This is the first in a short series of six investigations I am doing on movable assemblages in 3D printing, as part of an Independent Study course at OCAD University. My process involves researching a particular joint or basic assembly followed by the creation of a 3D print incorporating the element in a creative manner, with documentation images and notes along the way.
Ball Joints consist of a ball and a socket connected in a manner that allows for rotation around the centre of the ball, like that of the hip joint in the human body. Depending on how tightly the pieces are connected, the ball can move loosely or be secured into customizable positions. To further extend the range of movement, ball joints can be connected in series or with other joints.
Figure 1: “Types of Joints” Smartdraw (Lippincott Williams & Wilkins, [1989-2001]) https://www.smartdraw.com/skeletal-system-diagram/examples/types-of-joints/
Figure 1 shows six examples of joints in skeletal system anatomy. The plane joint facilitates planar movement, the hinge and pivot joints can rotate on one axis, the saddle joint ‘rocks’ along a surface rail or (virtually) identical joint, and the condyloid joint is effectively an ovaloid version of the ball and socket joint (and thereby lacks the ability to pivot). The plane, saddle, and condyloid joints seem to be more esoteric variations in 3D printing, presumably due to their reliance in the human body on surrounding tissue to keep their elements secured. I plan to cover hinges in a later post.
The two most common varieties of Ball Joints seem to be the Slotted Ball Joint and the Threaded Ball Joint:
The Slotted Ball Joint in Figure 2 is a simple variant of the ball joint that takes advantage of the flexibility of plastics to allow the ball component snap in and out of place with the socket component, which is comprised of four leaflets. These leaflets also restrict the motion of the ball to the paths between the leaflets. While looking for examples, I have encountered several variations where the ball is completely enclosed without separate leaflets (i.e. printed in place) and examples where the leaflets come in different numbers and shapes.
The Threaded Ball Joint in Figure 3 uses threaded components in its socket in order to secure the ball into place. The base of the socket is divided into sections which compress and apply pressure to the ball when the cap is secured and tightened around it. Unlike the Slotted Ball Joint, this variant allows for adjustments in how freely the ball moves in the socket through the adjustment of the tightness of the joint but seems to have less freedom-of-movement due to the restrictiveness of the threaded element.
One unusual variant I came across was this ball joint in Figure 4, which seems like it would facilitate movement for a connecting element on both sides of the socket.
As an experiment, I decided I would create a Polyhedra with faces that could pivot and rotate in place using ball joints connected by elongated stems to a spherical core.
I found a Rhinoceros plug-in called RhinoPolyhedra that could do the work of creating the complex geodesic geometry I had in mind. I chose to build a Dual Geodesic Icosahedron (Pattern 1) which has 32 faces, which I felt was a nice quantity for the task I had in mind. Because of the Polyhedra’s patterned qualities, I was able to save some time drawing the stems in 3D by taking into account the shape’s symmetry and using Mirror and Rotation commands.
The distance from the core of the solid to the centre of its faces for Pentagons were 2mm longer than the Hexagons. I reduced the length further when setting up the core’s geometry, which would eventually hold these stems in place.
I designed and printed several variants of the ball and socket joint to test its properties. In some cases, the joint was too loose, or tight enough to damage the clasp. Adjustments were made accordingly, and I eventually settled on a variation of the threaded joint in Figure 3, with a reduced number of threads to better suit its dimensions.
The printing of the elements took about 5 days to complete: a sample print, 20 hexagons and 12 pentagons, plus 32 stems and 32 lock-in threads. Due to variations in the printing process, some of the stems had more secure fits than others, and in some cases the locking element failed to improve the tightness of the assembly altogether.
The end result was a peculiar ball with rather floppy faces. It was interesting to manipulate the faces into various configurations as demonstrated in Figures 16-18; the ball could be made to flower, resemble the initial geometry, or press flat against surfaces that it rolled along. The flatness of the faces made the ball slow itself dramatically when rolled. Ultimately, the physical prototype exhibited properties that were not evident or expected on paper or in virtual space, as well as several unexpected properties caused by variations in the printed materials, making for a very curious experimental piece.