4. Shape Function Orders and Additional Features#
Specifies finite-element shape function orders and additional features.
Input settings#
4_Shape_Function_Order_Added_Features JSON key: shape function orders and additional features object
NODE_ORDER(4) Nodal shape function order
EDGE_ORDER(4) Edge shape function order
METRIC_MOD(4) Eliminates magnetic-field issues near the center axis in axisymmetric two-dimensional analysis (GEOMETRY(2) = 2)
QUAD_TRI(4) Joins quadrilateral element faces (hexahedra) and triangular element faces (tetrahedra)
CALC_IND(4) Calculates COIL(17.1) inductance internally
THIN_ELEMENT(4) Improves convergence for thin or elongated elements
PARALLEL_NO(4) Sets the degree of parallel calculation
PARALLEL_OPTION(4) Parallelization option for element-matrix creation
Input format#
Text-format section
- Type:
Fixed parameter set
- Number of rows:
1
- Number of parameters:
8
Text format#* NODE_ORDER * EDGE_ORDER * METRIC_MOD * QUAD_TRI * CALC_IND * THIN_ELEMENT * PARALLEL_NO * PARALLEL_OPTION * I I I I I I I I
JSON format#
"4_Shape_Function_Order_Added_Features": {
"NODE_ORDER": I,
"EDGE_ORDER": I,
"METRIC_MOD": I,
"QUAD_TRI": I,
"CALC_IND": I,
"THIN_ELEM": I,
"PARALLEL_NO": I,
"PARALLEL_OPTION": I
},
Important
In JSON format, default values are used when these parameters are not defined. See the detailed descriptions for default values.
Example: two-way parallel calculation#
Example settings for two-way parallel calculation
Text-format section
Text format#* NODE_ORDER * EDGE_ORDER * METRIC_MOD * QUAD_TRI * CALC_IND * THIN_ELEMENT * PARALLEL_NO * PARALLEL_OPTION * 1 1 0 0 0 0 2 0
"4_Shape_Function_Order_Added_Features": {
"NODE_ORDER": 1,
"EDGE_ORDER": 1,
"METRIC_MOD": 0,
"QUAD_TRI": 0,
"CALC_IND": 0,
"THIN_ELEM": 0,
"PARALLEL_NO": 2,
"PARALLEL_OPTION": 0
},
Detailed description#
- 4_Shape_Function_Order_Added_Features#
- JSON key:
Shape function orders and additional features object
NODE_ORDER#
- NODE_ORDER(4)#
Text-format section
Fixed parameter set 1#[row, col] = [1, 1]
Setting value |
Description |
|---|---|
1 |
Uses first-order elements. Default |
2 |
First-order elements in conductor regions of the mesh data are converted to second-order nodal elements. When edge shape functions also use second-order elements, the input mesh data must use second-order elements. |
Note
The default provides sufficient accuracy. Higher order slightly improves accuracy but increases computational cost.
EDGE_ORDER#
- EDGE_ORDER(4)#
Text-format section
Fixed parameter set 1#[row, col] = [1, 2]
Setting value |
Description |
|---|---|
1 |
Uses first-order elements. Default |
2 |
Second-order edge elements are used for hexahedra and tetrahedra. In this case, |
Note
The default provides sufficient accuracy. Higher order slightly improves accuracy but increases computational cost.
Warning
Gap elements, surface elements, and sliding motion do not support second-order elements. When these are used together, set NODE_ORDER(4) = EDGE_ORDER(4) = 1. Second-order elements cannot be used for two-dimensional mesh expansion (GEOMETRY(12) ≠ 0). Second-order nodal elements improve accuracy for eddy-current analysis with tetrahedral, triangular-prism, and pyramid elements.
METRIC_MOD#
- METRIC_MOD(4)#
Text-format section
Fixed parameter set 1#[row, col] = [1, 3]
Eliminates magnetic-field issues near the center axis in axisymmetric two-dimensional analysis (GEOMETRY(12) = 2). Except under specific conditions, this parameter is internally set to its default regardless of the user setting, so it can normally be 0. The default is 1 for axisymmetric two-dimensional analysis; -1 runs that analysis without METRIC_MOD for verification. The default is 0 for other analyses.
Setting value |
Description |
|---|---|
0 |
Does not apply |
1 |
Applies |
-1 |
Runs axisymmetric two-dimensional analysis without |
QUAD_TRI#
- QUAD_TRI(4)#
Text-format section
Fixed parameter set 1#[row, col] = [1, 4]
Joins quadrilateral element faces of hexahedra and triangular element faces of tetrahedra. Nodes must be shared at the interface, and the quadrilateral face must be divided into two triangular faces.
Setting value |
Description |
|---|---|
1 |
Joins quadrilateral and triangular element faces |
0 |
Set to 0 for an ordinary conforming mesh, which has no interfaces between quadrilateral and triangular element faces. If such an interface exists while set to 0, it is treated as a boundary surface. Default |
Caution
Searching for interfaces consumes computation time even when none exist. Set to 0 when the input mesh does not mix tetrahedra and hexahedra.
CALC_IND#
- CALC_IND(4)#
Text-format section
Fixed parameter set 1#[row, col] = [1, 5]
Calculates COIL(17.1) inductance internally. Input for 17. COIL integration element MESH(10.3) is required.
Setting value |
Description |
|---|---|
0 |
Does nothing. Default |
1 |
Calculates COIL inductance internally |
Warning
Set to zero when internal inductance, formerly entered through CIRCUIT(17.8) or NETWORK(17.9), is defined as external inductance.
THIN_ELEMENT#
- THIN_ELEMENT(4)#
Text-format section
Fixed parameter set 1#[row, col] = [1, 6]
Improves convergence for thin or elongated elements. Input data is required for 16. THIN_ELEMENT applicable element properties THIN_CRITERION(16). It is particularly effective when thin or elongated elements occur in air regions, but slightly increases memory use.
Details
- title:
Collapsed text-format section
Setting value |
Description |
|---|---|
0 |
Does nothing. Default |
1 |
Improves convergence for thin or elongated elements |
Tip
Because THIN_CRITERION(16) is effective in regions where magnetic flux density changes little within an element, it is less effective at conductor or magnetic-material surfaces where skin effect occurs.
PARALLEL_NO#
- PARALLEL_NO(4)#
Text-format section
Fixed parameter set 1#[row, col] = [1, 7]
Degree of parallelism (default: 1)
Setting value |
Description |
|---|---|
1 |
Does not parallelize (conventional sequential calculation). Default |
>1 |
Runs parallel calculation with the specified degree. If it exceeds the computer’s maximum parallelism (the number of installed cores per node), calculation uses the maximum instead. |
Caution
This is OpenMP parallel calculation and cannot use computers across different nodes or separate machines. The <PARALLEL module> is required.
PARALLEL_OPTION#
- PARALLEL_OPTION(4)#
Text-format section
Fixed parameter set 1#[row, col] = [1, 8]
Parallelization option for element-matrix creation (default: 0). Active when PARALLEL_NO > 1.
Setting value |
Description |
|---|---|
0 |
Parallelizes element coefficient-matrix creation. This requires large arrays for parallelization. Default |
1 |
Does not parallelize element coefficient-matrix creation. This minimizes additional arrays for parallelization. |
The <PARALLEL module> is required.
Tip
When PARALLE_NO > 1, a confirmation message is written to the check file.
Example: four-way parallel calculation
Text-format section
Text format#* NODE_ORDER * EDGE_ORDER * METRIC_MOD * QUAD_TRI * CALC_IND * THIN_ELEMENT * PARALLEL_NO * PARALLEL_OPTION * 1 1 0 0 0 0 4 0
**** Number of parallel threads: 4 ****
priority : speed