Computational design of advanced engineering materials

Recent advances in technology, among others the development of 3-D printing, opened the possibilities for synthetizing nearly any conceivable material. In order to efficiently explore these new possibilities, the rules are needed for designing materials with required mechanical (strength), thermal and electrical (conductivity) properties aa well as the specific chemical and biological characteristics (corrosion resistance, interactions with living cells). Fundamentally, such design rules are based on the understanding of the processes taking place in solids under the applied loads, which in turn depend on their chemical composition and on their structure, as defined by underlying type of the arrangements of the atoms (crystal-like, amorphous) and the specific point-like (e.g. vacancies), linear (e.g. dislocations), planar (e.g. grain boundaries) and volumetric deviations from these, termed as the defects.

It has been a great achievement of the modern materials science to show that there are properties of the engineering materials, which are governed by the atom-to-atom interactions (e.g. elastic modulus), by the arrangement of the atoms (e.g. elastic anisotropy) and by various deviations from the underlying pattern, (e.g. conductivity, plasticity, etc.). The respective relationships between the given type of the defects and the properties of interest are qualitatively quite well explained by the solid-state physics, mechanics and chemistry. The challenge is, however, to quantify these relationships, and in such a way that they can be used for predicting, and thus designing, materials of the desired functionalities. Because of the multi-length scale complexity of the modern engineering material this challenge can be efficiently addressed only by adopting a computational modelling approach.

In this context one should acknowledge that over the recent years a major progress has been made in the numerical methods for predicting the properties of the engineering materials. The methods are available for modelling the properties determined by arrangement of atoms in relatively small representative unit volume, – e.g. DFT calculations – and in large assemblies – e.g. molecular dynamics, MD, computations. The models have been also developed for explaining the properties of materials, structure of which can be defined only using relatively large representative unit volumes –e.g. finite element method combined with DFT and molecular dynamics.

The lecture provides examples of the application of the above-mentioned methods to modelling, predicting and optimizing (thus for designing) advanced engineering materials.

These examples include design of so called nano-materials, structure of which is controlled at the length scale below 100 nm. The examples given concern mechanical strength, thermal properties, thermal stability, diffusivity and conductivity of nano-metals and nano-composites, also with graphene. The results of the modelling are validated with the experimental results. Based on the results of modelling the rules of optimising/design of nano-materials are proposed.

Lecturer:

Prof K.J. Kurzydlowski

Warsaw University of Technology

Materials Science and Engineering

Materials Design Group

The 21st century will depend as much on photonics as the 20th century depended on electronics. The market for photonic technology is currently about one fifth of the global market for electronics but it is growing faster. And the prospect to shrink many optical functions on a single chip mimicking the path followed by integrated electronic is certainly appealing, although challenging. In photonics, integration technology is still in its infancy, very expensive and looking for large markets but it is developing rapidly.

The lecture aims to introduce, to a broad audience, the field of integrated photonics at large presenting the state of the art, potentialities and limits and a survey of applications in a multitude of fields, from telecommunication to biophotonics, from microwave photonics to quantum phenomena. The current enthusiasm for silicon photonics, the perception that photonic chips be suitable for consumer markets, the awareness that integrated photonics is a virgin territory for science and a gold mine for companies are boosting both science and industries towards new emerging initiatives around the globe. A short overview on roadmaps and visions in Europe and US will be described.

This is a Zhorez Alferov named lecture, organized by Dr. Pavel Cheben, National Research Council of Canada, in agreement with Facultad de Física and Sociedad Cubana de Física.

Compartir experiencia en investigación clinica acerca del uso de las Imagenes de Resonancia Magnética por Difusion Pesada (Diffusion-Weighted MRI) en diagnostico clínico (por ejemplo: cáncer de próstate), y seguimiento de tratamiento de inmuno terapia (metástasis hepática y/o cerebral de melanoma maligno).

El Prof. Daniel Stolik es profesor del Instituto de Ciencia y Tecnología de los Materiales y de la Facultad de Física, ambos de la UH. Es asesor del MINEM y gran promotor de la energía fotovoltaica. Es un experto reconocido por su dominio del estado del arte de la energía fotovoltaica en Cuba y en el mundo. En esta breve conferencia de 40 minutos el profesor nos actualizara sobre esta importante producción, limpia y rentable, de electricidad. Habrá oportunidad para preguntas y aclaraciones.

Electronic circuits composed of one or more elements with inherent memory -- memristors, memcapacitors and meminductors -- offer lower circuit complexity and enhanced functionality for certain computational tasks. Networks of these elements are proposed for novel computational paradigms that rely on information processing and storage on the same physical platform. We show a nanoscaled memdevice able to act as an electronic analogue of tipping buckets that allows reducing the dimensionality and complexity of a sensing problem by transforming it into a counting problem.
The device offers a well adjustable, tunable and reliable periodic reset that is controlled by the amounts of transferred quantum dot charges per gate voltage sweep. The presented memristor enables optical conductance control and may thus be considered for sensory applications in artificial neural networks as light-sensitive synapses or optically tunable memories.
This allows the integration of electrical and optical signals with a single memristor and enables the realization of complex logic functionalities with single devices. We also present the realization of four different learning rules with a quantum dot memristor synapse by tuning the shape, the magnitude, the polarity and the timing of voltage pulses.

Sistemas en los que una interacción atractiva de corto alcance compite con una interacción de largo alcance repulsiva,  bajo condiciones físicas apropiadas, desarrollan modulaciones en el parámetro de orden local.
Típicamente son observados diagramas de fases que contienen fases de fajas o fases de burbujas, dependiendo de parámetros como el campo externo aplicado o la temperatura.

En el primer seminario (Martes 10 de Enero) presentaremos una teoría para la fusión orientaciones de sistemas de fajas (stripes) clásicos, mostrando la existencia de una fase nemática estable en dos dimensiones, para sistemas en los que la interacción repulsiva es de suficiente largo alcance.

En el segundo seminario (Jueves 12 de Enero) se considera la presencia de fluctuaciones cuánticas en el sistema y sus efectos. De esta forma,  se estudia la transición de fase cuántica entre la fase ordenada orientacionalmente y la fase de “lí­quido de fajas”. Las propiedades críticas de esta transición son calculadas vía Grupo de Renormalización perturbativo.

Quantum computers are physical devices which, if they can be built, would allow to manipulate a global quantum state to accomplish some computational task. A main experimental challenge is how to keep the state isolated from the surrounding world so that it remains quantum, and does not turn classical.

This challenge comes with the theoretical problem of estimating the errors made by a quantum computer which is not perfectly isolated from the environment. In the quantum computing literature this has mainly been addressed in a factorized model instroduced by Aharonov, Kitaev and Nisan in 1998.

In this talk I will describe how to estimate the errors of general quantum computation by the Feynman-Vernon method, hence by-passing the assumptions in the Aharonov-Kitaev-Nisan theory. I will show how some simple estimates can be obtained for idealized systems, and how they can also be extended for more advanced schemes such as Kitaev's toric code. I will also discuss quantum error correction and error protection in such systems. The talk is mainly based on arXiv:1606.09407.

The last five years has seen a revolution in the ability to predict spatial contacts in protein structures from sequence data. Starting from many similar protein sequences and using them to learn Ising/Potts models of statistical physics leads to much better contact predictions than earlier bioinformatic methods, and there are also reports that it helps in full in silico protein structure prediction.

In this talk I will review these developments and show some examples of the progress made and methods used. I will also take the broader view and ask what is the underlying biological mechanism behind this success, and I will argue that it is a form of epistatis, or synergistic effects in genetic variation, which is captured by the inverse Ising / inverse Potts procedure. This suggests the same methods can be used to identify epistasis on the genome scale, and I will show that this is indeed so for the important human pathogen Spneumoniae.

This talk is built on joint work with many people, but especially on the two papers Feinauer et al PLoS Comput Biol (2014) and Skwark et al PLoS Genetics (2016, in press, bioRxiv: 071696)

We will review some of our latest results concerning the application of techniques from Statistical Mechanics to understand the behaviour of known algorithms and to the design of new ones. These algorithms are of relevance to the study of physical systems defined on a lattice, but also to solve problems of interdisciplinary character, from Bio-informatics to signaling processing. We will also sketch preliminary ideas about future research areas currently under development.

We present a new method to close the Master Equation representing the continuous time dynamics of Ising interacting spins. The method makes use of the the theory of Random Point Processes to derive a master equation for local conditional probabilities. We analytically test our solution studying two known cases, the dynamics of the mean field ferromagnet and the dynamics of the one dimensional Ising system. We then present numerical results comparing our predictions with Monte Carlo simulations in three different models on random graphs with finite connectivity: the Ising ferromagnet, the Random Field Ising model, and the Viana-Bray spin-glass model.