What infill percentage should you use? Strength, print time, and material cost

1. Treat infill as an internal structure, not a strength dial

Infill percentage describes how densely the slicer fills the space between the outer walls. At 0%, a suitable model can be mostly hollow. At 100%, the slicer attempts to fill the interior with adjacent extrusion lines. The percentage does not describe the whole part: wall count, top and bottom thickness, layer height, extrusion width, pattern, material, orientation, and bonding quality still matter.

Two models sliced at 20% can behave differently because one has three strong walls and a short load path while the other has thin walls, a stress concentration, or layers oriented across the force. Even the apparent density can differ between patterns and slicers. Compare settings using the same model, printer profile, material, and quality target.

For safety-critical, load-bearing, high-temperature, pressure-containing, or impact-sensitive parts, a generic percentage chart is not a validation method. Define the load case, print test pieces, inspect the failure mode, and apply an appropriate safety factor or use a qualified engineering process.

2. Increase walls before assuming that more infill is the best answer

Outer walls carry much of the tension and bending load because they sit farthest from the neutral center of a part. This is why adding perimeters can improve many functional prints more efficiently than filling the entire interior. Prusa’s guidance similarly emphasizes perimeter count when the goal is a stronger model, while noting that infill contributes to compression resistance.

Local geometry matters. A screw boss, snap fit, hinge, narrow neck, or hole can fail before the broad interior does. Extra walls, modifiers around fasteners, fillets, a better print orientation, or a redesigned load path can place material where it works harder. Raising global infill may add weight far away from the actual failure point.

Do not optimize only one setting. Top and bottom layers need enough support and thickness, walls need reliable bonding, and the interior must keep broad surfaces from buckling or sagging. A well-designed shell with moderate infill can outperform a poorly oriented, nominally solid part.

• Identify tension, bending, compression, impact, and fastener loads.
• Check wall thickness and local features before increasing global infill.
• Orient layers so the weakest inter-layer direction is not exposed unnecessarily.
• Use modifier regions when only one area needs additional material.

3. Use ranges as test starting points, not universal recommendations

Low-density infill can be adequate for visual models whose walls and top surfaces only need internal support. General prototypes often begin in a moderate range. Functional parts may justify more internal structure, but their required density depends on geometry and the direction and duration of the load. A percentage copied from another model is evidence about that model—not yours.

The ranges below are deliberately broad. Slice at the low end first, inspect the preview for unsupported spans, and test the finished part. Move upward only when the failure mode or surface quality shows that more internal structure is useful. If the walls or layer interface fail first, changing infill alone may not solve the problem.

4. Choose the pattern for the load and the printing behavior

A rectilinear or grid-style pattern is easy to understand and can be efficient, but its response depends on direction. Three-dimensional patterns such as cubic or gyroid distribute structure through multiple directions and can be useful when loads are not limited to one axis. Pattern names alone do not guarantee a stronger part: density, wall connection, extrusion quality, and geometry change the result.

Patterns also change the toolpath. Crossing lines, continuous curves, short accelerations, and repeated direction changes can affect noise, vibration, print time, and the chance that the nozzle contacts already printed material. Lightning-style infill is designed mainly to support upper surfaces with little material; it is not a substitute for structural interior support.

Pick one or two patterns that your printer handles reliably, then compare the slicer’s grams and time for the actual model. A pattern that is theoretically efficient but creates poor intersections or unstable tall features on your machine is not the lowest-cost option.

5. Check top surfaces before reducing infill too far

The first top layers bridge across the spaces in the infill. If those gaps are too wide for the material, cooling, speed, and layer height, the strands can sag and leave rough or incomplete upper surfaces. Increasing top thickness may hide some defects, but the underlying span still needs to be printable.

Large flat roofs, curved upper surfaces, and thin decorative shells may need a different balance than compact parts. Gradual infill, local modifiers, additional top layers, or a pattern with better support can add material only where the surface needs it. The lowest global percentage is not economical if it creates rejected prints.

Inspect the sliced layer preview rather than judging the percentage field alone. Look at infill spacing under broad roofs, connections to walls, narrow regions the slicer treats as solid, and areas where supports or internal bridges change the material estimate.

6. Compare slicer output because percentage does not map directly to cost

More infill generally adds extrusion distance, material, and print time, but the relationship is not perfectly linear. Walls, top and bottom layers, travel, acceleration, minimum layer time, and solid internal regions remain even when sparse infill changes. A small part dominated by walls may show little difference; a large enclosed volume may change dramatically.

Create a controlled comparison by slicing the same orientation and profile at 10%, 20%, and 40%. Record total filament grams and estimated time for each version. If your slicer can export project files, save the comparison so later profile changes do not silently alter the baseline.

Put the reported grams and a time estimate corrected from actual printer records into the cost calculator. Material cost follows grams. Machine depreciation and electricity follow hours. Hands-on labor may stay almost unchanged unless denser infill makes removal, cooling, or post-processing different.

7. Validate the cheapest setting that passes the part’s real job

Begin with purpose and failure criteria. A display model may only need clean surfaces and normal handling. A fixture may need repeatable dimensional stability. A bracket may need to survive a known load for a defined duration. Write the test before changing the percentage.

Slice a baseline, inspect the interior, and print a representative sample. If it fails, identify where and how. Increase walls for shell failure, change orientation for layer separation, reinforce a local feature for stress concentration, or add infill when the interior crushes, buckles, or fails to support surfaces. Change one major variable at a time.

Record the final profile with its model revision, material, printer, nozzle, layer height, walls, pattern, percentage, weight, time, and test result. That turns an arbitrary preset into a repeatable production decision and gives future quotes a defensible cost basis.

• Define the part’s purpose and unacceptable failure.
• Slice at several densities with every other setting fixed.
• Compare grams, corrected print time, and surface support.
• Print and test the geometry in its real load direction.
• Keep the lowest-cost profile that consistently passes.