All workMaterial research / 2025

NetWork · Bio-gradient tactile library

What if the material refused to be a block?

A failed bioplastic cast became the beginning of a family of mesh-reinforced skins. NetWork turns recovered nets, a starch binder and repeated small tests into one tactile surface.

Experimental material studyPhysical samples + 20 × 20 in / 50 × 50 cm surface

The assembled NetWork surface, isolated for the project presentation · Photograph of physical outcome

A presentation cutout of the photographed, assembled surface—not a digital material render.

The question

Can household waste become a material worth touching?

The starting point was a kitchen-scale material inquiry, not a finished-product brief. I looked at how starch, glycerin, acid and recovered fibres behaved under small changes in recipe, thickness and drying. The research drew on DIY-Materials, material-driven design and the practice of keeping failed samples as evidence.

I planned controlled one-variable trials alongside experiments that let flow, shrinkage and chance reveal a direction. That distinction matters: the early notes describe an intended method; the photographs and later batch record show what I actually made and observed.

Project statusIndependent material study · Physical samples · 2025

My role: Research, formulation, hands-on testing, sample making, documentation and final presentation.

Working principleFollow the material’s behaviour before deciding what form it should take.

Project boundary. This is an independent concept — not a claim of commissioned, built, tested, or approved work beyond what this page documents. Environmental performance and future applications were not tested.

The work in one line

01CastTry solid forms.
02ObserveRecord cracking, warp and moisture.
03ReframeLet the mesh carry the skin.
04ComposeBuild a gradient of hand-feel.

Before there was a NetWork

Start with a table of possibilities.

The first inquiry was called Looking 4 Plastic? I cast small starch mixes with changes in glycerin, pectin, paper, fabric, coffee and colour. Some were brittle; some stayed damp. I kept them side by side to learn what each change did before deciding what the material should become.

First-stage casts, film and aggregate studies arranged for comparison · Edited project presentation photograph

An edited project photograph of first-stage casts and film; these are exploratory samples, not the finished surface.

Recovered produce nets and fabric offcuts sorted before coating · Original process photograph

Recovered nets and fabric offcuts laid out before coating. Their open weave later changed the direction of the work.

How the search was organized

Controlled intention. The early research outline proposed changing one variable at a time and recording tactile and curing observations. The surviving material shows recipe variants and photographs, not a complete laboratory log for every cast.

Chance with a purpose. Letting mixtures flow or dry in different ways exposed unexpected surface qualities. By the final stage, the designer’s judgement of when a sample could flow or fuse mattered more than random selection.

01 / Evidence changed the form

The solid version did not hold.

Thicker casts shrank and fractured during curing. Some retained moisture; thin unsupported films could warp. Rather than hide those pieces, I kept them in the record because they exposed a mismatch between the binder and the rigid object I was trying to force it to become.

An early kitchen-scale matrix of cast samples, ingredients and colour trials · Original archive photograph

The starting matrix kept different ingredients, thicknesses and results in one view.

A dark early cast that did not become a viable solid surface · Original archive photograph

This solid test did not yield a viable skin.

An early thick starch cast broke into fragments while curing · Original process photograph

The cast broke apart. This is a source photograph, not a polished proposal image.

An early thin film warped and cracked as it dried · Original process photograph

Thin film alone was not yet the answer.

A red mesh-backed trial with visible cracking · Original process photograph

Adding net was promising, but early reinforced trials could still crack.

Design decision

Stop forcing the material to behave like a brick. Work with the binder as a thin skin, and use reclaimed mesh as visible reinforcement.

02 / From solid to skin

Let the structure show.

Recovered onion and fruit nets, reusable-bag mesh and fibrous additions became more than filler. Their open weave gave the binder somewhere to grip while leaving the texture legible to the hand and eye. The material’s irregularity became the design language.

The flexible red net skin held in one hand after the solid-cast pivot · Original archive photograph

The hand-scale photograph makes the shift to a flexible membrane immediately visible.

Red recovered net retained as a visible structural layer · Photograph of physical sample

A net-backed sample rather than a deep cast.

A mesh-backed sample being handled to inspect its surface and flexibility · Original process photograph

Handling helped compare flexibility and surface quality. This was an informal observation, not a standardized strength test.

Small coloured binder and mesh samples arranged for comparison · Photograph of physical samples

The samples revealed multiple textures, not one uniform material.

What stayed

Starch-based binder, household-scale making and waste-stream material as an input.

What changed

Thickness, reinforcement, transfer method and the ambition: from rigid object to tactile library.

What I did not claim

A certified textile, structural performance or a compostable whole-object material.

03 / A repeatable starting point

Measure, dip, cure, compare.

After the pivot, the recipe became a baseline for comparison—not a claim that every sample had the same formula. I used measured liquids and mesh squares, then changed support and aggregate to see how the hand-feel and curing response shifted.

Recorded baseline / later trials

Water: 125 ml
Cornstarch: 1 tbsp
Glycerin: 3.6 ml
Vinegar: 7.3 ml
Pectin: Approx. 0.45 ml
Mesh square: Approx. 16 × 16 cm (6 × 6 in)

From the second-stage written method. Measurements describe a baseline batch, not a tested manufacturing specification.

Measuring vessels and tools used in the kitchen-scale material trials · Original process photograph

Small volumes and labelled trials made changes easier to compare.

Starch binder and aggregate being mixed in a pan · Original process photograph

The warm binder was mixed and transferred to mesh or filler.

A red and green mesh trial curing in a lined tray · Original process photograph

A mesh-backed trial curing in a lined tray.

Separate coloured samples assembled on a board · Original process photograph

Individual studies were arranged before becoming a continuous field.

An orange mesh layer and binder during the surface build · Original process photograph

A close view of the binder and mesh relationship during construction.

Selected batch observations

The useful answer was not one perfect sample.

These are my recorded observations, not independent laboratory results. The failures were as informative as the flexible pieces.

B–01 / onion netHigh-glycerin, 12-hour trial remained flexible in the recorded inspection.

B–02 / coffee, no meshCurled and cracked; evidence against a binder-only route.

B–04 / hybrid seamRed-blue join held under an informal hand pull.

B–05 / kinetic sandFailed to bond, likely because of its coated surface.

See the other recorded tests

B–03 / blue mesh + eggshellChalky, rigid surface; no recorded warp in the short observation.

B–06 / granulated sandStiffer but brittle.

B–07 / scaled sandToo thick and rigid for the intended skin.

B–08 / planned onlyProposed in the batch sheet; no completed result was documented.

Process / original footage

Hand assembly, not a simulation.

A short excerpt from the project’s own making video. The full camera roll is kept in the archive; this edit shows the working scale without asking you to open a presentation.

Twenty-four-second original process excerpt showing hand assembly of mesh and binder samples; no sound · Process footage

04 / The physical result

A surface that reads through the hand.

The final presentation reports a 20 × 20 inch (50 × 50 cm) continuous piece assembled from distinct tactile zones: open net, denser coated mesh, soft colour, rougher eggshell and contrasting joins. The earlier approximately 40 × 20 cm (16 × 8 in) study was a separate prototype ambition, not the dimension of this result.

The assembled tactile surface on its work surface, before the presentation cutout · Original outcome photograph

The work in the studio, with its scale and making context visible.

Blue, orange and green zones meeting in the final surface · Original outcome detail

The connection between zones remains legible rather than being hidden under a uniform finish.

Close view of translucent colour and the net structure · Original outcome detail

Colour and translucency shift across the layered mesh.

The material surface in a full studio view, showing its patchwork scale · Original outcome photograph

A second physical view of the assembled work.

Final presentation plate mapping the tactile gradient between sample zones · Original project presentation

The final presentation’s tactile-gradient map; an interpretation of the physical sample, not another object.

NetWork’s strongest outcome is a change in material thinking: the binder did not become the rigid block first imagined. It became a way to join and differentiate recovered nets into a touchable field.

The final reflection / The Adaptive Alchemist

Making is a cycle, not a straight line.

Alongside the material surface, I wrote a five-stage stewardship manifesto. It names the human part of this process: exhaustion, renewed attention and the judgement that grows from repeated handling. This is my design reflection—not a test result or a claim about the material’s environmental performance.

FatigueNotice the cost of trying to make use of what a waste system discards.
SparkLet an ordinary onion net become interesting again when light passes through it.
CycleKeep failure, breakdown and rest in the process—not only successful samples.
InstinctMove beyond chance alone; use the hand to judge when a surface can flow or join.
ValueAsk what another material life could be, without claiming that it has been proven.

See the original manifesto diagram

The original five-stage stewardship diagram from the final presentation · Original project presentation

Original final-presentation plate. The on-page wording above is an editorial paraphrase of its five stages.

05 / What this does—and does not—prove

Still a material study.
Not yet a material product.

The samples and surface establish a documented design direction. They do not establish long-term strength, washable use, biological breakdown, safe disposal or a lower environmental footprint. Some meshes are synthetic; a starch binder does not make the full composite compostable.

The presentation speculates about clothing, screens and shelter. Those are prompts for future research, not fabricated or performance-tested applications.

Research, tools and source boundaries

Research lens

I worked from the DIY-materials tradition — small, documented recipe tests — and treated chance as a tool, not an accident. These sources informed the tests; they are not evidence of environmental performance.

Tools with a purpose.

Kitchen-scale formulation — measured batches made recipe changes comparable.
Mesh cutting and layering — recovered mesh carried the skin and built the surface from distinct tactile zones.
Physical sample comparison — failed samples were kept as evidence, not discarded.
Photography and video — documented each batch and the final surface.
Word and PowerPoint — the written method and final presentation.

Where AI enters the process.

No AI tools were used on this project. The opening image is a presentation cutout of the photographed physical surface.

See the Selected Four