How small can a robot get? According to a team of researchers at Georgia Tech, really, really small. Described in the July 23 issue of the journal Soft Matter, the Georgia Tech team has been running complex computational models of swimming robots on the micron (0.001 mm or about 0.000039 inches) scale. At this microscopic level, water takes on very different properties from those of the human scale, but despite these challenges the team believes that such robots could have fascinating practical applications.
Designed by team leader Alexander Alexeev, assistant professor in the George W. Woodruff School of Mechanical Engineering, Hassan Masoud and Benjamin Bingham, the simulated microorganism-sized robots faced unusual challenges. Swimming on so tiny a scale isn’t like paddling about in a pool. At that size, water is as thick as honey and a micro-robot hasn’t any inertia to move it forward, so it isn’t so much swimming as crawling through glue. That means more was involved than just modelling a tiny robot and sticking a propellor on. It had to be designed from scratch.
The basis for the micro-robots was hydrogels. A hydrogel is a sort of super sponge. It’s a network of polymer chains that trap water much in the way that proteins in cooked egg whites do. In fact, hydrogels trap water so well that a blob of saturated hydrogel is over 99 percent water, and looks alarmingly like a blob from outer space. Hydrogels have a wide variety of applications from bioengineering to keeping lawns moist between watering and it's even used in disposable diapers.
By making a robot out of hydrogel, the Georgia Tech thinks that it could use expanding and contracting the hydrogel as a “chemical engine” to move tiny flaps that would propel the swimmer.
The micro-robot currently used in the models is about ten microns long and has a flap on either side of its body. A third flap projects forward. This is a steering flap that responds to light, heat, chemicals or other stimuli. The oscillating volume that the robot uses for propulsion is set off by changes in its environment, such as temperature shifts, chemical reactions or oscillating electrical fields. Meanwhile, the front flap acts as a sort of rudder. The robot can swim, though not very fast. Top speed is estimated to be only a few micrometers per second.
“The combination of these flaps and the oscillating body creates a very nice motion that we believe can be used to propel the swimmer,” said Alexeev. “To build a device that is autonomous and self-propelling at the micron-scale, we cannot build a tiny submarine. We have to keep it simple.”
Modelling is very important because there are so many variables involved. Different flaps and bodies needed to be studied, for example, and it’s only in a computer that these can be handled practically.
“We have captured the solid mechanics of the periodically-oscillating body, the fluid dynamics of moving through the viscous liquid and the coupling between the two," says Alexeev. "From a computational fluid dynamics standpoint, it’s not an easy problem to model at this scale.”
The hope is that eventually the team’s modelling work will be of benefit to engineers building the first micro-robot prototypes. The feedback between the simulations and practical testing would make development much faster and easier.
The team also hopes that the micro-bots will find practical applications. They are particularly keen to see them used to move cargo through microfluidic chips, operating lab-on-a-chip devices and maybe acting as swarms of tiny construction robots building components on a tiny scale impossible with today’s techniques.
Source: Georgia Tech