Repository navigation
docs: GoSPL coupling uses instantaneous velocity after any remesh (DynEarthSol #112) - #95
Merged
Merged
Conversation
…nEarthSol #112) A remesh renumbers surface nodes, so the old node-count check could miss a remesh that happened to leave the count unchanged and difference unrelated nodes. The fix compares var.nremesh instead, so any remesh now falls back to the instantaneous velocity; update the known-limitations caveat, the log line format, and the build note (conda environment no longer needs to be deactivated) to match. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01MyXyjvBfCbpMgZ6ttvJXf3
echoi
pushed a commit
that referenced
this pull request
Oct 29, 2026
* docs: add skeleton for tectonic rifting and subduction tutorial New tutorial page following the layout of the GoSPL coupling page: physics background, a rifting part built on examples/rifting-2d.cfg, an open subduction part, exercises and troubleshooting. Section bodies to be written are marked with MDX TODO comments so the page renders while incomplete; figure links are commented out until the images exist in docs/tutorial/img/tectonics/. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: write Part 1 of the tectonics tutorial around core-complex.cfg Part 1 now uses examples/core-complex.cfg: full config walkthrough, most-changed parameters, run and visualize steps, an optional MMG box using the smallest_size/largest_size size band, three exercises (seed dip, surface diffusivity, MMG refinement) and a pointer to rifting-2d.cfg. The shared rheology table gains a "Used in" column. Part 2 carries a visible work-in-progress notice; all authoring notes are MDX comments so nothing marked TODO renders. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: flag unmerged MMG dependency in tectonics Part 1 A caution box at the top of Part 1 says the MMG box and Exercise 3 use parameters from a pull request not yet merged, while the stock run works on the current release. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: add strain-softening and rolling-hinge figures to tectonics Part 1 Two hand-written SVGs in the style of the GoSPL tutorial graphics. The strain-softening sketch plots cohesion and friction against plastic strain for core-complex.cfg and marks where the weak-zone seed starts. The rolling-hinge figure animates the setup (walls, seed, free surface, Winkler foundation) through fault breakthrough, footwall uplift and fault flattening over a 14 s SMIL loop. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: update tectonics Part 1 for merged DynEarthSol PR #95 MMG is now a submodule built with make usemmg=1, and the conservative remeshing parameters are on master, so the unmerged-PR caution is removed, the build wording follows the README, and the parameter reference points at examples/defaults.cfg and --help until the MMG page catches up. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: write tectonics Part 2 around examples/subduction.cfg Part 2 follows examples/subduction.cfg on the DynEarthSol branch feature/virtual-topology: settings table, poly geometry with a region table, abridged config walkthrough, the phase_change_option = 3 rules and the virtual-topology grid, run and smoke-run instructions, what to look for in the reference run, and two exercises. Adds the geometry sketch (SVG) and a material/viscosity evolution figure from the 5 Myr run, and stores the SVG generator scripts under img/tectonics/_src. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: align tectonics Part 2 with DynEarthSol PR #110, drop the friction exercise The caution box names the open pull request. The rule table and the virtual-topology paragraph follow the PR head: a grid cell is crust when more than half its markers are oceanic crust or sediment, serpentinite dehydrates back above 550 C, reach is counted in whole cells, and the rules read the element mean stress. The interface-friction exercise is removed; the one Part 2 exercise is now the serpentinite layer thickness at 0, 5 and 10 km, with 0 as has_serpentinization = no since a positive thickness is required. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: draw the subduction halo region as the poly file bounds it The refined-mantle halo polygon detoured through the slab end (nodes 11 and 10) and so omitted the wedge beyond the slab tip that segment 14-15 of poly-subduction.poly encloses. It now follows the outer boundary of the file, 12-8-9-13-15-14, with the slab drawn on top. Node coordinates, dip (30.00 deg) and thickness (7.00 km) were checked against the parsed file and already matched. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: tectonics Part 2 exercise and figures from the PR #110 reference runs Exercise 3 (serpentinite layer none / 5 km / 10 km) is filled from three 5 Myr runs on the PR #110 head binary. Without the layer the interface locks, the slab stalls near 75 km and necks off, the plates buckle and the run ends at 3.45 Myr in a remeshing failure; with 10 km the trench is 1.5 km shallower and eclogite drops by 40 %. The evolution and thickness figures are regenerated from those runs and restyled for the page: cropped to the slab, surface drawn, corner boxes with tip depth and marker counts, common 3.4 Myr comparison time. Reference-run numbers in the text follow the new figure. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: tectonics Part 2 figures with velocity arrows, remeshing history and a restart case The evolution and thickness figures gain time-averaged velocity arrows, the thickness figure spans the whole domain with its time-series row aligned to the panels above, and the reference-run numbers follow the figures. Two figures are new: the remeshing history of the reference run (elements by plastic-strain class with each remesh marked by its trigger, and cumulative counts by trigger) under a "How the mesh keeps up" subsection, and a troubleshooting case after Exercise 3 showing the locked no-serpentinite run restarted from its 3.0 Myr checkpoint with max_boundary_distortion halved to 0.5, which reaches 5 Myr with the physics unchanged. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: place the tectonics tutorial before GoSPL, fix its links, finish the thickness figure The tutorial moves to sidebar position 7, after rate-and-state friction, and the GoSPL page to 8. Prerequisites are rewritten around the pages that now precede it. Internal links become file-path links, since the installation page is published at /docs/usage and the bare URL form was a dead link on the deployed site. "What you will learn" covers phase changes, remeshing triggers and the restart case. The thickness figure shows the 5 and 10 km runs at 5 Myr and the failed run where it stopped, with its restart as a dashed history in the time series. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: tectonics Part 1 figures and Exercise 1 from runs of core-complex.cfg Three 100 kyr runs of examples/core-complex.cfg on an MMG build at the PR #95 merge commit back Part 1. The evolution figure shows plastic strain on the mesh beside strain rate with velocity arrows at 25, 50, 75 and 100 kyr, then the surface profiles and relief and fault-band dip against time (the hinge rolls from 46 to 35 degrees). Exercise 1 is filled with its own figure built on passive horizons from the coord0 field: the fault and the exhumation are identical at diffusivities of 1e-7, 1e-6 and 1e-5 m2/s, only the topography differs. Step 2 says the stock file stops at 50 kyr and gives the one-line change for 100 kyr; the cost paragraph quotes the measured run. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: add a geometry-and-mesh figure for core-complex.cfg, make the hinge animation a pure sketch A new SVG beside the Part 1 settings table shows the initial geometry and mesh of core-complex.cfg with the parameters that set them: the box, the refined zone with a 1 km element against a 3 km one outside, the planar weak-zone seed drawn from the formula in ic.cxx (dipping toward -x), the wall velocities, the free surface and the Winkler bottom. The rolling-hinge animation loses its setup annotations so the two no longer overlap, and moves up to the intro paragraph that names the mechanism. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: tectonics tutorial uses core-complex.cfg alone and states the MMG build requirement The Part 1 MMG box no longer lists how core-complex-mmg.cfg differs; it gives the size band the stock parameters imply and compares the Triangle and MMG builds of the same file. Prerequisites name the MMG build as the one requirement, and the build tip says Part 1 runs on either build while Part 2 needs MMG for its restored side walls. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: tectonics tutorial in one style with a progressive workflow; Exercise 2 on smallest_size Both parts now share four steps (geometry and mesh, configuration, run, what to look for) followed by exercises; a roadmap table after "What you will learn" shows what Part 2 adds to Part 1. The remaining TODOs are written out: continental temperature options, troubleshooting rows marked by what was observed, and a References section. Exercise 2 changes smallest_size to 1, the stock 0.01 and 0.001, all three run: at 1 the band cannot sharpen and the run is ten times faster; at 0.001 the mesh is the same as stock because the refinement target, not the floor, limits it, and the run takes 60 % more steps. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: Exercise 2 adds a stronger-target run and splits its figure; legends moved below the panels A fourth run, smallest_size = 0.001 with mmg_metric_refine_coeff = 10, shows that the floor alone did not limit the band: with the doubled target the band holds 584 elements at a median edge of 0.29 km against 0.38, so the two parameters act as a pair. The exercise table is trimmed to one point per row, the figure is split into the four meshes and a second figure with topography, time series and a stacked bar of elements by plastic-strain class, and the diffusivity figure's legend moves below its bottom row. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: tectonics tutorial ships a workshop package Adds static/downloads/des-tectonics/DES-Tectonics-Tutorial.zip (11 MB), a 90-minute class package like the RSF one: a Jupyter notebook that plots every figure of the tutorial from DES output, the configs of both parts and all exercise variants behind a one-command run.sh launcher, a short build recipe, and reference results of every run stripped to the fields the notebook plots so the class never waits for a run. The failed no-serpentinite run ships its frame-60 checkpoint so the restart troubleshooting case can be tried. tectonics.md links the package in a tip box after the introduction. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> * docs: move the GoSPL coupling tutorial after the tectonics tutorial The tectonics tutorial sits at position 7, after Rate-and-state friction; GoSPL coupling, which builds on models like these, follows it. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> --------- Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
This was referenced Oct 31, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
var.nremeshinstead, so any remesh now falls back to the instantaneous velocity for the next coupling event.docs/tutorial/couplinggospl.md: rewrites the "Remeshing between coupling events" known-limitation to describe the fix, matches the new log-line wording, notes the remesh fallback in the coupling-loop walkthrough, and drops the now-unnecessaryconda deactivatebuild step.Note: DynEarthSol #112 is still open upstream (not yet merged) as of this PR. The content here matches the PR as currently written; flagging in case it changes before merge.
Test plan
yarn buildsucceeds🤖 Generated with Claude Code
https://claude.ai/code/session_01MyXyjvBfCbpMgZ6ttvJXf3