Repository navigation
fix(outputs): the tangent outputs of a finite-EI line take its tangent DOF - #114
Conversation
|
@codex review |
Codex Review Summary: ✅ CompletedThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
Codex Review: Didn't find any major issues. Bravo. Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
…t DOF The declination and azimuth of a finite-EI node were taken from a difference of node positions (central inside, one-sided at an end), the direction of a chord. A finite-EI node has its own tangent, the Hermite DOF m = dr/ds. Where the line bends within an element the two differ: at an end clamped by an END CONNECTIONS Rigid row the end tangent read 171.15 deg for a vertical (180 deg) clamp, 163.83 deg for 165 deg and 148.87 deg for 135 deg, up to 14 deg off; with a current the vertical clamp of a cable read 167.08 deg. line_geom_value and line_geom_values now take the angles of a finite-EI node (stride-6 state) from its tangent DOF, normalised and oriented End A -> End B (cable_node_tangent), and the range graph's declination of a cable reads the same. This is the shared evaluator of every finite-EI route, so L<L>N<J>Dec/Azi, FairDecl/AnchDecl, FairIncl/ AnchIncl and FairAngle/AnchAngle, the Declination, Inclination and Azimuth columns of .static.out, the range graph and the line tangent of the initialisation reports change together on the standalone, mixed and multibody routes, through the C API (CableDyn_GetLineValues, channels, notes) and in the OpenFAST shell, which reads them from the aggregate. EI = 0 lines, whose segments are straight, keep the difference of node positions. outputs.rst defines the node tangent of both. Tests: end_tangent_channels clamps a cable at End A at declinations of 120 to 180 deg and azimuths of 0 and 20 deg and pins End B. At the clamp the channels equal the clamp's angles to 1.4e-14 deg (14.1 deg off before), at the pinned end and every other node the angle of the solution's tangent DOF to 2.8e-14 deg (5.4 deg off before). driver_end_tangent_channels checks the channels, the static profile, the range graph and the reports of the standalone and mixed routes, and test_end_tangent.py the C API line values. Examples: the finite-EI tangent outputs change, by up to 3.1 deg in the line tangent report of buoy_clamped_cable.dat (now the clamp direction turned with its buoy), 1.8 deg in the static Declination of lozon_gomex80_power_cable_motion.dat, 1.4 deg in that of iea15mw_umaine_mixed_cabledyn.dat, 0.63 deg in the FairIncl1 history of lazy_wave_vessel_motion.dat and 0.49 deg in the range graph of torsion_lazy_wave_hangoff_twist.dat; 0.2 deg or less elsewhere. Positions, forces, tensions, curvatures and every EI = 0 example are bit-identical. Signed-off-by: Jae Hoon Seo <jaehoon.seo@inha.ac.kr>
e9cec8a to
3d42c12
Compare
|
@codex review |
The [Unreleased] entries of the open changes are added together in one change, so that they do not conflict with each other. Signed-off-by: Jae Hoon Seo <jaehoon.seo@inha.ac.kr>
|
@codex review |
|
Codex Review: Didn't find any major issues. Keep them coming! Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
|
Codex Review: Didn't find any major issues. Bravo. Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
The declination, azimuth and inclination outputs of a finite-EI line took their direction from differences of node positions, not from the node's tangent. That is the direction of a chord, not the line's tangent.
doc/outputs.rstalways describes these outputs as the node's axial tangent, so this PR makes the code match its documentation.At a clamped end the gap was large:
The largest error was 14.1° at a clamp and 5.4° at other nodes. Both are now at round-off.
Change
line_geom_valueandline_geom_values(the shared channel evaluator) now take a finite-EI node's direction from its tangent DOF, oriented End A → End B. EI = 0 lines keep the position difference, since their segments are straight.Outputs that change together:
L<L>N<J>Dec/AziFairDecl/AnchDecl,FairIncl/AnchIncl,FairAngle/AnchAngle.static.outDeclination, Inclination and Azimuth columnsRoutes covered: standalone, mixed and multibody runs; the C API's
CableDyn_GetLineValues; and OpenFAST through the aggregate.CableDyn_OF.f90itself is unchanged.doc/outputs.rstgains a node-tangent definition, cross-referenced from those outputs.Effect on results
Positions, forces, tensions and curvatures are unchanged. 40 of the 54 examples are bit-identical; the other 14 differ only in angle columns and report lines:
buoy_clamped_cablelozon_gomex80_power_cable_motioniea15mw_umaine_mixed_cabledynlazy_wave_vessel_motionFairIncl1historyTests
Each new test failed first on the unfixed code (982 assertions in
end_tangent_channels).end_tangent_channels(Fortran, aggregate route): clamp at 120–180°, azimuth 0 and 20°. The clamp-end outputs equal the clamp angles; every other node equals the solution's own tangent DOF.driver_end_tangent_channels: channels, static profile, range graph, init report and the mixed route.python/tests/test_end_tangent.py: the C API's line values.Test status
-W: clean.