Repository navigation
fix(stiffener): no negative EI where two STIFFENERS rows meet inside an element - #129
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. Keep it up! 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 |
ab3cc07 to
1a317d1
Compare
|
Codex Review: Didn't find any major issues. You're on a roll. 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". |
…an element
CD_Stiffener_Overlay takes an element's bending stiffness as the compliance
average l / (sum of covered parts / EI_sleeved + bare part / EI_cable). Two
rows meeting inside an element (a cylinder, then a cone) cover it in two
parts whose sum can exceed the element length by round-off, and the bare part
l - covered then came out a tiny negative length. Divided by the cable's EI it
is negligible for an ordinary cable, but a soft one (EI = 1e-12 or 1e-300
N m^2) or a very stiff sleeve turned it into a negative compliance and a
negative element EI: the static solve then failed with an internal message
("finite-EI hierarchy preflight failed: CD_HermiteCable_Sequence_Coarse_Max_Length:
l0>0, EA>=0, EI>=0 required"). The bare part is now never below zero, and
the bend-stiffener example with a cable EI of 1e-12 or 1e-300 runs.
A STIFFENERS RootOD or TipOD is now at most 1e3 m, the bound of a line type's
Diam, so a sleeve of 1e6 m is refused by name instead of failing the statics;
the Python deck validator applies the same bound.
On the shipped bend-stiffener example (2 s) the outputs change at the static
solver's tolerance: 7.5e-8 of the column maximum on the main output, 1.0e-8 in
the static profile and range graph; the per-element extrema file moves the
position of the minimum along elements whose axial force is flat to 1e-7.
Tests: test_bend_stiffener sweeps the cable EI (1e-300 to 2e4), the element
length (three meshes) and the root OD (0.45 m to 1e6 m) over two rows meeting
inside an element, requiring every element EI at least the cable's and finite
(27 of 45 failed before); the parity tests check the OD bound against the
validator and the driver.
Signed-off-by: Jae Hoon Seo <jaehoon.seo@inha.ac.kr>
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 |
1a317d1 to
5c99461
Compare
|
Codex Review: Didn't find any major issues. Nice work! 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". |
CD_Stiffener_Overlaysums the length of each element that the STIFFENERS rows cover. Where two rows meet inside an element the sum can exceed the element length by round-off, so the bare length came out a tiny negative number; divided by a soft cable's EI (1e-12 or 1e-300) or under a very wide sleeve, it made the element's EI negative and the run failed inside the mesh sequencing. The bare length is now clamped at zero.STIFFENERS RootOD and TipOD must also be at most 1e3 m, the bound on a line type's
Diam; the solver and the Python validator both refuse a larger one by name.Tests. A unit sweep over EI, element length and OD (27 of its 45 cases failed before); validator/driver parity cases for the OD bound. CTest 369/369 on both BLAS kernels; pytest passes.
Result changes. The bend-stiffener example moves by up to 7.5e-8 of the column maximum on the main output, and by 1e-8 in the static and range files, which is within the static solver's tolerance. Every other example is unchanged.