ChuckLi® Composites

Material experiments by Chuck Driesler and Li Jin

[wearetyvek-blog]

[wearetyvek-blog]

Method Abstract

Our material process involves the sandwiching of different materials between fiberglass resin sheets to generate a thin, programmable surface. Different materials are used for different uses - solid chipboard for structural layers, thin paper or film for translucent effects, and sometimes even printed images, etc. Overall forms are made by leaving some of these sheets not covered by resin during the first pass, folding or holding the pieces into the actual form, and then coating the seams/surfaces in resin. The shape of the structural layer comes from force diagrams in fusion, and lately we’ve been experimenting with weaving different layers/material through the holes that leaves behind.

[wearetyvek-blog]

[youaretyvek]

[youaretyvek]

[youaretyvek]

[youaretyvek]

[youaretyvek]

[youaretyvek]

[youaretyvek]

[wearetyvek-blog]

R003:0219 a broken column : translation process

(5/5)

This diagram shows how we translated the force-path geometry from Li’s earlier experiments into the geometry of this revision. We hoped that adapting the process to an arbitrary early form would help test and exhibit its versatility. But joining disparate pieces, not initially cast onto the same strip, was extremely difficult. For the next revision, we plan to experiment with the construction without a specific form in mind. If we create a much larger strip, and start folding it onto itself and casting it into solid form, we’ll hopefully discover a better way to construct things in the future. We may also play with the geometry of the rigid-cast rib.

[youaretyvek]

[wearetyvek-blog]

R003:0219 a broken column : the break

(4/5)

After the pieces were cast with zones of exposed fiberglass, we tested a series of ways to create the edge connection. Gravity, pinching, and twisting were a few of the larger operations. The results had a small amount of flexibility, but were consistent enough to control the geometry.

[youaretyvek]

[wearetyvek-blog]

R003:0219 a broken column : mold process (results)

(3/5)

[youaretyvek]

[wearetyvek-blog]

R003:0219 a broken column : mold process

(2/5)

The mold process from R002 was maintained. Many of the issues we ran into with the domes were avoided because the new geometry was smaller and linear, with less surface area.

[youaretyvek]

[wearetyvek-blog]

R003:0219 a broken column : concept sketches

(1/5)

We chose to move forward with the rib joinery from R001. The evolution this week reduced the various molds to a more universal geometry: the straight line. Various structural shapes can be created by freeform arrangement.

In order to make the rigid composite piece more adaptable in construction, some of the bottom layer of fiberglass is left not covered by resin as the piece cures. The hope is that we could then hold the linear bars in the arrangement we need (likely using gravity to our advantage for curvature), and cover the free fiberglass zone in resin.

[wearetyvek-blog]

R002:0215 bunker inversion : joinery tests

(8/8)

We tested a number of knot types, using thin elastic string and TPU filament, in search of the best joinery. None were as rigid as we needed, and using the failed pieces as a testing ground only muddled our ability to evaluate.

Next week, we plan to both scale up the pieces and test joinery on properly constructed domes. We hope that enough pieces will proved a more realistic tension scenario. Issues that emerge (we’re guessing self-weight, end pieces, and ground connection) will also be addressed. The spheres may need to be variable in size, thickness, or opening.

[wearetyvek-blog]

R002:0215 bunker inversion : speculative assembly (7/8) We returned to the geometry sketches, and speculated on how the hemisphere shells could be networked. Right now, we are proposing joinery by elastic string, with the strength of the knot controlling how much the wall can bend.

[wearetyvek-blog]

R002:0215 bunker inversion : successful production process

(6/8)

We took a step back and made a list of every issue facing us in the mold process:

-the desire for a clean edge at the end of the hemisphere shell -a mold with even pressure applied to the fiberglass-resin sheet -the consistent mapping of a square sheet of fiberglass to a spherical mold -a mold material that was difficult to remove easily, and without damage

And proposed the following solutions, piecemeal:

-trim the excess fiberglass BEFORE curing, while pressed -the re-insertion of the foam hemisphere to allow weight application -symmetrical cuts in the fiberglass sheets to allow controlled and limited layering (top left image. initially, we were trying to remove all overlapping) -application of mold release (not a total solution)

These changes gave us clean and consistent results. If more of our molds had survived the earlier carnage, production would be fast and reliable.