Skip to content

update master#2499

Merged
pcarruscag merged 448 commits intomasterfrom
develop
May 4, 2025
Merged

update master#2499
pcarruscag merged 448 commits intomasterfrom
develop

Conversation

@pcarruscag
Copy link
Member

@pcarruscag pcarruscag commented May 4, 2025

No description provided.

pcarruscag and others added 30 commits January 1, 2025 14:57
@pcarruscag
missing initialization
cf4f609
@pcarruscag
time average fix
ae84d15
@pcarruscag
fix...
13b08f5
@pcarruscag
Merge pull request #2372 from su2code/feature_small_restart
f4cda52 
Write compact restart files
@bigfooted
Merge branch 'develop' into fix_cornernode_normal
4e27793
@Vaish-W
Add nodal temperature integration in Compute_StiffMatrix
323c27b
@Vaish-W
Merge branch 'coupled_thermoelasticity' of github.com:Vaish-W/SU2 int…
46249b6 
…o coupled_thermoelasticity

Add nodal temperature integration in Compute_StiffMatrix
@Vaish-W
Handle temperature from the heat_solver
60d7b6c
@Vaish-W
Output solution for the heat solver in CElasticityOutput
b1499dd
@Vaish-W
Remove Tutorials folder from the repository
0298a31
@Vaish-W
Add a check for the heat solver
3c83bbe
@bigfooted
cleanup old implementation, start new
f5f2de9
@bigfooted
small update
47f3839
@pcarruscag
Merge pull request #2404 from Vaish-W/coupled_thermoelasticity
1fc88ac 
[WIP] Coupled solver for thermoelasticity
@pcarruscag
instantiate numerics class
77eaa57
@pcarruscag
Merge branch 'tmp' into coupled_thermoelasticity
707ab9f
@bigfooted
update
55d2163
@bigfooted
update
535c841
@bigfooted
update
4b30110
@bigfooted
cleanup
bd45235
@bigfooted
fix (#2415)
77a5b14
@bigfooted
projection
31bd154
@bigfooted
cleanup
1f1a4fe
@pcarruscag
simplify
7d55b44
@bigfooted
Merge branch 'develop' into fix_cornernode_normal
cb29d29
- Added Non-dimensionalization to translation velocity
140dec2
- fixed unsteady names
72d8f30
@rois1995
Merge pull request #2417 from su2code/fix_TranslVel
2b6d97a 
Fix Non-dimensionalization to translation velocity
@rois1995
Merge branch 'develop' into fix_UnstNames
110773c
@bigfooted
Merge branch 'develop' into fix_cornernode_normal
1b43468
pcarruscag and others added 27 commits April 6, 2025 15:43
@pcarruscag
more updates
f81e979
@EvertBunschoten
Removed restart check from spark initation (#2487)
b5f6ac3
@pcarruscag
tune settings of translating naca
c8de598
@pcarruscag
better convergence with inconsistent Jacobian
33c866d
@pcarruscag
disable enormous file diffs
85af966
@pcarruscag
Merge remote-tracking branch 'origin/develop' into symmetry_jacobian
b0a7fec
@pcarruscag
more updates
e402203
@pcarruscag
more updates
1943709
@pcarruscag
Merge branch 'develop' into SU2_master
367ee26
@pcarruscag
Merge pull request #2478 from Bot-Enigma-0/SU2_master
8becc90 
2D DSMA661 Airfoil Near-Wake V&V case
@oleburghardt
Merge branch 'develop' of https://github.com/su2code/SU2 into fix_sma…
a0d2d4a 
…ll_AD_issues
@oleburghardt
Merge pull request #2472 from su2code/fix_small_AD_issues
c9f098d 
Fix preaccumulation error, improve recording of RANS solvers
@pcarruscag
fix adj dyn fsi case
a537aa1
@pcarruscag
Merge remote-tracking branch 'origin/develop' into symmetry_jacobian
a38005e
@pcarruscag
update more fsi cases
94df327
@pcarruscag
change large filediffs to residuals
c995769
@pcarruscag
last updates?
551d7c5
@pcarruscag
Merge pull request #2486 from su2code/symmetry_jacobian
02be3f6 
Improve the Euler wall Jacobian
@pcarruscag
Merge branch 'develop' into coupled_thermoelasticity
e3fdfc2
@pcarruscag
refactor output
881ed0c
@pcarruscag
testcase and handle non dimensional mode
5f67a0f
@pcarruscag
update tests
e0354b2
@pcarruscag
Merge pull request #2429 from su2code/coupled_thermoelasticity
386d1ef 
Coupled thermoelasticity
@pcarruscag
2024-2025
b74d69a
@pcarruscag
Merge pull request #2497 from su2code/update_copyright
150727b 
2024->2025
@actions-user
Changing version number to 8.2.0
6fb2011
@pcarruscag
Merge pull request #2498 from su2code/release_v8.2.0
a600604 
Changing version number to 8.2.0
github-advanced-security[bot]
github-advanced-security bot found potential problems May 4, 2025
View reviewed changes
Common/include/geometry/elements/CElement.hpp
* \brief Shape function Jacobian (dNi) evaluated at point Xi,Eta.
*/
inline static void ShapeFunctionJacobian(su2double Xi, su2double Eta, su2double dNi[][2]) {
inline static void ShapeFunctionJacobian(const su2double& Xi, const su2double& Eta, su2double dNi[][2]) {

Check notice

Code scanning / CodeQL

No raw arrays in interfaces

Raw arrays should not be used in interfaces. A container class should be used instead.

Copilot Autofix

AI 9 months ago

To fix the issue, replace the raw array parameter su2double dNi[][2] in the ShapeFunctionJacobian method with a safer container class, such as std::array. Specifically:

  1. Use std::array<std::array<su2double, 2>, 4> to represent a 4x2 matrix, as the size of the array is fixed and known at compile time.
  2. Update the method signature to accept the container by reference.
  3. Modify the method implementation to work with the std::array structure instead of raw arrays.

This change ensures type safety, prevents array decay, and makes the code more expressive and maintainable.

Suggested changeset 1
Common/include/geometry/elements/CElement.hpp

Autofix patch

Autofix patch
Run the following command in your local git repository to apply this patch
cat << 'EOF' | git apply
diff --git a/Common/include/geometry/elements/CElement.hpp b/Common/include/geometry/elements/CElement.hpp
--- a/Common/include/geometry/elements/CElement.hpp
+++ b/Common/include/geometry/elements/CElement.hpp
@@ -33,2 +33,3 @@
 #include <vector>
+#include <array>
 
@@ -828,3 +829,3 @@
    */
-  inline static void ShapeFunctionJacobian(const su2double& Xi, const su2double& Eta, su2double dNi[][2]) {
+  inline static void ShapeFunctionJacobian(const su2double& Xi, const su2double& Eta, std::array<std::array<su2double, 2>, 4>& dNi) {
     dNi[0][0] = -0.25 * (1.0 - Eta);
EOF
@@ -33,2 +33,3 @@
#include <vector>
#include <array>

@@ -828,3 +829,3 @@
*/
inline static void ShapeFunctionJacobian(const su2double& Xi, const su2double& Eta, su2double dNi[][2]) {
inline static void ShapeFunctionJacobian(const su2double& Xi, const su2double& Eta, std::array<std::array<su2double, 2>, 4>& dNi) {
dNi[0][0] = -0.25 * (1.0 - Eta);
Copilot is powered by AI and may make mistakes. Always verify output.
Unable to commit as this autofix suggestion is now outdated
Common/src/CConfig.cpp
"of entries (the number of materials).", CURRENT_FUNCTION);
if (nMaterialThermalExpansion == 0) {
nMaterialThermalExpansion = 1;
MaterialThermalExpansion = new su2double[1]();

Check warning

Code scanning / CodeQL

Resource not released in destructor

Resource MaterialThermalExpansion is acquired by class CConfig but not released anywhere in this class.

Copilot Autofix

AI 9 months ago

Copilot could not generate an autofix suggestion

Copilot could not generate an autofix suggestion for this alert. Try pushing a new commit or if the problem persists contact support.

Common/src/toolboxes/fem/CFaceOfElement.cpp
Comment on lines +148 to +232
switch (nCornerPoints) {
case 2: {
/* Element is a line. Check if the node numbering must be swapped. If so
also the element information must be swapped, because element 0 is to
the left of the face and element 1 to the right. */
if (cornerPoints[1] < cornerPoints[0]) {
std::swap(cornerPoints[0], cornerPoints[1]);
swapElements = true;
}
break;
}

case 3: {
/* Element is a triangle. The vertices are sorted in increasing order.
If the sequence of the new numbering is opposite to the current
numbering, the element information must be exchanged, because
element 0 is to the left of the face and element 1 to the right. */
unsigned long nn[] = {cornerPoints[0], cornerPoints[1], cornerPoints[2]};
unsigned short ind = 0;
if (nn[1] < nn[ind]) ind = 1;
if (nn[2] < nn[ind]) ind = 2;

unsigned short indm1 = ind == 0 ? 2 : ind - 1; // Next lower index.
unsigned short indp1 = ind == 2 ? 0 : ind + 1; // Next upper index.

if (nn[indp1] < nn[indm1]) {
/* The orientation of the triangle remains the same.
Store the new sorted node numbering. */
cornerPoints[0] = nn[ind];
cornerPoints[1] = nn[indp1];
cornerPoints[2] = nn[indm1];
} else {
/* The orientation of the triangle changes. Store the new
sorted node numbering and set swapElements to true. */
cornerPoints[0] = nn[ind];
cornerPoints[1] = nn[indm1];
cornerPoints[2] = nn[indp1];
swapElements = true;
}

break;
}

case 4: {
/* Element is a quadrilateral. The vertices are sorted in increasing order
under the condition neighboring vertices remain neighbors. If the
sequence of the new numbering is opposite to the current
numbering, the element information must be exchanged, because
element 0 is to the left of the face and element 1 to the right. */
unsigned long nn[] = {cornerPoints[0], cornerPoints[1], cornerPoints[2], cornerPoints[3]};
unsigned short ind = 0;
if (nn[1] < nn[ind]) ind = 1;
if (nn[2] < nn[ind]) ind = 2;
if (nn[3] < nn[ind]) ind = 3;

unsigned short indm1 = ind == 0 ? 3 : ind - 1; // Next lower index.
unsigned short indp1 = ind == 3 ? 0 : ind + 1; // Next upper index.
unsigned short indp2 = ind >= 2 ? ind - 2 : ind + 2; // Opposite index.

if (nn[indp1] < nn[indm1]) {
/* The orientation of the quadrilateral remains the same.
Store the new sorted node numbering. */
cornerPoints[0] = nn[ind];
cornerPoints[1] = nn[indp1];
cornerPoints[2] = nn[indp2];
cornerPoints[3] = nn[indm1];
} else {
/* The orientation of the quadrilateral changes. Store the new
sorted node numbering and set swapElements to true. */
cornerPoints[0] = nn[ind];
cornerPoints[1] = nn[indm1];
cornerPoints[2] = nn[indp2];
cornerPoints[3] = nn[indp1];
swapElements = true;
}

break;
}

default: {
std::ostringstream message;
message << "Unknown surface element type with " << nCornerPoints << " corners." << std::endl;
SU2_MPI::Error(message.str(), CURRENT_FUNCTION);
}
}

Check notice

Code scanning / CodeQL

Long switch case

Switch has at least one case that is too long: [4 (35 lines)](1).

Copilot Autofix

AI 9 months ago

To address the issue, we will refactor the logic in the case 4: block into a separate function named ProcessQuadrilateralCase. This function will encapsulate the logic for handling quadrilateral elements, making the switch statement more concise and easier to follow. The function will take the necessary parameters (e.g., cornerPoints, swapElements) and return any modified values or flags.

The changes will involve:

  1. Extracting the logic from case 4: into a new function ProcessQuadrilateralCase.
  2. Replacing the case 4: block with a call to the new function.
  3. Ensuring that the new function has access to all required variables and updates them correctly.
Suggested changeset 1
Common/src/toolboxes/fem/CFaceOfElement.cpp

Autofix patch

Autofix patch
Run the following command in your local git repository to apply this patch
cat << 'EOF' | git apply
diff --git a/Common/src/toolboxes/fem/CFaceOfElement.cpp b/Common/src/toolboxes/fem/CFaceOfElement.cpp
--- a/Common/src/toolboxes/fem/CFaceOfElement.cpp
+++ b/Common/src/toolboxes/fem/CFaceOfElement.cpp
@@ -191,34 +191,3 @@
     case 4: {
-      /* Element is a quadrilateral. The vertices are sorted in increasing order
-         under the condition neighboring vertices remain neighbors. If the
-         sequence of the new numbering is opposite to the current
-         numbering, the element information must be exchanged, because
-         element 0 is to the left of the face and element 1 to the right. */
-      unsigned long nn[] = {cornerPoints[0], cornerPoints[1], cornerPoints[2], cornerPoints[3]};
-      unsigned short ind = 0;
-      if (nn[1] < nn[ind]) ind = 1;
-      if (nn[2] < nn[ind]) ind = 2;
-      if (nn[3] < nn[ind]) ind = 3;
-
-      unsigned short indm1 = ind == 0 ? 3 : ind - 1;        // Next lower index.
-      unsigned short indp1 = ind == 3 ? 0 : ind + 1;        // Next upper index.
-      unsigned short indp2 = ind >= 2 ? ind - 2 : ind + 2;  // Opposite index.
-
-      if (nn[indp1] < nn[indm1]) {
-        /* The orientation of the quadrilateral remains the same.
-           Store the new sorted node numbering. */
-        cornerPoints[0] = nn[ind];
-        cornerPoints[1] = nn[indp1];
-        cornerPoints[2] = nn[indp2];
-        cornerPoints[3] = nn[indm1];
-      } else {
-        /* The orientation of the quadrilateral changes. Store the new
-           sorted node numbering and set swapElements to true. */
-        cornerPoints[0] = nn[ind];
-        cornerPoints[1] = nn[indm1];
-        cornerPoints[2] = nn[indp2];
-        cornerPoints[3] = nn[indp1];
-        swapElements = true;
-      }
-
+      ProcessQuadrilateralCase(cornerPoints, swapElements);
       break;
@@ -242,2 +211,36 @@
   }
+}
+
+void CFaceOfElement::ProcessQuadrilateralCase(unsigned long* cornerPoints, bool& swapElements) {
+  /* Element is a quadrilateral. The vertices are sorted in increasing order
+     under the condition neighboring vertices remain neighbors. If the
+     sequence of the new numbering is opposite to the current
+     numbering, the element information must be exchanged, because
+     element 0 is to the left of the face and element 1 to the right. */
+  unsigned long nn[] = {cornerPoints[0], cornerPoints[1], cornerPoints[2], cornerPoints[3]};
+  unsigned short ind = 0;
+  if (nn[1] < nn[ind]) ind = 1;
+  if (nn[2] < nn[ind]) ind = 2;
+  if (nn[3] < nn[ind]) ind = 3;
+
+  unsigned short indm1 = ind == 0 ? 3 : ind - 1;        // Next lower index.
+  unsigned short indp1 = ind == 3 ? 0 : ind + 1;        // Next upper index.
+  unsigned short indp2 = ind >= 2 ? ind - 2 : ind + 2;  // Opposite index.
+
+  if (nn[indp1] < nn[indm1]) {
+    /* The orientation of the quadrilateral remains the same.
+       Store the new sorted node numbering. */
+    cornerPoints[0] = nn[ind];
+    cornerPoints[1] = nn[indp1];
+    cornerPoints[2] = nn[indp2];
+    cornerPoints[3] = nn[indm1];
+  } else {
+    /* The orientation of the quadrilateral changes. Store the new
+       sorted node numbering and set swapElements to true. */
+    cornerPoints[0] = nn[ind];
+    cornerPoints[1] = nn[indm1];
+    cornerPoints[2] = nn[indp2];
+    cornerPoints[3] = nn[indp1];
+    swapElements = true;
+  }
 }
EOF
Copilot is powered by AI and may make mistakes. Always verify output.
Unable to commit as this autofix suggestion is now outdated
@pcarruscag pcarruscag merged commit 5d71ae2 into master May 4, 2025
56 of 70 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

changelog:none

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.