Fused filament fabrication

An ideal 20.0 mm build with a 0.20 mm first layer takes 100 layers with 0.20 mm subsequent layers or 199 layers with 0.10 mm subsequent layers. Use the slicer's actual preview and account for travel, extrusion and cooling when quoting cycle time. Set dimensional acceptance from part measurements.

In fused filament fabrication (FFF), thermoplastic roads bond during material extrusion. NIST ties that weld to temperatures of the extrudate and previous layer. Verify bonding separately from shape. NIST, Thermography and Weld Strength Characterization of Thermoplastic Extrusion 3D Printing, May 23, 2016

Define the material and manufacturing route

FFF describes the filament-fed process, while material extrusion is the broader category. NISTIR 8059 identifies Fused Deposition Modeling (FDM) as Stratasys terminology and discusses the influence of polymer characteristics, temperature and bead geometry on fusion. A pellet-fed extrusion system or a metal-filled feedstock requiring later debinding/sintering should not inherit properties from an ordinary thermoplastic FFF part. NISTIR 8059, Materials Testing Standards for Additive Manufacturing of Polymer Materials, May 2015

Specify the exact filament grade and supplier, color/fillers where relevant, lot, nominal filament diameter, storage state, printer/nozzle and approved profile. “PLA,” “nylon” or “carbon-filled” alone is not a complete manufacturing specification. Distinguish a prototype used to check fit from a production part required to sustain load, temperature, chemicals or repeated use.

Keep the released model and drawing separate from the manufacturing file. Record units, model revision, slicer/version, orientation, layer strategy, walls, infill, supports and any local modifiers. Review the sliced result: thin walls, holes and small text may be represented differently from their ideal CAD surfaces.

Orientation, roads and layer bonding

Choose orientation against the actual loads and functional surfaces. Stratasys identifies FDM parts as anisotropic. Orientation changes loading relative to the deposited structure, support placement and layer stepping. A tensile result in one orientation is not a general material allowable. Stratasys ABS-M30 data sheet, printed p. 8

Nozzle temperature, material delivery and cooling are coupled to bonding. NIST’s 2016 study compares extruder temperature and feed rate with cooling histories and weld fracture strength. This supports controlling the thermal route; it does not justify “hotter is always stronger” or one temperature for every grade, nozzle and geometry. NIST weld-strength study

For qualification, identify walls, raster/infill arrangement, seams and top/bottom construction as well as orientation. An infill percentage is a slicing input, not a direct percentage of tensile strength or a guarantee of uniform internal density. Test representative construction when those details matter.

Moisture: storage and drying are different

Prusa’s drying guidance describes moisture-related surface problems and, in some materials, reduced layer adhesion, blobs or bubbling during extrusion. Sensitivity differs between filaments. Its drying recommendations are material-specific and also consider the spool’s temperature resistance. Do not apply a nylon drying cycle to a different polymer or assume a spool can withstand any temperature suitable for its filament. Prusa: drying filament

The Prusa USS Drybox illustrates the distinction: its desiccant enclosure slows moisture uptake during printing, but the manufacturer explicitly says it does not dry already moist filament. Agree the conditioning/storage route and time exposed before printing; a low air-humidity reading in a box is not itself proof of filament moisture content. Prusa USS Drybox

When evaluating a failed print, compare material condition, extrusion behavior and profile records rather than attributing every string or weak layer to moisture. Drying is one controlled variable, not proof that the printer, filament and thermal settings are otherwise correct.

Supports and functional access

PrusaSlicer’s support documentation describes generating structures beneath unsupported regions and suggests changing orientation or splitting a model to reduce support demand. Support removal must be planned before printing. Identify contact surfaces, removal access and whether a passage can actually be cleared. A supported underside may need a different appearance or dimensional requirement from an upward-facing surface. Prusa: support material

Specify any soluble support as an exact compatible material/process combination, including its removal and subsequent drying or conditioning. Do not assume that a cavity is clean merely because soluble material was selected. Inspect critical channels, mating faces and thin features after supports and finishing are complete.

Layer height is not a dimensional tolerance

Prusa defines layer height as slice thickness and explains that it mainly changes vertical resolution and print time. It does not change horizontal resolution in the same way; nozzle diameter and extrusion strategy also matter. Its nozzle/layer guidance is specific process guidance rather than a tolerance statement. Prusa: layers and perimeters

A small layer height does not establish hole diameter, flatness, shrinkage compensation or final size accuracy. Define dimensions on the finished part, including the measurement state and condition. Trial mating features and critical surfaces using the actual orientation and route rather than extrapolating from a calibration cube.

Original bounded layer-count example

Assume a 20.0 mm vertical model, a 0.20 mm first layer, and uniform subsequent layers ending exactly at the top:

• At 0.20 mm subsequent height: 1 + (20.0 − 0.20)/0.20 = 100 layers.
• At 0.10 mm subsequent height: 1 + (20.0 − 0.20)/0.10 = 199 layers.

The second plan has 99% more layers. This does not prove a 99% longer cycle: travel, extrusion rate, cooling constraints, first-layer settings and other operations affect time. It also does not establish ±0.10 mm accuracy. These are invented arithmetic assumptions, not an instruction for a specific slicer; variable layers or a non-divisible model height require its actual preview.

Named property example and acceptance

Stratasys’ ©2025 sheet MDS_FDM_ABS-M30_0925a, Table 4, reports ABS-M30 Black on an F900 with T16 tip, at the printed 0.25 mm (0.010 in.) layer setting. Its ASTM D638 tensile-at-break results are 28.1 MPa in XZ and 26.8 MPa in ZX; elongation at break is 8.1% and 1.8%, respectively. Table 5 concerns a different printer/head. The cover labels the data typical and unsuitable for design specifications or quality control; retain the machine, orientation and test context. ABS-M30 sheet, printed pp. 1, 9–10 and 13

Define dimensional, surface and material-performance acceptance separately. For service parts, identify the required load duration, environment, interfaces and failure consequences, then obtain evidence for that application. A short tensile test does not establish creep life, fatigue, impact performance, chemical compatibility or leak tightness.

A useful handoff includes the released geometry, exact material and conditioning, printer/nozzle/profile, build orientation, internal construction, support/removal route, finishing and inspection stage. Agree which changes require a new trial. The cited Prusa pages are live documentation without a separately verified technical issue date here; the Stratasys footer verifies its identifier and ©2025, not an exact publication day.

Related reference

Manufacturing processes