Changes

Summary

1. updated outputfld dump use in a test case.  add momentum flux correction to old PDEIBL2D formulation but commented out (details)
2. Added jta29@psu.edu to DeveloperEmailAddresses (details)
3. Added DumpSolurion to Pseudo Time (details)
4. Added jta29@psu.edu to DeveloperEmailAddresses (details)
5. Editing DumpSolution for Psuedo Time (details)
6. updated gitignore to ignore VS Code stuff (details)
7. Added new 3D IBL turbulent closure models (details)
8. Packed turbulent closure model into a dedicated file to allow for mucking with different closure models (details)
9. fixed vera warning: standard library format (details)
10. Added new neural net closure models with built-in symmetry.  Verified on 2Dness on a flat plate ZPG turbulent case. (details)
11. updated notation for Re_theta: Ret -> Rth (details)
12. added updated 3D laminar IBL closures (a mix of neural nets and simple curvefits).  checked 2D ZPG flat plate case (details)
13. added an option to toggle viscous-inviscid coupling mode at runtime for 3D IBL+panel. some minor refactoring of torpedo case (details)
14. removed old neural net closure models without built-in symmetry (details)
15. Added H*1_diff and Cf1_diff clousre models. Updated H22 closure to be the lastest trained on data without duplicates for crossflow --> closure model is essentially the same.  Tested quasi-2D turbulent BL on ZPG flat plate. (details)
16. minor cleanup of IBL3 pde (details)
17. minor cleanup (details)
18. completed masterState for PDEIBL3FourEqn3D so that PTC solver can be used (details)
19. fixed erroneous rh0ref parameter calculation in setting up the case (i.e. isentropic relation). minor cleanup on other parts of the test case as well (details)
20. turned on both amp and lag equations for IBL3 four equation formulation.  Refactored ntinit condition and ntcrit attractors for flat plate case (details)
21. mucked with 2D ZPG flat plate case with free transition: including initial solution for nt (details)
22. bug fix in rh0ref/rhref isentropic relation when setting up test case (details)
23. added a 3D turbulent BL torpedo case setup (with forced transition near the inlet) (details)
24. minor refactoring of transition model parameters (details)
25. minor update to IBL3 torpedo case (details)
26. trying fully turbulent BL runs for IBL3 on torpedo case.  also some auto-formatting from VS Code (details)
27. minor fix (details)
28. added missing include (details)
29. updates on PETSc settings etc. (details)
30. added a new set of primary variables (i.e. lg(theta) and H's).  They don't quite help with linear solve as intended though... (details)
31. renamed functions (details)
32. renamed variables for better clarity (details)
33. turned on LF flux for 4-eqn IBL3 formulation, which (partially) fixed crossflow instability issues (details)
34. refactored IBL3 LF flux dissipation coefficient to avoid zero stabilization in the case of qe be tangent to element interface (details)
35. turned on mass jump for 3D IBL3 4-eqn's LF flux: providing stabilization for flow transition (checked on ZPG flat plate transition case) (details)
36. removed outdated closure models for 4-eqn IBL3 laminar flow (details)
37. removed includes of outdated closure models (details)
38. mucked with 4-eqn IBL3 on flat/torpedo case (details)
39. intermediate commit...mucking with a sandbox case (details)
40. added knob for controling added dissipation in LF flux for IBL3 (details)
41. minor refactoring (details)
42. make IBLVarDataType3D templated, to allow for both Real and Surreal etc. (details)
43. added symmetry (i.e. no crossflow) BC for 4-eqn IBL3 (details)
44. refactored conservative variables, which is the sames as the previous version for incompressible flow (details)
45. scaled ke and lc eqns by 1/qe^2, but it didn't fix the symmetry plane instability (details)
46. added an exception for stagnation point in IBL3 4-eqn formulation (details)
47. added easier macro toggle for turning on/off amp+lag equations (details)
48. fix formatting (details)
49. add exception for corner case (details)
50. minor refactoring of torpedo case for IBL3 in sandbox (details)
51. added but commented out a hack for trying power-law crossflow velocity distribution (details)
52. added BC dict missing from a previous commit.  minor cleanup of 4-eqn IBL3 stuff (details)
53. updated H22 turb closure (details)
54. Added Space Time Case (details)
55. reverted turbulent Cf1 and Hs1 closure to 2D (details)
56. minor tweaks/comments (details)
57. Started another set of turbulent closure models by copying over existing ones (details)
58. added turb Cf1 closure (details)
59. added turb H*1 closure (details)
60. Updated Insantiation (details)
61. updated CDc closure (details)
62. updated turb Cf2 closure (details)
63. updated turb Hc2 closure (details)
64. fixed function signature (details)
65. fixed typo in function call. switched default to using new regularized turb closure (details)
66. switched back to using 2D turb H*1 closure (details)
67. Remove trailing white space (details)
68. Removed trailing white space (details)
69. added output of closure model derivatives for IBL3 flux system eigen-analysis (details)
70. added prescribed CD(x,y) field (details)
71. added optional vertex coordinates for single triangle/tet constructors (details)
72. Added prescribed CD field for inversion of CDiss correction factor beta (details)
73. minor tweaks of torpedo case IBL3 setup (details)
74. Mucking with a correction multiplier term to the dissipation relation (details)
75. fixed long lines (details)
76. mucking with dissipation relation inversion (details)
77. Use MKL_PARDISO for PETSc DirectLU solver (details)
78. Only use one thread for DirectLU MKL_PARDISO (details)
79. move the PETSc solve time print (details)
80. Add PETSc option to scale by diagonal (details)
81. Added but commented out hacks with inflow BC bump/step features (details)
82. Added a few prescribed CD(x,y) field based on RANS data (details)
83. fillSystemVector syncs MPI DOF. Fix Sensor symmetry BC for 3D. More parallel testing with NaN's (details)
84. added new beta closure (details)
85. Fix compile errors (details)
86. Fix compile errors (details)
87. Refactor isValidState checks in parallel (details)
88. Compile fix (details)
89. One more isValidState fix (details)
90. Add positive preserving source for 3D. Dissabled for now. (details)
91. Compile fix (details)
92. Add some asserts to 'Node Rank' for xfld tecplot dump (details)
93. Also dump out xfld spirit (details)
94. Add set_dot and continuously relax sewFace tolerances for CamberBlade LE and TE surfaces. (details)
95. Format fix (details)
96. Fix the last fix (details)
97. Actually fix the fix of the last fix for the previous fix (details)
98. Add UpdateFraction to Galerkin Stabilized (details)
99. Added macro to toggle the use of prescribed CD(x,y) field (details)
100. Allow MaxChangeFraction to work with SteveLogic (details)
101. Skip zombie DOF for UpdateFraction check (details)
102. fixed outdated class template argument (details)
103. More fixes to allow UpdateFraction with SteveLogic (details)
104. Don't mix units (details)
105. fixed more outdated class template argument (details)
106. Fix unit tests (details)
107. EmbeddedCGField fixes (details)
108. Fix one more missing EmbeddedCGField (details)
109. Updated Bernstein bases for use in CG fields. We dont have Hessians implemented, so stabilized CG needed a change to only calculate Hessians if we have stabilization turned off. To go above p3 in 3D CG, we need to change the basis functions to be match our implied ordering. Since Bernstein polynomials are modal and not hierarchical, we need to solve a linear system to do projections between orders. (details)
110. Fixed unit tests.  Turned off the correction when ctau surpasses ctau_Tinit in 3D IBL_foureqn (details)
111. Added 4D Bernstein bases (linear) on pentatope, with CG and DG fields (details)
112. updated pyrite files for IBL3 four eqn laminar (details)
113. fixed undefined references (details)
114. git was not happy about the commit message...trying to fix it here..this commit is about adding definition of static class members in cpp file (details)
115. Adding adapt case with P1 corner singularity divergence weirdness. P2 and 3 are fine, but P1 error blows up 4 orders of magnitude.... (details)
116. Update tolerances on L2 project tests for Bernstein bases. Fix NaN issue with auxiliary problem for Homotopy. (details)
117. Update tolerances for intel (details)
118. Add Lagrange Node tecplot dump for CG fields (details)
119. Fix last commit (details)
120. added PyDict parameter (StatusOKatMaxIterations) to change status when solver hits (details)
121. Updated tecplot CG dump (details)
122. Compile fix (details)
123. Allow zero bodies in a model (details)
124. Add Cylindrical system for Coles profile (details)
125. Maybe compile fixes (details)
126. Fix CG tecplot dump (details)
127. Default coordinate system for Coles profile (details)
128. 1st try at 3D NS/RANS no-slip isothermal/adiabatic BCs via Robin sans-Lagrange (details)
129. added -DUSE_GRID_LAPLACIAN to makefiles in all directories containing files where the #ifdef is used (details)
130. make FieldWeighted consistent with BasisWeighted (mostly involved adding param to PDE calls) (details)
131. replaced hard-coded false to PDE check for including viscous terms in boundary flux (details)
132. instantiations needed for uniform source in sensor PDE for artificial viscosity (details)
133. added #pragma message to all *.h files that use USE_GRID_LAPLACIAN; this will (details)
134. removed explicit #define/#undef USE_GRID_LAPLACIAN from unit tests that use this (details)
135. added Tq & Tg template parameters for sourceTrace (details)
136. added p_ref() function to density model to provide a means of accessing reference (details)
137. changed SANS_ASSERT to SANS_ASSERT_MSG to output message on error (details)
138. fixed typo: D1 -> D2 (details)
139. added PyDict option to specify Dirichlet data for Robin BC for AV; useful for (details)
140. added FieldWeighted unit test for sans-Lagrange Dirichlet BCs for Euler (details)
141. added test for jacobian/metric of an equilateral triangle (details)
142. exact data explicitly provided, rather than determined by do-loop dot product accumulation (details)
143. added unit grid where coordinates of all 4 corners can be input; facilitates making (details)
144. typo fix in unit test (details)
145. turn off clang warnings for '#pragma message' (details)
146. More tecplot dump fixes (details)
147. fixed typo in unit test (details)
148. vera: lines too long (details)
149. Compile fix (details)
150. replace several repeats of explicit pow(*, 2) coding with call to private function (details)
151. new MatrixS matmul instantiations needed for sans-Lagrange Robin BCs of no-slip isothermal NS/RANS 3D (details)
152. added Robin_sansLG to VMSD and VMSD-BR2 (details)
153. Hopefully finally got the link problems fixed (details)
154. more #include's for Robin_sansLG BCs (details)
155. added/deleted #include's for IntegrandBoundaryTraces; added Robin_sansLG where needed, (details)
156. new MatrixS matmul instantiations needed for Robin_sansLG BCs (details)
157. Trying to fix vera errors (details)
158. Add missing USE_GRID_LAPACIAN. Minor PDERANSSA formatting fixes. Fix for Coles RANSSA. (details)
159. Add instantiation (details)
160. Use H instead of H^2 in Robin BCs for AV when USE_GRID_LAPLACIAN is defined. (details)
161. Update BCRANSSA3D (details)
162. Fixing errors (details)

updated outputfld dump use in a test case.  add momentum flux correction to old PDEIBL2D formulation but commented out

Added jta29@psu.edu to DeveloperEmailAddresses

Added DumpSolurion to Pseudo Time

Added jta29@psu.edu to DeveloperEmailAddresses

Editing DumpSolution for Psuedo Time

updated gitignore to ignore VS Code stuff

Added new 3D IBL turbulent closure models

Packed turbulent closure model into a dedicated file to allow for mucking with different closure models

fixed vera warning: standard library format

Added new neural net closure models with built-in symmetry.  Verified on 2Dness on a flat plate ZPG turbulent case.

updated notation for Re_theta: Ret -> Rth

added updated 3D laminar IBL closures (a mix of neural nets and simple curvefits).  checked 2D ZPG flat plate case

added an option to toggle viscous-inviscid coupling mode at runtime for 3D IBL+panel. some minor refactoring of torpedo case

removed old neural net closure models without built-in symmetry

Added H*1_diff and Cf1_diff clousre models. Updated H22 closure to be the lastest trained on data without duplicates for crossflow --> closure model is essentially the same.  Tested quasi-2D turbulent BL on ZPG flat plate.

minor cleanup of IBL3 pde

minor cleanup

completed masterState for PDEIBL3FourEqn3D so that PTC solver can be used

fixed erroneous rh0ref parameter calculation in setting up the case (i.e. isentropic relation). minor cleanup on other parts of the test case as well

turned on both amp and lag equations for IBL3 four equation formulation.  Refactored ntinit condition and ntcrit attractors for flat plate case

mucked with 2D ZPG flat plate case with free transition: including initial solution for nt

bug fix in rh0ref/rhref isentropic relation when setting up test case

added a 3D turbulent BL torpedo case setup (with forced transition near the inlet)

minor refactoring of transition model parameters

minor update to IBL3 torpedo case

trying fully turbulent BL runs for IBL3 on torpedo case.  also some auto-formatting from VS Code

minor fix

added missing include

updates on PETSc settings etc.

added a new set of primary variables (i.e. lg(theta) and H's).  They don't quite help with linear solve as intended though...

renamed functions

renamed variables for better clarity

turned on LF flux for 4-eqn IBL3 formulation, which (partially) fixed crossflow instability issues

refactored IBL3 LF flux dissipation coefficient to avoid zero stabilization in the case of qe be tangent to element interface

turned on mass jump for 3D IBL3 4-eqn's LF flux: providing stabilization for flow transition (checked on ZPG flat plate transition case)

removed outdated closure models for 4-eqn IBL3 laminar flow

removed includes of outdated closure models

mucked with 4-eqn IBL3 on flat/torpedo case

intermediate commit...mucking with a sandbox case

added knob for controling added dissipation in LF flux for IBL3

minor refactoring

make IBLVarDataType3D templated, to allow for both Real and Surreal etc.

added symmetry (i.e. no crossflow) BC for 4-eqn IBL3

refactored conservative variables, which is the sames as the previous version for incompressible flow

scaled ke and lc eqns by 1/qe^2, but it didn't fix the symmetry plane instability

added an exception for stagnation point in IBL3 4-eqn formulation

added easier macro toggle for turning on/off amp+lag equations

fix formatting

add exception for corner case

minor refactoring of torpedo case for IBL3 in sandbox

added but commented out a hack for trying power-law crossflow velocity distribution

added BC dict missing from a previous commit.  minor cleanup of 4-eqn IBL3 stuff

updated H22 turb closure

Added Space Time Case

reverted turbulent Cf1 and Hs1 closure to 2D

minor tweaks/comments

Started another set of turbulent closure models by copying over existing ones

added turb Cf1 closure

added turb H*1 closure

Updated Insantiation

updated CDc closure

updated turb Cf2 closure

updated turb Hc2 closure

fixed function signature

fixed typo in function call. switched default to using new regularized turb closure

switched back to using 2D turb H*1 closure

Remove trailing white space

Removed trailing white space

added output of closure model derivatives for IBL3 flux system eigen-analysis

added prescribed CD(x,y) field

added optional vertex coordinates for single triangle/tet constructors

Added prescribed CD field for inversion of CDiss correction factor beta

minor tweaks of torpedo case IBL3 setup

Mucking with a correction multiplier term to the dissipation relation

fixed long lines

mucking with dissipation relation inversion

Use MKL_PARDISO for PETSc DirectLU solver

Only use one thread for DirectLU MKL_PARDISO

move the PETSc solve time print

Add PETSc option to scale by diagonal

Added but commented out hacks with inflow BC bump/step features

Added a few prescribed CD(x,y) field based on RANS data

fillSystemVector syncs MPI DOF. Fix Sensor symmetry BC for 3D. More parallel testing with NaN's

added new beta closure

Fix compile errors

Fix compile errors

Refactor isValidState checks in parallel

Compile fix

One more isValidState fix

Add positive preserving source for 3D. Dissabled for now.

Compile fix

Add some asserts to 'Node Rank' for xfld tecplot dump

Also dump out xfld spirit

Add set_dot and continuously relax sewFace tolerances for CamberBlade LE and TE surfaces.

Format fix

Fix the last fix

Actually fix the fix of the last fix for the previous fix

Add UpdateFraction to Galerkin Stabilized

Added macro to toggle the use of prescribed CD(x,y) field

Allow MaxChangeFraction to work with SteveLogic

Skip zombie DOF for UpdateFraction check

fixed outdated class template argument

More fixes to allow UpdateFraction with SteveLogic

Don't mix units

fixed more outdated class template argument

Fix unit tests

EmbeddedCGField fixes

Fix one more missing EmbeddedCGField

Updated Bernstein bases for use in CG fields. We dont have Hessians implemented, so stabilized CG needed a change to only calculate Hessians if we have stabilization turned off. To go above p3 in 3D CG, we need to change the basis functions to be match our implied ordering. Since Bernstein polynomials are modal and not hierarchical, we need to solve a linear system to do projections between orders.

Fixed unit tests.  Turned off the correction when ctau surpasses ctau_Tinit in 3D IBL_foureqn

Added 4D Bernstein bases (linear) on pentatope, with CG and DG fields

updated pyrite files for IBL3 four eqn laminar

fixed undefined references

git was not happy about the commit message...trying to fix it here..this commit is about adding definition of static class members in cpp file

Adding adapt case with P1 corner singularity divergence weirdness. P2 and 3 are fine, but P1 error blows up 4 orders of magnitude....

Update tolerances on L2 project tests for Bernstein bases. Fix NaN issue with auxiliary problem for Homotopy.

Update tolerances for intel

Add Lagrange Node tecplot dump for CG fields

Fix last commit

added PyDict parameter (StatusOKatMaxIterations) to change status when solver hits
MaxIterations without converging; setting StatusOKatMaxIterations=true sets
status.converged=true, so subsequent solver doesn't error out; default is
StatusOKatMaxIterations=false

Updated tecplot CG dump

Compile fix

Allow zero bodies in a model

Add Cylindrical system for Coles profile

Maybe compile fixes

Fix CG tecplot dump

Default coordinate system for Coles profile

1st try at 3D NS/RANS no-slip isothermal/adiabatic BCs via Robin sans-Lagrange
formulation; should be neutrally stable for inviscid terms and consistent with
Clasius-Duhem for viscous; adiabatic not yet functional; isothermal completely untested

added -DUSE_GRID_LAPLACIAN to makefiles in all directories containing files where the #ifdef is used

make FieldWeighted consistent with BasisWeighted (mostly involved adding param to PDE calls)

replaced hard-coded false to PDE check for including viscous terms in boundary flux
for BCNone

instantiations needed for uniform source in sensor PDE for artificial viscosity
applied to two-phase porous media; currently commented out; to activate, uncomment #define
in each

added #pragma message to all *.h files that use USE_GRID_LAPLACIAN; this will
output compile time message as to whether USE_GRID_LAPLACIAN is or is not defined