Wikipedia Deep Dive

3D printing

15 min read

The Machine That Builds Itself

In 2005, a British engineer named Adrian Bowyer had a radical idea. He designed a 3D printer that could manufacture seventy percent of its own components. The machine could, in essence, reproduce. He called it RepRap, short for Replicating Rapid-prototyper, and he gave the plans away for free.

This was the moment 3D printing stopped being an industrial curiosity and became something closer to a revolution.

But the story of how we got there—how a technology dreamed up in science fiction magazines of the 1940s became the machine sitting on desktops in garages and workshops around the world—is a tale of abandoned patents, ignored inventors, and the gradual realization that we'd been thinking about manufacturing all wrong.

Building Up Instead of Cutting Away

To understand what makes 3D printing special, you first need to understand what it replaced.

For most of human history, we made things one of two ways. We could cast them—pouring molten material into a mold and letting it solidify. Or we could carve them—starting with a block of material and removing everything that wasn't the final object. Sculptors did this with marble. Machinists did it with metal lathes and milling machines. The computer-controlled version, called CNC machining (that's Computer Numerical Control), became the backbone of modern manufacturing.

Both approaches have a fundamental limitation. With casting, you need a mold, which means you need to make something before you can make the thing. With machining, you can only create shapes that a cutting tool can reach. Try carving the inside of a hollow sphere, or creating internal lattice structures, and you'll quickly discover what's impossible.

3D printing flips the entire logic. Instead of removing material, you add it. Layer by microscopic layer, you build up an object from nothing. The technical term for this is additive manufacturing—manufacturing by addition rather than subtraction.

This simple inversion unlocks geometries that were previously unthinkable. You can print objects with internal honeycombs that reduce weight while maintaining strength. You can create interlocking parts that move freely without ever having been assembled. You can build shapes that exist only because a computer can calculate them, shapes no human hand could carve.

The Science Fiction Writers Got There First

Before anyone built a working 3D printer, writers imagined them.

In 1945, Murray Leinster published a short story called "Things Pass By" in which he described a device that sounds remarkably like a modern 3D printer. The machine, he wrote, worked by feeding "magnetronic plastics" through a moving arm that traced patterns in the air. Plastic emerged from the tip of the arm and hardened as it came out, following digital drawings.

Five years later, Raymond F. Jones described something similar in Astounding Science Fiction magazine. He called his version a "molecular spray."

These weren't engineering proposals. They were flights of imagination. But they captured something essential about the technology that would eventually emerge: the idea that you could build solid objects the way an inkjet printer builds images on paper, one tiny dot at a time, one layer at a time.

The Patent Nobody Wanted

The actual invention of 3D printing happened multiple times, in multiple places, with multiple inventors who often didn't know about each other's work.

In 1971, a man named Johannes Gottwald received a US patent for something he called the Liquid Metal Recorder. His device used continuous jets of molten metal to build up three-dimensional patterns on a reusable surface. When you were done with the object, you could melt it down and print something new. The patent explicitly described what we would now recognize as rapid prototyping and on-demand manufacturing.

Gottwald's patent went nowhere.

In 1974, David Jones described the concept of 3D printing in his column in New Scientist magazine. His column was called Ariadne, named after the figure from Greek mythology who gave Theseus the thread that guided him through the labyrinth. It was an apt name for a column about following technological ideas to their logical conclusions.

The real breakthroughs came in the 1980s.

In April 1980, Hideo Kodama of the Nagoya Municipal Industrial Research Institute in Japan invented two methods for creating three-dimensional plastic objects using light-sensitive polymers—materials that harden when exposed to ultraviolet light. He filed a patent in 1981. His research papers were published that same year.

Nobody cared.

Kodama's annual research budget was 60,000 yen—roughly $545. His boss showed no interest. His device wasn't valued in his own laboratory. He abandoned the patent, and the project died.

Meanwhile, in France, three inventors—Alain Le Méhauté, Olivier de Witte, and Jean Claude André—filed their own patent for stereolithography, the same basic process Kodama had invented. The French General Electric Company reviewed their application and rejected it. The official reason: "lack of business perspective."

The technology that would eventually transform manufacturing was being invented and abandoned, invented and abandoned, over and over again.

Chuck Hull Makes It Stick

The person who finally made 3D printing work as a commercial technology was an American engineer named Charles Hull—Chuck, to those who knew him.

In August 1984, Hull filed a patent for stereolithography fabrication. His system worked by shining ultraviolet light onto a vat of liquid photopolymer—a substance that hardens when exposed to light. A computer controlled the light beam, tracing out the shape of one layer of an object. That layer hardened. Then a platform lowered the just-completed layer slightly into the vat, fresh liquid flowed over it, and the laser traced the next layer. Repeat this hundreds or thousands of times, and a solid three-dimensional object emerges from the liquid.

Hull didn't just invent the machine. He invented the ecosystem around it.

He created the STL file format, which stands for stereolithography and is still the standard way to describe 3D objects for printing. He developed the digital slicing algorithms that convert a 3D model into the series of 2D layers a printer can actually create. These innovations are what made 3D printing practical.

In 1986, Hull received his patent. He founded a company called 3D Systems Corporation. In 1987 or 1988—records vary slightly—the company released the SLA-1, the first commercial 3D printer.

It cost over $300,000. In today's money, that's about $650,000.

The Desktop Revolution Takes Its Time

Through the late 1980s and 1990s, 3D printing remained firmly industrial. You needed a budget in the hundreds of thousands of dollars. You needed engineers who understood the equipment. You used it for prototyping—making quick test versions of parts before committing to expensive manufacturing tooling.

This is why the technology was initially called rapid prototyping. The word "rapid" seems almost ironic now, but at the time it was revolutionary. Traditional prototyping could take weeks or months. 3D printing could produce a test part in hours or days.

In 1988, S. Scott Crump invented a different approach called Fused Deposition Modeling, or FDM. Instead of using lasers and liquid polymers, FDM works by pushing heated plastic through a nozzle, like a hot glue gun controlled by a computer. The nozzle moves in precise patterns, laying down thin roads of molten plastic that cool and solidify almost instantly.

Crump founded a company called Stratasys to commercialize his invention. By 1992, they were selling FDM machines.

FDM would eventually become the most common form of 3D printing, largely because it's the simplest and cheapest. The raw material is just plastic filament, wound on spools like fishing line. The mechanism is straightforward. When patent protections on FDM expired in 2009, the technology exploded into the consumer market.

But in the 1990s, all of this was still the province of engineers and researchers. The machines were expensive. The software was arcane. The materials were limited.

The Metal Problem

Printing in plastic is one thing. Printing in metal is another entirely.

Metal has a much higher melting point. Metal conducts heat, which makes precise control difficult. Metal parts need to be strong enough for actual use, not just visual prototyping. The challenges are fundamentally different.

Through the 1980s and 1990s, various metal printing processes emerged, each with its own awkward name. Selective Laser Sintering (SLS) used lasers to fuse metal powder without fully melting it. Direct Metal Laser Sintering (DMLS) did something similar with different parameters. Selective Laser Melting (SLM) fully melted the powder for denser, stronger parts.

The Fraunhofer Society, a German research organization, developed SLM in 1995. This process could produce metal parts with properties approaching those of traditionally manufactured components—parts strong enough to actually use, not just look at.

Today, metal 3D printing is used to manufacture jet engine components, medical implants, and automotive parts. But the technology took decades to mature from laboratory curiosity to industrial tool.

The Inkjet Connection

One of the less-told stories of 3D printing involves its connection to ordinary 2D inkjet printers—the kind that have been printing documents and photographs since the 1970s.

In 1983, a company called Howtek (originally R.H. Research) began developing color inkjet printers that used hot-melt thermoplastic inks—plastics that liquify when heated and solidify when they cool. This is essentially the same principle that would later drive FDM 3D printing.

Several engineers who worked on Howtek's inkjet technology later founded or joined 3D printing companies. Richard Helinski formed a company called Visual Impact Corporation. Herbert Menhennett started HM Research and later worked with Bill Masters at Ballistic Particle Manufacturing. The technical knowledge of how to precisely deposit tiny droplets of molten material flowed from 2D printing into 3D printing.

The connection even shows up in the name. The term "3D printing" originally referred specifically to a process developed at MIT in 1993 by Emanuel Sachs, which used modified inkjet print heads to deposit binding material onto layers of powder. The powder that got hit with the binder stuck together; the rest remained loose and could be brushed away, revealing the finished object.

This is why we call it printing at all. The technology borrowed not just concepts but actual hardware from the paper printing industry.

When the Patents Expired

In 2009, something transformative happened. The key patents on Fused Deposition Modeling expired.

Suddenly, anyone could build an FDM printer without paying licensing fees to Stratasys. The RepRap project, which had been designing self-replicating printers since 2004, accelerated. The Fab@Home project, started in 2006 by Evan Malone and Hod Lipson, made their designs freely available.

Most importantly, the software was open source. Anyone could download it, modify it, and share their improvements. A global community of tinkerers, hackers, and enthusiasts began iterating on 3D printer designs at a pace no single company could match.

The price of a basic 3D printer dropped from hundreds of thousands of dollars to thousands, then to hundreds. Today, a functional FDM printer can be purchased for less than the cost of a modest smartphone.

What's Actually Happening Inside the Machine

When you print a 3D object, the process starts with a digital model—typically created in Computer-Aided Design (CAD) software or downloaded from an online repository. This model is then "sliced" by specialized software into hundreds or thousands of horizontal layers, each just a fraction of a millimeter thick.

The printer builds the object one layer at a time. In FDM, a plastic filament is pushed through a heated nozzle that moves along a precise path, depositing a thin line of molten material. When one layer is complete, the build platform moves down slightly (or the print head moves up), and the next layer is deposited on top.

The layers bond to each other because each new layer is deposited while still hot enough to partially melt into the layer below. This is also why 3D printed parts often have visible layer lines—each layer is a distinct horizontal band, like the rings of a tree trunk.

In stereolithography, the process is different but the layer-by-layer principle remains. A vat contains liquid photopolymer. A laser or projected light source traces the pattern of each layer, hardening the liquid into solid plastic. The build platform rises (or descends, depending on the machine design), and the next layer is traced.

Metal printing typically uses powder. A thin layer of metal powder is spread across a build platform. A laser or electron beam scans across the powder, melting it where the object should be solid. More powder is spread. The process repeats.

The common thread is layers. Always layers. The "additive" in additive manufacturing happens one horizontal slice at a time.

Why It Matters

3D printing changes the economics of manufacturing in a fundamental way.

Traditional manufacturing requires tooling—molds, dies, jigs, fixtures. This tooling is expensive to create but allows you to produce identical parts very cheaply in large quantities. The economics favor mass production: make a million of something, and the cost of the tooling gets spread across all million units.

3D printing has no tooling. The first unit costs the same as the hundredth or the thousandth. You can print one of something as cheaply as you could print many.

This inverts the traditional logic. Instead of designing products that can be manufactured cheaply in bulk, you can design products that are perfect for their specific purpose. Instead of maintaining warehouses full of spare parts, you can print parts on demand when they're needed. Instead of choosing from a catalog of available components, you can create exactly the component you need.

The phrase "on-demand manufacturing" captures this shift. In the old model, manufacturing required planning, forecasting, tooling, inventory. In the new model, you print what you need when you need it.

The Limitations Nobody Mentions

For all its promise, 3D printing has real limitations that enthusiastic coverage often glosses over.

Speed is the most significant. 3D printing is slow. A part that would take seconds to injection-mold might take hours to print. For mass production of simple shapes, traditional manufacturing remains far more efficient.

Material properties present another challenge. 3D printed parts are often weaker than their traditionally manufactured equivalents. The layer-by-layer construction creates potential failure points. In FDM printing, parts are typically weaker in the vertical direction (perpendicular to the layers) than in the horizontal plane.

Surface quality varies. The layer lines visible on most 3D printed parts are a cosmetic limitation. Post-processing—sanding, coating, or chemical smoothing—can improve surface finish, but it adds time and cost.

And the environmental story is complicated. Yes, 3D printing can reduce material waste compared to subtractive manufacturing. But the plastics most commonly used are petroleum-based polymers. Metal printing requires energy-intensive lasers. The distributed, on-demand model reduces shipping and warehousing, but enabling cheap production of plastic objects may simply encourage more consumption.

The Long Tail of Invention

What strikes me about the history of 3D printing is how long it took to matter.

The core concepts appeared in science fiction in the 1940s and 1950s. Working patents existed by the early 1970s. Multiple inventors independently created functional systems in the early 1980s. Commercial products launched in the late 1980s.

And yet the technology remained expensive, specialized, and obscure for another two decades. It wasn't until patents expired and open-source communities democratized the hardware that 3D printing became something ordinary people could use.

This pattern—invention, long incubation, sudden democratization—appears throughout technological history. The ideas come first. The practical implementations come second. The social and economic changes come last, and they come suddenly when the right barriers finally fall.

Hideo Kodama invented stereolithography in 1980 with a research budget of $545 a year. His boss didn't care. His patent was abandoned. Thirty years later, the technology he pioneered was being used to print everything from dental crowns to rocket components to plastic toys for children.

The future often arrives piecemeal, recognized by a few inventors who can't get anyone else to pay attention. Then, seemingly all at once, it's everywhere.

What Comes Next

3D printing continues to evolve in multiple directions.

Bioprinting uses modified 3D printers to deposit living cells, building up tissue structures layer by layer. Researchers have printed skin, cartilage, and blood vessels. Full organs remain a distant goal, but the trajectory is clear.

Construction-scale printing uses giant robotic arms to extrude concrete, building houses layer by layer. Several companies have demonstrated printed structures, though the technology remains more experimental than practical.

Multi-material printing allows objects to be made from different materials in different regions—hard and soft, conductive and insulating, opaque and transparent—all in a single print run.

And the resolution keeps improving. What once was measured in millimeters is now measured in microns. Nanoscale printing, building structures from individual atoms, remains in laboratories but points toward a future where the distinction between printing and chemistry blurs.

The question is no longer whether 3D printing will transform manufacturing. It already has, at least in certain domains. The question is how far the transformation will go.

When you can print anything, anywhere, on demand—when the plans for objects flow as freely as information flows today—what happens to the economy of physical things? What happens to the supply chains and factories and warehouses built for a world of mass production and long lead times?

Murray Leinster, writing in 1945, imagined a machine that drew shapes in the air and made them solid. Eighty years later, we have those machines. We're still figuring out what to do with them.