Skip to content

README.md

Latest commit

 

History

History
54 lines (39 loc) · 1.61 KB

README.md

File metadata and controls

54 lines (39 loc) · 1.61 KB
Raw

Testing nec2++

Quick start

Build a debug binary, then run the regression harness against the reference engines:

# 1. Build nec2++ and the nec2c reference engine
cmake -B build -DCMAKE_BUILD_TYPE=Debug
cmake --build build -j$(nproc)

# 2. Build nec2c (C reference)
make -C c_src

# 3. Symlink the binaries where the Makefile expects them
ln -sf ../build/src/nec2++ ../src/nec2++
ln -sf ../build/src/nec2diff ../src/nec2diff

# 4. Run all tests (default: herzian_dipole.nec)
make

Running a specific test

make DO_TESTS=data/gh_flat_spiral.nec

Adding a test

Drop a .nec input file into data/. The harness will run it through nec2++, nec2c, and FORTRAN (if available), then diff the outputs with nec2diff.

Available test models

The data/ directory contains over 45 models covering dipoles, patches, helices, Sommerfeld ground, arrays, and more.

Known-limitation decks

Some decks document behavior rather than assert correctness, so their output files record what every engine currently produces:

  • sommerfeld_near_ground.nec — a horizontal half-wave dipole 0.02 wavelengths above a GN 2 ground with sigma 1e10 S/m (an effectively perfect half-space). Over a true perfect conductor (GN 1) the same antenna gives R = 0.9168 ohms and average power gain 1.9953 (the lossless limit is 2), but every SOMNEC-derived engine — nec2++, nec2c and the original Fortran — returns R ≈ 1.759 ohms and gain ≈ 1.043 there. This is a shared NEC-2 Sommerfeld limitation near ground, not a translation bug (PR #137). Compare against the GN 1 variant of the deck to see the gap.