Scaling the Indian Buffet Process via Submodular Maximization

Inference for latent feature models is inherently difficult as the inference space grows exponentially with the size of the input data and number of latent features. In this work, we use Kurihara & Welling (2008)'s maximization-expectation framework to perform approximate MAP inference for linear-Gaussian latent feature models with an Indian Buffet Process (IBP) prior. This formulation yields a submodular function of the features that corresponds to a lower bound on the model evidence. By adding a constant to this function, we obtain a nonnegative submodular function that can be maximized via a greedy algorithm that obtains at least a one-third approximation to the optimal solution. Our inference method scales linearly with the size of the input data, and we show the efficacy of our method on the largest datasets currently analyzed using an IBP model.

Seafloor-riser interaction model

Fatigue stresses associated with extreme storms, vessel movements, and vortex-induced vibrations are critical to the performance of steel catenary risers. The critical location for fatigue damage often occurs within the touchdown zone, where cyclic interaction of the riser with the seabed occurs. Developing a model for seabed stiffness requires characterization of a number of complex nonlinear processes including trench formation, nonlinear soil stiffness, soil suction, and breakaway of the riser from the seafloor. The analytical framework utilized in this research considers the riser-seafloor interaction problem in terms of a pipe resting on a bed of springs, the stiffness characteristics of which are described by nonlinear load-deflection (P-y) curves. The P-y model allows for first penetration and uplift, as well as repenetration and small range motions within the bounding loop defined by extreme loading. The backbone curve is constructed from knowledge of the soil strength, the rate of strength increase with depth, trench width, and two additional parameters, while three parameters are necessary for the cyclic response. © ASCE 2009.

Seafloor interaction with steel catenary risers

This paper describes the key features of a seafloor-riser interaction model. The soil is represented in terms of non-linear load-deflection (P- y) relationships, which are also able to account for soil stiffness degradation due to cyclic loading. The analytical framework considers the riser-seafloor interaction problem in terms of a pipe resting on a bed of springs, and requires the iterative solution of a fourth-order ordinary differential equation. A series of simulations is used to illustrate the capabilities of the model. Thanks to the non-linear soil springs with stiffness degradation it is possible to simulate the trench formation process and estimate moments in a riser. Copyright © 2008 by The International Society of Offshore and Polar Engineers (ISOPE).

Numerical analysis of shallow foundations under combined horizontal and torsional loading

Foundations of subsea infrastructure in deep water subjected to asymmetric environmental loads have underscored the importance of combined torsional and horizontal loading effects on the bearing capacity of rectangular shallow foundations. The purpose of this study is to investigate the undrained sliding and torsional bearing capacity of rectangular and square shallow foundations together with the interaction response under combined loading using three-dimensional finite element (3D-FE) analysis. Upper bound plastic limit analysis is employed to establish a reference value for horizontal and torsional bearing capacity, and an interaction relationship for the combined loading condition. Satisfactory agreement of plastic limit analysis (PLA) and 3D-FE results for ultimate capacity and interaction curves ensures that simple PLA solution could be used to evaluate the bearing capacity problem of foundation under combined sliding and torsion.

Reversible CO2 absorption by the 6H perovskite Ba 4Sb2O9

A novel compound for carbon capture and storage (CCS) applications, the 6H perovskite Ba4Sb2O9, was found to be able to absorb CO2 through a chemical reaction at 873 K to form barium carbonate and BaSb2O6. This absorption was shown to be reversible through the regeneration of the original Ba4Sb 2O9 material upon heating above 1223 K accompanied by the release of CO2. A combined synchrotron X-ray diffraction, thermogravimetric, and microscopy study was carried out to characterize first the physical absorption properties and then to analyze the structural evolution and formation of phases in situ. Importantly, through subsequent carbonation and regeneration of the material over 100 times, it was shown that the combined absorption and regeneration reactions proceed without any significant reduction in the CO2 absorption capacity of the material. After 100 cycles the capacity of Ba4Sb2O9 was ∼0.1 g (CO 2)/g (sorbent), representing 73% of the total molar capacity. This is the first report of a perovskite-type material showing such good properties, opening the way for studies of new classes of inorganic oxide materials with stable and flexible chemical compositions and structures for applications in carbon capture. © 2013 American Chemical Society.

Revealing lithium-silicide phase transformations in nano-structured silicon-based lithium ion batteries via in situ NMR spectroscopy.

Nano-structured silicon anodes are attractive alternatives to graphitic carbons in rechargeable Li-ion batteries, owing to their extremely high capacities. Despite their advantages, numerous issues remain to be addressed, the most basic being to understand the complex kinetics and thermodynamics that control the reactions and structural rearrangements. Elucidating this necessitates real-time in situ metrologies, which are highly challenging, if the whole electrode structure is studied at an atomistic level for multiple cycles under realistic cycling conditions. Here we report that Si nanowires grown on a conducting carbon-fibre support provide a robust model battery system that can be studied by (7)Li in situ NMR spectroscopy. The method allows the (de)alloying reactions of the amorphous silicides to be followed in the 2nd cycle and beyond. In combination with density-functional theory calculations, the results provide insight into the amorphous and amorphous-to-crystalline lithium-silicide transformations, particularly those at low voltages, which are highly relevant to practical cycling strategies.

Computational modelling and characterisation of nanoparticle-based tuneable photonic crystal sensors

Photonic crystals are materials that are used to control or manipulate the propagation of light through a medium for a desired application. Common fabrication methods to prepare photonic crystals are both costly and intricate. However, through a cost-effective laser-induced photochemical patterning, one-dimensional responsive and tuneable photonic crystals can easily be fabricated. These structures act as optical transducers and respond to external stimuli. These photonic crystals are generally made of a responsive hydrogel that can host metallic nanoparticles in the form of arrays. The hydrogel-based photonic crystal has the capability to alter its periodicity in situ but also recover its initial geometrical dimensions, thereby rendering it fully reversible and reusable. Such responsive photonic crystals have applications in various responsive and tuneable optical devices. In this study, we fabricated a pH-sensitive photonic crystal sensor through photochemical patterning and demonstrated computational simulations of the sensor through a finite element modelling technique in order to analyse its optical properties on varying the pattern and characteristics of the nanoparticle arrays within the responsive hydrogel matrix. Both simulations and experimental results show the wavelength tuneability of the sensor with good agreement. Various factors, including nanoparticle size and distribution within the hydrogel-based responsive matrices that directly affect the performance of the sensors, are also studied computationally. © 2014 The Royal Society of Chemistry.