Overview

Ferrosketch was inspired by the Ferrolic ferrofluid display. The goal is to move ferrofluid with electromagnets.

The device has six subsystems: Display Box, Fluid, Solenoids, Frame, Drive Electronics, Network. Creating the device required the following skills learned during MAS863: Laser Cutting, Acrylic Welding, 3d Printing, Coil Winding, Board design and fabrication, Programming (ATTiny and PSoC).

Water-Tight Box

The front of the display is made of acrylic, laser-cut and then chemically bonded together. This is intended to make a water-tight seal.

I originally intended to use the "press-fit" techniques we learned earlier in the semester, but multiple people warned me off. Acrylic bonding works best on flat surfaces, and any interdigitation just adds potential failure points. Each wall is a simple rectangle.

The face is inset from the walls. This made using the 4-corner clamp relatively easy.

1) Place the front face down flat on the table.
2) Place the 4-corner clamp roughly in place.
3) Put the side walls in. Make sure they are flush with the table.
4) Acrylic cement these joints just enough to make them hold together.
5) Add spacers
6) Place the back face in. This needs to happen before everything fully sets.
7) Acrylic cement the back face. Be thorough!
8) Wait 30 minutes for this round to set.
9) Remove the 4 corner clamp.
10) Flip the piece so the front face is up.
11) Finish cementing the front face joints
12) Cement the side joints.
13) Wait an hour.
14) Water test!

Controlling the flow of the cement was difficult. The best way I found was to use the small syringe to apply the cement, and a separate small metal needle to guide it. Mildly pushing the wall away from the face can encourage whicking. You can see where the cement has whicked if you look from the opposite side.

Four Corner Clamp

Acrylic Cement seeped underneath and ruined the surface finish

Laser Cutting Acrylic

Water Test

Drive Electronics

Solenoid Construction

The original idea of using normal electromagnets was put on hold because it was difficult to create enough field to affect the ferrofluid. Instead, I'm using permanent magnets to move the fluid, and electromagnets to move the permanent magnet.

I chose to use 2 electromagnets driven unidirectionally instead of 2 electromagnets driven bidirectionally. This made winding them more difficult, but saved the use of an H-Bridge (that I hadn't ordered). If I were to re-do this project, I'd likely go with a single bi-directional coil.

I designed a bobbin to hold the three magnets together. The core is hollow to allow the permanent magnet to slide back and forth. There are three fins to hold the electromagnets' windings in place. The fins have additional features to allow wire routing.

I printed these bobbins on the Stratasys because the lathe was not working. It took a bit longer to clean than I expected!

Solenoid Testing

This video is of a successful test after the flux guides (see below) went in.

Surprise Cross Talk!

The magnets in adjacent tubes had too much affect on each other. Moving one magnet would move all the adjacent magnets as well! Sam Calisch suggested creating steel flux guides to isolate the effects. He graciously cut my design on the waterjet.

Retrospective

Here is a video of what I should have done:

This would have been much simpler and much more fun as a single magnet on an X-Y platform instead of coils galore.