Patreon for 3D printing creators — 2026 edition

FDM thermoplastic material science, layer adhesion anisotropy, G-code calibration, volumetric flow limits, SLA photopolymer chemistry, and the Apple Tax.

3D printing Patreons retain when they deliver the materials engineering and process physics layer that filament brand reviews and time-lapse print videos structurally compress away. Here is the technical substrate: PLA/PETG/ABS/ASA/TPU/Nylon/PC/PEEK glass transition temperatures and melt behavior, inter-layer Z-strength anisotropy mechanics, e-steps and pressure advance calibration, volumetric flow rate hotend limits, SLA cure depth equations, slicing parameter tradeoffs, post-processing from acetone vapor to epoxy XTC-3D, and exactly how much the Apple Tax costs a creator earning $200–$500 per month from a 70–78% iOS audience.

1. FDM thermoplastic material science

Every filament choice is a compromise between mechanical properties, temperature resistance, printing difficulty, and cost. Understanding the polymer physics behind each choice is the content layer that turns a subscriber into a patron.

PLA (poly-lactic acid) is semicrystalline — it has both an amorphous phase (Tg ~60°C) and crystalline domains that melt at 170–180°C. Derived from fermented corn starch or sugarcane, it prints at 190–230°C with no bed heating required. Young's modulus is ~3.5 GPa (stiff), UTS 50–65 MPa (strong), but it embrittles over time and begins to soften in a closed car on a summer afternoon. Its processing window is forgiving.

PETG (polyethylene terephthalate glycol-modified) is amorphous — glycol co-monomer disrupts crystallinity, producing a one-phase material that softens gradually above Tg ~80°C rather than melting sharply. Prints at 230–250°C. Inter-layer bonding is typically stronger than PLA because PETG remains viscous longer at the printing interface, allowing more chain diffusion. Excellent chemical resistance (dilute acids, many solvents). Slight stringing tendency due to high melt viscosity and adhesion to nozzle tip.

ABS (acrylonitrile-butadiene-styrene) is an amorphous terpolymer: acrylonitrile provides chemical resistance and heat resistance (Tg ~105°C), butadiene rubber domains provide impact toughness, styrene provides rigidity and processability. Prints at 230–260°C on a 100–110°C bed inside an enclosure — the enclosure prevents differential cooling between the hot newly-deposited layer and cold ambient air, which creates thermal stress sufficient to warp and delaminate the print from the bed. Soluble in acetone, enabling vapor smoothing. Emits styrene VOC — ventilation required.

ASA (acrylonitrile-styrene-acrylate) replaces the polybutadiene rubber phase in ABS with UV-stable polyacrylate rubber, eliminating the photooxidation pathway that causes ABS to yellow, chalk, and embrittle outdoors within 6–12 months. Tg ~100°C, similar processing to ABS. The correct choice for outdoor functional parts — automotive trim, garden fittings, enclosures.

TPU (thermoplastic polyurethane) is a segmented block copolymer: hard segments (urethane linkages from diisocyanate + chain extender) provide crystalline physical cross-links, soft segments (polyether or polyester polyol) provide elastic deformation. Shore A hardness 85–95 typical for printable grades — flexible but not rubber-soft. Elongation at break 450–600%. Requires direct drive extruder (the flexible filament buckles in a Bowden tube at any meaningful retraction or backpressure). Print at 220–240°C, retraction 0–1 mm, slow print speed 20–40 mm/s.

Nylon PA12 is semicrystalline (Tg ~50°C unfilled, Tm ~178°C). Extreme hygroscopicity: PA12 absorbs 1–3% moisture by weight when stored at ambient humidity, which causes hydrolysis of amide bonds at print temperature, producing steam bubbles, MW reduction, and severely degraded mechanical properties. Dry at 80°C for 8–12 hours before every print session. Post-print annealing (80°C, 2–4 hr) increases crystallinity, improving stiffness, dimensional stability, and creep resistance significantly.

PC (polycarbonate) is amorphous (Tg ~147°C), producing excellent optical clarity in natural grades. UTS 55–70 MPa, impact strength ~60 kJ/m² Izod — the best impact toughness of common engineering polymers. Prints at 270–310°C requiring an all-metal hotend (PTFE-lined hotends max at ~240°C before PTFE degrades). Bed at 100–120°C, enclosure critical. Hygroscopic — dry before printing.

PEEK (polyether ether ketone) is semicrystalline (Tg ~143°C, Tm ~343°C). E-modulus 3.6 GPa, UTS 90–110 MPa — approaches aluminum in specific stiffness. Chemical resistance to almost all solvents, radiation stability for medical and aerospace applications. Requires print temperature 360–400°C, bed 120–160°C, and a heated chamber at 90–120°C to prevent thermal shock cracking. Specialty printers only. Cost $200–400/kg vs $20–30/kg for PLA.

2. Layer adhesion and mechanical anisotropy

FDM parts are anisotropic by construction: strength differs between the XY print plane and the Z axis (build direction). This is the most structurally important fact about FDM and the one most creators fail to explain.

The extrudate cross-section is not circular — it is an ellipse (or stadium shape), wider than it is tall, produced as the nozzle compresses the melt against the previous layer. The inter-layer bond forms across the contact area between the flat top surface of the previous layer and the bottom of the fresh extrudate. That contact area is proportional to the layer height: at 0.2 mm layer height with a 0.4 mm nozzle, the contact width is approximately 0.35 mm; at 0.3 mm layer height, approximately 0.30 mm — actually slightly less per-unit because the extrudate is taller and thinner at the contact zone.

Bond formation requires polymer chain interdiffusion across the molten interface — the reptation mechanism. Chains must diffuse far enough across the interface (their radius of gyration ~10–20 nm for typical print polymers) before the interface solidifies. This time window is set by: (1) printing temperature — higher temperature keeps the interface above Tg longer; (2) layer height — thinner layers cool faster; (3) enclosure — slowing ambient heat extraction extends the reptation window. The practical result: printing ABS at 255°C in an enclosure at 50°C ambient produces dramatically better Z-strength than printing at 230°C in open air.

The 0.2 mm standard layer height recommendation emerged from the 0.4 mm nozzle — it represents 50% of nozzle diameter, balancing resolution and print time. But for maximum Z-bond, 0.28–0.32 mm (70–80% of nozzle diameter) produces superior inter-layer contact area while still providing adequate resolution for most functional parts. Fine detail modes (0.08–0.12 mm) sacrifice Z-bond area per unit height for surface resolution — prints are more brittle in Z.

3. G-code calibration and printer tuning

A well-tuned printer produces consistent, reliable results. The calibration stack — steps/mm, e-steps, PID, pressure advance, input shaping, bed mesh — builds from mechanical accuracy upward to motion quality.

Steps/mm calibration: for a leadscrew axis, steps/mm = (motor_steps/rev × microstepping) ÷ lead_mm. A 200-step motor at 16× microstepping on a 8 mm pitch lead screw gives 200 × 16 / 8 = 400 steps/mm. For belts: steps/mm = (200 × 16) / (2 mm pitch × 20 tooth pulley) = 80 steps/mm. These calculations assume no slip — verify with a digital caliper by commanding known moves and measuring actual displacement.

E-steps calibration: mark 120 mm on filament at the extruder entry point (cold-pull a reference mark, or use a piece of tape at exactly 120 mm). Command G1 E100 F200 (extrude 100 mm at 200 mm/min) with the hotend at temperature. Measure remaining distance from entry to the original mark. If only 18 mm remain, actual extrusion was 102 mm. Correction: new_e_steps = current_e_steps × (100 / 102). Repeat until actual extrusion matches commanded within 0.5%.

PID tuning: hotend PID controls the resistive heater element via PWM using the thermistor feedback. Run M303 E0 S210 C8 — Klipper or Marlin runs 8 heating cycles from ambient to target, measures overshoot and settling, outputs Kp (proportional), Ki (integral), Kd (derivative) constants. Accept with M301 P[Kp] I[Ki] D[Kd], save with M500. Poorly tuned PID causes temperature oscillations of ±5–10°C that appear as visible banding in printed walls.

Pressure advance (Klipper) / linear advance (Marlin): the melt zone inside the hotend acts as a spring — increasing print speed requires more filament pressure, stored as elastic deformation of the molten material in the melt zone. Without compensation, the printer prints the exact commanded feedrate but the actual extrusion lags (causing under-extrusion at corners when speed increases) and over-extrudes after corners when speed decreases and the built-up pressure releases. Pressure advance pre-computes how much extra extrusion pressure is needed for a given speed change and applies it ahead of time. Calibrate by printing a K-factor calibration object — a straight-wall rectangular tower with different K values printed at each height increment, then visually identifying the K value where corner sharpness is optimal.

Input shaping (Klipper SHAPER_CALIBRATE): mount an ADXL345 I2C accelerometer on the toolhead (typically taped to the hotend carriage). Run SHAPER_CALIBRATE — Klipper commands a sine sweep of increasing frequency in X (then Y), while the accelerometer records actual toolhead acceleration. FFT of the recorded signal reveals mechanical resonance peaks at the printer's natural frequencies (typically 20–80 Hz for X, 30–100 Hz for Y on a typical i3 or CoreXY). Klipper then selects a digital filter (ZV, MZV, EI, 2HUMP_EI, 3HUMP_EI) tuned to cancel the dominant resonance. The filter is applied as a pre-distortion of every motion profile — the commanded trajectory is modified so the printer's mechanical response to that profile produces the desired actual motion. Result: ghosting/ringing artifacts eliminated, allowing 200–500 mm/s print speeds on properly built machines.

Bed mesh leveling: virtually no printer bed is perfectly flat — thermal expansion, mounting point variations, and bed material irregularities create Z-height variations of 0.1–0.5 mm across the bed. A probe (CR Touch, BLTouch, eddy current probe, or virtual endstop via voltage sensing) measures bed height at grid points. Bilinear interpolation between 9 points (3×3) or bicubic through 25 points (5×5) produces a height map. The printer applies this map as Z compensation in real-time during the first 5–10 layers (fade_end parameter lets the compensation taper off as the print height increases beyond the first-layer critical zone).

4. Print speed and volumetric flow rate limits

Speed in FDM is not limited by motor acceleration — modern printers with input shaping reach 20,000 mm/s² trivially. The actual bottleneck is volumetric flow rate: how many cubic millimeters of plastic the hotend can melt per second.

Volumetric flow rate (VFR) = extrusion_width × layer_height × print_speed. For a 0.4 mm nozzle printing 0.4 mm wide lines at 0.2 mm layer height at 100 mm/s: VFR = 0.4 × 0.2 × 100 = 8 mm³/s — at the limit of a standard V6 hotend. At 150 mm/s the same parameters require 12 mm³/s — exceeding standard hotend capacity, producing under-extrusion. Switching to a 0.6 mm nozzle at 0.35 mm height at 100 mm/s: VFR = 0.6 × 0.35 × 100 = 21 mm³/s — requiring at minimum a Dragon HF class hotend.

The CHT (Ceramic Heating Technology) nozzle insert splits the melt path into 3 parallel channels inside the nozzle, tripling the surface area exposed to the heated brass nozzle tip. This alone increases max VFR by 30–50% at the same temperature without changing the hotend body.

CoreXY vs bed-slinger kinematics: bed-slinger printers (Creality Ender series, Prusa MK4) move the bed in Y and the toolhead in X+Z — the bed mass (400–1500 g including the print) limits Y-axis acceleration to 1,000–3,000 mm/s² before print adhesion fails or the print tips. CoreXY machines (Voron, Bambu, Prusa Core One) move the toolhead in both X and Y simultaneously with two belt drives in an H-configuration — toolhead mass only 150–350 g, enabling 10,000–25,000 mm/s² acceleration. The Bambu X1C achieves 500 mm/s perimeter speed with 20,000 mm/s² acceleration using input shaping and a 60 mm³/s hotend — at those speeds each layer takes 10–30 seconds regardless of print height, limited entirely by the number of perimeter passes and their total path length.

5. SLA resin photopolymer chemistry

Resin printing produces dramatically smoother surfaces and finer detail than FDM at equivalent cost, at the expense of messier post-processing, material toxicity, and brittleness for most standard resins.

MSLA (Monochrome Masked Stereolithography) is the dominant desktop resin printing technology: a UV LED array illuminates a monochrome LCD panel that acts as a photomask, selectively transmitting 405 nm light to cure the resin layer pressed against the build plate. Monochrome panels transmit approximately 8× more UV energy per pixel than the older RGB color LCD panels (which had heavy UV-blocking color filters), reducing exposure times from 8–15 seconds to 1–4 seconds per layer.

Resin formulation controls cure rate, cure depth, and final mechanical properties. Oligomers (urethane diacrylate or epoxy diacrylate, molecular weight 500–3000 g/mol) form the primary polymer backbone after curing, establishing stiffness and toughness. Monomers (HDDA hexanediol diacrylate is the most common cross-linking monomer) reduce viscosity to a printable 200–2000 mPa·s range and add cross-link density. Photoinitiators (TPO diphenyl(2,4,6-trimethylbenzoyl)phosphine oxide absorbs at 370–420 nm peak) cleave under UV to produce acyl and phosphinoyl radical pairs that initiate rapid acrylate chain polymerization. Pigments and UV absorbers (carbon black, organic dyes) limit the depth to which UV penetrates — a thicker absorber layer produces sharper XY detail but requires more exposure energy for the same cure depth.

The cure depth equation Cd = Dp × ln(E₀/Ec) governs every exposure decision: Dp is the penetration depth of UV light in the resin (a material constant — resins with higher absorber concentration have smaller Dp, perhaps 50–100 µm; translucent resins have Dp of 200–400 µm); E₀ is the delivered exposure energy in mJ/cm² (LED power × time); Ec is the critical exposure threshold at which the resin just polymerizes (material constant, 1–10 mJ/cm²). Practical implication: doubling exposure time does not double cure depth — it adds only Dp × ln(2) ≈ 0.69 × Dp additional depth. For a resin with Dp = 100 µm and Ec = 3 mJ/cm², printing at E₀ = 12 mJ/cm² gives Cd = 100 × ln(12/3) = 139 µm. This explains why overcuring (excessive exposure time) causes feature bleeding into gaps (XY resolution loss) while also increasing Cd beyond the nominal layer height.

The build plate release mechanism depends on the vat bottom membrane. PDMS (polydimethylsiloxane) is the original solution — freshly cast PDMS is very non-stick because uncured PDMS chains at the surface inhibit polymerization in the adjacent resin, leaving a thin non-adhered layer that peels easily. PDMS clouds and micro-tears after 500–1000 layers as the inhibition layer depletes. nFEP (fluorinated ethylene propylene film) replaced PDMS in consumer printers: more durable (2000–5000 layers), lower friction, but requires lower peel force speeds to avoid suction forces tearing thin features. The latest ACF (anti-adhesion coating) films achieve 10,000+ layer durability.

6. Slicing parameters and support strategy

The slicer converts a 3D model into toolpaths. Every parameter choice is a tradeoff between print time, material, surface quality, and mechanical properties.

Layer height: 0.05–0.12 mm for detail miniatures (SLA territory in FDM); 0.12–0.16 mm fine detail; 0.20 mm standard (50% of 0.4 mm nozzle); 0.24–0.32 mm quality-speed balance; 0.35–0.40 mm maximum for 0.4 mm nozzle before visible layering dominates. Each halving of layer height roughly doubles print time and improves vertical surface resolution at the cost of Z-bond area per unit height.

Infill pattern selection: gyroid is a triply periodic minimal surface pattern — mathematically, it has equal curvature in all spatial orientations, producing isotropic mechanical properties in XY, 45°, and Z. It excels for flexible materials (TPU) because the curved geometry distributes stress rather than concentrating it at straight-wall junctions. Honeycomb provides maximum XY stiffness-to-material but is weak diagonally (anisotropic). Grid/lines print fastest but offer the least uniform strength. Cubic subdivision (used in Prusa Slicer) produces a 3D cross structure visible internally — good Z-strength compared to 2D infill patterns.

Infill percentage tradeoffs: 10–15% for display objects and organics; 20% standard functional (most bracket and housing applications); 40% high-stress joints; 80–100% maximum compression only — beyond 40–50% infill, adding perimeter walls (0.4 mm each) is more efficient per unit strength than increasing infill because perimeters are fully solid continuous lines while infill has void boundaries.

Overhang angle: in FDM, each layer must be supported by the layer below. At 0° overhang (vertical wall), the layer is fully supported. At 45°, each new layer projects half its width beyond the previous layer's edge — the overhanging half is unsupported but the prior layer provides enough thermal mass and mechanical support for the extrudate to bridge. Above 45°, the unsupported fraction increases: at 60°, each layer projects 57% beyond the prior edge and begins to droop, producing ridged or fuzzy surfaces. Above 70–80°, support generation is required. Cooling fan speed strongly affects overhang quality — faster cooling solidifies the extrudate before it can sag.

Support types: normal supports (pillars at regular grid spacing) are material-efficient but leave a rough interface on the supported surface. Tree supports (Bambu, Prusa, OrcaSlicer organic supports) branch organically, touching the model at small points to minimize contact area — easier removal and better surface finish, especially for curved surfaces. Support interface layers (1–3 layers of dense grid or solid contact material) sit between support and model, set with a Z-gap of 0.10–0.20 mm — the gap means the interface layer separates cleanly from the model surface rather than fusing, producing a much smoother support-contact area.

Seam placement: the Z-seam is the point where each layer's perimeter loop starts and ends — a small nub or artifact is inevitable at this transition. Aligned seam places it at the same XY position every layer, producing one predictable vertical line that can be oriented toward a hidden face. Scattered seam randomizes placement, distributing artifacts so no single line forms but producing a stippled surface texture. Rear seam always places the seam behind the part relative to the printer's front — useful when front-facing surfaces are important.

Ironing: a post-infill toolpath pass where the nozzle travels over the top solid surface at 70–90% reduced flow rate, heating and reflowing the top layer to eliminate the ridge pattern produced by the fill lines. Adds 10–30% to print time for the top layers but produces a remarkably flat, near-injection-molded top surface — critical for functional mating surfaces (lids, buttons, flat-display pedestals).

7. Post-processing and surface finish

Raw FDM parts have visible layer lines at 0.1–0.3 mm pitch and a matte thermoplastic surface. Post-processing techniques range from sanding (accessible, reliable) to chemical smoothing (fast, chemistry-dependent) to epoxy coating (finish-level quality).

Sanding progression: start at 80 grit for heavy material removal (thick ABS layer lines), 120–180 grit for moderate removal (standard PLA layers), 220–400 for smoothing sanding marks, then wet sanding at 800 → 1200 → 2000 → 3000 grit with water lubrication to prevent heat buildup (thermoplastics are friction-sensitive). Dry with compressed air and check with raking light — remaining scratches show as reflective lines. Total material removed: 0.1–0.3 mm on ABS/ASA, 0.05–0.15 mm on PLA (PLA is harder, slower to sand but chips rather than clogging sandpaper).

Acetone vapor smoothing: works on ABS, ASA, and HIPS — polymers soluble in acetone. Technique: place 15–25 mL acetone on paper towels at the bottom of a sealed glass or acrylic chamber (never polystyrene — acetone dissolves it). Suspend the print on a wire rack 10–20 cm above the acetone. The vapor rises, condenses on the cold print surface, and dissolves the outer 0.05–0.3 mm of polymer, which then reflows by surface tension and re-solidifies as a glossy surface that entirely obscures layer lines. Exposure time: 5–10 minutes for light smoothing, 20–30 minutes for high gloss (risk of detail loss in fine features). Safety: acetone flash point −20°C — fire risk; styrene VOC emission from ABS; operate in a fume hood or outdoors.

Epoxy coating (XTC-3D, AeroMarine 300/21): two-part epoxy mixed per manufacturer specifications, brush-applied in thin coats at 0.2–0.5 mm per application. Epoxy self-levels by surface tension, filling layer lines on horizontal and near-horizontal surfaces. Vertical surfaces receive less fill — thin coats on multiple passes required. Cures fully in 4–6 hours at room temperature (24 hours for full mechanical strength). Coverage: 50–100 g kit covers 12–20 average-sized prints. Advantages: works on all FDM materials (PLA, PETG, ABS, TPU after priming), adds meaningful impact resistance improvement, excellent paintability.

UV resin dip/brush for miniatures: prime with rattle-can grey filler primer (blocks UV transmission into the thermoplastic substrate, ensures adhesion). Brush a very thin layer of UV resin (0.1–0.2 mm) over all surfaces using a soft brush. Cure with a 405 nm UV lamp for 60–120 seconds. The UV resin fills micro-scratches and micro-layer-line ridges at the detail level without obscuring fine surface detail like heroic minature faces. Chain the sanding → UV resin → sand 1200-grit → UV resin → paint sequence for museum-quality miniature finishes.

Painting: adhesion primer is essential — rattle-can grey filler-primer creates a tooth for subsequent coats. Acrylic paint adhesion to primed surfaces is excellent and permanent. For airbrushing: gravity-feed airbrush, 0.3–0.5 mm needle size, PSI 15–20 for detail, 20–30 for base coats, reduce paint to 40–60% water/medium by volume. Enamel paint (oil-based) requires acetone-proof primer on PLA — standard rattle-can primer is fine on ABS but dissolves PLA slightly (use lacquer primer instead). UV-clear coat for outdoor parts using ASA or ABS (Rust-Oleum 2X Ultra Cover Clear, or Krylon UV-Resistant Clear).

8. The Apple Tax

On November 1, 2026, Apple's 30% App Store fee applies to all in-app purchases on iOS — including Patreon subscriptions processed through the Patreon iOS app. The mechanism: when an iOS Patreon patron pays their pledge through the app, Apple takes 30% of the gross payment before Patreon receives anything. Patreon then takes their 8–12% platform fee from the remaining 70%, and passes the rest to the creator. The creator receives approximately 60–62% of what the patron paid — roughly half the expected income from iOS patrons.

3D printing content skews heavily iOS: tutorial videos on Bambu, Prusa, Creality settings drive 65–78% iOS viewership on YouTube (where most 3D printing instruction lives). CAD creators explaining Fusion 360, OnShape, or Blender modeling techniques see 70–82% iOS. Miniature printing — presupported STL file creators, Warhammer/DnD terrain designers — see the highest iOS rates at 75–88%, because the gaming community consuming miniature content is overwhelmingly mobile-first.

A 3D printing creator earning $200/month with 70% iOS patrons loses $42/month ($504/year) to Apple's cut from November 2026. At $350/month with 74% iOS: $77.70/month ($932.40/year). At $500/month with 78% iOS: $117/month ($1,404/year). These are amounts the patron paid that never reached the creator — not Patreon's 8–12% fee, which comes out of the amount Apple left.

The fix is web-only billing: patrons who subscribe through patreon.com in a mobile browser or desktop browser bypass the App Store entirely. Patreon's own documentation confirms that web-subscriptions do not trigger Apple's IAP requirement. The operational challenge is convincing iOS patrons to re-subscribe via browser — requiring a campaign of direct messages, a pinned post explaining the change, and a short link to the web subscription page that bypasses the app.

For file creators (STL packs, presupported resin files, parametric models), the Apple Tax conversation is especially stark: the deliverable is entirely digital, delivered instantly, and the patron is already using a browser to download files. The Patreon iOS app subscription is the only iOS touch point in the entire transaction. Switching to web billing costs patrons nothing except the 30 seconds of re-subscribing. The creator keeps $42–$117/month depending on scale.