20 research outputs found
Air mediates the impact of a compliant hemisphere on a rigid smooth surface
No full textFleeting contact between solids immersed in a fluid medium governs the response of critically important materials, from coffee to soil. Rapid impact of soft solids occurs in systems as diverse as car tires, soft robotic locomotion and suspensions, including soil and coffee. In each of these systems, the dynamics are fundamentally altered by the presence of a fluid layer mediating solid contact. However, observing this class of interactions directly is challenging, as the relevant time and length scales are extremely small. Here we directly image the interface between a soft elastic hemisphere and a flat rigid substrate during rapid impact over a wide range of impact velocities V at high temporal and spatial resolution using the Virtual Frame Technique (VFT). In each experiment, a pocket of air is trapped in a dimple between the impactor and the substrate, preventing direct solid-solid contact at the apex of the hemisphere. Thus, unlike the quasi-static Hertzian solution where contact forms in an ellipse, in each rapid air-mediated impact, contact forms in an annular region which rapidly grows both inward toward the impact axis, and rapidly outward away from the impact axis. We find that the radius of initial contact varies non-monotonically with V, indicating a transition between elastically dominated dynamics to inertially dominated dynamics. Furthermore, we find that for slower impact speeds, where the outer contact front cannot outpace the Rayleigh velocity, contact expands in a patchy manner, indicating an elasto-lubricative instability. These behaviors, observed using the VFT, occur in regimes relevant to a wide variety of soft systems, and might modulate frictional properties during contact. The size of the air pocket varies with V and impactor stiffness. Our measurements reveal an unanticipated, sudden transition of the air pocket's size as V increases beyond 1 m s(-1) and multiple modes of air entrainment at the advancing solid-solid contact front that depend on the front's velocity.EMS
Demonstration of Decentralized, Physics-Driven Learning
No full textIn typical artificial neural networks, neurons adjust according to global
calculations of a central processor, but in the brain neurons and synapses
self-adjust based on local information. Contrastive learning algorithms have
recently been proposed to train physical systems, such as fluidic, mechanical,
or electrical networks, to perform machine learning tasks from local evolution
rules. However, to date such systems have only been implemented in silico due
to the engineering challenge of creating elements that autonomously evolve
based on their own response to two sets of global boundary conditions. Here we
introduce and implement a physics-driven contrastive learning scheme for a
network of variable resistors, using circuitry to locally compare the response
of two identical networks subjected to the two different sets of boundary
conditions. Using this innovation, our system effectively trains itself,
optimizing its resistance values without use of a central processor or external
information storage. Once the system is trained for a specified allostery,
regression, or classification task, the task is subsequently performed rapidly
and automatically by the physical imperative to minimize power dissipation in
response to the given voltage inputs. We demonstrate that, unlike typical
computers, such learning systems are robust to extreme damage (and thus
manufacturing defects) due to their decentralized learning. Our twin-network
approach is therefore readily scalable to extremely large or nonlinear networks
where its distributed nature will be an enormous advantage; a laboratory
network of only 500 edges will already outpace its in silico counterpart.Comment: 11 pages 7 figure
Physical learning beyond the quasistatic limit
No full textPhysical networks, such as biological neural networks, can learn desired
functions without a central processor, using local learning rules in space and
time to learn in a fully distributed manner. Learning approaches such as
equilibrium propagation, directed aging, and coupled learning similarly exploit
local rules to accomplish learning in physical networks such as mechanical,
flow, or electrical networks. In contrast to certain natural neural networks,
however, such approaches have so far been restricted to the quasistatic limit,
where they learn on time scales slow compared to their physical relaxation.
This quasistatic constraint slows down learning, limiting the use of these
methods as machine learning algorithms, and potentially restricting physical
networks that could be used as learning platforms. Here we explore learning in
an electrical resistor network that implements coupled learning, both in the
lab and on the computer, at rates that range from slow to far above the
quasistatic limit. We find that up to a critical threshold in the ratio of the
learning rate to the physical rate of relaxation, learning speeds up without
much change of behavior or error. Beyond the critical threshold, the error
exhibits oscillatory dynamics but the networks still learn successfully.Comment: 26 pages, 5 figure
Cornerstones are the Key Stones: Using Interpretable Machine Learning to Probe the Clogging Process in 2D Granular Hoppers
No full textThe sudden arrest of flow by formation of a stable arch over an outlet is a unique and characteristic feature of granular materials. Previous work suggests that grains near the outlet randomly sample configurational flow microstates until a clog-causing flow microstate is reached. However, factors that lead to clogging remain elusive. Here we experimentally observe over 50,000 clogging events for a tridisperse mixture of quasi-2D circular grains, and utilize a variety of machine learning (ML) methods to search for predictive signatures of clogging microstates. This approach fares just modestly better than chance. Nevertheless, our analysis using linear Support Vector Machines (SVMs) highlights the position of potential arch cornerstones as a key factor in clogging likelihood. We verify this experimentally by varying the position of a fixed (cornerstone) grain, and show that such a grain dictates the size of feasible flow-ending arches, and thus the time and mass of each flow. Positioning this grain correctly can even increase the ejected mass by over 50%. Our findings demonstrate that interpretable ML algorithms like SVMs can uncover meaningful physics even when their predictive power is below the standards of conventional ML practice.16 pages, 11 figure
Seismological Stress Drops for Confined Ruptures Are Invariant to Normal Stress
No full textAbstract Seismic moment and rupture length can be combined to infer stress drop, a key parameter for assessing earthquakes. In natural earthquakes, stress drops are largely depth‐independent, which is surprising given the expected dependence of frictional stress on normal stresses and hence overburden. We have developed a transparent experimental fault that allows direct observation of thousands of slip events, with ruptures that are fully contained within the fault. Surprisingly, the observed stress drops are largely independent of both the magnitude of normal stress and its heterogeneity, capturing the independence seen in nature. However, we observe larger, normal stress‐dependent stress drops when the fault area is reduced, which allows slip events to frequently reach the edge of the interface. We conclude that confined ruptures have normal stress independent stress drops, and thus the depth‐independent stress drops of tectonic earthquakes may be a consequence of their confined nature
Stochastic dynamics of granular hopper flows: a configurational mode controls the stability of clogs
No full textGranular flows in small-outlet hoppers exhibit several characteristic but
poorly understood behaviors: temporary clogs (pauses) that last for an extended
period before flow spontaneously restarts, permanent clogs that last
indefinitely, and non-Gaussian, non-monotonic flow-rate statistics. These
aspects have been extensively studied independently, but a model of hopper flow
that includes all three has not been formulated. Here, we introduce such a
phenomenological model that provides a unifying dynamical explanation of all
three behaviors: a coupling between the flow rate and a hidden configurational
mode that controls the stability of clogs. In the theory, flow rate evolves
according to Langevin dynamics with multiplicative noise and an absorbing state
at zero flow, conditional on the hidden mode. The model fully reproduces the
statistics of pause and clog events taken from a large ( flows)
experimental dataset, including non-exponentially distributed clogging times
and non-Gaussian flow rate distribution, and explains the stretched-exponential
growth of the average clogging time with outlet size. Further, we highlight
several microscopic configurational features of this hidden mode, including
size and smoothness of the static arch structure formed during pauses and
clogs. Our work provides a unifying framework for several poorly understood
granular phenomena, and suggests numerous new paths toward further
understanding of this complex system.Comment: 7 pages, 4 figures; SM included as an ancillary PDF fil
Beyond Quality and Quantity: Contact Distribution Encodes Frictional Strength
No full textClassically, the quantity of contact area between two bodies is
considered a proxy for the force of friction. However, bond density across the
interface - quality of contact - is also relevant, and contemporary debate
often centers around the relative importance of these two factors. In this
work, we demonstrate that a third factor, often overlooked, plays a significant
role in static frictional strength: the distribution of contact. We perform
static friction measurements, , on three pairs of solid blocks while
imaging the contact plane. By using linear regression on hundreds of
image- pairs, we are able to predict future friction measurements with 3
to 7 times better accuracy than existing benchmarks, including total quantity
of contact area. Our model has no access to quality of contact, and we
therefore conclude that a large portion of the interfacial state is encoded in
the spatial distribution of contact, rather than its quality or quantity.Comment: 6 pages 4 figure
Virtual Frame Technique: Ultrafast Imaging with Any Camera
No full textMany phenomena of interest in nature and industry occur rapidly and are difficult and cost-prohibitive to visualize properly without specialized cameras. Here we describe in detail the Virtual Frame Technique (VFT), a simple, useful, and accessible form of compressed sensing that increases the frame acquisition rate of any camera by several orders of magnitude by leveraging its dynamic range. VFT is a powerful tool for capturing rapid phenomenon where the dynamics facilitate a transition between two states, and are thus binary. The advantages of VFT are demonstrated by examining such dynamics in five physical processes at unprecedented rates and spatial resolution: fracture of an elastic solid, wetting of a solid surface, rapid fingerprint reading, peeling of adhesive tape, and impact of an elastic hemisphere on a hard surface. We show that the performance of the VFT exceeds that of any commercial high speed camera not only in rate of imaging but also in field of view, achieving a 65MHz frame rate at 4MPx resolution. Finally, we discuss the performance of the VFT with several commercially available conventional and high-speed cameras. In principle, modern cell phones can achieve imaging rates of over a million frames per second using the VFT
Training self-learning circuits for power-efficient solutions
No full textAs the size and ubiquity of artificial intelligence and computational machine learning models grow, the energy required to train and use them is rapidly becoming economically and environmentally unsustainable. Recent laboratory prototypes of self-learning electronic circuits, such as “physical learning machines,” open the door to analog hardware that directly employs physics to learn desired functions from examples at a low energy cost. In this work, we show that this hardware platform allows for an even further reduction in energy consumption by using good initial conditions and a new learning algorithm. Using analytical calculations, simulations, and experiments, we show that a trade-off emerges when learning dynamics attempt to minimize both the error and the power consumption of the solution—greater power reductions can be achieved at the cost of decreasing solution accuracy. Finally, we demonstrate a practical procedure to weigh the relative importance of error and power minimization, improving the power efficiency given a specific tolerance to error
