12.03.18 09:56

Physics: Article in Nature Communications

Dancing robots follow the whip

By: Editorial staff / A.C.

Physicists from Friedrich Alexander University Erlangen-Nuremberg (FAU) and Heinrich Heine University Düsseldorf (HHU) have proved that systems of oppositely rotating mac-roscopic particles separate and form homogenous fractions of particles that rotate either clockwise or counter-clockwise. For their experiment, the researchers used mini robots fabricated by means of 3D printing. The results have now been published in the renowned scientific journal Nature Communications.

The oppositely rotating robots separate over time: State after 10 seconds (left), 60 seconds (centre) and 900 seconds (right). The formation of homogenous domains can be clearly seen (Photos: Christian Scholz / Institute for Multiscale Simulation of Particulate Systems).

Black robot fabricated by means of 3D printing. The four satellites can be seen on the top edge.

The phenomenon is well-known: Biological organisms, such as bacteria, and artificial active particles tend to organize themselves in swarms and patterns. What has, by contrast, scarcely been researched so far is how exactly this self-organization works and what forces have an effect. Experiments on the dynamics of microscopic particles are difficult to conduct; simulations too are pushed to their limits if fundamental interaction mechanisms are not fully understood.

Physicists at FAU and HHU have now managed to observe experimentally the self-organization of rotating particles. To do so, they placed small robots, each about 1.5 centimetres in size and equipped with seven tilted legs, on a vibrating plate. The legs act as elastic springs and transform the vibration impulse into a rotational motion. To intensify their interaction, the robots fabricated in the 3D printer were fitted with four small satellites. This makes them behave like cogwheels that engage with each other. "The setup is actually quite simple," explains Professor Thorsten Pöschel from FAU's Institute for Multiscale Simulation of Particulate Systems. "In a ring, we placed a mixture of 210 clockwise and 210 counter-clockwise spinning rotors and arranged them in a chequered configuration. Then we activated the vibrating table and watched what happened."

The result surprised even the researchers: After just a minute, distinct domains could be clearly recognized and after 15 minutes the robots had separated almost completely. "This segmentation is not intuitive," says Dr. Christian Scholz from the Department of Theoretical Physics II at HHU, who was primarily responsible for designing and performing the experiments and for the analysis of the results. "We might have expected that above all oppositely rotating particles would remain together because their satellites do not become entangled - similar to a chain of rotating cogwheels that alternately rotate clockwise and counter-clockwise." But the opposite is the case: Rotors turning in the same direction block each other and form joint fractions. By tracking the individual robots, the researchers were also able to observe super-diffusive edge currents: Particles close to the interface are far more agile than those in the centre of the domains.

Numerous repetitions show that the results of the experiments are very robust - after about 1000 seconds the rotors had mostly formed three to four separate domains. In corresponding simulation runs on the basis of what are referred to as Langevin equations, the robots even separate completely into two fractions. "That the variations in the trial runs are greater than in the simulation could be due to irregularities in our 3D-printed rotors and the influence of gravitation, because we are unable to adjust the vibrating table so that it is 100% level," explains Professor Michael Engel from FAU's Institute for Multiscale Simulation of Particulate Systems.

Both methods - experimental trials with physical rotors as well as Langevin simulation - are ideally suited for describing the collective dynamics and phase separation of rotating particles. The researchers hope that this work will make a contribution to the further study of active soft matter and microscopic or even molecular particles.  

Original publication

Christian Scholz, Michael Engel and Thorsten Pöschel, Rotating Robots Move Collectively and Self-Organize, Nature Communications, 2 March 2018

DOI: 10.1038/s41467-018-03154-7

Online: https://www.nature.com/articles/s41467-018-03154-7


 Editor / Arne Claussen

Responsible for the content: E-MailStabsstelle Presse und Kommunikation