Released · 0.2.11
Rotor modal demonstration
A bounded experimental serial route for rigid disk gyroscopic modal analysis, limited mode tracking, and Campbell results. It is not a general rotor framework.
Project roadmap
Future paths, guided by evidence.
This public roadmap shows the direction of travel through V1.0 and V2.0. It is a planning view, not a schedule: scope and order can change as verification results reveal what needs more work.
Published planning document · revision 1 · reviewed 2026-10-17. Download Markdown · Download source JSON
Release path
One visible path, with clear gates.
Versions are milestones for planning. A feature moves forward only when its own implementation and evidence support it.
- 01
- 02Next · 0.2.12
Stabilization and targeted corrections
Close current investigations, add focused regression coverage, and keep each capability claim matched to its validated route. Timing depends on results.
- 03Planned · 0.3.0
Broader engineering verification
Expand selected engineering cases and strengthen traceable verification and validation evidence, with incomplete or failed gates kept visible.
- 04Target · V1.0
A dependable open engineering core
Build a wider, clearly bounded FEM scope with reviewable V&V evidence and carefully verified nonlinear routes. The version number does not imply blanket qualification.
- 05Horizon · V2.0
Advanced mechanics and data driven methods
Explore broader nonlinear and rotordynamic models alongside PINNs and neural operators, each assessed against classical formulations and its own evidence gates.
Every route has its own formulation checks, validation evidence, and release decision. A roadmap item does not change today’s published capability status.
Published next plan · revision 1
0.2.12 · planning target: Stabilization and targeted corrections
Reviewed 2026-10-17. This persistent plan is the current planning record. Local drafts do not change it until a maintainer reviews and publishes an update.
- Version or horizon
- 0.2.12 · planning target
- Planning status
- Proposal
- Next proposed item
- Stabilization and targeted corrections
- Purpose
- Close current investigations, add focused regression coverage, and keep each capability claim matched to its validated route.
- Required evidence
- List affected routes, define focused regression checks, record results, and review whether public claims match their evidence.
- Dependencies
- Resolve the investigations relevant to the selected changes and identify the release scope before implementation.
- Open questions
- Which investigations are ready to close? Which route specific checks are still missing?
Branching workstreams
Four paths grow from the same evidence base.
These branches can progress at different speeds. Their shared foundation is transparent assumptions and route-specific verification.
Shared foundation · open source, traceable V&V, explicit boundaries
FEM & V&V
Broaden the verified core
Add engineering cases with declared assumptions, comparable reference results, regression checks, and visible limits for each route.
Evidence gate: comparable inputs, observables, and tolerances
Nonlinear mechanics
Extend one formulation at a time
Advance material, geometry, and contact routes incrementally. A new implementation stays experimental until its numerical checks and validation evidence are reviewed.
Evidence gate: formulation, convergence, and route specific tests
Rotordynamics
Grow from bounded demonstrations
Use the current rigid disk modal route as a starting point for additional rotor models, mode tracking, and response workflows.
Evidence gate: model assumptions and independent checks
PINNs & neural operators
Evaluate data driven solvers carefully
Research PINNs and neural operators as complementary methods. Compare scope, error behaviour, and generalization against defined reference problems.
Evidence gate: reproducible datasets and held out cases
Keep the next step concrete
Turn a direction into a reviewable plan.
Read the published plan, then prepare a proposal with its validation gate, dependencies and open questions.
Contribute to the planning
Prepare a reviewable proposal.
State the affected routes, expected outcome, dependencies and evidence needed. Your draft stays on this device until you submit it to GitHub.
What remains bounded
Some boundaries are features.
Leaving a capability unqualified is part of the trust model, not a hidden omission.
General nonlinear production usePublished evidence is route-specific; it does not qualify general large-strain plasticity, frictional contact, or finite sliding.
General HPC, GPU, or mixed scalingRecorded PETSc/MPI workloads do not establish hardware-independent performance, mixed distributed support, or nonlinear scaling.
Research kernelsResearch implementations remain visible for traceability and are not presented as normal-use or production-ready capabilities.