Metal & ceramic sintering filaments (FDM feedstocks)

Published · Researched 2026-09-20

Overview — what it is, best uses, key context

Metal/ceramic FDM filaments let an ordinary filament printer make real metal or ceramic parts: the filament is a highly loaded polymer–powder composite (typically ~80% metal powder in a polymer binder). You print a "green part" oversized, then debinding + sintering (catalytic debind, then furnace sintering near the metal's melting point) burns out the binder and fuses the powder into dense metal — with significant shrinkage. Best uses: small stainless/bronze/copper parts, tooling, jewelry, functional prototypes that need real metal properties — without a $100k+ metal printer. The green part is not a metal part. Debinding/sintering requires a furnace cycle (often outsourced to a service bureau) and the workflow has real cost, shrinkage compensation, and safety considerations. This does NOT overlap specialty-filaments.md (decorative metal-effect/metal-filled PLA that never becomes metal) — those are aesthetic filaments; these are foundry feedstocks.

Current 2026 products and brands

• BASF Forward AM Ultrafuse 316L (stainless steel): 3 kg $565.00 → $188.33/kg computed — matterhackers.com (fresh), obs. 2026-09-20; LulzBot listing $234.00 (spool size not stated — UNVERIFIED). ~80% 316L powder in polymer binder (figure varies by source: "more than 80%" vs "80%" — flagged). Catalytic debinding + sintering required.
• BASF Forward AM Ultrafuse 17-4 PH (precipitation-hardening stainless): 1 kg €161.70 (temporarily out of stock) / 3 kg €399.30 → €133.10/kg computed — anodas.lt (fresh), obs. 2026-09-20; US reseller FBRC8 1 kg $160.30 sale / $229.00 regular (167-day crawl — aging).
• The Virtual Foundry "Filamet" (shop.thevirtualfoundry.com, fresh 6-day crawl), obs. 2026-09-20 — the hobby-accessible line:
  • Copper: from $142.99 (size UNVERIFIED)
  • Bronze: 0.5 kg $179.99 → $359.98/kg computed
  • 316L stainless: 0.5 kg $210.99 → $421.98/kg computed
  • 17-4 PH: 0.5 kg $208.99 → $417.98/kg computed
  • M300 tool steel: 1 kg $332.99
  • Tungsten: $716.99 (size UNVERIFIED); Inconel 718: $294.99 (size UNVERIFIED); Titanium: $290.99 (size UNVERIFIED)
  • Ceramics: Zirconium Silicate from $186.99, Pyrex $88.99, Silicon Carbide $156.99, Porcelain from $87.99 (sizes UNVERIFIED)
• Ceramic sintering filaments are the same workflow with ceramic powder (zirconia-silicate, porcelain) — decorative/technical ceramic parts after kiln firing.

Key specs (documented)

What's needed (hardware)

• Hardened steel nozzle (metal powder is highly abrasive); print the green part oversized per the sintering shrinkage factor.
• Debinding + sintering: catalytic debind furnace cycle for BASF grades; sintering furnace (~1300 °C+ class for stainless). Most hobbyists outsource to a sintering service — factor this into part cost.
• Ventilation for binder burnout; standard metal-powder handling care (don't inhale dust).
• Patience: the full cycle is measured in days, not hours.

Brand/support reality

• BASF Forward AM (Ultrafuse): the industrial standard; best powder loading and sintering documentation; sold via MatterHackers (US, $188/kg at 3 kg), LulzBot, EU resellers (Anodas). Debind/sinter via BASF's partner network or third-party services.
• The Virtual Foundry: the hobby-accessible brand; smaller spools (0.5 kg), widest metal menu (copper, bronze, brass, aluminum 6061, iron, steels, tungsten, Inconel, titanium) plus ceramics; premium per-kg pricing ($360–422/kg at 0.5 kg) vs BASF bulk. Direct + MatterHackers.

Community verdict

Thin in this pass (FB zero-yield; no dedicated sintering thread pulled — coverage gap, recorded). General practitioner consensus from the category's public documentation: results are genuinely impressive when the sinter cycle is right (dense, machinable metal), but dimensional accuracy depends on shrinkage compensation and furnace control — this is a process, not a filament. Hobbyist outcomes vary widely; professional debind/sinter services are the difference between "metal part" and "expensive paperweight." Copper/bronze are the friendliest entry metals (lower sinter temps).

Pricing — observed 2026-09-20 (per kg as listed)

Per-kg ladder (normalized): BASF 17-4 PH €133 (3 kg) < BASF 316L $188 (3 kg) < VF M300 $333 (1 kg) < VF bronze $360 < VF 17-4 $418 < VF 316L $422 (0.5 kg). Bulk BASF is dramatically cheaper per kg than hobby-size Filamet.

Sources consulted

• Google web search 2026-09-20: matterhackers.com Ultrafuse 316L listing + MH Virtual Foundry category (fresh); anodas.lt Ultrafuse 17-4 PH 1 kg/2.2 kg listings (fresh); shop.thevirtualfoundry.com Filamet collection (6-day crawl); buy.lulzbot.com BASF 316L (image page).
• Facebook groups: zero relevant posts (2026-09-20 pass).
• Reddit/forums/YouTube: no sintering-specific thread pulled this pass (coverage gap, recorded).

Image notes

• ../../images/sintering-filaments-01.jpg — BASF Forward AM Ultrafuse 316L metal filament 3 kg spool (top3dshop; replacement — original zortrax.com URL failed to connect; substitution documented in manifest).
• ../../images/sintering-filaments-02.jpg — BASF 17-4 stainless filament printed gate-lock part (3dprint.com).
• ../../images/sintering-filaments-03.jpg — BASF Ultrafuse metal filament product image (crea3d). All decoded via PIL, converted to RGB JPEG (quality 88), verified non-corrupt.

Open questions / UNVERIFIED

• Exact sintering shrinkage factors per alloy — UNVERIFIED in this pass (needed for dimensional-accuracy claims).
• Powder-loading % for Ultrafuse 17-4 PH and Filamet grades — UNVERIFIED.
• Green-part print temperatures per BASF profiles — UNVERIFIED.
• Debind/sinter service pricing and turnaround — UNVERIFIED.
• Virtual Foundry "from" prices' spool sizes — UNVERIFIED (per-kg figures computed only where size stated).

Deepening pass (2026-09-20)

Social layer: Facebook groups zero-yield (8 keyword queries across "3D Printing" 17k and "3D Printing World" 10k — polypropylene, POM, sintering, PVA, BVOH, PEEK, ABS-like resin, silk PLA — none returned relevant posts). Threads keyword search: zero results for all queries. Instagram: no public topic search via the CLI — no IG signal, noted honestly. Signal came from the Metal FDM 3D Printing Group (FB group exists — "Metal FDM 3D Printing Group, by The Virtual Foundry"), retailer listings, and YouTube practitioner videos.

Critical correction — do NOT dry Filamet: Virtual Foundry explicitly warns that Filamet spools should not be dried — filament dryers can adversely affect the metal-loaded filament. This contradicts the general "dry your filament" instinct and any blanket drying guidance; it must be prominent in the review. (BASF Ultrafuse 316L guidance on drying was not re-verified this pass — keep the distinction.)

New product/process data (MatterHackers listings, Virtual Foundry):

• Powder loading: 316L Filamet = 85% metal + PLA; bronze 88–90%; titanium 78–82%; tungsten 91–93%. Filamet grades confirmed as a PLA-based binder system — consistent with the "prints like PLA" positioning.
• Print guidance: PLA-like profile, nozzle 205–235 °C, bed 40–50 °C, 0.6 mm stainless/hardened nozzle, and ~135% starting flow (the high metal loading needs the extra push).
• Sintering/shrinkage guidance (Virtual Foundry): ~7% shrink can yield ~80–85% density; ~20% shrink reaches low-90% density. Z shrinks slightly more than XY (gravity). Starting targets: ~5% isostatic shrink for copper/bronze, ~10% for steels. These are vendor workflow targets, not independent measurements — dimensional control remains the hard part.
• The two-track market is confirmed and stable: Virtual Foundry Filamet (print like PLA → debind → sinter, hobby-accessible green parts) vs BASF Ultrafuse 316L/17-4 PH (catalytic debinding, needs industrial partner — the professional track).

New failure modes: (1) drying metal filament in a standard dryer can damage it (Filamet); (2) wrong-nozzle wear — brass nozzles die fast on 85–93% metal loading, 0.6 mm hardened/stainless is specified; (3) density-vs-shrinkage tradeoff — chasing full density means accepting ~20% shrinkage and the dimensional pain that comes with it; (4) green-part fragility before debinding is under-discussed in vendor copy.

Brand vs brand: Filamet (hobby-accessible, PLA-binder, user-managed sintering via service/kiln, lower density ceiling) vs Ultrafuse (catalytic system, industrial debind/sinter partners, higher density, professional cost). No independent head-to-head density data found — the comparison remains workflow-tier vs workflow-tier.

Price re-verification (2026-09-20): no new cart prices moved this pass; first-pass figures stand as dated. The per-kg math remains brutal vs polymer filament — that's the category, not a finding.

Deepening pass 3 (2026-09-21)

New evidence (public search, observed 2026-09-21):

• Hobby-Machinist owner thread: one owner describes significant wasted time and material during process setup, severe shrinkage and porosity, and high furnace-temperature requirements. A sobering counterweight to vendor marketing — treat as one owner account.
• filament2print guidance: high metal loading makes feedstock fragile before melting (corroborates the brittle-feedstock note); keep the feed path straight; a filament warmer may help relax spool curvature; hardened nozzle, commonly ≥0.6 mm, with alloy-specific exceptions.
• Powder-loading figures vary across Virtual Foundry/reseller versions (316L/bronze/copper/tungsten values conflict with the existing item) — kept as version/source variation; do not merge into one definitive table.

Pricing dated 2026-09-21: no new per-brand prices beyond the existing item figures.

Persistent gaps: owner-reported sintering schedules remain absent (kiln-curve specifics are the single biggest evidence hole); no 2026 FB-group data; Virtual Foundry official TDS still not published.