Shared-domain assembly and conformity

Last changes: Documentation changelog

Problem statement

The OCC route creates magnetic and air volumes, fragments them together, then extracts one shared-domain mesh with component and interface provenance.

Governing equations

(1)\[\Omega=\Omega_a\cup\bigcup_i\Omega_i.\]

Symbols and SI units

Token

Meaning

SI unit

\(\Omega\)

shared computational domain

\(\mathrm{m^3}\)

\(\Omega_a\)

airbox volume

\(\mathrm{m^3}\)

\(\Omega_i\)

magnetic component volume

\(\mathrm{m^3}\)

Assumptions and validity

The union states the OCC assembly domain. Conformity follows the source’s occ.fragment realization, not merely the value of a compatibility policy.

Python API

Complete public signature and IR matrix

The following rows are the exhaustive public-signature contract for this page; each row mirrors one public_api.parameters entry in the source map.

Python

Type

Default

SI unit

Validation

Meaning

Backend support

ProblemIR

study.universe(mode=…)

str | None

None

\(1\)

forwarded to StudyUniverseConfig; omitted retains current/default mode

domain mode

FEM CPU/GPU capability-gated

runtime_metadata.study_universe.mode

study.universe(size=…)

Sequence[float] | None

None

m

coerced with as_vector3; explicit dimensions must be positive when realized

outer dimensions

FEM CPU/GPU capability-gated

runtime_metadata.study_universe.size

study.universe(center=…)

Sequence[float] | None

None

m

coerced with as_vector3

outer center

FEM CPU/GPU capability-gated

runtime_metadata.study_universe.center

study.universe(padding=…)

Sequence[float] | None

None

m

coerced with as_vector3; domain validation governs signs

directional padding

FEM CPU/GPU capability-gated

runtime_metadata.study_universe.padding

study.universe.mesh(cell_size=…)

Sequence[float] | None

None

m

positive 3-vector; mutually exclusive with FEM controls

structured-grid cell size branch

FDM only; not applicable to this FEM assembly page

runtime_metadata.common_fdm_cell_size

study.universe.mesh(hmax=…) / study.universe.mesh(maximum_element_size=…)

float | None

None

m

positive; canonical maximum size wins over alias

airbox coarse target

FEM CPU/GPU capability-gated

runtime_metadata.study_universe.airbox_hmax

study.universe.mesh(hmin=…) / study.universe.mesh(minimum_element_size=…)

float | None

None

m

positive and no greater than numeric maximum; canonical minimum size wins over alias

airbox lower clamp

FEM CPU/GPU capability-gated

runtime_metadata.study_universe.airbox_hmin

study.universe.mesh(growth_rate=…) / study.universe.mesh(maximum_element_growth_rate=…)

float | None

None

1

finite positive at most 2.5; canonical growth rate wins over alias

airbox growth target

FEM CPU/GPU capability-gated

runtime_metadata.study_universe.airbox_growth_rate

study.universe.mesh(grading=…)

str | None

None

1

forwarded without vocabulary validation by this handle

airbox grading request

FEM CPU/GPU capability-gated

runtime_metadata.study_universe.airbox_grading

generate_shared_domain_mesh_via_occ(geometries, hmax=..., order=1, airbox=..., options=None, object_regions=None) is internal. Public authoring is split between study.universe(mode, size, center, padding) and separate study.universe.mesh(...); StudyUniverseHandle.__call__ has no airbox argument.

Python

Type

Default

SI unit

Validation

Meaning

Backend support

ProblemIR

study.universe.mode

`str

None`

None

\(1\)

forwarded to StudyUniverseConfig

domain mode

FEM CPU/GPU capability-gated

study.universe.size

`Sequence[float]

None`

None

\(\mathrm{m}\)

as_vector3; explicit dimensions must be positive when realized

outer dimensions

FEM CPU/GPU capability-gated

study.universe.center

`Sequence[float]

None`

None

\(\mathrm{m}\)

as_vector3

outer center

FEM CPU/GPU capability-gated

study.universe.padding

`Sequence[float]

None`

None

\(\mathrm{m}\)

as_vector3; domain validation governs signs

directional padding

FEM CPU/GPU capability-gated

study.universe.mesh.maximum_element_size

`float

None`

None

\(\mathrm{m}\)

positive; wins over hmax

airbox coarse target

FEM CPU/GPU capability-gated

study.universe.mesh.minimum_element_size

`float

None`

None

\(\mathrm{m}\)

positive and no greater than numeric maximum; wins over hmin

airbox lower clamp

FEM CPU/GPU capability-gated

study.universe.mesh.maximum_element_growth_rate

`float

None`

None

\(1\)

finite positive at most 2.5; wins over growth_rate

airbox growth target

FEM CPU/GPU capability-gated

study.universe.mesh.grading

`str

None`

None

\(1\)

forwarded without vocabulary validation by this handle

grading request

FEM CPU/GPU capability-gated

# %%
import fullmag as fm

# %%
nm = 1.0e-9
study = fm.study("assembly_contract")
study.engine("fem")
study.device("cpu", precision="double")
study.mode("strict")
study.universe(
    mode="manual",
    size=(320 * nm, 200 * nm, 120 * nm),
    center=(0.0, 0.0, 0.0),
    padding=(0.0, 0.0, 0.0),
)
study.universe.mesh(maximum_element_size=40 * nm, minimum_element_size=5 * nm, maximum_element_growth_rate=1.4)
film = study.geometry(fm.Box(size=(100 * nm, 50 * nm, 5 * nm), name="film"), name="film")
film.mesh(maximum_element_size=8 * nm, minimum_element_size=4 * nm)
film.Ms = 800.0e3
film.Aex = 13.0e-12
film.alpha = 0.02
film.m = fm.texture.uniform(1.0, 0.0, 0.0)
study.exchange()
study.demag(realization="poisson_robin")
study.stages.add_relax(stage_id="equilibrium", dt=5.0e-13, max_steps=1)

ProblemIR

Geometry, universe airbox and mesh policy lower as requested intent. Component volume tags, interface surface tags and final cell blocks are realized artifacts produced after OCC fragmentation.

Round-trip and failure semantics

Requested intent contains geometries, airbox and regions. Resolved execution contains OCC tags and extracted topology. Validation errors include empty geometry, missing airbox, invalid region identity/owner and non-contained conformal region. Unsupported combinations fail before extraction; they are not represented as conformal success.

Discrete realization

The builder creates an FEM mesh artifact for CPU or capability-gated GPU solver lanes. FDM CPU/GPU are not applicable.

Implementation mapping

StudyUniverseHandle and GeometryMeshHandle own authored intent; _build_problem and Problem.to_ir lower metadata and assets; generate_shared_domain_mesh_via_occ owns OCC fragmentation; the domain asset pipeline records build provenance; planner/runtime consume the resulting topology under separate compatibility and execution gates.

Validation

Inspect final region/interface tags, non-empty volumes and conformity diagnostics. Runtime solver/device execution remains unverified by source inspection.

Limitations

The source map documents OCC assembly, not a general nonconforming coupling path.

Scientific bibliography

Geuzaine and Remacle, IJNME 79 (2009), doi:10.1002/nme.2579.

Contract source-code index

ID

Path

Symbol

Responsibility

Evidence

occ_shared

packages/fullmag-py/src/fullmag/meshing/_gmsh_occ.py

generate_shared_domain_mesh_via_occ

OCC fragment and shared extraction

source-inspected

public_study

packages/fullmag-py/src/fullmag/world.py

study

public study entry point

source-inspected

universe_authoring

packages/fullmag-py/src/fullmag/world.py

class StudyUniverseHandle

public universe authoring

source-inspected

mesh_authoring

packages/fullmag-py/src/fullmag/world.py

class GeometryMeshHandle

public object-mesh authoring

source-inspected

problem_lowering

packages/fullmag-py/src/fullmag/world.py

_build_problem

builder-state lowering

source-inspected

problem_ir

packages/fullmag-py/src/fullmag/model/problem.py

class Problem

ProblemIR and geometry-asset serialization

source-inspected

domain_realization

packages/fullmag-py/src/fullmag/meshing/asset_pipeline.py

realize_fem_domain_mesh_asset_from_components_with_report

shared-domain realization and report

source-inspected

build_report

packages/fullmag-py/src/fullmag/meshing/mesh_build_report.py

_build_shared_domain_build_report

realized method/fallback provenance

source-inspected

planner

crates/fullmag-plan/src/lib.rs

plan

ProblemIR planning and compatibility

source-inspected, runtime-unverified

runtime

crates/fullmag-runner/src/lib.rs

run_planned_problem

planned runtime dispatch

source-inspected, device-unverified

Last changes: 12:31 24.08.2026

A conforming shared mesh represents each physical interface once, with common facet nodes referenced by adjacent regions. This is the default contract for coupled FEM operators in Fullmag.

Implementation status

Conforming tetrahedral shared-domain assembly is source-backed. Nonconforming mortar/contact coupling is not implied by ordinary mesh import or coincident surfaces.

Scope and purpose

Use this page when two materials touch, a magnet is embedded in an airbox, or several geometry objects participate in one FEM solve. The objective is a partition—not overlapping meshes and not two independently meshed surfaces that merely occupy the same coordinates.

Scientific and numerical model

Scientific invariants

A finite-element mesh is not only a visualization asset. It defines the trial/test spaces used by exchange, anisotropy, DMI, magnetostatic and dynamic operators. The following conditions are therefore part of the numerical contract:

  1. Every magnetic volume has an unambiguous region marker and every exterior-air volume has the canonical air role.

  2. Interfaces used by coupled operators are conforming, or an explicitly supported nonconforming coupling operator is selected. Fullmag’s ordinary shared-domain path expects conformity.

  3. Cell orientation is valid: the element mapping has a positive Jacobian at all required evaluation points. Inverted or collapsed cells are build failures, not warnings to ignore.

  4. Requested topology, polynomial order, layer count and mesh-size controls are compared with the realized mesh. A topology change is legal only when the build mode permits fallback and the report names the actual method and reason.

  5. Mesh convergence is assessed on physical observables—energy, average magnetization, switching field, eigenfrequency, linewidth or field error—not only on element count.

For exchange-dominated variation, a useful starting scale is the magnetostatic exchange length

(2)\[\ell_{\mathrm{ex}}=\sqrt{\frac{2A}{\mu_0M_s^2}}.\]

Using an element size below roughly one half of the smallest relevant magnetic length scale is a common initial choice, not a proof of convergence. Curved boundaries, surface charges, DMI, defects, interfaces and through-thickness modes can demand a smaller local size.

For an internal facet \(F=\partial K_r\cap\partial K_s\), conforming topology means both cells reference the same global node IDs for \(F\). The adjacency degree is two for an ordinary manifold internal facet and one for an exterior facet. A discrete continuous field then has one nodal value on the interface, while material coefficients can remain region-dependent.

The assembly pipeline performs geometry boolean fragmentation before meshing, assigns physical groups, extracts typed cells/facets, reconstructs adjacency and derives roles such as outer, interface, periodic and provisional. Orphan/nonmanifold facets are reported explicitly.

Selection guide

Use case

Recommended choice

Reason

Magnet embedded in airbox

conforming fragment-and-mesh

Required by ordinary scalar-potential assembly

Two magnets separated by air

one shared exterior domain

Cross demag field is represented continuously

Two touching material regions

shared interface + distinct region markers

Coefficients can jump while topology remains conforming

Independent imported meshes

remesh/merge with certificate

Coincident coordinates alone are not conformity

Parameters

Python / IR key

Unit

Default

Validation

Numerical effect

enforce_conforming

1

True

Boolean

requires shared-node interfaces

interface_hmax_factor

1

0.5

0 < value <= 1

refines shared interfaces relative to object hmax

airbox_hmax_factor

1

3.0

positive

coarse exterior target relative to object size

region marker

1

generated

unique semantic mapping

selects material/physics coefficients

boundary marker/role

1

generated

complete and unambiguous

selects outer/interface/periodic conditions

geometry tolerance

m

backend derived

scale-aware

controls boolean and coincidence classification

Python API

Complete Python example

import fullmag as fm

nm = 1.0e-9
study = fm.study("shared_domain_reference")
study.engine("fem")
study.device("cpu", precision="double")
study.mode("strict")
study.universe(
    mode="manual",
    size=(700 * nm, 500 * nm, 260 * nm),
    center=(0.0, 0.0, 0.0),
    padding=(0.0, 0.0, 0.0),
)
study.universe.mesh(
    minimum_element_size=12 * nm,
    maximum_element_size=90 * nm,
    maximum_element_growth_rate=1.5,
    grading="geometric",
)

left = study.geometry(
    fm.Box(size=(180 * nm, 80 * nm, 10 * nm), name="left_geom").translate(
        (-120 * nm, 0.0, 0.0)
    ),
    name="left",
)
right = study.geometry(
    fm.Box(size=(180 * nm, 80 * nm, 10 * nm), name="right_geom").translate(
        (120 * nm, 0.0, 0.0)
    ),
    name="right",
)
for body, direction in ((left, (1.0, 0.0, 0.0)), (right, (0.0, 1.0, 0.0))):
    body.mesh(
        mesh_strategy="free_tetrahedral",
        minimum_element_size=4 * nm,
        maximum_element_size=8 * nm,
        interface_maximum_element_size=6 * nm,
        interface_thickness=15 * nm,
        transition_distance="airbox_boundary",
        transition_growth=1.4,
        order=1,
        compute_quality=True,
    )
    body.Ms = 800.0e3
    body.Aex = 13.0e-12
    body.alpha = 0.02
    body.m = fm.texture.uniform(*direction)

study.exchange()
study.demag(realization="poisson_robin")
study.build_domain_mesh()
study.stages.add_relax(
    stage_id="equilibrium",
    algorithm="llg_overdamped",
    dt=5.0e-13,
    tolA=1.0e-4,
    max_steps=20_000,
)

Control Room workflow

  1. Apply every object mesh policy and the Universe airbox policy.

  2. Use Apply & Build Shared-Domain Mesh rather than relying on object previews.

  3. In Mesh Details, inspect region inventory, interface inventory and facet-role counts.

  4. Check conformity/orphan/nonmanifold diagnostics and the scoped quality report for each side of every interface.

  5. Confirm that material and physics selections resolve to the intended region markers before run.

Conformity certificate

Require a complete cell-to-facet adjacency, exactly one geometric interface role per intended pair, zero unmatched/coincident duplicates, positive cell Jacobians and complete region ownership. For an interface field, compare traces/normals from both sides. For demag, check continuity of the scalar-potential representation and the expected normal-flux jump from magnetization.

Verification, quality and provenance

After every build, inspect the realized resource rather than assuming that the authored request was applied. The production check is:

  • geometry and mesh revisions match the current model;

  • requested and realized discretization/topology/order are recorded;

  • node, element and boundary-facet counts are nonzero for every required region;

  • region and boundary markers cover the complete topology;

  • inverted and degenerate element counts are zero;

  • interface diagnostics report no orphan, coincident, nonmanifold or unmatched facets;

  • local size distributions are consistent with the intended edge/interface/core grading;

  • any fallback or degradation has an explicit reason and an actual method;

  • a mesh-refinement sequence demonstrates convergence of the scientific observable.

MeshQualityReport exposes signed inverse condition number (SICN), gamma/radius quality, volume statistics and optional per-element arrays. The source constants gamma_min=0.08 and SICN p05=0.1 are implementation gates for named report paths; they are not universal physical acceptance thresholds for every element family or study.

Diagnostics and failure semantics

  • Adjacency degree 0 indicates an orphan facet; >2 indicates a nonmanifold facet unless explicitly modeled.

  • Two facets with different nodes but coincident coordinates are duplicate/nonconforming.

  • Boolean fragmentation can split one authored surface into several entities; semantic recovery must retain ownership.

  • Overlapping volume cells or ambiguous region markers are blocking.

  • A valid object mesh can become invalid after shared boolean assembly; inspect the final mesh.

Where this is implemented

Responsibility

Repository source

Stable owner / symbol

OCC shared assembly

packages/fullmag-py/src/fullmag/meshing/_gmsh_occ.py

geometry fragmentation and physical groups

Shared generator

packages/fullmag-py/src/fullmag/meshing/_gmsh_generators.py

shared-domain generation

Facet adjacency/roles

packages/fullmag-py/src/fullmag/meshing/_gmsh_extraction.py

_derive_facet_roles

Orphan diagnostics

packages/fullmag-py/src/fullmag/meshing/_gmsh_selectors.py

collect_orphan_entity_diagnostics

Conformity report

packages/fullmag-py/src/fullmag/meshing/mesh_build_report.py

conformity certificate

Mesh summary API

crates/fullmag-api/src/router_v2/handlers/meshing/mesh.rs

shared-domain summary/manifest

Implementation map reviewed against commit 5db00ccf0113b9756fec2d46feb36ade762b12c2 on 2026-08-24.

References

  • C. Geuzaine and J.-F. Remacle, “Gmsh: a three-dimensional finite element mesh generator with built-in pre- and post-processing facilities,” International Journal for Numerical Methods in Engineering 79 (2009), 1309–1331, doi:10.1002/nme.2579.

  • C. Abert, “Micromagnetics and spintronics: models and numerical methods,” European Physical Journal B 92, 120 (2019), doi:10.1140/epjb/e2019-90599-6.

  • Gmsh reference manual, mesh algorithms, size fields, extrusion and physical groups: gmsh.info/doc/texinfo.

Source-code index

  • Python contract source: packages/fullmag-py/src/fullmag/model/discretization.py and packages/fullmag-py/src/fullmag/world.py, where applicable. Runtime realization is owned by the relevant backends/fdm or backends/fem implementation; the page must not claim a symbol not named in its implementation mapping.