Gyroid Measurement Experiment

This project explores the optimization of 3D-printed gyroids for enhanced energy absorption. By crushing lattice structures across a varied design space, we identified the non-linear “sweet spot” for mechanical performance.

I owned the experimental design, MATLAB automation, and statistical modeling for the project.

Outcomes

• Identified statistically significant non-linear interactions (R²=0.94) between gyroid isovalue and density.
• Developed MATLAB scripts for automated STL generation and data-processing of Instron raw files.
• Published a Comprehensive Final Report (PDF) and Summary Slides (Google Slides).

Skills Demonstrated

• Experimental Design & Instrumentation. Calibrated and executed Instron compression tests, ensuring accurate force and displacement measurements.
• Data Analysis & Statistical Modeling. Processed load-displacement data, computed mechanical energy absorption, and applied regression modeling and hypothesis testing to confirm findings.
• Scientific Communication. Designed useful and compelling figures that communicate data insights and highlight key trends. Technical writing is also included in the presentation.

Experimental Design & Test Matrix

We mapped a 2D design space by varying isovalue (t) and unit-cell density. This 3×3 factorial design ensured we captured the full spectrum of the lattice’s mechanical response.

Runs	Isovalue (t)	Unit Cell Density
1–9	0.25	1, 2, 3 Units
10–18	0.50	1, 2, 3 Units
19–27	0.75	1, 2, 3 Units

Note: Each configuration was tested in triplicate (n=3) to ensure statistical significance and minimize noise from FDM print variances.

DAQ, Calibration, and Regression

Energy absorption (W) was calculated as the area under the force-displacement curve. Samples were compressed using an Instron Universal Testing Machine. The resulting data was processed through a quadratic regression model to visualize the performance “topography” of the gyroids.

Calibration Results

Result Topography

Mechanical Observations & Variability

While the quadratic model provided a high global fit ($R^2=0.94$), local deviations were observed in low-isovalue samples.

• Manufacturing-Induced Anisotropy. At lower isovalues, the FDM layer-height became significant relative to the strut’s cross-sectional area. This increased the “grain” effect of the print.
• Failure Mode Shift. While high-density samples failed through predictable plastic deformation, low-isovalue samples transitioned to brittle delamination. This suggests that as the feature size approaches the manufacturing resolution, the material’s inter-laminar bond strength becomes the dominant failure bottleneck rather than the lattice geometry itself.