Technology · 10 min read

Why Are 3D-Printed Objects Mostly Hollow Inside?

how does a 3d printer work?

That solid-looking 3D-printed part on your desk is probably 80% air inside. A 3D printer builds objects by melting a thin strand of plastic and tracing precise paths thousands of times, one layer at a time, with each layer just 0.2mm thick.

The core idea

Layer by Layer

A standard 10cm-tall print contains 500 layers, each just 0.2mm thick, stacked with 0.05mm precision.

Precision Melting

The nozzle heats plastic to 210°C and extrudes it through a 0.4mm opening, smaller than a pencil tip.

Digital to Physical

Slicing software converts a 3D model into thousands of tool-path instructions, controlling every millimeter of movement.

Key insight A 3D printer is essentially a computer-controlled hot-glue gun that moves with microscopic precision. It melts thermoplastic filament at around 200°C and deposits it in exact paths dictated by G-code, building objects from the bottom up. The secret is in the layer bond: each new line of molten plastic briefly re-melts the surface below it, allowing polymer chains to entangle across the boundary. This thermal bonding is why temperature control matters so much; too cool and the layers barely stick, too hot and the shape deforms.

You press "print," walk away, and come back six hours later to find a solid object sitting on a metal plate that did not exist when you left. No mold. No cutting. No assembly. Just a spool of plastic that is now a functioning gear, a phone stand, or a replacement part for your dishwasher. Something happened on that plate for six hours, layer by layer, in the dark. What exactly?

A 3D printer does not sculpt, mold, or carve. It is a computer-controlled hot-glue gun that traces precise paths thousands of times, stacking layers of melted plastic thinner than a sheet of paper.

Most people picture 3D printing as something like a Star Trek replicator: the machine somehow "forms" the object all at once, materializing it from raw material into its final shape. The reality is far more patient and mechanical. An FDM (Fused Deposition Modeling) printer, the type in over 95% of consumer machines, works exactly like a hot-glue gun mounted on a robotic arm that can move with 0.05mm precision. It melts a thin strand of plastic, squeezes it through a tiny nozzle, and traces a flat pattern. Then it moves up by a fraction of a millimeter, and traces the next pattern on top. The object does not exist as a complete shape until the very last layer is laid down. A 10cm-tall print at standard resolution is 500 separate layers, each just 0.2mm thick. Every single one had to be traced, cooled, and bonded to the one below.

The process starts long before the nozzle heats up. A 3D model, typically an STL file, gets loaded into slicing software like Cura or PrusaSlicer. The slicer does exactly what the name suggests: it cuts the 3D model into hundreds or thousands of horizontal cross-sections, each one layer height thick. For each slice, the software calculates the exact path the nozzle must follow, how fast it should move, how much plastic to push through, and where to place internal structure. The output is G-code, a line-by-line set of machine instructions that controls every millimeter of movement. A typical print job contains tens of thousands of G-code commands.

When printing begins, the extruder, a motor-driven gear system, grips the filament (a 1.75mm strand of thermoplastic, usually PLA) and pushes it into the hot end. The hot end is a precision thermal assembly with distinct zones: a heat sink at the top keeps the incoming filament solid, a narrow heat break creates a thermal barrier, and the heat block below brings the plastic to its melting point, around 210\u00B0C for PLA. The molten plastic is forced through a nozzle with a bore of just 0.4mm, smaller than a pencil tip. As the nozzle traces its programmed path across the build plate, it deposits a continuous line of soft plastic that begins cooling immediately.

The critical question is how these layers stay together. The answer is thermal bonding. When a fresh line of 210\u00B0C plastic lands on the previous layer, the heat briefly re-melts the surface beneath. During this window (just fractions of a second), polymer chains from the new layer diffuse across the boundary and entangle with chains in the old layer, creating an inter-molecular bond. This is why temperature matters so much: too cool and the chains barely entangle, creating weak inter-layer adhesion; too hot and the plastic stays molten too long, losing its shape before it can solidify.

Interactive -- the FDM printing process
HEATED BED 60°C HEAT SINK 210°C LAYER 0 / 0 PROGRESS explained.guide
Nozzle temperature 210°C
Layer height 0.3mm
Print speed 60 mm/s
333
Total layers
2.8 hrs
Est. print time
Good
Layer adhesion
Draft
Surface quality
The hot end is the melting chamber at the heart of every FDM printer. Solid filament enters through a heat sink (which keeps it rigid), passes through a narrow heat break (a thermal barrier), and enters the heat block where a 40W heating cartridge brings it to 200-260°C. A thermistor monitors temperature within 1°C accuracy. The molten plastic is forced through a brass nozzle with a 0.4mm bore. The entire assembly must maintain a sharp thermal gradient: solid filament just millimeters from molten plastic.
Standard draft settings: 0.2mm layers at 60 mm/s. Visible layer lines but reasonable print time. Good layer adhesion at this temperature and speed combination.

The secret inside every printed object

Here is the fact that surprises most people: the solid-looking bracket, case, or figurine that comes off a 3D printer is almost certainly not solid inside. Pick up any 3D-printed part and you will notice it is lighter than you expect. That is because the slicer fills the interior with a sparse geometric pattern called infill, not solid plastic. A typical household print uses just 15 to 20% infill, meaning the inside is 80 to 85% air, structured in a honeycomb, gyroid, or grid pattern that provides strength while using a fraction of the material.

This is not a cost-cutting trick. It is smarter engineering. A part printed at 20% gyroid infill weighs roughly one-fifth of what a solid part would weigh, prints in half the time, uses far less filament, and retains most of its structural strength. Going from 20% to 100% infill only adds about 15 to 20% more tensile strength while tripling the material cost and print time. The real strength of a printed part comes from its walls (the solid outer perimeters), not its interior fill. Adding one extra wall perimeter often does more for strength than doubling the infill percentage.

Interactive -- inside a 3D-printed part
CROSS-SECTION VIEW TOP SOLID LAYERS (3) BOTTOM SOLID LAYERS (3) WALL WALL 10 cm 20% INFILL MATERIAL 23% used explained.guide
Infill percentage 20%
Wall count 2
48g
Part weight
1.8 hrs
Print time
72%
vs. solid strength
$0.96
Filament cost
Infill is the internal geometric pattern that fills the space between the solid outer walls and top/bottom layers. Common patterns include grid (criss-cross lines), gyroid (smooth wave surface), and honeycomb. At 20% infill, the interior is 80% air, yet the part retains roughly 70% of the tensile strength of a solid part. Going above 60% infill provides diminishing returns: only 15 to 20% more strength for double the material and print time. Gyroid infill distributes stress most evenly in all directions.

The layer height dilemma

Every 3D print is a negotiation between speed and resolution. Cut the layer height in half and you double the detail, but you also double the print time. There is no setting that gives you both.

A 3D printer's resolution is defined by its layer height. At 0.3mm, layers are clearly visible to the naked eye, giving the surface a staircase texture. At 0.1mm, the layers nearly disappear, producing a smooth finish suitable for display models or miniatures. But here is the cost: a 10cm print at 0.1mm needs 1,000 layers; at 0.3mm, it needs only 333. Same object, triple the time. Professional users learn to match layer height to purpose: 0.3mm for structural brackets nobody will see, 0.2mm for everyday items, 0.12mm for detailed figurines. There is no universal "best" setting.

FDM is also not the only way to build objects layer by layer. SLA (stereolithography) printers use a UV light source to cure liquid resin, producing layers as fine as 0.025mm with dramatically smoother surfaces. SLS (selective laser sintering) fuses powdered nylon with a laser and needs no support structures at all, because the unmelted powder holds the part in place during printing. Each technology trades something different: FDM is cheapest and prints largest, SLA produces the finest detail, and SLS makes the strongest parts.

Interactive -- FDM vs SLA vs SLS
FDM
Fused Deposition Modeling
\u25BC melts + deposits
0.1mm
min layer height
SLA
Stereolithography
UV LCD LIQUID RESIN \u25B2 cures layer
0.025mm
min layer height
SLS
Selective Laser Sintering
LASER fuses powder no supports needed
0.06mm
min layer height
0.1mm
Min layer (selected)
~$300
Entry price
Good
Part strength
Layered
Surface finish
FDM (Fused Deposition Modeling) melts thermoplastic filament and extrudes it through a nozzle, building parts layer by layer. It is the most accessible 3D printing technology: printers start at around $200, filament costs $15 to 40 per kilogram, and the build volumes are the largest of any consumer technology (up to 300x300x400mm). The tradeoff is surface finish: visible layer lines are inherent to the process, and overhangs steeper than 45 degrees require support structures that leave marks when removed.

Every manufactured object you have ever owned was made by a process you could not control: injection molding in a factory, CNC machining in a workshop, casting in a foundry. A 3D printer changes the equation. For the first time, the entire chain from idea to physical object fits on a desk. You design the shape, the slicer calculates the toolpath, and the printer executes it layer by layer while you sleep. The result is not as strong as injection-molded plastic, not as smooth as machined metal, and not as fast as mass production. But it is yours, made exactly to your specifications, in your home, from a $20 spool of plastic. The next time you hold a 3D-printed part, run your finger along the surface. Those tiny ridges you feel are the layers, each one 0.2mm thick, each one precisely placed, each one thermally bonded to the one below. That is the entire secret: patient, precise, repetitive stacking of melted plastic, hundreds of times, until nothing becomes something.

The parts that make it work

Hot End

The heated tip that melts plastic and shapes it through a tiny nozzle.

The melting chamber where solid filament transforms into liquid plastic. A heating cartridge brings the heat block to 200-260°C while a thermistor monitors temperature within 1°C accuracy. The nozzle (typically 0.4mm brass) shapes the extruded stream.

Extruder

The motor-driven gear that feeds plastic into the hot end.

The feed mechanism that grips filament with a toothed drive gear and pushes it into the hot end at precisely controlled rates. Direct-drive systems mount on the print head for better retraction control; Bowden systems push filament through a PTFE tube from the frame.

Heated Build Plate

The warm surface where the first layer sticks and the print grows.

The foundation surface where the first layer adheres. Heated to 50-110°C depending on material, it prevents warping by keeping the base layers warm and dimensionally stable. PEI, glass, or textured surfaces provide grip.

Stepper Motors

Precision motors that position the nozzle with pinpoint accuracy.

Four precision motors (X, Y, Z, and extruder) move in 1.8° increments with 200 steps per revolution, positioning the nozzle with 0.05mm accuracy. X and Y axes use timing belts for speed; the Z axis uses lead screws for precision.

Control Board

The brain that reads instructions and coordinates every movement.

The brain running firmware that interprets G-code instructions, manages PID temperature loops, calculates acceleration curves, and coordinates all motor movements in real time. Modern boards also run auto bed-leveling mesh compensation.

Cooling System

Fans that quickly cool each layer so it holds its shape.

A part-cooling fan solidifies each deposited layer within seconds, preserving detail and preventing deformation. The hot-end heat sink and its dedicated fan keep the upper filament solid, preventing heat creep from softening plastic prematurely.

Nozzle Temperature by Filament Material

PLA (easiest to print) 210°C
PETG (strong, flexible) 235°C
ABS (heat-resistant) 245°C
Nylon (durable, wear-resistant) 260°C
Polycarbonate (strongest) 290°C

Tips & maintenance

  1. Level your bed before every print session. Even 0.1mm of deviation between the nozzle and build plate can cause first-layer adhesion failure, which is the number one cause of failed prints.
  2. Store filament in sealed containers with silica gel packets to keep humidity below 30%. Moisture-contaminated filament causes popping, bubbles, and weak layer bonds that can ruin multi-hour prints.
  3. Use 15 to 20% gyroid infill for most household prints. Going above 60% infill adds only 15 to 20% more strength while doubling print time and filament usage.
  4. Print a temperature tower for every new filament roll. Optimal temperature varies by brand and color; finding the sweet spot (usually within a 10°C range) dramatically reduces stringing and improves layer adhesion.
  5. Orient your part so the weakest axis (between layers) runs perpendicular to the primary load direction. FDM parts are two to three times stronger along the layer plane than across it.

Common questions

Consumer FDM printers work with thermoplastic filaments. PLA is the most popular: it prints at 210°C, produces minimal fumes, and comes in hundreds of colors. PETG offers better strength and slight flexibility. ABS is heat-resistant but requires an enclosure to prevent warping. TPU produces rubber-like flexible parts. Nylon delivers exceptional durability but absorbs moisture aggressively. Specialty filaments include carbon-fiber-reinforced, wood-fill, and glow-in-the-dark variants.

Print time varies enormously based on size, layer height, and speed. A small keychain takes 30 minutes to 2 hours. A standard test print (the "Benchy" boat) takes 2 to 3 hours on a regular printer or 35 minutes on a high-speed machine. Large items like helmets or enclosures can take 20 to 60+ hours. The biggest time factor is layer height: printing at 0.1mm (fine) takes three times longer than 0.3mm (draft) for the same object.

FDM parts are anisotropic, meaning they are strongest along the layer plane but weakest between layers, where adhesion depends on thermal bonding. A well-printed ABS part reaches 25 to 35 MPa tensile strength compared to 40 to 48 MPa for injection-molded ABS. You can maximize strength by adding more wall perimeters (often more effective than increasing infill), orienting the part so loads run parallel to layers, and printing at higher temperatures to improve inter-layer chain diffusion.

FDM melts plastic filament through a nozzle; it is the cheapest and most common, great for functional parts and prototypes. SLA uses UV light to cure liquid resin, producing smoother surfaces and finer details (0.025mm layers vs. FDM's 0.1mm), ideal for miniatures and jewelry. SLS fuses powdered nylon with a laser and needs no support structures, producing the strongest parts. SLA and SLS require more post-processing: alcohol washing and UV curing for SLA, powder removal for SLS.

The most common cause is poor first-layer adhesion: if the bed is not level or the nozzle is too far from the surface, the print detaches mid-job. Other frequent failures include nozzle clogs from debris or moisture-damaged filament, warping from uneven cooling (especially with ABS), and layer shifting from loose belts or the print head hitting a curled section. Most failures happen in the first few layers, which is why experienced users watch the start of every print.

Yes. "Print-in-place" designs use carefully calculated air gaps (typically 0.2 to 0.3mm) between interlocking components. During printing, these gaps prevent the molten plastic from fusing the parts together. Once complete, the parts can rotate or slide freely without assembly. Common examples include working hinges, planetary gear sets, and chain links. The key is precise calibration: too little clearance and the parts fuse; too much and they feel loose.