How to Choose Infill Settings | Density and Pattern Strength Comparison

# How to Choose Infill Settings | Density and Pattern Strength Comparison When it comes to 3D printing infill, the instinct is to crank up density for more strength. But with functional parts, that approach alone falls short surprisingly often. In my own setup, bumping density from 20% to 40% didn't change the failure mode at all -- prints kept splitting along layer lines. The moment I increased wall count from 2 to 4, though, the part suddenly felt like something I could actually rely on. This article is for anyone using Cura 5.x, PrusaSlicer 2.x, or Bambu Studio who isn't sure how to pick the right infill density and pattern. The core idea: start around 20% with Gyroid or a Cubic variant, then adjust wall count as needed before touching density. We'll cover density ranges as a starting point -- 10-15% for decorative prints, 20-30% for general use, 40-60% for load-bearing parts -- along with a quick comparison of Gyroid, Grid, Cubic, Adaptive, and Lightning patterns, followed by step-by-step instructions for each slicer. Density starting points are 10-15% for decorative prints, 20-30% for general use, and 40-60% for durability-focused parts. From there, we'll compare Gyroid, Grid, Cubic, Adaptive, and Lightning patterns briefly, then walk through the actual steps to change settings in each slicer so you can reproduce the results.

What Is Infill? The Basics of Density and Pattern

Infill is the internal fill structure that determines how the inside of a 3D-printed model gets filled. You can't see it from the outside, but it has a major influence on how a print behaves. Density (%) is the proportion of the interior that gets filled with material. Walls/perimeters form the outer shell thickness. Top/bottom layers are the solid layers that close off the upper and lower surfaces. Pattern is the geometric shape used to fill the interior. As the Prusa Knowledge Base's infill documentation explains, infill affects not just strength but also material consumption, print time, appearance, and how well top surfaces are supported.

This setting is easy to misjudge if you treat it as "more is always stronger." Pick up a low-density part and you'll notice immediately how light it is -- tap it with your fingernail and you get a distinctly hollow sound. That's expected when the interior has a lot of open space. But reduce top layers while running low density, and pressing on a broad flat surface produces a slight flex. I read that flex not as a sign of insufficient density, but as a sign that the scaffolding underneath the top surface is too sparse. Infill serves double duty: it's a structural member and a foundation for the layers above it.

Density and Pattern Serve Different Roles

Density controls how much material goes inside. Pattern controls how it gets arranged. At the same 20%, Gyroid, Grid, and Cubic variants produce very different internal geometries. Gyroid is a continuous 3D curved structure that handles loads from multiple directions well, making it a strong general-purpose candidate. Grid creates an easy-to-understand lattice, though it has crossing points within the same layer. Cubic and Adaptive Cubic build three-dimensional cells -- Adaptive variants concentrate density where it's needed most, which saves material and time on larger prints.

One thing that's easy to overlook is the connection to top surface quality. Infill divides the open space beneath broad top surfaces into shorter spans, reducing the bridge distance that top layers have to cross unsupported. When density drops too low, those spans widen and top surfaces start to sag or ripple. On the other end, patterns like Support Cubic and Lightning are designed primarily to efficiently support the areas directly beneath top surfaces rather than maximizing overall internal rigidity. They work well for box-shaped prints and display models, but they aren't designed as primary load-bearing structures.

Think of 0% and 100% as Extreme Settings

Setting infill to 0% leaves the interior completely hollow. The sides might look fine, but any shape with a top surface loses its support scaffolding almost entirely, causing top layers to collapse. Broad flat surfaces in particular become nearly impossible to bridge cleanly. Outside of vase-mode-style geometries, 0% has very limited practical use.

At the other end, 100% might seem like the ultimate strength setting, but the slicer behavior is worth understanding. In PrusaSlicer, 100% infill automatically switches the pattern to Rectilinear. When filling the interior completely, straight-line paths make more sense than patterns designed around air gaps. So even if you've selected a different pattern on screen, the actual toolpath changes at 100%. It's more accurate to think of 100% not as "your chosen pattern made denser" but as a separate complete-fill mode with its own logic.

Cross-Sections Make the Roles Clearer

Words only go so far, so a cross-section view helps organize the concepts. The outer shell is the perimeters, the solid plates on top and bottom are the top/bottom layers, and the framework in between is the infill.

Cross-Section Diagram

 ┌────────────────────┐  ← Top layers (solid layers on the upper surface)
 │████████████████████│
 │█                  █│  ← Walls / perimeters (outer shell)
 │█   /\/\  /\/\     █│
 │█  /    \/    \    █│  ← Infill (internal fill structure)
 │█  \    /\    /    █│     Density = how much of this interior is filled
 │█   \/\/  \/\/     █│     Pattern = the shape of the fill
 │█                  █│
 │████████████████████│  ← Bottom layers (solid layers on the lower surface)
 └────────────────────┘

Legend:
- Walls / perimeters = outer shell surrounding the model
- Top / bottom layers = solid layers closing the top and bottom
- Infill = internal fill structure
- Density (%) = fill ratio
- Pattern = fill geometry

With this cross-section in mind, it becomes clear why low-density prints can feel surprisingly sturdy on the sides (if walls are thick enough) while top surfaces deteriorate quickly when internal support is lacking. Whenever I'm unsure about infill settings, I picture this cross-section first. Where does the load land? Where are the top surfaces? Answering those questions tells me whether to increase density, change the pattern, or -- as mentioned earlier -- add more walls.

The Short Answer: Quick Reference for Density and Pattern

If you need to decide fast, here are solid starting points: 10-15% for decorative prints, 20-30% for general use, 40-60% for parts that need durability. The Prusa Knowledge Base infill documentation also notes that low to medium density works for most prints, and going above 30% isn't commonly necessary. In practice, prototypes and enclosures tend to work fine in the 10-20% range, while functional brackets and jigs are easier to dial in starting from 20-30%. Starting a load-bearing part at 10% tends to leave both top surface support and internal rigidity lacking.

For pattern selection, sorting by role makes things straightforward. Gyroid is the top general-purpose pick. Cubic variants balance strength and efficiency. Grid is a familiar baseline. Lightning and Support Cubic lean toward top surface support. Gyroid's continuous 3D structure handles multi-directional loads well, making it hard to go wrong with for either functional or decorative prints. In my experience, switching from Grid to Gyroid at the same 20% density noticeably smooths out head movement -- the small vibrations from tight direction changes settle down. The interior also looks more "filled" even at identical density values, giving prints a more solid feel. That's the kind of practical difference that doesn't show up in theoretical strength comparisons.

Meanwhile, Cubic and Adaptive Cubic are strong candidates when you need strength but want to keep time and material in check. Cubic builds three-dimensional cells, so it pairs well with parts that receive loads from unpredictable directions. Adaptive variants concentrate density where it matters, and the efficiency gains become more apparent on larger hollow shapes. Grid's lattice structure is intuitive and easy to read in cross-section, making it a good learning pattern for beginners who want to understand how setting changes affect the interior.

Getting the use case right for Lightning and Support Cubic is critical. Both prioritize supporting top surfaces with minimal material rather than strengthening the entire interior. They excel at box shapes, display models, and lightweight parts with broad top surfaces, printing faster and using less filament. But they aren't the right choice when mechanical durability is the goal. For strength-focused parts, start with Gyroid or a Cubic variant instead.

Here's a comparison table for quick initial decisions:

A practical starting workflow: set Gyroid at around 20%, bump to 25-30% if you want more rigidity, and consider 40%+ for parts under real load. Going from 20% to 40% roughly doubles the material used for infill alone. On paper that's obvious, but the weight difference is striking when you hold the same model side by side. That's exactly why the 40-60% range should be reserved for parts that genuinely need the durability, not used as a comfort blanket.

These density ranges and patterns are strong starting candidates, not final answers. The optimal setting shifts with material, geometry, and load direction, so treat them as reliable defaults to build from rather than a fixed ranking.

How Density Affects Your Prints: 10%, 20%, 40%, and 60%+

Density isn't a case of "higher is better." A more useful lens is where improvements start to kick in and where diminishing returns set in. 20% works as a practical baseline because parts without extreme strength demands often hold up fine at this level. Internal support is adequate, and weight plus print time stay reasonable. I typically start prototypes and small jigs around 20%, then work from there: "if it's not strong enough, go higher" and "if the top surface is weak, check top layers and walls too." That progression tends to produce the fewest surprises.

10-15%: Decorative and Prototype Territory

The 10-15% range suits visual mockups, display pieces, and dimensional check prototypes. Prints come out light and finish quickly, making it a go-to for getting shapes into your hands fast. The Prusa Knowledge Base also notes that typical models can print successfully in this density range.

That said, this range prioritizes feasibility over rigidity. Box shapes with broad top surfaces or models with large upper areas are prone to the top layer scaffolding being too thin, resulting in slight surface sag or roughness. The light weight and fast print times are appealing, but for parts where you want clean top surfaces, you'll need to account for support beyond just density.

20-30%: Why This Range Becomes the Default

20-30% is the most versatile range for everyday mechanical parts and test prints. The interior is connected well enough, top layers get adequate support, and the increases in material and time are still modest. The reason 20% gets called the baseline is that it sits at the balance point of strength, time, and material. If strength requirements aren't extreme, 20% handles most situations, and you can fine-tune to 25% or 30% from there.

From a hands-on perspective, 20% hits that sweet spot of "not too weak, not too heavy, not too slow." Running the same model at 20%, 40%, and 60% back-to-back, the time difference from 20% to 40% doesn't feel as bad as the slicer numbers suggest. It increases, sure, but it's still tolerable once the print is running. Push to 60%, though, and there's a palpable shift to "this is never going to finish." That gap feels bigger than the numbers alone suggest, which is another reason 20% makes sense as a starting point.

40-60%: Durability-Focused, But Not Free

40-60% is the range for parts where durability matters. You can feel the density when you grip them, and flex resistance improves noticeably. Going from 20% to 40% nearly doubles the material used for infill, and the weight increase is unmistakable. That trade-off means this range is worth reserving for the parts that actually need it.

The practical catch is that above about 40%, time and material costs start outpacing strength gains. Up to 40%, "a bit heavier, a bit longer" is manageable. As you approach 60%, head travel and fill volume spike, and print time becomes hard to ignore. With less empty space inside, the "work required to fill" ramps up steeply even for the same pattern.

As a side note, some community observations suggest that in Cura's higher density ranges (roughly 60%+), toolpath generation behavior changes. These are version- and plugin-dependent observations, so if citing specific cases, include the primary source URL. The takeaway for practical purposes: at high densities, it's worth thinking about slicing efficiency and toolpath behavior alongside raw strength.

100%: Not "Maximum Strength" -- A Different Mode Entirely

100% fills the interior completely, but it doesn't belong on the same continuum as 60% or 80%. PrusaSlicer automatically switches to Rectilinear at 100% infill. You're not pushing your selected Gyroid or Cubic to the extreme -- the slicer enters a separate complete-fill logic.

At this level, material consumption is heavy and print times balloon. Heat buildup inside the part increases, shrinkage stress rises, and the practical handling gets noticeably harder. Solid fill doesn't automatically mean bulletproof, and most functional parts don't actually need 100%. Before jumping to complete fill for strength, it's usually more rational to explore what wall thickness and orientation changes can achieve.

A summary table makes density decisions much faster:

Framed this way, 20% as the baseline isn't just convention -- it's the density where parts work well enough without getting heavy. Up to 40%, the strength-to-cost ratio is still readable. Past 60%, diminishing returns become obvious. Knowing where those thresholds fall eliminates most of the guesswork.

Pattern Strength Comparison: Gyroid, Grid, Cubic, Adaptive Cubic, and Lightning

Ranking patterns by strength alone oversimplifies things. The real evaluation depends on which direction loads come from, how cleanly you need top surfaces supported, and how much time and material you're willing to spend. My approach starts by separating "internal structures designed for mechanical strength" from "material-saving structures designed to support top surfaces." That distinction alone prevents the confusion of comparing Gyroid and Lightning as if they're competing for the same job.

Here's a practical comparison of the major patterns:

Gyroid

Gyroid forms a continuous 3D curved surface throughout the interior. Its key advantage is balanced support across all directions rather than favoring any single axis. That versatility makes it a safe pick whether you're printing functional components or decorative cases.

Beyond the theory, Gyroid is just pleasant to work with. The print head moves without harsh stop-and-reverse motions, the internal structure avoids directional bias, and the balance between strength, appearance, and print stability is easy to maintain. For parts where "I don't know yet which direction the force will come from," Gyroid takes the guesswork out far better than Grid or Lines.

Top surface support is also well-behaved. The continuous structure creates consistent scaffolding beneath top layers, so broader top surfaces hold up better than with many alternatives. Gyroid's strength isn't about dramatic improvements -- it's about a high floor that's hard to get wrong.

Grid

Grid is a familiar lattice with crossing lines within each layer. The internal structure is intuitive, it works well as a comparison benchmark, and it's been a default choice for good reason. Strength performance is middle-of-the-road -- not weak by any means.

That said, Grid's defining characteristic is also its limitation. Same-layer crossing points mean the nozzle has to navigate intersections, which can affect travel smoothness and surface consistency. Compared to Gyroid, you sometimes notice the difference in how settled the print head's motion feels. Structurally intuitive, yes, but the crossing points make it a step behind in both mechanical behavior and print stability.

So Grid is a "reasonable default" but not something you'd choose aggressively today. Its best use cases are when you want to intuitively understand the internal structure or when you're building a reference print for comparison. For strength alone, the more three-dimensionally connected Gyroid and Cubic variants are usually easier to recommend.

Rectilinear / Lines

Rectilinear and Lines are straightforward patterns built from straight passes. The standout feature is simple: speed. Toolpaths are clean, wasted motion is minimal, and for prototypes or form checks, they're extremely practical.

Strength-wise, they're directional for better and worse. Straight lines running parallel provide good rigidity along one axis but less resilience perpendicular to that direction. This isn't a balanced all-direction pattern -- it's a clean baseline with directional bias.

What makes this pattern important is its relationship to 100% infill behavior. The Prusa Knowledge Base shows that complete fill switches to Rectilinear automatically. When filling every gap, simple straight lines make more sense than patterns designed around air space. Rectilinear isn't a budget option -- it's the high-speed baseline and the standard for complete fill.

Cubic / Adaptive Cubic

Cubic variants create three-dimensional cells, pairing well with parts that receive loads from multiple directions. Where Gyroid supports through continuous curves, Cubic takes more of a 3D framework approach. For brackets or parts where you can't predict the exact load direction, it provides reliable stability.

Prusa documentation and forum discussions show examples of Adaptive variants saving material and time. Specific savings percentages vary significantly by source and conditions, so cite primary sources when quoting numbers.

From hands-on experience, Adaptive Cubic really shines on larger prints. Where uniform 20% Gyroid or Grid fills every corner dutifully, the Adaptive approach keeps support where it's needed -- near top surfaces and outer walls -- while thinning out deep interior zones. It's not sacrificing strength so much as prioritizing where support matters. For large parts where material savings are a goal, this is a smarter approach than simply dropping density uniformly.

Support Cubic / Lightning

Support Cubic and Lightning belong in strength comparison tables but operate on a different philosophy. Their primary purpose is supporting top surfaces and minimizing material, not maximizing the mechanical strength of the entire part. For box shapes, display models, and cosmetic prototypes where top surfaces just need to close cleanly, they're highly effective.

PrusaSlicer and various forum posts include case studies showing time and material savings with Adaptive Support and Support Cubic variants, though reported values vary by environment and model. The key point is that these patterns are designed around "efficiently supporting top surfaces" rather than "maximizing overall rigidity." When citing specific percentages, always include the primary source URL.

This means choosing Lightning for mechanical strength is a category error. For lightweight mockups, display pieces, form checks, and large shell shapes that only need a closed top, it's excellent. For brackets, jigs, or anything under real load, it belongs lower on the priority list.

A Note on Numerical Evidence

Numbers vary by source and conditions, so reference primary sources wherever possible. When no source is available, label data as "observed example" with appropriate caveats. The most usable numbers relate to density ranges and material savings. The Prusa Knowledge Base infill documentation, for instance, shows that typical models work well at lower densities and that high density isn't always necessary, with Adaptive Cubic and Adaptive Support Cubic offering time and material reduction examples. These figures aren't proof of absolute pattern strength rankings -- they're indicators of which design philosophies save what and where.

The practical takeaway: Gyroid for multi-direction balance, Grid as a crossing-based baseline, Rectilinear as the fast standard, Cubic for multi-directional load cases, Adaptive and Support Cubic for material savings, Lightning not for mechanical duty. With this framework, pattern comparison becomes purpose-driven selection rather than a hunt for "the strongest."

インフィル ｜ Prusa Knowledge Base

インフィルの主な目的は、上層部に内部サポートを提供することです。これがないと、上層部は空間に対してブリッジしてプリントする必要があります。インフィルは、プリント速度、構造強度、フィラメント消費量、さらにはプリントされたオブジェクトの外観にも

help.prusa3d.com

When You Need More Strength, Check These Settings First: Walls, Top Layers, and Local Reinforcement

When a part isn't strong enough, the first impulse for many people is to increase infill density. But as Prusa's official documentation emphasizes, strength is primarily determined by wall count -- shell thickness -- and keeping that in mind dramatically reduces decision fatigue. Increasing walls from 2 to 3 or 3 to 4 provides more resistance to bending and twisting than going from 20% to 40% density. This is especially true when failures occur at stress concentrators like corners, holes, and edges rather than across broad surfaces.

I saw this clearly while prototyping a shelf bracket. Keeping infill at 20%, I changed wall count from 2 to 4 and top layers from 3 to 6. The cracking that had been happening when tightening screws stopped completely. What became obvious was that thickening the outer shell and strengthening the load paths through it mattered far more than packing the interior slightly denser. It might sound like an exaggeration to say one setting change transformed the part, but with functional prints, that's genuinely what happens.

Start with Wall Count

A clean decision rule: if the part is cracking, twisting, or splitting around screw holes, look at wall count first. Going from 2 to 3, or 3 to 4 walls is straightforward and impacts side-wall rigidity more than the numbers suggest. Brackets, case mounting ears, screw holes, and impact-prone corners -- anything where force hits the outside -- benefit substantially.

Screw holes are a textbook example. Fastener forces concentrate at the hole perimeter, and having adequate wall thickness around that hole matters far more than high overall density. Hook features and snap tabs follow the same logic: if walls are thin, the part fractures from the outer surface inward regardless of how dense the interior is. "Thicken the walls first" is a consistently reproducible approach to strength problems.

Top Layer Issues? Fix with Top Layers

When flex or surface roughness on the top face is the concern, adjusting top/bottom layer count is usually a more direct fix than increasing infill density. Symptoms like sinking top surfaces, visible line gaps, or a "springy" feel when pressing on a flat top typically mean the closing layers are insufficient. Adding top layers addresses this more efficiently.

For targeted top-surface support, patterns like Support Cubic offer an alternative approach. Rather than strengthening the whole part, they place support structure directly beneath the top surface to prevent thin ceilings from sagging. This works well for box lids and parts with large internal cavities. Trying to solve top surface issues by pushing density alone tends to burn material and time while the actual improvement comes slowly.

Increase Density Only Where Needed

When the entire part doesn't need to be heavier, local reinforcement is a highly practical strategy. PrusaSlicer's modifier meshes, process splits, or slicer features that allow per-region infill settings let you increase density only around screw holes, insertion points, and load-bearing areas while keeping the rest of the part at standard density.

The advantage is clear: you can treat areas that need strength differently from areas that are simply empty space. A large cover part, for instance, gains nothing from high density in its broad hollow center, but the mounting bosses and screw holes absolutely need reinforcement. Instead of pushing the whole part to 40%, targeting just the fastener areas with higher density or extra walls keeps weight and print time down.

Screw hole reinforcement pairs naturally with this approach. Adding a few millimeters of high-density material around hole perimeters, or thickening walls locally around bosses, reduces cracking and buckling during fastening. Functional part failures almost always happen at a single point that gives way first, so targeting that point with reinforcement makes far more sense than uniformly increasing density everywhere.

In diagram form, the priority reads: Increase walls -> Add top/bottom layers -> Local high-density reinforcement -> Raise overall infill density. This arrow sequence captures the section's intent. Infill density matters, but as a first move for strength improvement, it belongs at the back of the line for functional parts.

Slicer Settings Walkthrough: Cura 5.x / PrusaSlicer 2.x / Bambu Studio

The General Workflow

Establish a consistent workflow and slicer differences stop being confusing. The process is: load a model, change Infill Density and Infill Pattern, check the preview to verify top surface support and material/time changes, then run a small test print to evaluate. Don't stop at the settings panel -- previewing is part of the workflow.

Recording changes in a consistent format helps too. Note density as "20% -> 30%" and pattern as "Grid -> Gyroid" with before-and-after pairs. Keeping these records for the same model lets you isolate whether you gained actual improvement or just added time.

In the preview, focus on whether the area directly beneath the top surface has adequate support rather than examining the infill structure itself. If the space below a broad flat surface looks sparse, either density is too low or the pattern doesn't fit the use case. I personally find switching to Gyroid in preview quite satisfying -- the layer-to-layer continuity is visually traceable, making it easy to understand how the internal structure supports itself. During printing, the head avoids harsh direction reversals and acceleration stays gentle, which is reassuring to watch.

For test prints, 20mm calibration cubes or duplicated small models work well. Run separate comparisons: one changing only density, another changing only wall count. Track at minimum weight (g), print time, and failure mode (where it broke). Pairing numbers with failure analysis makes setting changes significantly more meaningful.

Cura 5.x

In Cura 5.x, place your model, open the print settings on the right side, and navigate to the infill section. Edit Infill Density and Infill Pattern. The exact UI layout varies by profile view, but in practice you're working within the infill settings group.

A useful first comparison: change Infill Density from 20% to 30% and Pattern from Grid to Gyroid. Grid gives you a readable baseline, while switching to Gyroid shows improved internal continuity that tends to work better for general-purpose parts. In Cura's preview, Gyroid's layer connections are easier to follow than Grid's, and head movement appears more settled.

After making changes, slice immediately and check the preview. Three things to look for: whether top surfaces have sufficient support underneath, how much material usage increased, and whether the print time increase is acceptable. Moving from 20% to 30% definitely increases interior fill, but whether the improvement you need is top surface quality or overall rigidity determines how much that change actually helps. Tracing layers just below the top surface in the layer view makes the difference visible.

Note that community reports suggest Cura's pattern handling may change at higher density levels (e.g., community observations). Don't rely on the pattern name alone -- verify actual toolpath generation in the preview. Where possible, add URLs to relevant posts or test articles.

PrusaSlicer 2.x

In PrusaSlicer 2.x, load the model, open print settings on the right, and find the infill section to change Infill Density and Fill Pattern. PrusaSlicer's pattern names are relatively clear, making it straightforward to choose between Rectilinear, Gyroid, Cubic, and Adaptive Cubic.

A productive comparison starting point: change Infill Density from 20% to 40% and Pattern from Rectilinear to Adaptive Cubic. Rectilinear provides a readable baseline. Adaptive Cubic concentrates density where it's needed, and the difference becomes apparent on models with substantial internal space. Prusa's documentation highlights Adaptive Cubic's material efficiency, so you get a "strong-looking" interior without uniformly filling every corner.

In the preview, look beyond just whether the interior got denser. Check how the area beneath the top surface is supported and how empty space distributes layer by layer. Adaptive Cubic doesn't fill uniformly -- it concentrates around top surfaces and geometry changes, which can look "lighter" if you're expecting a regular grid. As long as support is concentrated where it's needed, that's the intended behavior.

One PrusaSlicer-specific behavior to understand: at 100% infill, the pattern automatically switches to Rectilinear. Even if you've selected a different pattern, complete fill uses straight-line paths. When including 100% in comparisons, treat it as a separate complete-fill category rather than an extension of the 20%-40% pattern continuum.

Prusa Knowledge Base

All information you need to know about Original Prusa 3D printers. Assembly manuals, print quality troubleshooting, cali

help.prusa3d.com

Bambu Studio

Bambu Studio follows the same general workflow: place the model, open infill settings, change Infill Density and pattern, then check the preview.

A useful starting comparison: change Infill Density from 15% to 25% and Pattern from Grid to Cubic or a Subdivision variant. 15% represents the lightweight end, 25% is a practical test density for functional parts, and even this difference produces noticeable changes in top support and interior fill. Bambu Studio's preview is well-designed for side-by-side comparisons.

When choosing Cubic variants, pay attention to naming. Bambu Studio lists Cubic, Subdivision, Support Cubic, and Lightning as separate options with similar-sounding names but different purposes. For strength applications, start with Cubic or Subdivision variants. For top surface support and material savings, evaluate Lightning and Support Cubic separately. Lightning in particular produces a visually very light interior since it builds only the minimum structure needed to support top surfaces.

In the preview, look at how much toolpath density concentrates directly beneath top surfaces rather than just the overall interior fill. Support Cubic and Lightning variants may sound robust, but their actual purpose is top surface support, not overall rigidity. Confusing the two leads to parts that used less material but feel unreliable for their intended function.

Screenshot Notes

Effective screenshots for this section follow a one settings screen + one preview screen format for each slicer. For Cura 5.x, PrusaSlicer 2.x, and Bambu Studio, show the screen where Infill Density and Pattern are visible first, followed by the post-change preview.

Settings screenshots should make before-and-after values clear with density numbers and pattern names visible. For example: Cura showing 20% to 30% and Grid to Gyroid, PrusaSlicer showing 20% to 40% and Rectilinear to Adaptive Cubic, Bambu Studio showing 15% to 25% and Grid to Cubic or Subdivision.

Preview screenshots should select a layer position that reveals top surface support. A full overview alone won't convey the differences well -- use a cross-section or layer slider position that shows how infill sits beneath the top surface. If material usage and print time are visible on the same screen, keep those numbers in frame too.

A supplementary comparison of 20mm test cubes printed under identical conditions -- one set varying only density, another varying only wall count -- works well as supporting imagery. In photo captions, briefly note weight, print time, and where failure occurred to visually reinforce how setting changes translate to results.

Recommended Settings by Use Case

When the use case is clear but settings aren't, separating where force is applied from how top surfaces need to be supported makes decisions much easier. The Prusa Knowledge Base notes that general printing stays in the low-to-medium density range with wall count being a major strength factor. In my own experience, "increasing walls and top layers" outperforms "increasing interior density" for targeted strength improvements more often than not.

Here's a quick reference organized by application:

Figurines / Display Bases

Figurine bodies and display bases work well starting at 10-15%, which balances light weight against appearance. Pairing this density range with Gyroid produces natural internal connectivity that supports thin outer shells evenly.

For bases with large hollow interiors, the finish depends more on how the top surface is supported than on overall internal strength. This is where Lightning comes in handy -- it creates minimal internal structure focused on top surface support, saving both material and time. That said, top surface quality depends on more than just infill, so keeping top layers slightly thicker prevents the "light but flimsy-looking surface" problem. Broad flat bases are especially sensitive here.

Figurines tend to be more affected by base settings than body settings. Rather than packing material into the invisible interior, giving the top layers a little extra thickness avoids the cheap feel when you pick the piece up.

General Small Items / Prototypes

Cases, small covers, dimensional-check prototypes, and everyday functional bits below jig-level demands do best at 20-30%. Gyroid or Cubic at this range provides a reliable combination -- enough internal support without unnecessary density, balancing top surface support and everyday rigidity well.

A common prototype trap is over-filling parts where the goal is just visual and dimensional verification, adding print time and weight for no benefit. I usually start in the low 20s and only push toward 25-30% if flex or feel becomes an issue. Most general small items settle well at this stage, and the results are more efficient than just making the interior denser.

Gyroid offers versatile, directionless rigidity that feels consistent no matter how you handle the part. Cubic tilts slightly more toward a "mechanical component" feel. At the prototype stage, where the goal is "not as strong as the final part but not flimsy either," both are reliable choices.

Jigs and Brackets

Load-bearing jigs, L-brackets, and mounting fixtures enter the 40-60% territory. For functional strength, Cubic or Adaptive Cubic combined with 3-4 walls is a strong starting point. Parts under load depend heavily on shell thickness for impact resistance, not just interior fill.

An often-missed opportunity is local reinforcement -- making only the screw hole surroundings high-density rather than the entire bracket. This saves material while producing more controlled failure behavior. Adaptive Cubic is particularly suited to this since it naturally concentrates density without requiring uniform fill across the entire part.

A failure pattern I see regularly: hook-shaped or clip-style brackets where increasing density from 40% to 60% barely extends part life. These parts typically fail because layer orientation is unfavorable for the load direction, and thin walls can't arrest crack propagation regardless of how dense the interior is. Reorienting the part so layers aren't perpendicular to the pulling force, combined with thicker walls, produces dramatically better durability at the same density. That improvement dwarfs what density increases alone can achieve.

Screw bosses and tapped holes deserve separate attention within this category. 4+ walls as a baseline, with 50-60% density localized around fastener areas, reduces cracking noticeably. If direct-tapping into plastic keeps failing, heat-set inserts are often the more stable design choice. The solution to fastener cracking isn't making the whole part harder -- it's adding wall thickness and local density around the fastener.

TPU Parts

TPU diverges from PLA and PETG because the settings depend entirely on whether you're using the flexibility or the compression resistance. For flexibility, 10-20% is the better range. Cable holders, simple bumpers, and parts designed to flex benefit from lower density that lets TPU move the way it's supposed to.

For parts that need to resist crushing -- foot pads, high-load bumpers, parts that should compress and spring back -- 50%+ provides stability. Low-density TPU is light and supple but deflects more than expected under concentrated loads. Higher density gives the interior enough structure to maintain the part's shape under pressure.

What makes TPU interesting is how much the same density percentage can feel different depending on the pattern. I lean toward Gyroid for parts meant to bend, and Cubic for parts meant to resist compression. The former produces a supple feel, the latter a "won't quite bottom out" quality. TPU isn't a material where high density automatically equals correct -- it's better understood as a material where you pick from two extremes based on the required deformation behavior.

Box Shapes with Broad Top Surfaces

Storage cases, electronics enclosures, and box shapes with integrated lids -- anything with a wide top surface -- call for a slightly different approach. The priority here is closing the top surface cleanly rather than overall structural strength. Start at 20-30% with Support Cubic or Lightning as the pattern. This combination is built for top-surface-focused applications.

Box shapes can get away with low density on the side walls, but without adequate scaffolding beneath the ceiling, top layers droop and show visible lines. Rather than filling the entire interior, choosing an infill pattern designed to support top surfaces provides the efficiency you need. Support Cubic offers straightforward top-layer support, while Lightning minimizes material aggressively. Neither pattern is designed for making the box itself tough, though -- for a robust enclosure, thicken walls and top layers alongside the infill pattern choice.

Broad lid surfaces are where adding top layers produces the most visible improvement. When a box print has "ugly top only" or "center sags slightly," the cause is almost always insufficient top-layer support plus too few closing layers, not density alone. For these shapes, approaching settings from "how do I support the top" rather than copying jig-level Cubic settings produces far better results.

Common Mistakes and Pitfalls

Top Surface Roughness at Low Density

The classic low-density symptom: the top surface closes, but it's rough, slightly sunken in the center, or shows gaps between lines. The cause is straightforward -- scaffolding beneath the top layers is too sparse, forcing the extruded lines to bridge too far unsupported. With box shapes and broad lid surfaces especially, the issue isn't density itself but bridge distance that's become too long.

Cranking up overall density fixes this but inefficiently. The more effective first move is adding top layers, followed by switching to a top-support-oriented pattern like Support Cubic or Lightning to add support directly beneath the top surface. This targets the problem without making the whole part heavier. In my own workflow, when a case has rough top surfaces, I adjust top layers before touching density. Surface roughness is much more often a "support deficit" than a "strength deficit."

Strength Issues with Lightning

Lightning is a genuinely useful pattern, but role confusion leads to disappointment. It's designed to close top surfaces while minimizing material, not to provide mechanical reinforcement for jigs or brackets. Light, fast, and able to produce clean top surfaces with minimal material -- yes. Able to uniformly stiffen the entire interior -- no.

Display pieces, cosmetic prototypes, and lightweight mockups are Lightning's home turf. Load-bearing parts that need real mechanical strength will feel "emptier than expected" with Lightning. Support Cubic follows the same logic: its focus is efficient top surface support, not overall rigidity enhancement. An impressive-looking infill pattern in the preview doesn't mean the strength center of gravity is there.

Material Differences

Infill settings don't transfer one-to-one across PLA, PETG, and ABS. Copying numbers between materials produces noticeably different results. PLA offers good rigidity and works well for prototypes and indoor jigs, but it's heat-sensitive, limiting where you can use it. PETG provides toughness and is versatile for functional parts. ABS is a candidate for heat-resistant applications but brings its own printing challenges.

Automotive interior use with PLA is a case where material trumps settings. I once left a PLA jig in a parked car during summer. It didn't visibly collapse, but when I tried to reinstall it, the dimensions had shifted just enough to create play. That kind of deformation is worse than breaking -- the part looks fine but no longer fits properly. Higher density doesn't overcome PLA's glass transition temperature weakness. For car interiors, sunny windowsills, or areas near heat sources, material selection matters more than any density setting.

Diminishing Returns at High Density

Higher density provides psychological comfort, but past a certain point, the additional material and time outpace the strength improvements. General prints work fine at low to medium density, and the Prusa Knowledge Base doesn't assume high density as a default. Even with functional parts, if the failure mode is layer separation, packing the interior denser won't change the failure mechanism.

Additionally, slicer behavior can shift at high densities. Some community observations report toolpath changes around 60% and slicer time spikes around 70%. These are environment-specific observations -- slicer version and pattern choice matter -- so cite primary sources when referencing them.

When high-density parts still break, the controlling factors are almost always wall count and layer orientation. If layers are oriented unfavorably relative to the pulling direction, the part splits along layer lines regardless of fill density. Increasing wall thickness and reorienting the part to avoid unfavorable layer alignment is more effective. Infill matters, but when the fracture origin is at the outer shell or between layers, density alone won't solve the problem.

100% Infill Behavior

100% is better understood as a complete-fill mode rather than "the strongest normal setting." PrusaSlicer locks the pattern to Rectilinear at 100%, so your selected pattern doesn't carry through. You're not getting "Gyroid at maximum density" -- the slicer switches to gap-free filling logic.

At this level, side effects become hard to ignore. Heat retention increases, shrinkage stress intensifies, and parts become significantly heavier. Small block shapes may hide these issues, but parts with larger surface areas are prone to warping and dimensional tightness. "Went to 100% but it's not as indestructible as expected, and actually harder to work with" is a common experience. When pursuing high strength, checking what wall thickness and orientation changes can achieve before jumping to complete fill usually produces better outcomes.

Wrapping Up: Default Settings When You're Unsure

When in doubt, start at about 20% infill with Gyroid or a Cubic variant. If strength is lacking, resist the urge to immediately increase overall density. Check wall count first. The effective sequence is: wall count, top/bottom layers, local high-density reinforcement, then overall density adjustment. Following this order lets you diagnose weak points without unnecessarily inflating print time or weight.

I frequently run the same model with just 2-3 variable changes, and even quick A/B tests reveal "what actually worked" through differences in failure mode. Recording print time, weight, and where the part broke makes setting decisions noticeably faster the next time around.

[Editorial note: Internal link additions]

• No internal links have been added as no other articles currently exist on this site. Candidate links for future editorial reference when building internal links after publication:
• setup-infill-basics (Infill fundamentals guide)
• troubleshoot-top-layer (Top surface roughness and sag fixes)
• material-guide-pla-petg (PLA/PETG material selection guide)

These are editorial candidate slugs. When publishing, replace with the actual article slugs and URLs.