Sand 3D Printer vs Metal 3D Printer: Which One Should a Foundry Buy?

📅 September 24, 2026 💬 By Jeff

A sand 3D printer and a metal 3D printer are not rival versions of the same machine, and the budget question is not “which one do I buy”. One prints molds; the other prints parts. A metal 3D printer (SLM/DMLS) melts metal powder into finished parts, practical up to roughly a 500 mm build envelope, at a per-kilogram cost that stays high even at volume. A sand 3D printer prints molds and cores that you pour in your own foundry, producing real castings in any castable alloy, up to 4000×2000×1000 mm on the J4000Pro — and larger still with segmented tooling. The decision rule is short: if the drawing calls for a casting, the alloy is castable, or the part is bigger than half a meter, the sand route wins; if the part is small, high-value titanium or nickel aerospace work with topology-optimized geometry, and no casting capability exists on site, metal AM wins outright. And many foundries that run the numbers honestly end up buying both.

Foundry production floor with a sand 3D printer

The Decision Rule

The machine that wins is the one matched to the part, not the one with the more impressive spec sheet. A sand printer and a metal printer cost similar money and both carry the label “3D printer”, but they buy completely different capabilities, so the decision collapses to three questions: how big is the part, what alloy does the drawing specify, and is the final part supposed to be a casting?

Part larger than roughly 500 mm, or a castable alloy (gray and ductile iron, steel, aluminum, bronze, nickel-base cast grades): buy the sand printer, feed your existing melt shop, and print molds and cores.
Small, high-value parts in powder-only alloys such as Ti-6Al-4V or nickel superalloys, topology-optimized or lattice geometry, and no foundry on site: the metal printer is the honest answer.
Both in the mix: buy both, or start with the route that matches the majority of your jobs and outsource the rest until volume justifies the second machine.

The most common mistake is comparing spec sheets — build volume against build volume, laser power against print speed — as if the two machine classes were the same kind of purchase. A sand printer with 4000 mm of build volume and an SLM machine with 400 mm of build volume are not competing on size; the first is measured in molds per day, the second in kilograms of finished part per day. Comparing them on print speed is like comparing a core shooter with a lathe. The honest comparison is on the output: castings per week versus finished metal parts per week, and the cost per part each route delivers at your quantities.

The rest of this guide walks each of those criteria in order: the budget logic, how each process works, a decision table, the cases where metal AM wins outright, the cases where sand printing wins, and the hybrid floor where both machines earn their keep.

The Real Question: Two Routes, One Budget

The confusion is understandable, because both purchases sit in the same budget line, cost six figures, and get the same label — “3D printer for metal”. A foundry director with one line in the capital plan sees two machines and assumes they are alternatives. They are not. They sit on opposite sides of the casting supply chain, and the one you need is decided by what the melt shop can and cannot do.

Route A — metal AM — makes the machine the foundry. Powder goes in, finished parts come out. You need an inert-gas chamber, powder handling and sieving, stress relief or hot isostatic pressing, and CNC finishing, but no melt, no pouring, no shakeout. Route B — sand 3D printing — replaces the pattern shop instead: molds and cores come out of the printer and go straight onto the existing molding line, where the foundry already knows how to make money per kilogram.

Budget the same way for either route: the machine is the easy line, and the costly line is everything around it. The sand route needs sand handling, reclamation, and an operator who can read a core print; the metal route needs powder management, sieving, an argon supply, heat-treatment capacity, and stricter safety procedures for reactive powders such as titanium. Shops that forget these second-order costs misjudge the payback of both routes, which is exactly why the per-part math in the cost guides matters more than the sticker price.

The split between the routes is driven by the alloy book.

If the foundry pours iron, a metal printer is dead weight: there is no SLM process for gray or ductile iron. If the foundry is an aluminum or investment shop, the metal route’s alloy range — aluminum, stainless, titanium, nickel — becomes a real argument. The market context is the same everywhere: from EFF-member jobbing foundries in Europe to the small iron shops that dominate the JFS membership in Japan, where three-quarters of iron foundries employ 30 people or fewer, the order book is low-volume, high-mix, core-heavy. That mix is exactly what patternless mold printing feeds, which is why the sand route carries most of the volume in real foundry buying decisions — while the metal route keeps winning the small, high-value part.

How Each Route Works: Molds vs. Parts

Sand 3D printing is binder jetting. A recoater spreads a layer of sand roughly 0.28–0.4 mm thick, the printhead deposits binder where the mold or core must be solid, and the process repeats layer by layer until the tool is complete. Loose sand is then removed, the mold is assembled, and metal is poured. The result is a mold or core set that is dimensionally faithful to the CAD file, including complex internal cores that no core box could draw. The full step-by-step is in our how sand 3D printing works guide.

Metal AM via laser powder bed fusion (SLM/DMLS) is a different machine class. A laser melts each layer of powder inside an argon-filled chamber; the part grows inside the powder bed; overhangs need printed supports; and the finished part carries residual stress, so it typically needs stress relief, sometimes hot isostatic pressing, then CNC finish. Metal binder jetting is a third variant — a green part that goes through a sintering furnace, closer to powder metallurgy than to welding — and the choice between those two metal routes is covered in our metal binder jetting vs. SLM guide.

A typical sand route job is short: the engineer exports the mold and core geometry as a printable file, the printer runs it overnight, and the mold is assembled and poured within days of the CAD being final — the same geometry that would have waited weeks on a patternmaker. The printed core also arrives as one piece where a conventional core box would force the core to be built from several sand pieces, which is how printed cores win on both lead time and dimensional consistency.

That difference — tool versus part — decides everything downstream.

The sand route’s output is a tool that makes castings, so it inherits the economics and metallurgy of casting. The metal route’s output is the part itself, so it inherits the economics and metallurgy of powder metallurgy: expensive input powder, slow build rates, and per-kilogram costs that stay high at any quantity.

The Decision Table

Put the two routes side by side on the seven dimensions that actually decide the purchase — not build-speed marketing numbers, but the constraints a foundry director lives with:

Dimension	Sand 3D printer + casting	Metal 3D printer (SLM/DMLS)
Part size	Molds up to 4000×2000×1000 mm (J4000Pro); larger via segmented molds	Typical envelope under 500 mm; larger formats get exponentially more expensive
Final part properties	Real castings: cast microstructure, per-spec casting alloy properties, iron damping, pressure-tightness	Near-full-density powder-metallurgy structure; anisotropy and build orientation matter
Material choice	Any castable alloy: gray/ductile iron, steel, aluminum, bronze, nickel-base cast grades	Powder alloys only: Ti-6Al-4V, nickel superalloys, 316L/17-4PH, AlSi10Mg, tool steels; no cast iron
Cost per part curve	High fixed cost, low per-mold cost; falls with series volume and amortization	High per-kg cost that stays high at volume; machine time and powder dominate
Batch path	Prototype → pilot → series; switch to a conventional pattern when volume justifies	One-off and small-batch; per-part cost roughly flat with quantity
Infrastructure needed	Existing foundry: melt, pour, shakeout, reclamation (95%+ sand reuse)	Inert-gas chamber, powder handling, stress relief/HIP, CNC finish; no melt shop
Lead time	Mold in days after CAD, pour same week; printer delivery 6–8 weeks (J1800)	Build in days, plus powder, heat treatment, support removal, and machining

Read the table the way a buyer should: the sand column is about leverage — one machine, unlimited mold geometry, feeding an existing melt shop — and the metal column is about directness, powder to part with no casting step. Both are real; they just answer different questions.

When Metal AM Genuinely Wins

Being honest about this matters, because 3DPTEK sells both lines. There are parts where a metal printer is not just competitive — it is the only rational answer.

First, titanium and nickel aerospace work. Machining a titanium bracket from billet discards most of the material; SLM builds a near-net-shape part and cuts the buy-to-fly ratio dramatically, and at small quantities that wins on cost and material utilization. Aerospace primes run SLM for brackets, housings, and heat exchangers in Ti-6Al-4V and nickel superalloys at sizes that fit comfortably in a sub-300 mm envelope.

Second, topology-optimized and lattice geometry. Casting and machining are shape-bound; metal AM can print internal lattice, organic fillets, and cooling channels that collapse under conventional processes. For lightweighting-critical parts where the geometry is the whole point, metal AM is the only route that delivers the design as drawn.

Third, the shop with no foundry. An OEM, a machine builder, or a medical device firm without casting capability can put a metal printer on the floor and make production metal parts in-house. There is no melt shop to feed, so the sand route has nothing to offer them — the direct metal route is their digital manufacturing.

And fourth, small high-value series: tooling inserts with conformal cooling channels, repair builds on expensive components, and one-off engineering parts where powder cost is irrelevant next to lead time. The metal route also wins on detail and surface complexity in small parts, where casting would need draft and minimum wall thicknesses the design cannot afford. For these jobs, the metal machine is the right purchase, and the cost structure is laid out honestly in our metal 3D printer cost guide.

When Sand Printing Wins

Flip the coin, and the sand route wins on the dimensions where foundries actually make money: size, alloy range, and the batch path from prototype to series.

Size first. The largest practical metal build envelope is typically under 500 mm, and going bigger multiplies the machine price. A sand printer starts at 1800×1000×800 mm on the J1800 and goes up to 4000×2000×1000 mm on the J4000Pro, with molds segmented and assembled for castings beyond even that. A wind-turbine housing, a machine-tool bed, a GIS enclosure — these are sand-printed mold jobs, full stop.

Alloy range second. Casting covers the entire alloy universe, including cast iron, which no metal printer will ever produce. Gray iron’s graphite structure gives it vibration damping; ductile iron gives castings ductility; steel, aluminum, bronze, and nickel-base cast grades all pour into printed molds. The question of which alloys actually work with printed molds is covered in our what can you cast with 3D printed sand molds guide. And the parts that come out are real castings — cast microstructure, per-spec mechanical properties, pressure-tightness — not powder-metallurgy approximations.

Batch path third. A sand printer runs the full prototype-to-production ladder: print one mold, validate the part, print pilot molds for first articles, then either keep printing production molds or hand the job to a conventional pattern once volume justifies it. That pattern is the cost the printer eliminates at low volume — a machined pattern runs roughly €3,000–15,000 and takes 4–8 weeks — so the printer pays for itself on exactly the jobs foundries quote every week. Sand reclamation keeps the consumable side honest at 95%+ reuse.

The industry evidence is not hard to find. BMW’s Landshut plant has printed sand cores for engine components; the Toyota supply chain uses printed cores to consolidate water jackets and oil galleries; pump and valve makers cast impellers and valve bodies that would never be SLM’d in production. Walk INTERMOLD Nagoya and the foundry aisles are full of patternless mold machines serving exactly these shops. The limits are real too — surface finish, minimum wall thickness, and draft all have floors, detailed in our sand 3D printing limitations guide — but for castable parts the sand route is the volume answer.

The Hybrid Reality: Both on One Floor

3DPTEK sells the J-series sand printers and the metal 3D printer line (SLM and metal binder jetting), and the honest answer for a growing share of foundries is that both belong on the same floor. The metal machine takes the small, high-value parts in powder alloys; the sand printer takes the large and castable work; and the metal machine also makes foundry tooling — gating, chill inserts, fixtures — that improve the casting process the sand printer feeds. Each route earns its keep on different parts, and the mix is decided by size, alloy, and quantity, not by brand loyalty.

If you are not sure which route dominates your mix, sequence the purchase instead of guessing. Run the sand route first when your parts are castable and a melt shop exists, because it touches the largest share of the order book and the utilization risk is lower; run the metal machine first when your work is small, high-value powder-alloy jobs and there is nothing to pour. And both routes can be validated without buying anything: print molds through a sand casting service and parts through a metal printing service, collect real costs on your own parts, then let the data justify the machine.

If you are building the business case, our sand 3D printing cost guide and metal 3D printer cost guide give the full numbers for both routes, and the product line pages list the machine classes available today. For most foundries the practical start is one machine matched to the majority of the order book, with the second justified by the parts the first cannot touch.

Request a formal quotation with your part drawings and annual volumes, and we will route each part — casting or direct metal — and quote the machine that matches. That is the decision rule in practice: let the parts decide.

Written by the 3DPTEK engineering team — the same team that operates our own foundries and sand 3D printers in daily production. About 3DPTEK.

Both routes in practice: the 3DPTEK J2500 for sand molds and cores, and the AFS-M120 or 3DPTEK J400P for direct metal parts.

Frequently Asked Questions

Can a sand 3D printer replace a metal 3D printer?

No — they are different machines doing different jobs. A sand 3D printer makes molds and cores that you pour to produce castings; a metal 3D printer makes finished metal parts directly. A sand printer replaces pattern tooling and core boxes, not the part-making step, so it only makes sense where a foundry and a melt shop exist. If you need a finished small part in a powder alloy and have no casting capability, a metal printer is the tool; if you need a casting, the sand printer is the tool.

Which is cheaper: sand printing or metal printing?

Both are six-figure capital purchases, so the comparison only means something per part. Sand printing feeds an existing melt shop and produces castings at normal casting economics once the pattern is removed — a machined pattern typically costs €3,000–15,000 with a 4–8 week lead time, and the printed mold eliminates it. Metal printing’s per-kilogram cost stays high at any quantity because expensive powder and slow machine hours dominate. The full breakdowns are in the sand 3D printing cost guide and the metal 3D printer cost guide.

Can I cast the same alloys metal printing offers?

Partially. Stainless steel, aluminum, nickel, and titanium exist in both cast grades and powder form, and many parts could be made either way. But each route has alloys the other cannot touch: casting covers cast iron and the full family of castable alloys, while metal AM handles certain high-strength and wrought-style formulations that are impractical to cast. The alloy set is dictated by the drawing and the specification, not by the machine.

How big can each process make parts?

Metal printers typically build within a sub-500 mm envelope, and the machine price climbs steeply beyond that. Sand printers make molds up to 4000×2000×1000 mm on the J4000Pro, and segmented molds can be assembled for castings larger than the build volume. If the part is a large casting, the sand route is the only one of the two that even competes.

Do I need a foundry for sand 3D printing?

Yes. The printer produces molds and cores, but you still need melt, pouring, shakeout, and finishing. That is why sand 3D printing is a foundry technology: it replaces tooling inside an existing casting operation. If no casting capability exists, the direct metal route — or a casting service — is the alternative.

When should I buy both?

When your order book spans both regimes: small, high-value parts in powder-only alloys (titanium or nickel aerospace brackets, topology-optimized components) plus large or castable-alloy parts in series. Many foundries run a J-series sand printer for molds and cores and an AFS metal machine for inserts, repair, and tooling on the same floor. Route each part by size, alloy, and quantity, and let the mix justify the second machine.