A flexible robot finger needs to stretch after it is printed, but its uncured material must also flow through the printing process.
The KAIST team used digital light processing (DLP), which exposes a liquid formulation to light to cure it into a chosen shape.
According to KAIST, increasing stretchability and durability tends to raise viscosity until the liquid becomes hard to print; thinning it improves flow while sacrificing stretchability and strength.
KAIST formulation measurements and machine-learning analysis
The researchers cured several liquid formulations in small molds and measured stretchability, hardness, light-curing speed, and flow.
Their training data included both printable formulations and highly viscous ones that were difficult to print.
Machine learning examined how the formulations related to performance and identified a combination aimed at both printability and high stretchability.
The broader workflow KAIST describes still asks researchers to verify AI-suggested candidates through tests.
The printed actuator and assembled hand in KAIST’s demonstration
KAIST reports that the selected formulation printed reliably on a DLP printer and extended to more than six times its original length when pulled without easily tearing.
The team printed a soft actuator that expanded with air and bent like a finger.
A hand assembled from several actuators lifted a 1 kg water bottle and grasped eggs, glass bottles, an egg carton, and a computer mouse in the reported demonstration.
Form-fitting wearables and custom medical devices are potential future applications named in the KAIST article.
In this KAIST DLP study, the liquid formulation needed to flow well enough to print, while the cured material needed high stretchability.
Sources:
Reader links:
Top comments (0)