Computational Design of Kinesthetic Garments

Abstract

Figure 1: Given a pose and garment, our method leverages an FEM simulation of the garment (light blue) as it deforms over the body (a) and optimizes the placement of a reinforced material (dark blue) by maximizing energy density for a given area budget. Our algorithm produces a continuous range of designs with various amounts of reinforcement coverage (b,c), allowing designers to explore the trade-off between instrumentation and performance. The selected design is then manufactured using a standard HTV (heat-transfer-vinyl) process. The above design is a posture-correcting shirt that pulls back on the user when slouching forward.

Computational Design of Active Kinesthetic Garments

“…Closer to our work, topology optimization has moved into the on-body domain where it has been used for personalized cast design [43] and casts designed for thermal comfort [42]. Most recently, Vechev et al demonstrated the design of kinesthetic garments, which are passively reinforced garments designed to resist a single motion [40]. However, this work only formulates a single compliance minimization objective, and thus cannot be used in a setting that leverages active components.…”

“…Finding a passive mechanical structure that optimally connects electrostatic clutches placed by the user is a key challenge in the design of active kinesthetic garments. Recent work by Vechev et al [40] demonstrated a method for on-body topology optimization using a single compliance-minimization objective (summarized in Appendix C). Such an objective cannot be applied in our setting, as it has no notion of component states, and the single objective does not sufficiently capture the high-level goal of minimizing motion interference when components are inactive.…”

“…We combine our active component model, with the garmenton-body model described in [40]. The terms 𝐸 garment (x, 𝑑), 𝐸 body (𝑣, x), and 𝐸 attach are summarized in Appendix B.…”

“…As our connecting material, we attached a layer of polyurethane (Siser EasyWeed Stretch) onto 100% cotton fabric. This material combination enables much higher forces than in [40]. As base garments we used stretchable GripGrab UV sleeves and Nike Dri-Fit Pro Compression shirts.…”

“…As our garment model, we use a compressible neo-Hookean material model [6] adapted with a relaxed energy under wrinkling as in [40]. This allows the garment to wrinkle under compression without producing geometric artifacts.…”