Skip to content
Converraby Avila LabsOpen workbench

Open-source REBCO magnet design

Explore your
REBCO winding.

Compare winding options against a fixed magnet specification.

Set the field, aperture, temperature, and screening limits. Converra records candidate costs, operating margins, and comparisons with your baseline.

Browser calculations run on your device

CONVERRA / DESIGN EXPLORATION01
Conductor-pack design explorationA schematic racetrack winding with nested conductor paths around an open aperture. Dimensions and winding choices are evaluated against fixed requirements. This is an illustration, not a simulated field plot.WINDING CHOICESFIXED REQUIREMENTSCOMPARE COST & MARGIN
You defineField. Aperture. Limits.
Converra exploresTurns. Tapes. Geometry.

Schematic winding illustration

Design workflow

Compare winding options.

Hold the declared requirements fixed as you compare turns, tape counts, and supported geometry. Keep the inputs and calculation record with each result.

01 / DEFINE

Set the requirements.

Record the field, aperture, operating point, baseline, material data, and cost basis for the comparison.

A versioned case
02 / EXPLORE

Search the winding choices.

Compare supported turn counts, tape counts, and geometry. Review candidates that pass, fail, or remain unresolved.

Candidate results and limits
03 / REVIEW

Inspect the evidence.

Review the cost ledger, material coverage, model versions, input identities, and acceptance checks before using a result.

A portable review record

Benchmarks and evidence

Read each result in scope.

Converra’s public benchmarks cover numerical checks, conductor comparisons, bounded searches, and workflow behavior. Their assumptions and outcomes differ; some comparisons remain inconclusive.

The 5 T, 80 mm demonstration includes a model-extended conductor dataset. At some geometries the declared baseline is not feasible or remains inconclusive, and refined acceptance does not support a general savings claim.

Read the evidence summary

Checks and limitations

Keep assumptions visible.

Measured data, model extensions, and synthetic prices have different evidence status. Converra records which basis applies to a result.

PASS

Met the declared checks.

FAIL

Violated a declared requirement.

INCONCLUSIVE

Available model or data could not resolve it.

NOT EVALUATED

The check was not performed.

Calculation records

Cases and result records identify inputs, datasets, model versions, checks, and costs. Export them for review or reproduction.

Screening is not qualification

A result applies to its recorded assumptions and sampling plan. Engineering, manufacturing, thermal, structural, protection, and quench review remain separate work.

Inspect the evidence summary

Get started

Choose your workflow.

Try the hosted workbench, download a desktop or CLI release, or build the current source. The handbook explains which features each workflow supports.

Research or collaboration question? Contact Avila Labs.

Before you begin

Common questions

Full evidence summary
Which magnet designs are supported?

Converra supports planar racetrack, circular, and piecewise line-and-arc windings. Non-planar helix cases can be screened when the case supplies a compatible Cartesian field map. General CAD geometry and D-shaped toroidal-field coils are outside the current built-in field-model scope.

What can I try in the browser?

The hosted workbench opens, revises, and searches cases in a background browser worker. Calculations run on your device. It supports CSV, TSV, and XLSX material-table imports and stores draft recovery in that browser. Folder libraries, directory bake-offs, watch mode, and queues require the desktop build.

Which version is in the desktop download?

The latest published desktop and CLI archives are v0.2.0. Those archives predate several workflows described in the current handbook, including spreadsheet imports, browser draft recovery, and scenario-study features. The hosted workbench and current source contain newer workflows; check the handbook for interface-specific availability and the release notes for the exact archive contents.

Does a passing result qualify a design for manufacture?

A pass means a candidate met the declared checks under the selected models and sampling plan. Review material coverage, model extensions, unresolved checks, and the separate engineering work required for structural, thermal, hydraulic, quench, and manufacturing decisions.

Where does design data go?

The browser workbench runs calculations locally, and the desktop and CLI run on your machine. You choose which case files, records, and reports to export.