axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER > Class Template Reference

Preprocesses NURBSPatch geometry for GWN evaluation, and performs the calculation on the DOFs of an input MFEM mesh. More...

#include </home/docs/checkouts/readthedocs.org/user_builds/axom/checkouts/develop/src/axom/quest/GWNMethods.hpp>

Public Types
using	BoxType = axom::primal::BoundingBox< double, 3 >
using	GWNMoments = axom::quest::GWNMomentData< double, 3, ORDER >
using	PatchType = axom::primal::NURBSPatch< double, 3 >
using	PatchArrayType = axom::Array< PatchType >
using	NURBSCacheManager = typename axom::primal::nurbs_cache_3d_traits< ExecSpace >::type

Public Member Functions
	NURBSPatchGWNQuery ()=default
void	preprocess (const PatchArrayType &input_patches, bool use_direct_eval=false, bool use_memoization=true)
	Process input patches, optionally building a BVH. More...
void	query (mfem::DataCollection &dc, const primal::WindingTolerances &tol, const double slice_z=0.0)
	Evaluate the GWN for a query grid at the DOFs of the dc query mesh. More...
Setters for misc algorithm parameters
void	setSubdivisionBboxThreshold (double subdivision_bbox_threshold)
void	setSubdivisionMaxPasses (int subdivision_max_passes)
void	setSubdivisionMaxNumPatches (int subdivision_max_patches)

Detailed Description

template<typename ExecSpace, int ORDER = 2>
class axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >

Preprocesses NURBSPatch geometry for GWN evaluation, and performs the calculation on the DOFs of an input MFEM mesh.

Template Parameters

ExecSpace	The execution space for the algorithm.
ORDER	If agglomeration is used, this is the order of the Taylor expansion.

Possible evaluation modes are use_direct_eval : If true, evaluation is done patch-by-patch. If false, evaluation is sped up with agglomeration via Taylor-expansion use_memoization : Caches and re-uses trimming curve quadrature data for patch evaluation

Member Typedef Documentation

BoxType

template<typename ExecSpace , int ORDER = 2>

using axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >::BoxType = axom::primal::BoundingBox<double, 3>

GWNMoments

template<typename ExecSpace , int ORDER = 2>

using axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >::GWNMoments = axom::quest::GWNMomentData<double, 3, ORDER>

PatchType

template<typename ExecSpace , int ORDER = 2>

using axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >::PatchType = axom::primal::NURBSPatch<double, 3>

PatchArrayType

template<typename ExecSpace , int ORDER = 2>

using axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >::PatchArrayType = axom::Array<PatchType>

NURBSCacheManager

template<typename ExecSpace , int ORDER = 2>

using axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >::NURBSCacheManager = typename axom::primal::nurbs_cache_3d_traits<ExecSpace>::type

Constructor & Destructor Documentation

NURBSPatchGWNQuery()

template<typename ExecSpace , int ORDER = 2>

axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >::NURBSPatchGWNQuery ( )

default

Member Function Documentation

setSubdivisionBboxThreshold()

template<typename ExecSpace , int ORDER = 2>

void axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >::setSubdivisionBboxThreshold ( double  subdivision_bbox_threshold )

inline

setSubdivisionMaxPasses()

template<typename ExecSpace , int ORDER = 2>

void axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >::setSubdivisionMaxPasses ( int  subdivision_max_passes )

inline

setSubdivisionMaxNumPatches()

template<typename ExecSpace , int ORDER = 2>

void axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >::setSubdivisionMaxNumPatches ( int  subdivision_max_patches )

inline

preprocess()

template<typename ExecSpace , int ORDER = 2>

void axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >::preprocess ( const PatchArrayType &  input_patches, bool  use_direct_eval = false, bool  use_memoization = true  )

inline

Process input patches, optionally building a BVH.

Processing involves "cleaning" input surfaces for more robust GWN evaluation by

• Normalizing the parameter space of each surface according to the number of knot spans
• Ensuring each input is represented as a trimmed surface
• Clipping the parameter space of each surface to the visible (i.e. trimmed) portion

Parameters

[in]	input_patches	A view to the input trimmed NURBS surfaces
[in]	use_direct_eval	If false, use accelerated agglomeration algorithm via BVH
[in]	use_memoization	If true, allocate a per-thread cache for each patch

References AXOM_ANNOTATE_SCOPE, AXOM_LAMBDA, axom::spin::BVH< NDIMS, ExecSpace, FloatType, BVHImpl >::getTraverser(), axom::spin::BVH< NDIMS, ExecSpace, FloatType, BVHImpl >::initialize(), SLIC_WARNING, axom::quest::subdivide_patches(), and axom::Array< T, DIM, SPACE, StoragePolicy >::view().

query()

template<typename ExecSpace , int ORDER = 2>

void axom::quest::NURBSPatchGWNQuery< ExecSpace, ORDER >::query ( mfem::DataCollection &  dc, const primal::WindingTolerances &  tol, const double  slice_z = 0.0  )

inline

Evaluate the GWN for a query grid at the DOFs of the dc query mesh.

Parameters

[in]	dc	A query grid to be evaluated at the DOFs
[in]	tol	A collection of possible tolerances for GWN evaluation
[in]	slice_z	If the dc mesh is 2D, the GWN will be evaluated on a slice parallel to the x-y plane with this offset on the z-axis

References AXOM_ANNOTATE_SCOPE, axom::primal::WindingTolerances::disk_size, axom::primal::WindingTolerances::edge_tol, axom::primal::WindingTolerances::EPS, axom::quest::fast_approximate_winding_number(), axom::spin::BVH< NDIMS, ExecSpace, FloatType, BVHImpl >::getTraverser(), axom::spin::BVH< NDIMS, ExecSpace, FloatType, BVHImpl >::isInitialized(), axom::primal::WindingTolerances::ls_tol, axom::primal::WindingTolerances::quad_tol, SLIC_WARNING, and axom::primal::winding_number().

---

The documentation for this class was generated from the following file:

• /home/docs/checkouts/readthedocs.org/user_builds/axom/checkouts/develop/src/axom/quest/GWNMethods.hpp