Skip to content

Pack Primary Phases

Group (Subgroup)

Synthetic Building (Packing)

Description

If there are primary Phases Types in the volume, then this Filter will place primary Features with the sizes, shapes, physical orientations and locations corresponding to the goal statistics. All primary phase types will placed at once.

First, the Filter will determine the available volume for placing primary Features. This is accomplished by querying the Feature Ids array for the number of Cells not currently assigned to a valid Feature (Feature Id > 0). Then, the available volume is divided amongst the primary phase types according to their relative volume fractions. The size distribution of each primary phase type is sampled until the necessary volume of Features is generated. After each primary phase type has a list of Feature sizes from sampling the size distribution, the shapes, number of neighoring Features and physical orientations are sampled from distributions that are correlated to the size distribution for that primary phase type. At this point, the Features are fixed in their definition and are placed randomly in the volume. Once all Features, from all primary phase types, are placed, the packing is assessed on two criteria: 1. How well do the Features fill space (i.e .minimal overlaps and gaps) and 2. How well do the neighborhoods of Features match the neighbor statistics distributions. For a fixed number of iterations (100 * number of Features), the Features are moved and swapped while trying to optimize against the two criteria mentioned previously. If a move or swap improves the packing, it is accepted and if it does not it is rejected. During this process, the Features are not actually placed and are not filling space, but rather being represented analytically. Once the itrative process is finished, the Features are locked at their current location and they begin to grow from their centroid location according to their size, shape and orientation. The growth rates are defined such that the Features grow as the Shape Type they are (i.e. ellipsoid, superellipsoid, cube-octaheron, cylinder, etc), in the orientation they were placed and at a speed relative to their size. This growth continues until Features impinge and until all available Cells from the initial check are consumed.

The user can specify if they want periodic boundary conditions. If they choose periodic boundary conditions, when the Features are being placed and when they are growing, if a Feature attempts to extend past the boundary of the volume, it wraps to the opposing face and is placed on the opposite side of the volume.

The user can also specify if they want to write out the goal attributes of the generated Features. The Features, once packed, will not necessarily have the exact statistics (size, shape, orientation, number of neighbors) as sampled from the distributions. This is due to the use of non-space-filling objects in the packing process. The overlaps and gaps that occur after packing, must be assigned and will cause the Features to deviate from the intended goal (albeit hopefully in a minor way). Writing out the goal attributes allows the user to then calculate the actual attributes and compare to determine how well the packing algorithm is working for their Features.

The user can specify if they want to use a mask when building the volume. If the user chooses to use a mask, the they will have specify a boolean array that defines the volume that Features can be placed in (=true) and the volume that Features cannot grown past (=false).

The user can also choose to read in a list of Features with their locations and size and shape descriptions already determined. If this option is choosen, the Filter will skip the steps of generating the Features and iteratively placing them and will begin growing the Features defined in list. The format of the Feature Input File is:

Number of Features
Phase X Y Z A B C O3 Phi1 PHI Phi2
Phase X Y Z A B C O3 Phi1 PHI Phi2
Phase X Y Z A B C O3 Phi1 PHI Phi2
.
.
.
Phase X Y Z A B C O3 Phi1 PHI Phi2

where (X,Y,Z) are the coordinates of the Feature's centroid, (A,B,C) are the major, mid and minor principal semiaxis lengths of the Feature, O3 is the Omega 3 value of the Feature and (Phi1,PHI,Phi2) are the Euler angles that describe the Features's orientation in the sample reference frame.

For more information on synthetic building, visit the tutorial.

How to write out the Goal Attributes

In previous versions of DREAM.3D for this filter there was an option to "Write Goal Attributes". This has been replaced with the feature that allows the user to save the Shape Parameters to another Attribute Matrix. To regain this functionality the user can enable this option from a drop down combo box in the Filter Parameter input widget, then add the "Export Feature Data as CSV File" later in the pipeline. In that added filter, the user can select the Feature Attribute Matrix where they saved the Shape Parameters.

Parameters

Name Type Description
Periodic Boundaries bool Whether to wrap Features to create periodic boundary conditions
Use Mask Boolean Whether there is an array that defines where the Features can be placed and where they cannot grow past
Already Have Featrues bool Whether the user already has the final location and the size and shape definition of the Features and can skip the Feature generation and iterative placement process
Feature Input File File Path Path to the file that contains the description and location of the Features the user wishes to use (only necessary if Already Have Featrues is true)
Save Shape Description Arrays Int 0=Do not Save, 1=Save to New Attribute Matrix, 2=Append to existing AttributeMatrix
New AttributeMatrix DataArrayPath AttributeMatrix to save the Shape DescriptionArrays into

Required Geometry

Image

Required Objects

Kind Default Name Type Component Dimensions Description
Attribute Matrix CellData Cell N/A Cell Attribute Matrix in which to pack Features
Ensemble Attribute Array Statistics Statistics Object (1) Statistics objects (depending on Phase Type) that store fits to descriptors such as size distribution, shape distribution, neighbor distribution, ODF, MDF, etc.
Ensemble Attribute Array PhaseTypes uint32_t (1) Enumeration specifying the phase type of each Ensemble
Ensemble Attribute Array ShapeTypes uint32_t (1) Enumeration specifying the type of shape to place for each Ensemble

Created Objects

Kind Default Name Type Component Dimensions Description
Cell Attribute Array FeatureIds int32_t (1) Specifies to which Feature each Cell belongs
Cell Attribute Array Phases int32_t (1) Specifies to which Ensemble each Cell belongs
Attribute Matrix CellFeatureData Cell Feature N/A Feature Attribute Matrix for the created Feature Ids
Feature Attribute Array Phases int32_t (1) Specifies to which Ensemble each Feature belongs
Attribute Matrix CellEnsembleData Cell Ensemble N/A Ensemble Attribute Matrix for the created phases
Ensemble Attribute Array NumFeatures int32_t (1) Specifies the number of Features in each Ensemble

Shape Description Arrays

Kind Default Name Type Component Dimensions Description
Cell Attribute Array AxisEulerAngles float (3) Morphological Orientation in Bunge Convention (ZX'Z'')
Cell Attribute Array AxisLengths float (3) The Ellipsoid Aspect Ratios[a, a/b, c/a]
Cell Attribute Array Centroids float (3) The XYZ value for each ellipsoid
Cell Attribute Array EquivalentDiameters float (1) Equivalent Spherical Diameter
Cell Attribute Array Neighborhoods int32_t (1)
Cell Attribute Array Omega3s float (1) The Shape Parameter
Cell Attribute Array Volumes float (1) The Volume of the ellipsoid

Example Pipelines

  • (02) Single Hexagonal Phase Equiaxed
  • (03) Single Cubic Phase Rolled
  • InsertTransformationPhase
  • (01) Single Cubic Phase Equiaxed
  • (04) Two Phase Cubic Hexagonal Particles Equiaxed
  • (06) SmallIN100 Synthetic

Please see the description file distributed with this Plugin

DREAM.3D Mailing Lists

If you need more help with a Filter, please consider asking your question on the DREAM.3D Users Google group!