fix: Complete stress extraction fix for NX Nastran OP2 files

THREE critical fixes applied:

1. API Access Pattern
   - Support dotted attribute names (e.g., 'stress.chexa_stress')
   - Compatible with newer pyNastran versions (NX 2412.5)
   - Fallback to older API formats for compatibility

2. Correct Von Mises Index
   - Solid elements (CHEXA, CTETRA, CPENTA): index 9
   - Shell elements (CQUAD4, CTRIA3): last column
   - Data structure: [oxx, oyy, ozz, txy, tyz, txz, o1, o2, o3, von_mises]

3. Units Conversion (CRITICAL)
   - NX Nastran outputs stress in kPa, not MPa
   - Apply conversion: kPa / 1000 = MPa
   - Example: 113094.73 kPa -> 113.09 MPa

Test Results:
- Before: 0.00 MPa (FAIL)
- After:  113.09 MPa at element 83 (SUCCESS)

Files modified:
- optimization_engine/result_extractors/op2_extractor_example.py

Test files added:
- examples/test_stress_direct.py
- examples/test_stress_fix.py
- examples/debug_op2_stress.py
- STRESS_EXTRACTION_FIXED.md
- TESTING_STRESS_FIX.md

examples/debug_op2_stress.py

@@ -0,0 +1,88 @@
+"""
+Deep diagnostic to find where stress data is hiding in the OP2 file.
+"""
+
+from pathlib import Path
+from pyNastran.op2.op2 import OP2
+
+op2_path = Path("examples/bracket/bracket_sim1-solution_1.op2")
+
+print("="*60)
+print("DEEP OP2 STRESS DIAGNOSTIC")
+print("="*60)
+print(f"File: {op2_path}")
+print()
+
+op2 = OP2()
+op2.read_op2(str(op2_path))
+
+# List ALL attributes that might contain stress
+print("--- SEARCHING FOR STRESS DATA ---")
+print()
+
+# Check all attributes
+all_attrs = dir(op2)
+stress_related = [attr for attr in all_attrs if 'stress' in attr.lower() or 'oes' in attr.lower()]
+
+print("Attributes with 'stress' or 'oes' in name:")
+for attr in stress_related:
+    obj = getattr(op2, attr, None)
+    if obj and not callable(obj):
+        print(f"  {attr}: {type(obj)}")
+        if hasattr(obj, 'keys'):
+            print(f"    Keys: {list(obj.keys())}")
+            if obj:
+                first_key = list(obj.keys())[0]
+                first_obj = obj[first_key]
+                print(f"    First item type: {type(first_obj)}")
+                if hasattr(first_obj, 'data'):
+                    print(f"    Data shape: {first_obj.data.shape}")
+                    print(f"    Data type: {first_obj.data.dtype}")
+                if hasattr(first_obj, '__dict__'):
+                    attrs = [a for a in dir(first_obj) if not a.startswith('_')]
+                    print(f"    Available methods/attrs: {attrs[:10]}...")
+
+print()
+print("--- CHECKING STANDARD STRESS TABLES ---")
+
+standard_tables = [
+    'cquad4_stress',
+    'ctria3_stress',
+    'ctetra_stress',
+    'chexa_stress',
+    'cpenta_stress',
+    'cbar_stress',
+    'cbeam_stress',
+]
+
+for table_name in standard_tables:
+    if hasattr(op2, table_name):
+        table = getattr(op2, table_name)
+        print(f"\n{table_name}:")
+        print(f"  Exists: {table is not None}")
+        print(f"  Type: {type(table)}")
+        print(f"  Bool: {bool(table)}")
+
+        if table:
+            print(f"  Keys: {list(table.keys())}")
+            if table.keys():
+                first_key = list(table.keys())[0]
+                data = table[first_key]
+                print(f"  Data type: {type(data)}")
+                print(f"  Data shape: {data.data.shape if hasattr(data, 'data') else 'No data attr'}")
+
+                # Try to inspect the data object
+                if hasattr(data, 'data'):
+                    print(f"  Data min: {data.data.min():.6f}")
+                    print(f"  Data max: {data.data.max():.6f}")
+
+                    # Show column-wise max
+                    if len(data.data.shape) == 3:
+                        print(f"  Column-wise max values:")
+                        for col in range(data.data.shape[2]):
+                            col_max = data.data[0, :, col].max()
+                            col_min = data.data[0, :, col].min()
+                            print(f"    Column {col}: min={col_min:.6f}, max={col_max:.6f}")
+
+print()
+print("="*60)