In this week I developed a new design for pneumatic curling. Instead of the spines being internal, I tried to extrude them. First I developed a CAD model for the mold which ended up being split into three parts:

When constructed the parts look like this for the mold:

Finally after the silicone has cured and been removed from the mold, it looks like:

The movement of this mold is much more extreme than the previous design. Specifically it appears to take much less air to cause a full curvature of the silicone. Also of note this design appears to be more manageable for air leaks and fixes, and the air channels are less likely to clog. Overall, however, the need to remove air bubbles is crucial for both this design and the previous silicone designs, as air is most like to alter a uniform distribution and focus to weak areas of the silicone.

For these two weeks, I’ve been working on incorporating the servos directly into the body of my robot. The servos will fit inside the larger rounded end of the body and control the ears through strings running through channels to the surface of the body. I started by quickly creating a model of the rear structure of the body to make sure that I leave enough space to accommodate the servos.

Based on this model’s measurements, I’m planning on creating an internal box that will hold the servos in place. The future body will have the bottom face of the body missing to allow me to put the servos in. Time and Solidworks abilities permitting, I will try to create a bottom that I can attach to the body which will make the body robust enough to be carried rather than only resting on the table.

Outside of body design, I’m starting to think about ways to redesign the ears to make them more robust, since the current version relies on barbecue skewers for attachment.

The new model had some serious problems around measurements. The slots which the ears would have attached to were printed too small, either as a result of poor measurement or lower than expected printing resolution, probably a combination of the two. To fix this, I printed several variants of the ear slots to help me determine what size they will need to be in the final version.

In doing this, I also realized that the structure I created was not as reliable as I expected. The very circular internal chamber which the screw head rotates within makes it difficult  to lock the ears into place. Future versions of the body will incorporate a different attachment mechanism that should let the ears lock into place better.

Pneumatics Progress

These past couple weeks I designed and 3D printed two separate molds for soft bodies. One is a gripper (obtained from instructables with some slight alterations), and the other is just a tube like body with an air chamber in it. The first is to test movements as a result of different air chamber architectures, how much air push is necessary, how loud the movements are, etc. The second is to test another concept I found called fabric reinforcement, where by wrapping fabric along the outside of the body in different patters, one can obtain different movements.

1). The Gripper:

Here is an image of the 3D printed mold:

Video to Come

2). The tube movement WITHOUT fabric reinforcement:

Here is an image of the soft body mold:

Here is a video of the mold moving (with sound) when air is blown into it:

3). The tube movement WITH fabric reinforcement:

Here is an image of the soft body mold:

Here is a video of the mold moving (with sound) when air is blown into it:

Summary of Results:

This seems like a completely viable option to pursue for movements, however I will need to be very strict in my architecture and soft body creation to get the exact movements that I want. In terms of the fabric reinforcement, there is certainly some viability to this idea as the video shows the beginning of a “twisting” motion due to the wrapped string. More research will be needed to to discover exactly what patterns/ degrees of angles are needed to get particular movements.