Discovering causal structure with reproducing-kernel Hilbert space ε-machines.

We merge computational mechanics' definition of causal states (predictively equivalent histories) with reproducing-kernel Hilbert space (RKHS) representation inference. The result is a widely applicable method that infers causal structure directly from observations of a system's behaviors whether they are over discrete or continuous events or time. A structural representation-a finite- or infinite-state kernel ϵ-machine-is extracted by a reduced-dimension transform that gives an efficient representation of causal states and their topology. In this way, the system dynamics are represented by a stochastic (ordinary or partial) differential equation that acts on causal states. We introduce an algorithm to estimate the associated evolution operator. Paralleling the Fokker-Planck equation, it efficiently evolves causal-state distributions and makes predictions in the original data space via an RKHS functional mapping. We demonstrate these techniques, together with their predictive abilities, on discrete-time, discrete-value infinite Markov-order processes generated by finite-state hidden Markov models with (i) finite or (ii) uncountably infinite causal states and (iii) continuous-time, continuous-value processes generated by thermally driven chaotic flows. The method robustly estimates causal structure in the presence of varying external and measurement noise levels and for very high-dimensional data.

Transparent experiments: releasing data from mechanical tests on three dimensional hydrogel sphere packings.

We describe here experiments on the mechanics of hydrogel particle packings from the Behringer lab, performed between 2012 and 2015. These experiments quantify the evolution of all contact forces inside soft particle packings exposed to compression, shear, and the intrusion of a large intruder. The experimental set-ups and processes are presented and the data are concomitantly published in a repository (Barés et al. in Dryad, Dataset 10.5061/dryad.6djh9w0x8, 2019).

Development and characterization of bioactive edible films from spider crab (Maja crispata) chitosan incorporated with Spirulina extract.

Active food packaging films based on crab chitosan and Spirulina extract (SE) were developed. The effects of the SE incorporation at different levels on physical (color, opacity water vapor and oxygen permeability) and mechanical (tensile strength and elongation at break) properties of chitosan films were investigated. FTIR was carried out to observe the potential modifications of the chitosan films when incorporated with SE. The obtained results suggested that incorporation of SE into chitosan films improved mechanical and barrier properties. The antioxidant activity of the chitosan/SE films was characterized by means of three different analytical assays (DPPH, FRAP and FIC). Crab chitosan edible films containing SE showed higher antioxidant activity, regardless concentrations and methods assayed. Furthermore, the antioxidant activity occurred in a concentration-dependent manner. The agar disc diffusion method was used to determine the antibacterial activities of chitosan edible films against Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa, Listeria monocytogenes, Salmonella typhimurium, Bacillus subtilis and Bacillus cereus. The chitosan/SE films were more effective (p<0.05) against five of the seven tested bacteria. The obtained crab chitosan edible films incorporated with SE showed great potential to be used for active food packaging due to its excellent antioxidant and antibacterial activities.

Refractive index matched scanning and detection of soft particles.

We describe here how to apply the three-dimensional imaging technique of refractive index matched scanning to hydrogel spheres. Hydrogels are water based materials with a low refractive index, which allows for index matching with water-based solvent mixtures. We discuss here various experimental techniques required to handle specifically hydrogel spheres as opposed to other transparent materials. The deformability of hydrogel spheres makes their identification in three-dimensional images non-trivial. We will also discuss numerical techniques that can be used in general to detect contacting, non-spherical particles in a three-dimensional image. The experimental and numerical techniques presented here give experimental access to the stress tensor of a packing of deformed particles.

Multiple-contact discrete-element model for simulating dense granular media.

This article presents a new force model for performing quantitative simulations of dense granular materials. Interactions between multiple contacts (MC) on the same grain are explicitly taken into account. Our readily applicable MC-DEM method retains all the advantages of discrete-element method simulations and does not require the use of costly finite-element methods. The new model closely reproduces our recent experimental measurements, including contact force distributions in full 3D, at all compression levels of the packing up to the experimental maximum limit of 13%. Comparisons with classic simulations using the nondeformable spheres approach, as well as with alternative models for interactions between multiple contacts, are provided. The success of our model, compared to these alternatives, demonstrates that interactions between multiple contacts on each grain must be included for dense granular packings.

Spanning the scales of granular materials through microscopic force imaging.

If you walk on sand, it supports your weight. How do the disordered forces between particles in sand organize, to keep you from sinking? This simple question is surprisingly difficult to answer experimentally: measuring forces in three dimensions, between deeply buried grains, is challenging. Here we describe experiments in which we have succeeded in measuring forces inside a granular packing subject to controlled deformations. We connect the measured micro-scale forces to the macro-scale packing force response with an averaging, mean field calculation. This calculation explains how the combination of packing structure and contact deformations produce the observed nontrivial mechanical response of the packing, revealing a surprising microscopic particle deformation enhancement mechanism.

Self-assembly in a near-frictionless granular material: conformational structures and transitions in uniaxial cyclic compression of hydrogel spheres.

We use a Markov transition matrix-based analysis to explore the structures and structural transitions in a three-dimensional assembly of hydrogel spheres under cyclic uniaxial compression. We apply these methods on experimental data obtained from a packing of nearly frictionless hydrogel balls. This allows an exploration of the emergence and evolution of mesoscale internal structures - a key micromechanical property that governs self-assembly and self-organization in dense granular media. To probe the mesoscopic force network structure, we consider two structural state spaces: (i) a particle and its contacting neighbours, and (ii) a particle's local minimal cycle topology summarized by a cycle vector. In both spaces, our analysis of the transition dynamics reveals which structures and which sets of structures are most prevalent and most likely to transform into each other during the compression/decompression of the material. In compressed states, structures rich in 3-cycle or triangle topologies form in abundance. In contrast, in uncompressed states, transitions comprising poorly connected structures are dominant. An almost-invariant transition set within the cycle vector space is discovered that identifies an intermediate set of structures crucial to the material's transition from weakly jammed to strongly jammed, and vice versa. Preferred transition pathways are also highlighted and discussed with respect to thermo-micro-mechanical constitutive formulations.

Fixed-bed adsorption of toluene on high silica zeolites: experiments and mathematical modelling using LDF approximation and a multisite model.

The adsorption of toluene (TOL) as a target volatile organic compound has been studied experimentally and modelled on various hydrophobic zeolites: Faujasite (FAU), ZSM-5 (Z) and Mordenite (MOR). The influence of the nature of the compensating cation (H+ or Na+) has also been investigated for ZSM-5 zeolite, which is known to possess three kinds of adsorption sites (sinusoidal channels, straight channels and intersections). Type I isotherms observed on FAU, Na-Z and MOR fitted well with the Langmuir model. A deviation from a type I isotherm was observed for H-Z, because of the structure of this zeolite. The Successive Langmuir Model was more successful to fit the 'bump' of the experimental curve than the Double Langmuir. Classical shapes were found for MOR, FAU and Na-Z breakthrough curves that were fitted with good accuracy using the Linear Driving Force (LDF) approximation. In the case of H-Z, a change of profile was observed during the dynamic adsorption and the differences seen between the Na-Z and H-Z behaviours were explained by the strong interactions between Na+ and adsorbed TOL at the intersection sites. The Na+ cations prevented reorientation of TOL molecules at the intersection and thereby avoided the filling of the sinusoidal channel segments. Thus, a specific model was developed for fitting the breakthrough curve of H-Z. The model developed took into account these two types of adsorption sites with the overall uptake for each site being given by an LDF approximation.

Shallow granular flows down flat frictional channels: steady flows and longitudinal vortices.

Granular flows down inclined channels with smooth boundaries are common in nature and industry. Nevertheless, flat boundaries have been much less investigated than bumpy ones, which are used by most experimental and numerical studies to avoid sliding effects. Using numerical simulations of each grain and of the side walls we recover quantitatively experimental results. At larger angles we predict a rich behavior, including granular convection and inverted density profiles suggesting a Rayleigh-Bénard type of instability. In many aspects flows on a flat base can be seen as flows over an effective bumpy base made of the basal rolling layer, giving Bagnold-type profiles in the overburden. We have tested a simple viscoplastic rheological model [Nature (London) 441, 727 (2006)] in average form. The transition between the unidirectional and the convective flows is then clearly apparent as a discontinuity in the constitutive relation.

Removal of volatile organic compounds by heterogeneous ozonation on microporous synthetic alumina silicate.

A hybrid process combining adsorption and ozonation was examined as an alternative treatment for odorous volatile organic compounds (VOCs). Methyl ethyl ketone (MEK) was chosen to study the influence of operating parameters. Two synthetic aluminosilicates (faujasite-Y and ZSM-5) were tested for adsorption and reactivity with ozone. The adsorption equilibrium measurement on both adsorbents showed that adsorption performance depends on temperature but is not sensitive to relative humidity, due to the hydrophobic properties of the materials. Adsorbed VOCs were oxidized at low temperature when ozonated flow was sent to the reactor. Regeneration of the fixed bed was achieved at the same time, releasing mainly CO(2) and H(2)O. Intermediates of oxidation, such as 2,3-butanedione and acetic acid, were identified, leading to incomplete mineralization. The influence of concentration and humidity are discussed. Four successive cycles were tested: after the first adsorption/ozonation cycle, the adsorption efficiency was not affected during subsequent cycles. These results show that the same sample of adsorbent can be used in the treatment process for a long time. Ozonation regeneration is a promising process for VOC removal.