Full-Layer Metal 3D Printing: What LLNL's Filing Means

Full-layer metal 3D printing is the idea Fabbaloo pulled from a Lawrence Livermore patent application on October 1, 2026. A diode array and a digital mask would expose a chosen powder pattern in one shot, instead of walking a laser spot across the layer. For a lab or a shop class, the takeaway is narrow: this is a patent, not a priced machine, and earlier experiments already show why flashing a whole layer is harder than the drawing.

What's Happening

Kerry Stevenson at Fabbaloo reported on October 1, 2026 that application US20260273840A1 proposes replacing the usual raster scan in laser powder bed fusion. In a conventional machine, a focused beam traces the current slice. A layer may be only 50 to 100 microns thick, but the spot still has to visit every location that should become solid. That travel dominates build time on anything wider than a coupon.

The filing, as Fabbaloo describes it, starts with a diode array rather than a single fiber laser. Light from diode bars is arranged into a two-dimensional pattern and sent through a computer-controlled mask. One embodiment is a liquid-crystal module with a polarizing mirror. Pixels block light in some regions and pass it in others. Exposed powder melts or sinters. Unexposed powder can act as support. The claims Fabbaloo highlights cover generating that pattern, changing it each layer, and setting power density for the powder in the bed.

Power is what keeps this off a classroom bench. Fabbaloo says the filing lists calculated average flux from 12 watts per square centimeter for aluminium to 2,900 for tungsten, assuming 30-micron powder in a 90-micron layer. The diode array is described as at least about 10 kilowatts per square centimeter, with a maximum above 100 at a two percent duty cycle. Those are patent calculations, not a product test. Aluminium and tungsten do not absorb heat the same way, so one exposure recipe cannot cover every alloy.

Fabbaloo also notes a relay that tiles light across the bed, talk of scaling toward a square meter, and a 25 centimeter square as the largest rastered part the filing cites. It sketches multiple nozzles laying different metals before one exposure. None of that is a machine you can buy.

Why Full-Layer Metal 3D Printing Matters for Labs and Shops

Metal powder-bed work is still a specialist process. Powder handling is a safety job, and a tall build can run for days because the laser is drawing the layer rather than flashing it. Anyone teaching additive manufacturing, or quoting metal parts, already feels that bottleneck. A mask-based exposure would attack time and, on paper, build area together. That is why the filing belongs in a lab notebook even if nobody in the room will buy one.

The practical filter is simpler. Ask whether the source is a paper with a build, a patent with a calculation, or a printer with a manual. Livermore's earlier work, covered below, did print with a related optical trick. The October filing goes further on power control, bed size, and multi-metal dosing. It does not show a finished one-meter machine, a repeatability study, or a price. Students get misled when those three documents are treated as the same kind of evidence.

For a desktop shop the distance is clearer still. Continuous carbon fiber on a polymer printer is for a bracket, fixture, or jig that has to carry a load without a metal powder cell. A diode-mask architecture, if it ever leaves the patent, would sit in a facility with inert gas, powder logistics, and post-machining. Knowing which tool belongs to which job is the skill this story trains.

How Full-Layer Metal 3D Printing Compares on Paper

The table separates three public descriptions. The patent row is what Fabbaloo reports from the filing. The diode-area row is from a published titanium study.

Approach	How a layer is exposed	What the public record shows
Raster laser powder bed fusion	One spot traces the slice	Matthews et al. (2017) call this slow relative to conventional manufacturing because the beam must visit every location
Diode area melting	Several low-power diodes melt a wide track	Caglar et al. (2024): nine 4 W, 450 nm lasers, tracks wider than 1 mm, samples above 99% density; multi-layer melt depth was uneven
LLNL application US20260273840A1	Diode array through a digital mask	Fabbaloo (Oct 1, 2026): calculated flux about 12 to 2,900 W/cm² by alloy; not a catalog product

What the Research Says

Matthews and colleagues at Livermore showed the optical idea in a build, not only in a filing (Matthews et al., 2017). Their Optics Express paper uses an optically addressable light valve as a photomask so an entire powder layer can be melted at once. The sheet combined a diode pulse on the order of 5 kilowatts and 20 milliseconds with a Q-switched pulse on the order of 1 megawatt and 7 nanoseconds. They patterned the near infrared by imaging 470-nanometer light onto a liquid-crystal valve. The paper's point is time: raster scanning makes selective laser melting slow next to most traditional manufacturing, often hours to days. That 2017 run is the lab ancestor of the new filing. It is not evidence that a square-meter bed already works.

Caglar and colleagues later melted a wide track with diodes instead of one spot (Caglar et al., 2024). Nine 450-nanometer lasers at 4 watts each processed Ti6Al4V. Tracks came out wider than 1,000 micrometers, so a layer needed fewer passes, and they reported samples above 99 percent density. The catch for any full-layer claim: a wider pool did not stay even through the depth, which made density harder in multi-layer parts. Rescanning cut roughness and raised density, while hardness fell by about 40 percent. Wider exposure saves passes. It does not automatically save the microstructure.

Frequently Asked Questions

What is full-layer metal 3D printing?

It means melting a chosen pattern across a powder layer in one exposure, instead of tracing that pattern with a moving laser spot. Lawrence Livermore described that idea in a patent application Fabbaloo reported on October 1, 2026. A 2017 lab paper already melted small layers this way. The filing is not a printer a shop can order today.

Can I buy an LLNL diode-mask metal printer?

No. Fabbaloo's October 1, 2026 report covers patent application US20260273840A1, which describes a diode array and a digital mask for powder-bed metal printing. It does not announce a product, a price, or a ship date. University work on diode area melting is also lab equipment, not a catalog machine. Treat the filing as a research signal.

How is diode area melting different from a normal metal printer?

A normal laser powder-bed printer moves one spot across the layer. Diode area melting uses several low-power diode lasers at once so the melted track is much wider. In a 2024 titanium study, nine 4-watt, 450-nanometer lasers made tracks wider than one millimeter and dense single tracks, but keeping an even melt depth through many layers was still difficult.

Fibricate's Place in This Story

Fibricate does not sell metal powder-bed systems, and this filing does not change that. The split is between a research architecture and a desktop tool you can run. A desktop continuous carbon fiber 3D printer such as the FibreSeeker 3 co-extrudes continuous fiber with polymer on a 300 by 300 by 245 millimeter bed, for loaded polymer parts rather than titanium powder. When a job needs more fiber in the bead, a continuous carbon fiber spool is the feedstock. If the part must be metal, the next step is a qualified metal shop. A Livermore filing is homework, not that shop.

What to Watch Next

Watch whether this application is granted, narrowed, or followed by a methods paper with a measured build. Patent drawings can describe a square-meter diode array years before anyone holds the thermal uniformity. The useful follow-ups are density through many layers, residual stress when a whole layer sees energy at once, and how the mask behaves from aluminium to tungsten. Caglar's hardness drop after rescanning is a reminder that a faster layer and the same metal are different claims.

Over the next year or two, diode-array powder-bed work will keep showing up as papers and filings before it shows up on a quote. Labs should file those sources next to the raster machines they already teach. Shops should keep quoting the process they can inspect. If a vendor later offers a mask-based metal printer, ask for layer-wise density and a powder specification, not a patent number.

References & Further Reading

1. Matthews, M. J., et al. (2017). Diode-based additive manufacturing of metals using an optically-addressable light valve. Optics Express.
2. Caglar, H., Liang, A., Groom, K., et al. (2024). Multi-laser powder bed fusion of Ti6Al4V: diode area melting utilizing low-power 450 nm diode lasers. Journal of Materials Processing Technology.
3. LLNL Patent Replaces LPBF Rastering With A Digital Light Mask. Fabbaloo. Retrieved October 1, 2026.