Optimize ccoor*.F files with array operations and intrinsic functions

[Copilot]

Description of the feature or the bug

The ccoor*.F files contained repetitive hand-coded loops for array assignments and manual distance calculations that could be replaced with more efficient Fortran intrinsics and array operations.

Description of the changes

Replaced hand-coded loops with array slice operations and intrinsic functions across four files:

Array slice operations:

• Connectivity/ID assignments: IX1(LFT:LLT)=IXC(2,LFT:LLT) instead of DO loops
• Scalar assignments: DT1C(JFT:JLT)=DT1 instead of element-by-element assignment
• 3D velocity/rotation: VL1(I,1:3)=V(1:3,IXCTMP2) instead of 3 separate lines per node
• Zeroing operations: VL1(I,1:3)=ZERO instead of 3 separate assignments

Intrinsic functions:

• Surface normal magnitude: HYPOT(HYPOT(E3X,E3Y),E3Z) for better numerical accuracy
• Kept squared distance calculations unchanged (XN* variables used in comparisons)

Code cleanup:

• Removed redundant loop in ccoor3.F that set IXCTMP* variables without using them

Example transformation:

! Before (48 lines):
VL1(I,1)=V(1,IXCTMP2)
VL1(I,2)=V(2,IXCTMP2)
VL1(I,3)=V(3,IXCTMP2)
VL2(I,1)=V(1,IXCTMP3)
...
VRL4(I,3)=VR(3,IXCTMP5)

! After (8 lines):
VL1(I,1:3)=V(1:3,IXCTMP2)
VL2(I,1:3)=V(1:3,IXCTMP3)
VL3(I,1:3)=V(1:3,IXCTMP4)
VL4(I,1:3)=V(1:3,IXCTMP5)
VRL1(I,1:3)=VR(1:3,IXCTMP2)
VRL2(I,1:3)=VR(1:3,IXCTMP3)
VRL3(I,1:3)=VR(1:3,IXCTMP4)
VRL4(I,1:3)=VR(1:3,IXCTMP5)

Net reduction: 171 lines → 71 lines (~58% reduction)

Original prompt

Can you have a look at ccoor*F files?
I would like to look for optimization opportunities. In particular by using intrinsic fuctions instead of home made code snippet, and array operations instead of hand-made loops.

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.