| {"id": "formlabs-nylon12gf-tds", "title": "Formlabs Nylon 12 GF \u2014 TDS values and our measured baselines", "summary": "Published Fuse 1 TDS mechanical properties for Nylon 12 GF, next to what we measured on Fuse-printed control specimens and our best Inova batches. These are the optimization targets.", "content": "---\ntitle: Formlabs Nylon 12 GF \u2014 TDS values and our measured baselines\nsummary: Published Fuse 1 TDS mechanical properties for Nylon 12 GF, next to what we measured on Fuse-printed control specimens and our best Inova batches. These are the optimization targets.\n---\n\n# Formlabs Nylon 12 GF \u2014 TDS values and measured baselines\n\n## Published TDS (Fuse 1, Formlabs datasheet 2201635)\n\n| Property | Method | Value |\n|---|---|---|\n| Ultimate tensile strength | ASTM D638-14 Type I | 38 MPa |\n| Tensile modulus | ASTM D638-14 Type I | 2800 MPa |\n| Elongation at break | ASTM D638-14 Type I | 4% (X/Y), 3% (Z) |\n| Flexural modulus | ASTM D790-15 | 2400 MPa |\n\nSource: formlabs-media.formlabs.com/datasheets/2201635-TDS-ENUS-0.pdf\n\n## Our measured baselines (see `astm_query` for live numbers)\n\n- **Fuse-printed control specimens** (material_class `PA12GF_FL`, printed\n vertically, D638, n=5): tensile modulus **2599 \u00b1 98 MPa**. Below TDS \u2014\n consistent with the TDS quoting X/Y orientation while our controls were\n printed vertically (Z), the weak orientation in SLS.\n- **Best Inova batch to date** (batch H, D638, n=3): tensile modulus\n **2815 MPa** \u2014 at TDS level. Batches I/J/J_MB (later profiles) sit at\n 2245\u20132328 MPa.\n- Full progression A\u2192H spans 365 \u2192 2815 MPa as energy density and thermal\n settings improved; query `astm_query {group_by: \"profile\"}` for the\n profile-by-profile picture.\n- **Flexural (D790)** mirrors it: batch C 285 MPa \u2192 batch H **2272 MPa**\n (TDS 2400), controls 1950 MPa. Note: D790 moduli are chord values\n recomputed from load-deflection curves (span = 16 \u00d7 depth); values for\n very weak batches are lower bounds (crosshead compliance).\n\n## How to use these numbers\n\nTreat TDS values as the **target envelope** when optimizing Inova profiles\nfor PA12 GF (and as the pattern for any new powder: pull its TDS, put it in\nthis corpus, tune toward it). Compare like-for-like: orientation matters\n(X/Y vs Z), and our specimens' print orientation is recorded per batch in\nthe ASTM repo.\n"} |
| {"id": "fuse1-to-inova-translation", "title": "Fuse 1 \u2192 Inova MK1 parameter translation", "summary": "How to use Formlabs Fuse 1 material settings as starting points on the SLS4All Inova MK1 \u2014 machine differences, what transfers (normalized quantities), what doesn't (absolute settings).", "content": "---\ntitle: Fuse 1 \u2192 Inova MK1 parameter translation\nsummary: How to use Formlabs Fuse 1 material settings as starting points on the SLS4All Inova MK1 \u2014 machine differences, what transfers (normalized quantities), what doesn't (absolute settings).\n---\n\n# Fuse 1 \u2192 Inova MK1 parameter translation\n\nFormlabs publishes material settings and TDS values for the Fuse 1. We print\nthe same powders (e.g. Nylon 12 GF) on a SLS4All Inova MK1. **Absolute\nmachine settings do not transfer; normalized quantities do.**\n\n## Machine comparison\n\n| Property | Formlabs Fuse 1 | SLS4All Inova MK1 |\n|---|---|---|\n| Laser | 10 W ytterbium fiber | (fill in: see firmware config \u2014 not yet verified in this corpus) |\n| Laser spot size (FWHM) | 200 \u00b5m | (fill in) |\n| Layer thickness | 110 \u00b5m | 100 \u00b5m (all profiles to date) |\n| Build volume | 165 \u00d7 165 \u00d7 300 mm | see firmware `PrintableWidth/Height/Depth` |\n| Heating | quartz tube elements + PTC cartridges | halogen surface heating + zone heaters (powder/print chambers 1\u20134, beds) |\n\nSources: Formlabs Fuse 1 tech specs (formlabs.com/3d-printers/fuse-1/tech-specs).\n\n## What transfers between machines\n\n1. **Volumetric energy input, not laser power.** The Inova's profile knob is\n `LaserFillEnergyDensity` in **mJ/mm of scan path** (profiles are named by\n it: 14\u201340 mJ/mm tried so far, all at 100 \u00b5m layers). The Fuse 1's 10 W /\n 200 \u00b5m spot delivers energy per scan length that cannot be copied\n directly \u2014 different wavelength, spot, and scan speed. Treat energy\n density as the search variable on the Inova, using our own batch outcomes\n (see `astm_query` grouped by profile) as the map, not Fuse 1 settings.\n2. **Temperatures relative to material transitions.** PA12's melting point\n (~187 \u00b0C) and crystallization window are material properties. Our\n PA12 GF profiles hold the powder surface at 168\u2013176 \u00b0C\n (`SurfaceTarget` 168, `HeatingTargetPrint` 172,\n `BeginLayerTemperatureTarget` 176) \u2014 a few degrees below melt, the\n standard SLS sintering window. A new material's bed/chamber targets\n should be set the same way: from its DSC data or published sintering\n window, offset the same margins below melt.\n3. **Powder refresh ratio.** Material aging behavior transfers (it's powder\n chemistry, not machine). Formlabs specifies 30\u201350% refresh for their\n powders; our \"100% Recycled\" profile experiments track how the Inova\n tolerates deviation from that.\n4. **TDS mechanical values as targets.** ASTM D638/D790 results are\n machine-independent measurements. The Fuse 1 TDS values (see\n `formlabs-nylon12gf-tds`) are the benchmark our Inova profiles are tuned\n toward; batch H (36 mJ/mm era) already reached TDS-level tensile modulus.\n\n## What does NOT transfer\n\n- Laser power/speed settings (different laser physics entirely).\n- Layer thickness conventions (110 vs 100 \u00b5m \u2014 affects energy density per\n volume; when comparing mJ/mm figures across layer thicknesses, normalize\n to J/mm\u00b3: divide by hatch spacing \u00d7 layer thickness).\n- Chamber heating dynamics (heat-up rates, `HeatingMinimumTime`) \u2014 tune per\n machine.\n\n## Practical recipe for a new powder on the Inova\n\n1. Get the material's TDS + sintering window (melt point, recommended bed\n temp on any SLS machine \u2014 Fuse 1 settings are a fine source *for this*).\n2. Set surface/bed targets the usual margin below melt (compare our PA12 GF\n profile: melt 187 \u00b0C \u2192 surface 168\u2013176 \u00b0C, i.e. 11\u201319 \u00b0C below).\n3. Start energy density conservatively (our PA12 GF experience: 14\u201320 mJ/mm\n sinters but underperforms; 28\u201336 mJ/mm is the strong region; 40 mJ/mm\n explores the upper bound). For goal \"first print\": low-middle of that\n range with a small test object.\n4. Iterate using `astm_query` outcomes and `build_get` failure modes from\n prior builds as the evidence base.\n"} |
| {"id": "inova-profile-conventions", "title": "Inova print profile conventions", "summary": "How this lab's print profiles are named and structured, which JSON fields matter, and known naming pitfalls. Read before comparing profiles or proposing a new one.", "content": "---\ntitle: Inova print profile conventions\nsummary: How this lab's print profiles are named and structured, which JSON fields matter, and known naming pitfalls. Read before comparing profiles or proposing a new one.\n---\n\n# Inova print profile conventions\n\nProfiles live in the firmware (synced to `inova_print_profiles` in Postgres,\nfull JSON in the `profile` column). One profile \u2248 one experiment condition.\n\n## Naming\n\n`YYYY_MM_DD <energy> mJ/mm [variant] Formlabs PA12 GF` \u2014 the date is when\nthe profile was created, the energy is the `LaserFillEnergyDensity` value,\nvariants describe the deviation being tested (\"Higher Temp\", \"Extended\nCap\", \"25% Overlap\", \"100% Recycled\", \"1 Second Heat Time\").\n\n**Pitfall: names can lie.** The profile named \"2026_06_06 20mJ/mm\" actually\ncontains `LaserFillEnergyDensity: 15`. Always read the JSON field, never\ntrust the name \u2014 `db_query` on `inova_print_profiles` or\n`build_get.profile_key_params`.\n\n## Energy densities tried so far (all at LayerThickness 100 \u00b5m)\n\n14, 15, 20, 24, 28, 32, 36, 40 mJ/mm. Low end (14\u201320) sinters but\nunderperforms mechanically; 28\u201336 is the strong region; 40 explores the\nupper bound. `LaserFillEnergyDensity` is per mm of scan path \u2014 to compare\nacross layer thicknesses or hatch spacings, normalize to J/mm\u00b3.\n\n## Temperature fields (\u00b0C) for PA12 GF\n\n- `SurfaceTarget` / `SurfaceTarget2`: powder surface hold (168 / 172\n baseline; Target2 applies after `SurfaceTarget2Time`).\n- `HeatingTargetPrint` (172), `BedPreparationTemperatureTarget` (174),\n `BeginLayerTemperatureTarget` (176), `PrintCapTemperatureTarget` (174).\n- PA12 melt \u2248 187 \u00b0C; these sit 11\u201319 \u00b0C below (the sintering window).\n\n## Other fields that have mattered in experiments\n\n- `HotspotOverlapPercent` (25 baseline; the \"25% Overlap\" profile varies\n scan overlap), `OutlineCount` (2), `IsFillEnabled`.\n- `PrintCapThickness` / \"Extended Cap\" variants \u2014 powder above the last\n layer for thermal mass during cooldown.\n- `HeatingMinimumTime` (\"1 Second Heat Time\" variant collapses it),\n `PowderVolumePercent` (110 = 10% extra dose per recoat),\n `RecoaterPasses`, `RecoaterShakePercent`.\n- Durations are TimeSpan strings (\"00:45:00\"); factor-style runtime\n overrides multiply profile values (1.0 = neutral) \u2014 see AGENTS.md ground\n rules for the runtime knob semantics.\n\n## Linking profiles to outcomes\n\n`astm_query {group_by: \"profile\"}` gives mechanicals per profile;\n`builds_list {profile: \"36mJ\"}` finds the builds that ran it;\n`build_get` shows a build's `profile_key_params` and thermal tracking.\n"} |
|
|
