The Optimized Gyroid Checkbox Does Less Than the Headline Says

OrcaSlicer 2.4 added a checkbox called Gyroid Optimized, and the release notes that introduced it carried a number: up to 60% better compressive strength-to-mass. That number is real. It is also not describing the checkbox.

Here is what the checkbox does, in OrcaSlicer's own words. The Patterns page of the OrcaSlicer wiki describes the gyroid_optimized setting, which appears as “Optimize gyroid wave” in the release notes and as Gyroid Optimized in the wiki, as one that “tightens the gyroid wave along the Z (vertical) axis at low infill density to shorten the effective vertical column length and improve Z-axis compression buckling resistance.” It then lists three limits: filament use is preserved, there is no effect at roughly 30% sparse infill density and above, and it only applies when the sparse infill pattern is set to Gyroid.

That is the whole scope. If you print functional parts at 30% or 40% gyroid, which plenty of people do, ticking this box changes nothing at all.

Where the 60% figure comes from

The 2.4.0 Alpha release notes describe the feature, credit it to a contributor, and end with this sentence: “Winning architectures from this research beat commodity baselines by up to 60% in compressive strength-to-mass.” The research is the CRAMP project at Brown University, funded through a NASA Space Grant, in collaboration with ELEGOO and Polymaker.

Read that sentence carefully. It says winning architectures from a research programme beat commodity baselines. It does not say the slicer option delivers 60%. Orca's documentation nowhere claims a figure for what the checkbox itself produces, and as of this writing I could not find any published independent test of the option: no CNC Kitchen video, no lab data, no community teardown with numbers. The mechanism is documented. The magnitude is not.

This is the same shape of problem as a manufacturer's heat-resistance rating measured with no load on the part, which we went through with HT-PLA last week. The headline number is honestly derived from a real test. It just is not the test you are running.

What the option actually changes

Two of those rows matter more than the rest. “Byte-identical when unchecked” means the feature carries no risk to your existing profiles: if you turn it off, you get exactly the G-code you got before. And “Z-axis compression” is narrow. If your part fails in bending, in tension, or by layer separation, this setting has nothing to offer you.

Why the wavelength is the lever at all

The mechanism is not invented. In a 2025 paper in Progress in Additive Manufacturing, Lopez-Anido and colleagues at the University of Maine set out the geometry plainly: the relative density of gyroid infill is approximately proportional to the ratio of wall thickness to unit cell size. In fused filament fabrication the wall thickness is constrained by the size of the nozzle, so, as they put it, common slicer implementations rely on changing the unit cell size to achieve a desired density.

That is the constraint the Orca feature works inside. It cannot make the strands thicker. What it can do is change the shape and period of the wave, and specifically squash it vertically so each strand's unsupported vertical run gets shorter. For a slender column, the buckling load rises sharply as the effective length falls, which is the Euler-Bernoulli argument the release notes cite.

Whether that reshaping helps in practice is a separate question from whether the physics is coherent. The same Maine paper notes that other researchers have found the relationship between wall thickness and cell size affects gyroid compressive behaviour beyond relative density alone, citing work by Netto, Sardinha and Leite published in Mechanics of Materials in 2024. So there is reason to think the wavelength is not a free parameter. There is still no number attached to Orca's particular implementation.

The lever that does have numbers: density

The Maine study fitted a semi-empirical relationship for the effective compressive yield strength of gyroid infill as a function of relative density. Their form is:

Sg / Sm = D − (1 − D)^C₁ · D^C₂

where Sg is the yield strength of the gyroid structure, Sm is the yield strength of the neat material, and D is relative density. For compressive yield they found C₁ = 0.67 and C₂ = 1.2, with a root-mean-square error of 2.2% against their simulations.

That equation is easy to evaluate, and nobody seems to have put the output in front of hobbyists, so here it is. The middle column is the compressive yield strength of the gyroid lattice as a fraction of the solid material's yield strength.

Three caveats before you use this. It describes the infill lattice on its own, not a finished part with walls and solid top and bottom layers, and in most real parts the walls carry more of the load than the infill does. Their experiments used PETG gyroid cores printed on Ultimaker-5 machines in sandwich specimens tested to ASTM C393 for shear, and a tough PLA for the compression tests. And the authors are explicit that the curve under-estimates strength at high relative density, because at those densities the compressive strength of the material drives the behaviour and many materials are stronger in compression than in tension.

With those caveats, the shape of the curve is the point. Going from 15% to 25% raises the predicted compressive yield of the lattice by about 60%. Going from 20% to 40% more than doubles it. No wave-shape optimisation that preserves filament use is going to compete with that, because the optimisation is explicitly a rearrangement at the same material mass.

The four-unit-cell problem

One finding in that paper deserves more attention than it gets, and it applies to almost every bracket and mount people print.

The authors state that gyroid behaviour “does not converge to its periodic behaviour until the specimen consists of at least 4 unit-cells in each direction.” Below that, edge effects dominate and the part does not behave like bulk gyroid material at all. They designed their test specimens specifically to be large enough to avoid this.

At low densities the gyroid cell is large, because low density is achieved by stretching the cell. So the parts most likely to fall below four cells in a direction are exactly the small, sparse, low-infill parts that the Gyroid Optimized option is scoped to. If your infill region is, say, 15 mm across and the cell works out at 6 mm, you have two and a half cells, and neither the published curve nor anyone's strength intuition applies cleanly. This is a large part of why hobbyist infill tests scatter so badly.

What to reach for first

If the part is failing in compression, this is the order I would work through.

Walls, before anything internal. Stefan Hermann's gradient infill experiments at CNC Kitchen open with the reason: most mechanical parts are loaded highest on their outside and least in the middle, and there is a line through the centre of a bending beam carrying zero stress. His own results reinforce how situational internal changes are. On three-point bending bars his gradient infill was almost 30% stiffer at equal weight and almost 60% stiffer at equal print time. On his hook specimen the same technique showed no significant improvement over simply raising the infill percentage, because on a part that small it added material where it was not needed.

Density, second. See the table above.

Fill Multiline, third, and it is underrated. Orca's Infill wiki page documents this setting as generating up to 10 parallel extrusion lines per infill path while preserving both the set density and the overall material usage. The wiki explicitly notes Orca keeps the cross-section constant for the set density, where other slicers simply multiply the line count and the material with it. Its listed use cases include improving part strength and print speed without increasing material usage. Gyroid supports it, using the Classic closed-loop strategy.

Extra Solid Infill, fourth. Orca lets you insert solid layers at chosen intervals or specific layer numbers, so you can reinforce a mounting boss or a stress concentration without raising density through the whole part. The wiki warns those layers take noticeably longer to print and can cause z-banding-like bulges, so watch your cooling.

Then the Gyroid Optimized checkbox. It is free, it is reversible, and it might help. It is not where the strength is.

Testing it yourself, properly

If you want a real answer for your machine and your part, the discipline matters more than the fixture. Change one variable at a time, for the same reason calibration order matters so much. Print at least three of each condition from the same spool, because spool-to-spool variation will swamp the effect you are looking for. Make the specimen at least four gyroid cells across in every direction, or accept that you are measuring edge effects. And test in the direction the part is actually loaded, not the direction that is easy to fixture.

For coupons, a plain, affordable, consistent filament is worth more than a fancy one. Something like eSUN PLA+ at AED 70 a kilogram, with a published 63 MPa tensile strength, is a reasonable baseline that will not distort the comparison. If you are on a Bambu machine and want the profile to match the filament exactly, the Bambu Lab filament range removes one more variable.

The verdict

Turn Gyroid Optimized on if you already use gyroid below 30% density and your parts see vertical compression. It costs nothing, the output reverts exactly when you untick it, and the underlying mechanism is documented rather than invented.

Do not turn it on expecting 60%. That figure belongs to a research programme's best architectures, not to a checkbox in a slicer, and no one has published a measurement of what the checkbox itself is worth. Until someone does, the honest answer is that the option is a low-risk experiment with an unknown payoff, sitting several rungs below wall count and infill density on the list of things that will actually stop your bracket from crushing.

Sources

1. “Patterns” (Gyroid and Gyroid Optimized sections), OrcaSlicer Wiki, accessed 11 August 2026. https://github.com/OrcaSlicer/OrcaSlicer/wiki/strength_settings_patterns

2. “Infill” (Sparse infill density, Fill Multiline, Extra Solid Infill), OrcaSlicer Wiki, last edited 12 July 2026, accessed 11 August 2026. https://github.com/OrcaSlicer/OrcaSlicer/wiki/strength_settings_infill

3. “V2.4.0 Alpha” release notes (Optimized Gyroid infill, PR 13379), OrcaSlicer Wiki, accessed 11 August 2026. https://www.orcaslicer.com/wiki/releases/release_2_4_0_alpha

4. “V2.4.0” release notes, OrcaSlicer Wiki, accessed 11 August 2026. https://www.orcaslicer.com/wiki/releases/release_2_4_0

5. Lopez-Anido, R. A. et al., “Investigation of the nonlinear response of gyroid infills for prediction of the effective yield strength,” Progress in Additive Manufacturing, vol. 10, published online 24 June 2025 (issue cover date November 2025), DOI 10.1007/s40964-025-01200-7. https://link.springer.com/article/10.1007/s40964-025-01200-7

6. Hermann, S., “Gradient Infill for 3D Prints,” CNC Kitchen, 10 January 2020. https://www.cnckitchen.com/blog/gradient-infill-for-3d-prints

7. “3D Printer Calibration Order: What to Tune, and When,” Additronix blog, 4 August 2026. https://additronix.ae/blogs/additronix/3d-printer-calibration-order-what-to-tune-and-when

8. “eSUN PLA+” product page, Additronix, verified 11 August 2026. https://additronix.ae/products/esun-pla

9. “BambuLab Filaments” collection, Additronix, verified 11 August 2026. https://additronix.ae/collections/bambulab-filaments