xAxisColor

Color to use for the grid line that corresponds to the world y axis, or the horizontal axis if lockAxesToWorld is false.

yAxisColor

Color to use for the grid line that corresponds to the world y axis, or the vertical axis if lockAxesToWorld is false.

zAxisColor

Color to use for the grid line that corresponds to the world z axis.

lineWidth

Width of the grid lines, in pixels.

position

Translation position of the grid, in world coordinates.

orientation

Orientation of the grid, in world coordinates.

lockAxesToWorld

If true, forces the grid to stay aligned with the orientation and position of the world axes. In particular, the horizontal and vertical axes will always be parallel to one of the x, y, or z world axes, the grid center will be a multiple of major cell sizes from the origin, and axis labels will be set relative to the world origin.

useWorldOrigin

If true, causes the principal horizontal and vertical axes to be aligned with the world origin. Otherwise, the axes will be aligned with the grid center. This property can only be true if lockAxesToWorld is also true.

xAxisLabeling

Enables labeling of the x axis.

yAxisLabeling

Enables labeling of the y axis.

labelSize

’em’ size of the label text, in pixels.

labelColor

If set, specifies the color used to draw the label text. Otherwise, the major axis color is used.

To add a clipping plane, left click on the add clip plane button located on the viewer toolbar. This will create a clipping plane located in the plane perpendicular to the current line of sight.

It will also add to the viewer toolbar a clip plane icon for controlling the clipping plane. Right clicking on this icon will bring up an option menu.

To delete a clipping plane, right click on its icon and select Delete.

A clip plane is associated with a coordinate system and can be moved and/or rotated by dragging on the trans-rotate transformer located at its coordinate system origin. The clip region is the half space lying in the direction of the +z axis.

The transformer itself can be made invisible/visible by right clicking on the clip plane icon and selecting Hide transformer or Show transformer.

The clipping region is the half space lying in the direction of the +z axis of the plane’s local coordinate system. By default, clipping is actually offset by a small distance along the +z axis, so that small objects (such as points) lying in the x-y plane remain visible. The amount of this offset is controlled by the plane’s offset property, which is set to a nominal default value. To control this property directly, right click on the clip plane icon and select Set properties. This will bring up a panel which allows the offset to be adjusted.

Left clicking on the clip plane icon will enable/disable clipping. Disabling clipping allows the plane to be used as a regular movable grid. When clipping is disabled, the icon will change to the form .

Clipping planes can be placed in a slicing mode, whereby half-spaces in both the positive and negative z directions are clipped. The result is a small slice about the local x-y plane (Figure 6, right). The width of this slice is controlled by the plane’s offset property, as described above.

To enable or disable slicing, right click on the clip plane icon and select Enable slicing or Disable slicing.

Properties

Various properties associated with the plane, such as its color, line width, cell resolution, etc., can be set explicitly by the user. To do this, right click on the icon, select Set properties, and edit the resulting property panel. Most properties are the same as those described for the main viewer grid in 3.5.3.

Grid visibility

To make the grid invisible/visible, right click on the icon and select Hide grid or Show grid.

Alignment with world axes

The clip plane can be aligned so that it’s normal lies along the positive or negative direction of either the x, y, or z world axes. Right click on the icon and select the appropriate option. Clipping is performed so that the half-space lying in the direction of the normal is clipped.

Alignment with current line of sight

To align the clipping plane so that it is perpendicular to the current line of sight, right click on the icon and select Reset.

Model components may be assigned a string name; at the time of this writing names may not begin with a digit, have zero length, contain the characters ‘.’ or ‘/’, or equal the reserved word this. Components which do not have an assigned name are called nameless.

All components have a number, even if they do not have a name. The number is assigned automatically when the component is added to the parent, and is guaranteed to be persistent until the component is removed from the parent.

The names and/or numbers of a component’s ancestors can be used to form a component path name. This path has a construction completely analogous to Unix file path names, with the ‘/’ character acting as a separator. Absolute paths start with ‘/’ and begin with the root model. Relative paths omit the leading ‘/’ and can begin lower down in the hierarchy. The absolute path name of the red particle in Figure 7 would be

/Rigid Body Spring/models/msmod/particles/red

For nameless components in the path, their numbers can be used instead:

 /Rigid Body Spring/models/msmod/rigidBodies/1

Numbers can be used even for components that have names. Hence a path name consisting only of numbers, as in

 /0/0/0/3/1

is legal, although it most likely to appear only in machine-generated output.

In some cases, such as finite element models, the number of child components can be very large (on the order of thousands). In order to keep the navigation panel size manageable, the number of nameless children displayed is limited to a set number (currently 100). If the number of nameless children exceeds this number, the display will be augmented with an expand icon >>>. Clicking on this will expand the display to include all nameless components, and the expand icon will be replaced by a contract icon <<<. Clicking on the contract icon will cause the extra nameless components to be hidden again. This is illustrated in Figure 9.

Click and box selection are enabled by the arrow icon at the top of the selection toolbar:

Click selection involves left clicking on a component, causing it to be selected. Selection of multiple components is enabled by left clicking with a modifier key, which is usually CTRL but may be different for some legacy mouse bindings (Section 3.8).

Click selection selects only those components which are actually visible to the viewer; components which are hidden cannot be selected this way.

Box selection is effected by left-clicking and dragging in the viewer, causing the selection of all components rendered within the resulting drag box. Because this often results in the selection of more components than desired, it may be useful to employ a selection filter (Section 4.3.3). Any components within the drag box which are already selected will be deselected.

Box selection acts on all (filtered) components within the view frustum defined by the drag box, including those which are hidden from view.

Elliptic selection is enabled by the elliptic icon near the top of the selection toolbar:

This causes an additional elliptic cursor (which defaults to a circle) to be drawn around the mouse cursor. Selection is effected by dragging, and causes all visible objects within the ellipse to be selected. The selection process is cumulative, with subsequent drags selecting additional components. As with all selection operations, a filter can be set to restrict the components that are selected (Section 4.3.3). Generally, the drag select operation requires no modifier keys, although it may with some legacy mouse bindings (Section 3.8).

It is also possible to deselect components in the same way, by using the SHIFT modifier key to cause drag operations to cumulatively deselect components.

Elliptic selection selects only those components which are actually visible to the viewer; components which are hidden cannot be selected this way.

The elliptic cursor used for selection can be resized, either interactively, or by setting the ellipticCursorSize property of the viewer. To interactively change the cursor, initiate a drag operation with the CTRL and SHIFT modifiers. To set the ellipticCursorSize, invoke the context menu (usually right click) in the viewer when nothing is selected, and choose Set viewer properties. Finally, the ‘d’ key shortcut within the viewer will cause the cursor to be reset to its default size.

It is possible to limit viewer selection to components of a specific type. This can be done using the selection filter widget at the bottom left of the main ArtiSynth frame Figure 10.

To enable filtering, type into the widget text box the class name of the component type you wish to restrict filtering to. It is generally only necessary to enter the leaf name of the class (e.g., Particle or AxialSpring), and the system will then find the full class name by searching the ArtiSynth class path.

Once filtering is enabled, only components which are instances (including subclasses) of the specified type will be selectable.

Previously selected filters can be recalled using a history list accessible using the leftmost arrow button on the selection widget.

To remove selection filtering, enter the special filter *, either by typing this in the text box, or using the history list.

Under the default mouse bindings, pressing the SHIFT modifier key causes drag operations to be constrained to discrete step sizes. Rotation operations are constrained to intervals of five degrees, while translation operations are constrained to either the grid spacing (if a grid is selected, see 3.5), or to a suitable well-rounded number depending on the viewer’s zoom level.

Under the default mouse bindings, pressing the CTRL modifier key causes the dragger fixture to move independently of the selected objects. This allows its position and orientation relative to the selected objects to be changed. This is particularly useful for changing the orientation of the scaling directions in the scaling tool.

By default, a transformer is assigned a local coordinate frame for the object(s) that it is positioning, based on either the object’s own body frame (if it has one), or the objects’ bounding box. This frame will then move with the transformer, and may also move relative to the object(s) if the transformer is repositioned (Section 5.2.2).

Sometimes, it may be desirable to explicitly reset the transformer’s frame. This may be done using the following shortcut keys in the viewer:

w

Set the transformer frame to the world coordinate system, allowing subsequent transformations to be performed in world coordinates;

b

Reset the transformer frame to the original local frame for the object(s), based on either the object’s body frame or the objects’ bounding box.

Some properties are inheritable. The value of an inheritable property can be explicitly set or it can be inherited. If inherited, then it inherits its value from ancestor components further up the hierarchy. More specifically, if a property’s value is inherited, then the value is obtained from the nearest ancestor in which the same property exists and is explicitly set. If no such ancestor exists, then the property is set to a default value.

The inherited/explicit status of an inheritable property is controlled by an additional button placed at the left of the property widget (Figure 15). Clicking this button toggles the inherited/explicit status. If set to inherited, then the property’s value is determined from the hierarchy and the updated value is placed in the widget. Setting the value in the widget itself will cause the inheritable status to be set to explicit, and the value of inherited instances of the same property in descendant nodes will be updated accordingly.

There are a large number of render property settings. Loosely speaking, they are divided into generic settings, along with those related to faces, lines, and points. How these are used depends on what is being rendered. Mesh rendering typically uses the face settings, along with the line settings to render edges if the drawEdges property is set true. Line settings are also used for rendering axial springs and the edges of FEM elements. Point settings are used for rendering any subclass of Point, including Particle and FemNode.

Not all render properties may appear in a render panel; usually, only those properties relevant to the selected components will be presented.

Generic properties:

visible:

Whether or not the component is visible.

alpha:

The transparency for polygonal faces (range 0 to 1. Default is 1, for opaque).

shading:

How polygons are shaded (FLAT, SMOOTH, METAL and NONE). For viewer implementations there may be no difference between SMOOTH and METAL.

shininess:

Shininess parameter for polygons (range 0 to 128). Default is 32.

specular:

If not null, specifies the specular reflectance color.

Face related properties:

faceStyle:

Which polygonal faces are drawn (FRONT, BACK, FRONT_AND_BACK, NONE).

faceColor:

Color used for drawing faces.

backColor:

Color used for drawing backs of faces. If null, faceColor is used.

drawEdges:

If true, face edgesof the polygons are drawn explicitly.

Texture mapping properties:

colorMap:

If not null, specifies the image source file and properties for color mapping.

normalMap:

If not null, specifies the image source file and properties for normal mapping.

bumpMap:

If not null, specifies the image source file and properties for bump mapping.

Edge related properties:

edgeColor:

The color for edges.

edgeWidth:

Edge width in pixels.

Line related properties:

lineStyle:

How lines are drawn (CYLINDER, LINE, or SPINDLE).

lineColor:

The color for lines.

lineWidth:

Line width in pixels when LINE style is selected.

lineRadius:

Cylinder radius when CYLINDER or SPINDLE style is selected.

Point related properties:

pointStyle:

How points are drawn (SPHERE or POINT).

pointColor:

The color for points.

pointSize:

Point size in pixels when POINT style is selected.

pointRadius:

Sphere radius used when SPHERE style is selected.

A typical panel for editing render properties is shown in Figure 16. Texture mapping properties, if present, are normally hidden by default and can be exposed by clicking on the expand... button.

The play controls are in turn comprised of the following buttons:

The timeline proper is divided into a number of tracks. At the top is the waypoint track, which is used to arrange waypoints and breakpoints. Below that are a variable number of input and output tracks, which are used respectively for arranging input probes and output probes.

Waypoints save the model state when the simulation passes through them, allowing the timeline to be reset to that point (using the fast forward/backward buttons) at a later time without having to recompute the simulation from the beginning.

Waypoints are arranged along the waypoint track at the top of the timeline and are indicated by a small rectangular blue box (Figure 18). A solid box indicates a waypoint that contains a valid state and thus can be advanced to using the fast forward/backward buttons.

To add a waypoint, context click on the waypoint track. A popup menu will show a number of options, including Add waypoint here, which adds a waypoint at the current location of the time cursor, and Add waypoint(s), which will bring up a dialog prompting for a specific time to add a Waypoint. The "Add Waypoint" dialog also contains a repeat field, which will cause additional waypoints to be added with a spacing indicated by the time value, and an option to create breakpoints instead of waypoints.

Once created, waypoints can be moved by dragging them. They can also be deleted by context clicking on them and selecting the Delete waypoint option.

To delete all the waypoints, select Delete waypoints, either from the main File menu, or after context clicking on the waypoint track.

Breakpoints are waypoints that also cause the simulation to stop. They are displayed in red instead of blue.

Breakpoints can be added in the same way as waypoints, i.e., by right clicking on the waypoint track and selecting Add breakpoint here or Add waypoint(s).

Waypoints can be converted into breakpoints (and vice versa) through the context menu.

It is possible to save and load waypoints to and from an external file. The following menu options may be selected to do this, either from the main File menu, or after context clicking on the waypoint track:

Save waypoints as ...

Saves all waypoints and their state data to a given file, and sets this as the current waypoint file.

Save waypoints

Saves all waypoints and their state data to the current waypoint file.

Load waypoints ...

Loads waypoints and their state data from a given file, and sets this as the current waypoint file. The new waypoints are added to any existing ones, but previously waypoints are not deleted. Checks are made to help ensure that the waypoint data is consistent with the model’s current state.

Reload waypoints

Identical to Load waypoints ..., except that the current waypoint file is used.

Delete waypoints

Deletes all waypoints except the one at time 0.

Waypoint files are currently stored as binary files. The reason for this is that the required storage is about 1/2 of that required for text files, while the writing and parsing times are as much as faster.

New tracks may be added by context clicking in the waypoint track and selecting Add input track or Add output track.

The vertical location of a track can be moved by left clicking on it in the left panel and dragging it up or down to a new location.

A track can be deleted by context clicking on the track in the track panel and selecting Delete track.

A track can be muted or unmuted by clicking on its gray mute button in the track panel (Figure 19). All probes on a muted track are ignored during simulation.

All tracks can be muted, thus disabling all probes, by clicking on the mute all button in the gray panel above the tracks.

A track can be expanded or collapsed by clicking on its green expand button in the track panel (Figure 19). Expanding a track creates an additional area in which the data associated with the track’s probes may be displayed (see Figure 20). The way in which this data is displayed is probe-specific. Probes containing numeric data usually show their data graphically, as described in Section 7.4.

Multiple contiguous tracks can be selected by clicking on them while holding the control key. Furthermore, they can be grouped together or ungrouped by selecting the appropriate option from the context menu. Grouped tracks can be collapsed, moved simultaneously and muted together.

The range of the displayed data can be set by context clicking in the display and selecting Set Display Range . This pops up a dialog that lets the user either set the range explicitly, or enable automatic range computation by selecting the Auto range box.

The range of a display can also be fit to the current data set by context clicking in the display and selecting Fit Display Range.

As mentioned above, each entry in a numeric probe’s data vector is rendered in a different color (up to a limit). If the vector has a large dimension (say more than three), or if there is much overlapping of values, then the result can be difficult to visualize.

To manage this problem, numeric displays provide a legend tool that allows the user to control the color, drawing order, and visibility for each vector entry (Figure 21).

To create the legend tool, context click in the display panel and select Legend. Each row in the legend tool is associated with an entry in the data vector. The dimensions of the vectors are listed, followed by the color the entry is drawn in and a toggle controlling its visibility. Entries are rendered in bottom-to-top order, so those at the top will be most visible.

• •

  To change an entry’s color, click the Set button, which will bring up a color menu.

• •

  To make an entry visible or invisible, use the Visible toggle box.

• •

  To change the order in which an entry is drawn, click and drag the entry vertically within the panel.

Data for number input probes can be changed by adding, deleting, or moving knot points in the display.

• •

  To move a knot point, left click on the knot and drag it.

• •

  To add a knot point, place the mouse at the desired location and double left-click.

• •

  To delete a knot point, right click on the knot and select Delete Data Point.

• •

  To edit a knot point, right click on the knot and select Edit Data Point.

All knot points can be made invisible/visible by context clicking in the display and selecting (In)visible Data Points.

The data between knot points in a numeric probe is interpolated. Three types of interpolation are currently available: Step, linear, and cubic (Figure 22). For input probes, the interpolation can be set by right-clicking in the display and selecting the desired one.

A large display for a numeric probe can be created by context clicking on the probe icon and selecting Large Display. This will create a large numeric display in a separate panel, allowing more precise inspection of data (Figure 23).

In addition to the functionality of the smaller displays, large displays have additional buttons for controlling the domain and range of the displayed data. These are located on the left side of the display and include four mode control buttons:

On the top of the display are the fields for manually setting the displayed range and domain. There is once again the option to enable automatic range computation through the Auto range box.

It may be desirable to restrict rendering to view only the traces paths, perhaps along with a few selected model components. For instance, this might be necessary when producing graphs for a paper.

An easy way to do this is to edit the render properties for the desired points, and set the visible property to be explicitly true. Then, edit the render properties for the top-level model in the scene, and set its visible property to be explicitly false. This will make invisible all model components whose visible property is inherited. Since the traced points were explicitly set to be visible, they (and their traces) will remain visible. If the points themselves are rendered as spheres, they can be made to disappear by setting their pointStyle property to POINT, or by setting their pointRadius property to something very small.

Probes can be saved using the Save probes command from the File menu. This will save the current probe configuration in the probe directory (creating this directory if necessary). It will also save data for individual probes in their attached files, if such files are present.

A probe will not always have a file associated with it. Numeric probes will always have files associated with them; if a file is not named explicitly, a file name will be generated automatically when the probe is added to the system. This generated file name will be unique with respect to all probes currently configured and all files in the current probe directory.

To save probes in a different directory, one can use the command Save probes in from the File menu. This will prompt the user for a different directory, which will be created if necessary. The new directory will persist and the current probe directory.

Probes can be loaded form the current probe directory using the Load probes command in the File menu. The existing probe configuration will be cleared and the new configuration will be obtained from the file probeData.art, which must be present in the directory. Probes with attached files will be initialized from the data in those files.

To load probes from a different directory, one can use the command Load probes from in the File menu. This will prompt the user for a different directory, which must currently exist and which must contain the file probeData.art. The probe configuration will then be reset from the data in this new directory, which will persist as the new current probe directory.

Creating an output probe that simply records the value of a single numeric property is fairly easy. Starting with the output probe editor in Figure 24:

1. 1.

   Select a component, either externally through the navigation panel (Section 4.2), the viewer (Section 4.3), or the selection display (Section 4.4), or by entering its path name into the component/property widget. Once a component is selected, the left-most “up” arrow can be used to select that component’s parent.

2. 2.

   Select a property for the component from the property selection box at the right of the component/property widget.

3. 3.

   Adjust any required properties for the probe itself (Section 10.3).

4. 4.

   Click Done.

In general, output probes define a general map from the numeric values of n model component properties to a generalized output vector formed by the concatenation of m-subvectors formed by m formulae, as shown in Figure 25. In the simple case of Section 10.1.1, there is a single property, one sub-vector equal to the output vector, and a formula which is just the property value itself.

Each of the n properties has a numeric value which is represented by a variable and which is a vector of some dimension (scalars being considered vectors of dimension 1). The dimension of the depends on the property and is displayed to the right of the property selection box on the component/property widget.

Each of the m formulae is an arithmetic expression, implemented in Jython, which may involve one or more of the variables . The output from each formula is itself a vector of dimension , which is displayed at the right of each formula panel. The simplest formula is just a single variable , in which case equals the dimension of . The concatenation of all the output vectors from all the formulae produces the output vector associated with the probe, which has a dimension .

What the probe editor allows you to do is create the above mentioned map by selecting the properties of model components, assigning variable names to them, and creating formulae using these variables. When all the selected components form a coherent mapping, the Done button will be enabled and the probe can be completed. When one or more parts of the mapping is unspecified or incomplete, the associated widgets will be highlighted and the Done button will be disabled.

Extra properties can be added by requesting additional component/property widgets using the "+" button in the property panel. Similarly, extra formulae can be requested using the add button in the formula panel.

Properties and formulas can also be deleted; simply right click on the associated widget and choose delete.

Property variable names appear in a box at the left of the component/property widget. Variable names can be changed by editing this box. Name changes will be propagated into the formulae.

In order to streamline the probe creation process, the editor will try to guess certain desired actions. In particular, when the user chooses a property with a given component/property widget for the first time, the editor assigns a variable name to that property and creates a formula panel containing that variable. The variable name and formula panel can be changed if necessary.

Note:
The selection manager is connected to at most one component/property widget at a time. The component field for this widget is indicated with a blue border; external selections will affect only that widget. Left clicking on a component/property widget will cause it to be connected to the selection manager.

Creating an input probe that simply sets a single numeric property from the probe’s input vector is fairly easy, and is exactly analogous to creating a simple output probe (Section 10.1.1). Starting with the input probe editor in Figure 26:

1. 1.

   Select a component, either externally, or using the component/property widget.

2. 2.

   Select a property for the component from the property selection box at the right of the component/property widget.

3. 3.

   Adjust any required properties for the probe itself (Section 10.3).

4. 4.

   Click Done.

In general, input probes define a general map from the probe’s input vector (which is subdivided into n input variables of dimension ) to the values of m properties, where each value is determined by an independent formula based on the input variables (Figure 27). In the simple case of Section 10.2.1, there is one input variable which equals the input vector, and it drives a single property using a formula which is just the value of the input vector.

Each input variable is a vector of dimension (scalars being considered vectors of dimension 1), and the sum of all the equals the dimension of the input vector.

Each of the formulae controlling the properties is an arithmetic expression, implemented in Jython, which may involve one or more of the input variables . The output from each formula is itself a vector whose dimension must match the associated numeric property. The simplest formula is just a single variable , in which case its dimension equals the dimension of .

The probe editor allows you to create the above mentioned map by creating input variables, selecting properties, and creating formulae to drive these properties. When all the selected components form a coherent mapping, the Done button will be enabled and the probe can be completed. When one or more parts of the mapping is unspecified or incomplete, the associated widgets will be highlighted and the Done button will be disabled.

Extra properties can be added by requesting additional component/property widgets using the "+" button in the formula/property panel. Similarly, extra input variables can be requested using the add button in the formula panel.

Properties and input variables can also be deleted; simply right click on the associated widget and choose delete.

Each input variable is associated with a widget containing two text fields, the left one defining the variable’s name and the right one its dimension. The name or dimension can be changed by editing these fields. Changes will be propagated into the formulae; formulae that are found to be incompatible with the changes will be cleared.

In order to streamline the probe creation process, the editor will try to guess certain desired actions. In particular, when the user chooses a property with a given component/property widget for the first time, the editor creates an input variable whose dimension matches the property, and a simple formula consisting solely of the input variable. The input variable and formula can be changed if necessary.

Note:
The selection manager is connected to at most one component/property widget at a time. The component field for this widget is indicated with a blue border; external selections will affect only that widget. Left clicking on a component property widget will cause it to be connected to the selection manager.

A CompositeProperty is a Property which contains sub-properties. If a composite property is selected, the control panel will create a composite property widget, three of which are shown in Figure 31.

Clicking the set button of a composite property widget will create another panel presenting all the sub-properties. The composite property widget is disabled until this panel is closed.

In some instances, a composite property can be set to null. In such cases, the widget will provide an additional component, which

• •

  if the current property value is not null, will be a clear button which sets the value to null (second widget, <Figure 31).

• •

  if the current property value is null, will be a display label indicating this fact (third widget, <Figure 31).

It is possible to attach widgets to the sub-properties of a composite property, provided that the composite property has a non-null value.

In particular, when a non-null composite property is selected from the component/property widget, the user has the option of either

• •

  clicking the Done button and selecting the composite property, which will create a composite property widget as described in the previous section, or

• •

  selecting one of the composite property’s sub-properties.

When a non-null composite property is selected, the property’s name will move over into the component field of the component/property selector, and the combination box will be cleared and reset to allow the selection of the sub-properties.

For example, in Figure 32, we first select renderProps, which is a composite property of models/msmod/particle/2, and then (in the lower panel) select the sub-property visible. When renderProps is selected, its name is moved to the component panel, where it appears as

  models/msmod/particle/2:renderProps

Note:
The ‘:’ character is used to separate components from properties in component/property paths

A set of selected components can be deleted provided that

• •

  their parent components are editable

• •

  none of their ancestors are selected

If the currently selected components are deletable, then a delete option will appear in the context menu (obtained by right-clicking in the viewer or navigation panel). Selecting this will delete the components.

If the selected components are referred to by other components, then those components will be deleted also. In this case, a dialog will be presented to the user advising of this fact and requesting confirmation.

A set of selected components may be duplicated provided that

• •

  their parent components are editable

• •

  none of their ancestors are selected

• •

  they implement CopyableComponent

If the currently selected components are duplicatable, then a duplicate option will appear in the context menu. Selecting this will enable duplication of the components: the viewer cursor will change to cross-hairs, and the user may indicate the location for the duplicated components by left-clicking in the viewer (see Section 3.7). Duplication may be canceled by right-clicking.

Sometimes, when the components to be duplicated refer to other components, those referred components will be duplicated also. This is done when the referred components are required. For example, when duplicating an AxialSpring, the two points it is attached to will also be duplicated, because AxialSprings are not permitted to exist without attached points. Such cases are indicated to the user, after the duplicate option has been selected, by expanding the current selection to include all such additional components.

Many of the operations described here are undoable, by choosing the Undo option from the Edit menu. The menu option will indicate the name of the operation to be undone. Hitting the ‘z’ key from within the viewer (Section 3.9) will also perform undo operations.

To add a FemModel to a MechModel, select the MechModel and choose "Add FemModel ..." in the context menu. This will open the editing panel shown in Figure 35, which allows the user to provide information about the model’s properties and geometry.

Default values are provided for almost all of this information; the only information that must be specified by the user is the model’s position (corresponding to the origin of it’s volumetric mesh). This can be done either by left-clicking in the viewer (Section 3.7), or by entering coordinates in the position field of the Location subpanel. Once a position is specified, a wireframe preview of the FEM appears in the viewer (Figure 34), showing its geometry and allowing it to be moved or rotated using an attached transformer. The user is then free to continue editing the properties and geometry information, until the model is in the desired form, at which point it can be added to the MechModel by clicking the Add button.

From top to bottom, the FemModel panel contains:

• •

  An instruction box containing directions for the user.

• •

  A General Properties subpanel, which allows the user to set properties for the FemModel. For brevity, some of these properties are hidden and can be expanded by clicking the more... button.

• •

  A Location subpanel, allowing the position and orientation to be set manually. The position corresponds to the mesh origin, while the orientation is a rotation applied to the mesh, specified in axis-angle format (see Section 13.3).

• •

  A Geometry panel, allowing specification of the mesh geometry type and various properties specific to this type. Mesh types currently supported include Grid, Tube, Torus, Sphere, Extrusion, AnsysMesh, TetgenMesh and UCDMesh. For many of these, the associated element type can also be specified: Tet (tetrahedron), Hex (hexahedron), QuadTet (quadratic tetrahedron), QuadHex (quadratic hexadredron), and Wedge.

• •

  An option panel, containing the Add button, a Clear button which resets the displayed fields to default values, and a Cancel button which closes the panel without adding a FemModel.

To add a RigidBody to a MechModel, select the MechModel and choose "Add RigidBody ..." in the context menu to open an editing panel for rigid bodies, as shown in Figure 36.

As with adding FEM models, default values are provided for most information; the user must only specify the body’s position, either by left-clicking in the viewer (similar to Section 3.7), or by entering coordinates in the position field of the Location subpanel. Once a position is specified, a wireframe preview of the rigid body appears in the viewer (Figure 34), showing its geometry and allowing it to be moved or rotated using an attached transformer. The user is then free to continue editing the properties, geometry and inertia information, until the model is in the desired form, at which point it can be added to the MechModel by clicking the Add button.

From top to bottom, the Add RigidBody panel contains:

• •

  An instruction box containing directions for the user.

• •

  A General Properties subpanel, which allows the user to set properties for the body. For brevity, some of these properties are hidden; the panel can be expanded by clicking the more... button.

• •

  A Location subpanel, allowing the position and orientation of the body’s coordinate system to be set manually. Position is specified as a three-vector, while the orientation is given as a rotation in axis-angle format (see Section 13.3).

• •

  A Geometry And Inertia subpanel, which allows the user to specify the body’s surface mesh geometry and spatial inertia, using the same type of panel as described in Section 13.5.1.

• •

  An option panel, containing the Add button, a Clear button which resets the displayed fields to default values, and a Cancel button which closes the panel without adding a rigid body.

A FrameMarker is a massless Point attached to a RigidBody. It can be used for tracing motions of that body, or as an anchor point for attaching axial springs or other components.

To add one or more frame markers to the rigid bodies in a MechModel, you can select either the MechModel, or one of its rigid bodies, and then choose "Add FrameMarkers ..." in the context menu. This will open a FrameMarker editing panel, as shown in Figure 37. While this panel is open, frame markers can be added by using the viewer and left-clicking the mouse over the surface mesh of the rigid body at the location where you want the marker to be placed (see Section 3.7). The rigid body in question must belong to the MechModel that was originally selected; no marker will be added to bodies that belong to another MechModel or a MechModel which is a submodel of the current one.

From top to bottom, the FrameMarker editing panel contains

• •

  An Existing frame markers list, showing all the MechModel’s frame markers (expressed by their path names with respect to the MechModel). This list is connected to the selection manager and can be used to select one or more markers.

• •

  A name field that allows a name to be specified for the marker.

• •

  A Default marker properties panel, which allows the user to set properties for subsequent FrameMarkers that are added.

• •

  An instruction box containing directions for the user.

• •

  An option panel, which in this case contains a Done button which the user should click when finished.

A Particle is a dynamic component, with mass, derived from Point. It is usually connected to other components in a model with either axial springs (Section 13.4.5) or point-to-point attachments (Section 13.4.7).

To add one or more particles to a MechModel, select the MechModel in question and choose "Add Particles ..." in the context menu. This will open a Particle editing panel, as shown in Figure 38. While this panel is open, a particle can be added by left-clicking the mouse in the viewer at the location where you want the particle to be placed, using the constrain to plane option if necessary (see Section 3.7).

From top to bottom, the Particle editing panel contains

• •

  An Existing particles list, showing all the MechModel’s particles (expressed by their path names with respect to the MechModel). This list is connected to the selection manager and can be used to select one or more particles.

• •

  A name field that allows a name to be specified for the particle.

• •

  A Default particle properties panel, which allows the user to set properties for subsequent particles that are added. For brevity, some of these properties are hidden; the panel can be expanded by clicking the more... button.

• •

  An instruction box containing directions for the user.

• •

  A constrain to plane option.

• •

  An option panel, which in this case contains a Done button which the user should click when finished.

An AxialSpring is a point-to-point force effector that connects two Points and effects a force between them based on their separating distance. AxialSprings and its subclasses can be used to implement linear or nonlinear springs, as well as the subclass Muscle used to implement two-point muscles.

To add one or more axial springs to a MechModel, select the MechModel in question and choose "Add AxialSprings ..." in the context menu. This will open an AxialSpring editing panel, as shown in Figure 39. While this panel is open, axial springs can be added by selecting (using the viewer or any other selection mechanism) the two points to which the spring is attached. Points may include frame markers, particles, or FEM nodes. However, the points must be contained within the MechModel or one of its submodels.

By default, two points must be selected, in succession, for each axial spring added. Alternatively, by selecting add continuously at the bottom of the panel, a continuous sequence of springs will be created whereby the second point selected for a given spring becomes the first point for the spring following it.

From top to bottom, the AxialSpring editing panel contains

• •

  An Existing axial springs list, showing all the MechModel’s springs (expressed by the path names, with respect to the MechModel, of their points). This list is connected to the selection manager and can be used to select one or more springs.

• •

  A name field that allows a name to be specified for the spring.

• •

  A Spring type field that allows a specific subclass of AxialSpring to be selected.

• •

  A Default properties panel, which allows the user to set properties for subsequent springs that are added. The properties in question vary depending on the type selected in the Spring type field. For brevity, some properties may be hidden, in which case the panel can be expanded by clicking the more... button.

• •

  A progress field displaying the path names of the points as they are selected.

• •

  An instruction box containing directions for the user.

• •

  An add continuously option as described above.

• •

  An option panel, containing an Add/Stop button which is used to initiate or stop the addition of springs, and a Done button which the user should click when finished.

A BodyConnector is a component that implements constraint-based joints between either two rigid bodies, or between one rigid body and ground. The GUI currently allows two types of joints to be added: spherical and revolute.

To add one or more rigid body connectors to a MechModel, select the MechModel in question and choose "Add RigidBodyConnectors ..." in the context menu. This will open a RigidBodyConnector editing panel, as shown in Figure 40. While this panel is open, a connector can be added by selecting in succession (using the viewer or any other selection mechanism) the rigid bodies associated with it. For the case of a single rigid body connected to ground, the user clicks the Fixed button instead of selecting a second body.

After the bodies have been selected, the connector location must then be specified by left-clicking in the viewer (Section 3.7). By default, the orientation of the connector is aligned with the world axes. This can be adjusted later using the dragger fixtures (Section 5.1).

From top to bottom, the RigidBodyConnector editing panel contains

• •

  An Existing rigid body connectors list, showing all the MechModel’s connectors (expressed by the path names, with respect to the MechModel, of their rigid bodies). This list is connected to the selection manager and can be used to select one or more connectors.

• •

  A name field that allows a name to be specified for the connector.

• •

  A Connector type field that allows a specific connector type to be selected.

• •

  A Default properties panel, which allows the user to set properties for subsequent connectors that are added. The properties in question vary depending on the type selected in the Connector type field. For brevity, some properties may be hidden, in which case the panel can be expanded by clicking the more... button.

• •

  A progress field displaying the path names of the rigid bodies as they are selected.

• •

  An instruction box containing directions for the user.

• •

  An option panel, containing an Add/Stop button which can be used to initiate or stop the addition of connectors, a Fixed button used to indicate when a rigid body is to be connected to ground, and a Done button which the user should click when finished.

Within a MechModel, two particles or FEM nodes can be attached together, resulting in what is essentially a single particle that combines the dynamics of both original particles. In particular, these particle attachments are a convenient way to connect FEM models to other FEM models or to particles within a MechModel.

To attach particles contained within a MechModel, select the MechModel in question and choose "Attach particles ..." in the context menu. This will open a ParticleAttachment panel, as shown in Figure 41. While this panel is open, pairs of particles can be attached by selecting in succession (using the viewer or any other selection mechanism) the two particles to be connected.

From top to bottom, the ParticleAttachment panel contains

• •

  An Existing attachments list, showing all the MechModel’s attachments (expressed by the path names, with respect to the MechModel, of their particles). This list is connected to the selection manager and can be used to select one or more attachments.

• •

  A progress field displaying the path names of the particles as they are selected.

• •

  An instruction box containing directions for the user.

• •

  An option panel, containing an Attach/Stop button which can be used to initiate or stop the attachment process, and a Done button which the user should click when finished.

Within a MechModel, particles and FEM nodes can also be attached to rigid bodies. In particular, this provides a way to connect FEM models to rigid bodies within a MechModel.

To attach particles to a rigid body, select the rigid body in question and choose "Attach particles ..." in the context menu. This will open a ParticleRigidBodyAttachment panel, as shown in Figure 42. While this panel is open, particles can be attached to the body by selecting them in succession. By default, the attached particles remain where they are, so that the attachment point is determined by the current particle location relative to the rigid body’s coordinates. However, this will typically not coincide with the body’s surface mesh. By selecting project points onto body at the bottom of the panel, attached points will be relocated to the nearest location on the surface mesh as they are selected.

From top to bottom, the ParticleRigidBodyAttachment panel contains

• •

  An Existing attachments list, showing all the MechModel’s particle to rigid body attachments (expressed by the path names, with respect to the MechModel, of the particles and the bodies). This list is connected to the selection manager and can be used to select one or more attachments.

• •

  A progress field displaying the path names of the particles as they are selected.

• •

  A project points onto body field, described above.

• •

  An instruction box containing directions for the user.

• •

  An option panel, containing an Attach/Stop button which can be used to initiate or stop the attachment process, and a Done button which the user should click when finished.

In ArtiSynth, collision detection and handling can be enabled between rigid bodies (such as RigidBody), deformable bodies (such as FemModel3d), and more generally any body that implements the interface Collidable. Self intersection is not directly supported, but is indirectly supported for compound deformable bodies that contain sub-collidable components. For example, an FEM model is a compound collidable that may contain multiple surface meshes, some of which can be made to collide with each other. For more details on collision handling, see the “Collision Handling” section of the ArtiSynth Modeling Guide.

The collision response between any two pairs of bodies is determined by a CollisionBehavior component, which contains various properties controlling collision interactions. Two of these can be directly modified from the GUI:

enabled

whether or not collisions are enabled;

friction

the friction coefficient if collisions are enabled.

Collisions handling is managed by a CollisionManager component within each MechModel. Each MechModel provides four default behaviors that determine the default collision response for (a) rigid body pairs, (b) rigid-deformable body pairs, (c) deformable body pairs, and (d) deformable self-intersection. In addition to these, override collision behaviors can be specified for any pairs of bodies. In situations where a MechModel contains sub-MechModels, then the collision behavior for any pair of collidables is controlled by the lowest MechModel in the hierarchy that contains both.

There are several ways to edit collision behavior using the GUI.

• •

  For default behaviors, the enabled and friction properties can be edited by selecting the MechModel and then choosing "Set default collisions ..." from the context menu. This will open the dialog shown in Figure 43, allowing the enabled and friction properties to be adjusted. For the example shown, collisions are enabled between all rigid bodies, with a friction coefficient of 0.2, and between all rigid and deformable bodies, with a coefficient of 0.1. Other collisions are disabled.

• •

  If a user selects a particular set of rigid and/or deformable bodies, a specific collision behavior may be established among those bodies by choosing "Set collisions ..." from the context menu. That will open the dialog shown in Figure 44, allowing the enabled and friction properties for this behavior to be set.

• •

  If a user selects a single deformable body, a specific self-intersection behavior for that body may be established by selecting "Set self collision ...".

• •

  More detailed collision control can be realized by selecting the MechModel’s collision manager in the navigation panel (Section 4.2). Choosing Edit properties ... or Edit render props ... from the right context menu then allows other properties to be set to control either the collision behavior or the rendering of collisions. A sample property panel is shown in Figure 45, and these properties are described in the “Collision Handling” section of the ArtiSynth Modeling Guide.

• •

  For more fine-grained control, one may also use the navigation panel to select one or more of the behavior components located under the collision manager (see Figure 46). The first four of these control the default behaviors. Other behaviors, if any, are overrides that have been added either by application code, or through the GUI. Once selected, one can choose Edit properties ... or Edit render props ... from the right context menu to edit their properties. In the case of override behaviors, the context menu can also be used to remove them.

• •

  Finally, all override behaviors in a specific MechModel may be removed by selecting "Remove collision overrides". This will cause the collision behavior for all bodies to revert to default values.

Choosing "Edit geometry and inertia ..." in the context menu for a rigid body opens a geometry and inertia panel, as shown in Figure 47.

The upper part of the panel allows the user to set the mesh geometry, according to a type specified by the geometry type field. Changing the geometry type changes the panel to include fields for setting parameters appropriate to the type. Currently supported types include:

Box

An axis-aligned box, centered with respect to body coordinates, with the x, y, and z widths set by three numbers in a widths field.

Sphere

A sphere, centered with respect to body coordinates, with the radius and the number of vertical mesh slices given by fields radius and slices.

Mesh

A mesh read in from an Alias Wavefront .obj file, whose name is specified by a file name field. The read mesh can also be scaled, offset, and rotated using information provided by scale, offset, and rotation fields (see Section 13.3). The COM button causes the mesh to be offset so that its center of mass (assuming uniform density) is coincident with the origin of the body’s coordinate system (also causing the location of the center of mass to become zero).

Note:
At present, there appears to be a bug in the code that compute inertia from geometry, producing small errors in the center of mass calculation. That means that hitting the COM button will not cause the center of mass to become zero, but instead a small number that will converge to zero if COM is hit repeatedly.

The lower part of the panel sets the body’s spatial inertia. Spatial inertia for a rigid body can be set in three ways, corresponding to the value of the body’s inertiaMethod property:

Density

The spatial inertia is calculated from the density and the surface mesh geometry, with the assumption that the density is uniform.

Mass

The spatial inertia is calculated from the mass and the surface mesh geometry, with the assumption that the density is uniform. The density is computed by dividing the mass by the mesh volume.

Explicit

The spatial inertia is explicitly specified by entering values in the mass, inertia, and center of mass fields. The density is set to the average value obtained by dividing the mass by the mesh volume.

The inertia method can be set using the set inertia by field. Four other fields describe properties associated with the spatial inertia itself:

density

The mass divided by the volume

mass

The scale mass of the body

rotational inertia

The xx, yy, zz, xy, xz, and yz components of the rotational inertia tensor about the center of mass in body coordinates

center of mass

the position of the center of mass with respect to body coordinates.

Depending on the inertia method, the contents of these fields are either set by the user or updated automatically.

A FemMarker is a massless Point attached to a specific FemElement3d. It can be used for tracing motions within that element, or as an anchor point for attaching axial springs or other components. It is analogous to a FrameMarker for FEM elements.

To add one or more markers to an FEM model, you can select the FEM model in question and then choose "Add FemMarkers ..." in the context menu. This will open a FemMarker editing panel, as shown in Figure 48. While this panel is open, FEM markers can be added by left-clicking the mouse in the viewer over the location where you want the marker placed, using the constrain to plane option if necessary (see Section 3.7). An FEM marker is then created and attached to the nearest element in the FEM. If the marker position is outside the FEM, it is projected onto the closest point on the FEM surface.

In addition, the button Add Amira Landmarks allows a user to add a set of markers based on locations in an Amira landmark file.

From top to bottom, the FemMarker editing panel contains

• •

  An Existing markers list, showing all the FEM’s frame markers (expressed by their path names with respect to the FEM). This list is connected to the selection manager and can be used to select one or more markers.

• •

  A name field that allows a name to be specified for the marker.

• •

  A Default marker properties panel, which allows the user to set properties for subsequent markers that are added.

• •

  An instruction box containing directions for the user.

• •

  A constrain to plane option.

• •

  An option panel, containing an Add Amira Landmarks button, described above, and a Done button which the user should click when finished.

FemMuscleModel is a subclass of FemModel3d that supports muscle activation. A FemMuscleModel may contain muscle bundles (type MuscleBundle), each of which is composed of fibres and elements. The fibres are two-point muscles connecting nodes or markers within the FEM model, with activation provided by forces acting directly on the fibre end points. The elements are a set of references to existing elements within the FEM model, each combined with an activation direction. Each element reference within the bundle provides muscle activation behavior by superimposing a transversely isotropic material behavior on top of the underlying element’s material behavior.

Activation of a MuscleBundle can be effected by either the fibres or the elements, with the latter generally providing a superior simulation result. By default, the fibres are inactive, and are used simply to provide a good visual indication of the activation directions within the model, and a way to automatically compute the referenced elements and their directions (as described below). To make the fibres active, set the bundle’s fibresActive property to true. Conversely, to make the elements inactive, set the bundle’s muscleMaterial property to InactiveMuscle.

To add a MuscleBundle to a FemMuscleModel, select the model and then choose "Add MuscleBundle ..." from the context menu. This will immediately add a MuscleBundle to the model, and then open a MuscleFibre editing panel (see Section 13.7.1) to allow the user to add fibres to the model. The panel also contains two extra fields at the top: bundle name, allowing a name to be specified for the bundle, and bundle renderProps, allowing its render properties to be adjusted. At present, the panel does not contain a Cancel option. To remove the MuscleBundle, either select and delete it, or choose "Undo add MuscleBundle" from the Edit menu.

Fibres can be added to a MuscleBundle by selecting the bundle and then choosing "Edit fibres ..." from the context menu. This will open a MuscleFibre editing panel as shown in Figure 49.

The operation of this panel is essentially identical to the AxialSpring editing panel described in Section 13.4.5: fibres are added by successively selecting the points (which in this case must be FEM nodes or markers) which serve as the endpoints for the fibres in question.

The only difference from the AxialSpring panel is that the spring type is assumed to be a Muscle and there is no option to change this.

Elements can be added to a MuscleBundle by selecting the bundle and then choosing "Edit elements ..." from the context menu. This will open a MuscleElement editing panel as shown in Figure 50.

The operation of this panel is quite simple: one selects the elements that one wishes to add, and then clicks on the Add button to add them to the bundle. Elements which are already contained in the bundle will be excluded. Viewer-based element selection is described in more detail below.

From top to bottom, the MuscleElement editing panel contains

• •

  A list of element references already associated with the bundle (expressed by the elements’ path names with respect to the FEM muscle model). To remove element references from the bundle, one may select them in this list and then choose "Delete" from the context menu. It should be noted that this deletes the references for the elements within the bundle, and not the elements themselves from the FEM model.

• •

  Fields modelElementSize and bundleElementSize which control the size of the element widgets which are rendered for both the FEM muscle model and the bundle, as described below.

• •

  An instruction box containing directions for the user.

• •

  An option panel, containing an Add button which adds selected elements to the bundle, and a Done button which the user should click when finished.

Element selection is often done by clicking on an element widget in the viewer. An element widget is a simplified solid rendering of an element’s shape, with a size that varies from 0 to 1, with 0 being invisible and 1 being the full size of the element. Element widgets can be rendered for all the elements in an FEM model, with a size controlled by the model’s elementWidgetSize property. In addition, separate widgets can be rendered for the all the elements referenced by a muscle bundle, with a size controlled by the bundle’s elementWidgetSize property. In order to be able to see and select both the referenced elements in a bundle, and the other elements in the FEM model, one should set elementWidgetSize for the bundle and the model to values greater than zero, with the former larger than the latter. Figure 51 shows a simple example where referenced elements in a bundle are rendered using a widget size of 0.6, while the model elements themselves are rendered using a widget size of 0.5.

To facilitate element selection and visualization, the MuscleElement panel temporarily sets elementWidgetSize to 0.6 for the bundle and 0.5 for the FEM model. These values can then be adjusted as needed.

Since manually selecting elements and specifying their directions for a muscle bundle can be quite tedious, a number of methods exist to help do this automatically, using the easier-to-visualize information supplied by the muscle fibres. From the MuscleBundle context menu, one may select:

Compute element directions

Automatically computes directions for all referenced elements, using a Delaunay-based interpolation of the directions of the fibres which are closest to them.

Add elements neat fibres ...

Automatically adds to the set of referenced elements all elements whose centers are within a prescribed distance of one or more of the fibres.

Delete elements

Deletes all the element references for the bundle.

To remove specific fibres or element references, simply select them (using any of the selection mechanisms), and the choose "Delete" from the context menu.

The root element encapsulates the entire menu description. It must make reference to the schema for validation purposes. The following code snippet can be used as a template for creating a new menu file.

Note:
The second value of schemaLocation is the location of the schema file, which may need to be modified depending on its path relative to the current XML file.

Individual models are inserted into the menu using a model element. These specify a class to load when the menu entry is selected. The model must be a sub-class of RootModel.

The following snippet creates a menu entry titled "Muscle Arm", which will launch the model FemMuscleArm.

A separator is a horizontal line that visually separates menu entries. It can help distinguish groups of related menu items. The separator element has no attributes. The following snippet adds a separator line between models found in the artisynth.models.femdemos package and those in artisynth.models.inversedemos.

Labels are inactive text entries; they cannot be selected in the menu. They can be used to label a group of entries. To help distinguish a label from an active menu item, it is recommended to set the font style to "italic".

In the following code snippet, a label is added before the tongue tracking models.

Menu hierarchies can be created using the menu element. Apart from the root element, this is the only non-empty element type. It can contain any number of valid elements apart from the root element.

The following code snippet creates a sub-menu "FEM Models" with a specified icon, and lists all models found within the artisynth.models.femdemos package.

The package element can be used to include all models belonging to a particular Java package and its sub-packages. A base class can also be specified so that only models that are instances of that base class are added. For example to include all models that are sub-classes of HexTongueDemo, the following line can be used:

The models can be displayed in either a flat or hierarchical structure. To reduce the number of sub-menus, and to shorten some of the displayed text, a "compactness" level is introduced. Illustrative examples for the options are provided in Figure 54.

If font information is provided, it is applied to all entries.

The previous listing adds every single instance of RootModel found in the source tree to the menu. To get the menu to fit on the screen, compact="2" was used to strip away all package information from the displayed text, and a tiny font was applied to all entries.

Note:
The package element should be used sparingly for two main reasons:

1. 1.

   Each element adds to the ArtiSynth start-up time.

2. 2.

   Not all models are functional.

The reason for the start-up delay is that to populate the menu, the parser has to search through the supplied package, create an instance of each class, and test if it is a sub-class of base. The parser also has no way of detecting whether or not a model is functional, so all classes matching the criteria are added.

The XML format supports loading models from a demosFile that uses the original plain text format. Plain text files have the advantage of being more human-readable, and are easier to edit/comment lines out. Models are listed in the current sub-menu.

The following line loads all models from the original default ArtiSynth Models menu.

If font information is provided, it is applied to all model entries that are created.

If model history tracking is enabled, then a set of recent items can be added to the menu. To enable history tracking, pass the following command-line parameter to ArtiSynth on load:
      -historyFile <filename>
The history element then specifies the number of most recently loaded models to add to the history menu.

To facilitate modularity, the include element can be used to import all menu content from another XML menu file. All elements contained within the root element of the included file are inserted into the current sub-menu.

The following code imports the contents from two different XML menu files, and separates them with a separator item.

It is often convenient to have the ability to "comment-out" lines in any kind of coding system: the data remains in the file, but has no effect when processed. The typical method for commenting in XML is to use the <!-- --> tags. Unfortunately, this can be messy, and comments cannot contain other comments. For this reason, a special hidden element was created. Any entries within a hidden element are ignored by the parser.