Week 04 / How to Make Almost Anything

Updated

3D Printing
and Scanning

I designed and printed a small DNA model, then scanned a chair in my dorm with Polycam. Before printing my own design, I practiced with a 3DBenchy and a downloaded cable holder.

Prusa MK4S / PLAFusion / PrusaSlicerPolycam
Read the process ↓
04 / 3D PRINTING & SCANNINGFinished black DNA double helix standing upright on its tapered circular base
The finished DNA model after removing the supports.
Xiangtao Shi / Individual assignment← Back to all weeks
Printer practice3DBenchy and cable holder
My designA parametric DNA model
3D scanningA wooden dorm chair
01 / Printer training

Printing test objects

Our first training print was a 3DBenchy boat. The result had gaps and uneven layers.

I suspected under-extrusion, which means the printer deposits less plastic than expected. Possible causes included a nozzle temperature that was too low, a printing speed that was too high, or a mismatch between the installed nozzle and the slicer profile.

The photos below show the rough surface and gaps that made me question the extrusion settings. Prusa's under-extrusion guide explains the possible causes.

Front view of the rough black 3DBenchy training print
FIG. 01The training 3DBenchy, with gaps and uneven layers.
Side of the Benchy hull showing loose uneven extruded lines
FIG. 02A side view of the rough Benchy hull.

Printing a downloaded cable holder

I then printed a Cable Holder from Printables to get more familiar with the printer and slicer settings and try to improve the print quality. This was a downloaded design.

I selected the Original Prusa MK4S HF0.4 nozzle printer profile, the 0.20 mm SPEED print preset, and PLA filament. HF0.4 refers to a 0.4 mm high-flow nozzle. The cable holder printed successfully.

Original Prusa MK4S0.4 mm high-flow nozzle0.20 mm SPEED presetPLA
PrusaSlicer showing the cable holder and selected MK4S printing profile
FIG. 03The downloaded cable holder in PrusaSlicer.
Successful black printed cable holder with five cable slots
FIG. 04The completed cable holder.

Video / Cable holder print

The cable holder printing on the Prusa MK4S. Open video ↗
02 / My design

Designing a DNA model

I used parametric modeling in Fusion to make a simplified DNA double helix.

The model has two helical backbones offset by 180°, with connecting rods that represent base pairs. I added a tapered circular base, wider at the bottom, to help it stand upright. The overall dimensions are approximately 46 × 46 × 94 mm.

I used named parameters to control the diameter, height, number of turns, and rod size. I can change those values without rebuilding the model. Fusion's Parameters dialog also lets me define dimensions using expressions.

Fusion Parameters dialog showing helix dimensions, rod sizes, and base dimensions
FIG. 05Named dimensions and expressions in Fusion’s Parameters dialog.
Fusion viewport showing the two helical backbones and connecting rods on a circular base
FIG. 06The DNA model and its tapered base in Fusion.

Model dimensions

Parameters used in this version
Helix diameter30 mm
Helix height80 mm
Backbone section size4 mm
Connecting rod size2.5 mm
Turns / rods per turn1 / 6
Base height12 mm
Base top / bottom diameter36 mm / 46 mm

Pitch = helix height ÷ turns = 80 mm. Rod spacing = pitch ÷ rods per turn ≈ 13.33 mm, with a 60° angular step.

Video / DNA model

A closer look at the model before printing. Open video ↗

Explore the DNA model in 3D. Use the viewer controls to rotate, pan, and zoom.

Why 3D print it?

The intertwined backbones and connecting rods leave narrow spaces that are difficult for a milling tool to reach. Printing let me build the connected structure in one piece, with temporary supports for the overhangs.

03 / Print preparation

Exporting and slicing

I exported the model as an STL and prepared the print in PrusaSlicer.

  1. In Fusion, I opened Utilities → Make → 3D Print, selected Export and the model body, and chose STL (Binary) with the units set to Millimeter.
  2. I imported the STL into PrusaSlicer and used Rotate and Place on Face to put a suitable flat face on the build plate.
  3. I checked the orientation and support settings, sliced the model, and exported the G-code to a USB drive. I took the drive to the shop and inserted it into the printer.

STL stores a triangulated surface in a unitless format. PrusaSlicer showed imported dimensions of 46 × 46 × 94.16 mm.

PrusaSlicer showing the DNA model upright on its circular base
FIG. 07Orienting the STL on the build plate.
PrusaSlicer Support material settings with Organic selected
FIG. 08Selecting Organic as the support style.

Adding organic supports

The curved backbones and thin crossbars created several overhangs. I set supports to Everywhere and selected Organic under Print settings → Support material → Style.

Organic supports branch from thicker stems toward the areas that need support. They can use less material and be easier to remove than grid supports, depending on the model and settings. Prusa's organic support documentation explains how they work.

PrusaSlicer preview showing organic supports, 15 percent infill, and estimated time and material
FIG. 09The sliced model, supports, and print estimates.
DNA print setup and slicer estimates
Printer profileOriginal Prusa MK4S HF0.4 nozzle
Print preset0.20 mm SPEED (modified)
Filament profilePrusament PLA
Infill15%
SupportsEverywhere / Organic
BrimUnchecked
Estimated print time1 h 1 min
Estimated filament13.60 g / 4.56 m

The time and filament amount are estimates from PrusaSlicer, rather than measurements of the completed print.

04 / Troubleshooting

Fixing print problems

Stuck filament

Before I could start the print, the printer reported “Stuck filament detected.” I followed the screen instructions to unload the filament. The tip had an irregular shape, so I cut off the end and reloaded it. The warning cleared, and I continued.

Prusa MK4S screen displaying the stuck filament warning
FIG. 10The printer’s “Stuck filament detected” warning.
Hands trimming the unloaded filament end with cutters beside the printer
FIG. 11Trimming the irregular filament tip before reloading.

The base lifted off the bed

During another attempt, the edge of the base lifted from the build plate and the print could not continue properly. This is often called warping. I suspected poor first-layer adhesion.

I used compressed air to remove loose dust and debris from the plate, then increased the bed target from 60°C to 70°C. The next attempt adhered better, and I was able to finish the print.

Small black DNA base showing a lifted edge after an unsuccessful print attempt
FIG. 12The failed base with its edge lifted from the plate.
Compressed-air gun and hose on the workbench beside the printer
FIG. 13The compressed-air gun used to remove loose debris.
Prusa MK4S Tune menu showing a 70 degree Celsius bed target, 170 degree nozzle target, and 100 percent speed and flow
FIG. 14The bed target set to 70°C in the Tune menu.
For another attempt

A brim adds a flat border around the first layer to increase its contact area with the bed. I left it unchecked for this print, but it would be one option to try if the base lifted again. Compressed air removes loose debris; it does not remove grease. Prusa's first-layer guide explains surface cleaning and bed-temperature adjustments, and its brim guide explains the added contact area.

Recorded settings

I recorded PLA, a 70°C bed target, 100% print-speed multiplier, and 100% flow factor.

Settings recorded during the process
MaterialPLA
Bed target70°C
Print-speed multiplier100%
Flow factor100%
Nozzle temperature170°C / 230°C

05 / Print result

Removing the supports

The print finished with supports around the helix. I removed them carefully to reveal the model.

The thin printed rods were easy to damage, so I worked slowly around the backbones and connecting rods. Once the supports were off, I had a finished DNA model that stood on its circular base.

Printed DNA model standing on the printer bed among its organic supports
FIG. 15The completed DNA print with supports still attached.
DNA model with support branches remaining around the helical structure
FIG. 16The print during support removal.
Finished black DNA double helix standing upright on its tapered circular base
FIG. 17The finished DNA model after removing the supports.
06 / 3D scanning

Scanning a dorm chair

Sara, one of the TAs, recommended Polycam. I used it to scan a wooden chair in my Sydney-Pacific dorm.

I placed the chair in the middle of the room so I could walk around it and reach every side. I turned on all the lights, then moved around the chair at different heights to capture it from more angles.

I chose Photogrammetry because the chair had no transparent surfaces or strong reflections, and its shape did not depend on very fine details. Photogrammetry reconstructs a textured 3D mesh from overlapping photographs taken from different viewpoints.

Wooden chair placed in the dorm room before the Polycam scan
FIG. 18The wooden chair before scanning.

Preparing and capturing the object

Polycam app guide recommending access around the object and even lighting
FIG. 19Polycam’s preparation guide: object placement and lighting.
Polycam app guide illustrating photographs taken around an object at different heights
FIG. 20Polycam’s capture guide: viewpoints at several heights.
Polycam guide comparing photogrammetry and Gaussian splats for different object properties
FIG. 21The app’s guide to choosing a capture method.

Processing the scan

I captured 109 photos, chose Photogrammetry with the Dense processing setting, and enabled “Isolate object from environment.” This separated the chair from its surroundings in the processed result.

109 photosPhotogrammetryDense processingObject isolation enabled
Polycam processing screen showing the chair photos, photogrammetry, Dense quality, and enabled object isolation
FIG. 22The actual processing setup: 109 photos, Dense, and object isolation.
Textured 3D scan of the wooden chair on a plain background
FIG. 23The processed chair scan with the background removed.

Video / Chair scan

Viewing the processed chair scan. Open video ↗

Polycam produced a model of the chair. To improve the detail, I would add sharp, overlapping photos of the areas that were not captured well, especially around the legs and underside. Better coverage and consistent lighting matter more than simply increasing the photo count. Polycam's Object Mode guide explains these capture requirements.

Sources & credits

OpenAI Codex helped organize my notes, check technical references, polish the English, and create the HTML layout. The photographs and videos document my work.