3D-Printed Shadow Art With Lampshade Design

ISEF Category: Technology Enhances the Arts

Subcategory: 3D Modeling  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

A lampshade can act like a secret projector. Change the inside shape, and the shadow on the wall can become a face, a logo, or a pattern. Your job is to design that shape on purpose, not by luck. Then you can test how close the real shadow comes to your target image.

What Is It?

This project uses a simple idea from computer graphics and optics. You start with an image you want to see in shadow, then work backward to design a 3D shape that should cast that image under a specific light setup. That backward step is called inverse rendering. You are asking, “What shape should I print so the light behaves the way I want?”

Think of it like making a cookie cutter, but for light. A normal lamp spreads light out. Your design tries to steer that light so the shadow matches a target picture as closely as possible. The final test is not just whether it looks cool. You measure how close the projected shadow is to the target using PSNR, which stands for peak signal-to-noise ratio. Higher PSNR means the shadow image matches better.

Why This Is a Good Topic

This topic works well for a science fair because you can test it with clear numbers. You can change the target image, the wall distance, the light source, the shell thickness, or the print material, then compare how those changes affect shadow quality. The project connects art, CAD, simulation, and image analysis, so it feels creative and technical at the same time. You can also learn a real workflow from research and product design, from digital model to printed prototype to quantified result.

Research Questions

• How does the target image complexity affect shadow PSNR in a 3D-printed translucent lampshade?
• What is the effect of lamp-to-wall distance on shadow sharpness and PSNR?
• Does print layer height change the contrast and edge fidelity of the projected shadow?
• To what extent does translucent PLA opacity affect the accuracy of the final shadow image?
• Which inverse-rendering setting produces the best match between target image and projected shadow?
• How does the size of the lampshade model affect distortion near the edges of the shadow?

Basic Materials

• 3D printer with translucent PLA filament.
• Computer with CAD software and enough memory for mesh design.
• Digital camera or smartphone with manual exposure control.
• Tripod or fixed phone mount.
• Desk lamp or LED point light source.
• Plain white projection wall or matte poster board.
• Meter stick or tape measure.
• Dark room or light-controlled space.
• Image editing software for PSNR analysis.
• Basic calipers for checking print dimensions.

Advanced Materials

• 3D printer with interchangeable nozzle sizes.
• Translucent PLA, clear PETG, and opaque control filament.
• LUX meter or light sensor.
• High-resolution camera with RAW capture.
• Optical bench or fixed mounting rails.
• Calibration target for geometric alignment.
• Computer with Blender, MATLAB, or Python image-analysis workflow.
• Mesh processing software for topology optimization.
• Polarizing film for testing glare and diffusion.
• Spectroradiometer if available for light transmission measurements.

Software & Tools

• Blender: Builds and edits the lampshade geometry before printing.
• Python: Processes target images, aligns photos, and computes PSNR.
• ImageJ: Measures brightness maps and helps compare shadow edges.
• OpenSCAD: Creates parametric test shapes for fast design changes.
• MATLAB: Supports inverse-rendering experiments and matrix-based analysis.

Experiment Steps

1. Define the shadow target, the light source, and the fixed projection setup you will keep constant.
2. Choose the one design variable you will change first, such as shell thickness, mesh density, or print material.
3. Build a digital workflow that converts a target image into a printable surface model.
4. Plan a calibration method so every photographed shadow can be compared on the same scale.
5. Decide how you will score results with PSNR and at least one visual metric for edge quality.
6. Set controls that separate shape effects from lighting, camera, and printer effects.

Common Pitfalls

• Changing the lamp angle between trials, which shifts the shadow and ruins image comparisons.
• Printing a shade that is too thin, which leaks light and washes out the pattern.
• Using a glossy wall or background, which adds hotspots and lowers PSNR.
• Skipping image alignment before analysis, which makes a good shadow look worse than it is.
• Testing many design variables at once, which makes it hard to tell what actually improved the result.

What Makes This Competitive

A stronger project goes beyond making one cool shadow. You compare several inverse-rendering strategies, then explain which design choices improve PSNR and which ones fail. You also control the full imaging pipeline, from print settings to camera calibration to image registration. If you add a careful error analysis or test how the method changes across multiple target images, your project starts to look like real engineering research.

Project Variations

• Test how different target shapes, such as text, faces, or symbols, change the best lampshade design.
• Compare translucent PLA with another printable material to see how diffusion changes shadow fidelity.
• Keep the same target image, then vary the light source type, such as LED bulb, flashlight, or point-like lamp.

Learn More

• MIT OpenCourseWare, Computer Graphics: Search MIT OpenCourseWare for courses on rendering, inverse problems, and geometric modeling.
• NIST Digital Image Resources: Search NIST for guidance on image quality metrics and measurement practice.
• ImageJ Documentation: Use the official ImageJ guides to learn image measurement and analysis workflows.
• PubMed: Search for review articles on computational imaging, inverse rendering, and optical design.
• IEEE Xplore: Search for peer-reviewed papers on caustics, light field design, and 3D-printed optical structures.