# Studio walkthrough: screenshot record

These are current captures of the development application's real interface, including retained designs and studies. They are not rendered mockups. Refreshing the interface images does not refresh the underlying scientific benchmarks. Different views show different machines; the active design in the application header is not necessarily the source of a stored result.

| View | Source and context |
| --- | --- |
| Design | Working design `af50_14`: a YASA-style 50 mm, 14-pole / 12-slot machine. The wire-winding view shows the full machine with rotor discs hidden to reveal the stator and magnets. Geometry and feasibility warnings remain visible. |
| Simulate | The same working machine, configured for a native 3-D loaded study at 15 A RMS and 90 electrical degrees. The workload estimate indicates 57 solves. This is a configuration view, not a completed result or a runtime benchmark. |
| Sweep | Stored native 2-D sweep 22: 44 evaluated designs, 28 infeasible and 16 feasible, at 40 A RMS with a 26 V bus. Each candidate uses its own base speed and a single rotor angle. These are screening estimates, not measured prototype results. |
| Capability | Stored map 4: a GetDP 2-D, 20-point (4 by 5) map for a YASA-style 50 mm machine with four-turn round-wire windings, up to 212.13 A RMS and an 80 °C model condition, solved at 3,000 rpm. The displayed efficiency envelope uses a 48 V bus. Its source machine differs from the active design. There is no thermal-rating overlay; losses and efficiency are modelled, not hardware measurements. |
| Compare | Stored runs 102 and 93 for a Parviainen reference machine: 328 mm, 12 poles / 36 slots, 300 rpm and 8.4 A RMS. The native 3-D run uses 25 rotor angles; the quasi-3-D run uses 31. Run 102 does not resolve the ripple quantity. Differences in model and angular sampling remain part of the comparison. |
| Results | Stored native 3-D run 98: a one-sided 14-pole / 12-slot machine at 3,000 rpm, 40 A RMS and 49 rotor positions. The refreshed results interface shows the result library, key metrics and a torque waveform. Torque ripple is marked not assessed. |
| Visualize | Stored native 3-D run 106: signed axial flux density Bz, in tesla, on material surfaces of a partial sector. The rotor angle was not recorded, and the interface retains that qualification. The experimental 22-turn rectangular winding is infeasible at its recorded 80 A RMS; this is a field-inspection example, not a qualified operating point. |
| Report | A local HTML report preview generated from stored run 98, with an editable example title and the source study’s quality findings. Generating the document starts no solver job. |

The images are static files. The website does not connect to the application or to anyone's local jobs. Captures use a 1,600-pixel-wide dark-theme development view, with 1,000 or 1,200 pixels of height to retain plot axes and legends. Retained studies were loaded from a copied local database; the original data was not modified, and no new solver studies were needed for the capture session. Visible qualifications remain part of the examples, which do not establish a general accuracy range, measured efficiency or continuous rating.

The older `waveforms.jpg` remains available for the documentation's results-reading exercise and its annotated crops. The refreshed gallery results view is a separate asset, `results.jpg`.

The [reference-machine evidence register](reference-evidence.md) lists published comparisons and their conditions.

## Additional interface details

- [Geometry cross-section](design-section.jpg): the same AF50-14 working example as Design, with the radial/axial section shown to scale.
- [Torque-speed envelope](capability-envelope.jpg): map 4 with the same drive and temperature assumptions as Capability; electrical limits are shown without a continuous thermal overlay.
- [Electrical waveforms](electrical.jpg): run 98’s phase linkage and phase back-EMF at 3,000 rpm, with 49 rotor angles and RMS/peak conventions kept visible. The stored field files are unavailable for this run; these plots use its retained waveform data.

The application’s in-progress branding may differ from the final Axial Flux Studio name. Interface captures retain their original labels and result qualifications.