Curve

A material library for soft-touch silicones. Eighteen formulations, characterised across hardness, surface finish, and what people actually call them when they touch them.

Role

Course project — Materials Science

Year

2025

Tools
  • Shore-A durometer
  • Surface profilometry
  • Body-safe silicones
  • Blind-touch user study
HERO · 2025Replace
Curve
§ 01

Case study

The problem with "feel"

Engineers measure silicone in numbers. Shore-A hardness. Tear strength. Coefficient of friction. None of these numbers tell you whether the material feels good.

Curve started as a frustration. I'd spec'd a Shore-A 30 silicone for a prototype and the result felt cheap — exactly the opposite of what the spec sheet promised. I wanted to understand the gap between what the durometer says and what the hand says.

Method

I produced 18 sample tiles, varying:

  • Shore-A hardness (12, 20, 30, 45)
  • Surface finish (mirror, matte cast, micro-textured at three depths)
  • Outer-skin formulation (4 variants of platinum-cure silicone)

Each tile was characterised on the bench (durometer, profilometry, friction sled) and then run through a blind-touch user study with 24 participants. Participants rated each tile on a 7-point scale across nine descriptors I'd pulled from the marketing copy of every personal-product brand I could find: velvety, soft, premium, cheap, sticky, warm, dead, alive, like skin.

What came out

A correlation map showing which physical parameters drive which subjective descriptors — and importantly, where the correlation breaks. Premium correlates strongly with surface finish but only weakly with hardness. Like skin correlates with both, but only within a narrow hardness window (Shore-A 18–24).

"Soft" is not a property. It is a vocabulary problem the materials world refuses to take seriously.

This library is now my reference whenever I'm specifying outer-skin silicones for product work, and the framework — measure the bench number, then measure what people call it — is something I'd carry into any material I work with.

§ 02 — Plates
Sample tiles, batch 03Replace
PL. 02
Pl. 02Sample tiles, batch 03
Profilometry plotsReplace
PL. 03
Pl. 03Profilometry plots
Vocab map from user studyReplace
PL. 04
Pl. 04Vocab map from user study