Somewhere in a laboratory at the University of Pennsylvania, a fine copper thread no thicker than a human hair has been wound through a soft, springy fiber, giving it a quality that artificial muscles have long lacked: an awareness of its own shape. The fiber heats and contracts when current passes through, and the same copper reveals through its changing resistance how far it has moved, so the fiber can hold a chosen position rather than simply snapping between extremes. Humanity has spent decades teaching machines to move gently, and this small advance, in which sensing and acting become one
Graphene-Loaded Artificial Muscles Sense Their Own Contraction and Hold Positions on Demand
Somewhere in a laboratory at the University of Pennsylvania, a fine copper thread no thicker than a human hair has been wound through a soft, springy fiber, giving it a quality that artificial muscles have long lacked: an awareness of its own shape. The fiber heats and contracts when current passes through, and the same copper reveals through its changing resistance how far it has moved, so the fiber can hold a chosen position rather than simply snapping between extremes. Humanity has spent decades teaching machines to move gently, and this small advance, in which sensing and acting become one and the same thread, marks a patient step toward robots that can feel their own motion, though the work remains within the walls of the lab.
A copper coil 50 micrometers across runs the length of each artificial muscle fiber developed at the University of Pennsylvania, and it does two jobs at once. Current through the coil heats the fiber so it contracts, while the coil's electrical resistance reveals how far the contraction has gone. The fibers, called CoilLCE, combine a liquid crystal elastomer with graphene powder and the embedded copper. The team says this gives soft actuators something they have rarely had: the ability to sense their own state and hold intermediate positions on demand.
The sensing works because copper's resistance changes linearly with temperature, so no separate sensor is needed. Xuelin Wu, a co-author, described using that resistance as a feedback signal, adjusting heating power in real time. With a PID controller closing the loop, the fibers hit strain targets within about 2%. One held a 30% contraction under a 13 g load for 30 minutes, and all recovered within five seconds from blasts of liquid nitrogen and hot air.
A small amount of graphene, just 0.25% by weight, made a large difference. Toughness rose to 632 MJ/m³, more than three times that of pure LCE, while stiffness fell toward the range of skeletal muscle. The researchers credit graphene sheets that separate polymer chains, lowering crosslink density without disturbing the alignment that actuation depends on. More was not better: at 0.75%, alignment broke down and both strain and breaking stress declined.
In antagonistic pairs, where one fiber pulls against a passively stretched partner, the softer graphene-loaded fibers reached 16.1% actuation strain against 11.2% for pure LCE. The fibers, 0.55 to 0.73 mm across, achieved strains above 40% and held up through 116 heating and cooling cycles without measurable degradation.
Sung called this the first feedback control demonstrated on an LCE actuator, a necessary step toward future robotics. The work is still at laboratory scale, and the open question is how the approach performs beyond the demonstrations with the robotic eyes and hand.
A thread of copper, 50 micrometers across, runs the length of each artificial muscle fiber developed at the University of Pennsylvania. It does two jobs at once. Current through the coil heats the fiber and makes it contract, and the coil's own electrical resistance tells the system how far the contraction has gone.
The fibers are called CoilLCE. They combine a liquid crystal elastomer, or LCE, with multilayer graphene powder and the embedded copper coil, and they turn electrical signals into muscle-like motion. The team says the result is a capability that has largely eluded soft actuators: fibers that sense their own contraction state and can hold intermediate positions on demand.
The sensing works because copper's resistivity varies linearly with temperature. Resistance therefore reports the fiber's state without a separate sensor. Xuelin Wu, a co-author, described the approach: the team used the coil's resistance as a feedback signal and adjusted heating power in real time to keep resistance near a target value. With a PID controller closing that loop, the fibers hit strain targets to within about 2%. One held a 30% contraction under a 13 g load for 30 minutes. They also recovered within five seconds from blasts of liquid nitrogen and hot air.
The graphene is present in a very small amount, just 0.25 percent by weight. At that loading the composite reached a toughness of 632 MJ/m³, more than three times the 192 MJ/m³ of a pure LCE fiber. Its stiffness went down rather than up, bringing it closer to the Young's modulus of skeletal muscle, which runs from 0.1 to 0.3 MPa. The researchers attribute this to graphene sheets acting as physical barriers between oligomer chains. That lowers the effective crosslink density without disrupting the liquid crystal alignment the actuation depends on.
More graphene did not help. At 0.75 percent by weight, the alignment was substantially impaired, and both actuation strain and breaking stress declined.
The lower stiffness has a practical payoff in antagonistic pairs, where one fiber contracts against another that is passively stretched. A softer fiber pushes back less. In those pairs the graphene-loaded fibers reached an actuation strain of 16.1%, against 11.2% for pure LCE, an improvement of about 44%. The graphene also made heating more uniform along the fiber.
The fibers measure 0.55 to 0.73 mm across, comparable to small human muscle bundles. They deliver actuation strains above 40% at rates up to around 32% per second, with a work capacity of 343 J/kg. The team reported no measurable degradation over 116 heating and cooling cycles. Ziyun Zhang, co-lead author, is pictured with the robotic eyes and hand used to demonstrate the muscles.
Sung, quoted in the report, called this the first demonstration of feedback control on an LCE actuator, and a necessary step before such artificial muscles can be used in future robotic applications. The work is laboratory-scale, so the next thing to watch is how the feedback approach holds up beyond the demonstrations with the robotic eyes and hand.
Möchten Sie die ganze Geschichte?
Das Original lesen bei Graphene-Info ↗