An optimization method for conceptual design in engineering systems based on high order singular value decomposition

Se presenta un nuevo método de diseño conceptual en Ingeniería Aeronáutica basado el uso de modelos reducidos, también llamados modelos sustitutos (‘surrogates’). Los ingredientes de la función objetivo se calculan para cada indiviudo mediante la utilización de modelos sustitutos asociados a las distintas disciplinas técnicas que se construyen mediante definiciones de descomposición en valores singulares de alto orden (HOSVD) e interpolaciones unidimensionales. Estos modelos sustitutos se obtienen a partir de un número limitado de cálculos CFD. Los modelos sustitutos pueden combinarse, bien con un método de optimización global de tipo algoritmo genético, o con un método local de tipo gradiente. El método resultate es flexible a la par que mucho más eficiente, computacionalmente hablando, que los modelos convencionales basados en el cálculo directo de la función objetivo, especialmente si aparecen un gran número de parámetros de diseño y/o de modelado. El método se ilustra considerando una versión simplificada del diseño conceptual de un avión. Abstract An optimization method for conceptual design in Aeronautics is presented that is based on the use of surrogate models. The various ingredients in the target function are calculated for each individual using surrogates of the associated technical disciplines that are constructed via high order singular value decomposition and one dimensional interpolation. These surrogates result from a limited number of CFD calculated snapshots. The surrogates are combined with an optimization method, which can be either a global optimization method such as a genetic algorithm or a local optimization method, such as a gradient-like method. The resulting method is both flexible and much more computationally efficient than the conventional method based on direct calculation of the target function, especially if a large number of free design parameters and/or tunablemodeling parameters are present. The method is illustrated considering a simplified version of the conceptual design of an aircraft empennage.

Límites de ductilidad y de servicio al cálculo en carga última de estructuras de hormigón armado y de fábrica

En este trabajo se aborda una cuestión central en el diseño en carga última de estructuras de hormigón armado y de fábrica: la posibilidad efectiva de que las deformaciones plásticas necesarias para verificar un estado de rotura puedan ser alcanzadas por las regiones de la estructura que deban desarrollar su capacidad última para verificar tal estado. Así, se parte de las decisiones de diseño que mediante mera estática aseguran un equilibrio de la estructura para las cargas últimas que deba resistir, pero determinando directamente el valor de las deformaciones necesarias para llegar a tal estado. Por tanto, no se acude a los teoremas de rotura sin más, sino que se formula el problema desde un punto de vista elastoplástico. Es decir, no se obvia el recorrido que la estructura deba realizar en un proceso de carga incremental monótono, de modo que las regiones no plastificadas contribuyen a coaccionar las libres deformaciones plásticas que, en la teoría de rotura, se suponen. En términos de trabajo y energía, se introduce en el balance del trabajo de las fuerzas externas y en el de la energía de deformación, aquella parte del sistema que no ha plastificado. Establecido así el balance energético como potencial del sistema es cuando la condición de estacionariedad del mismo hace determinados los campos de desplazamientos y, por tanto, el de las deformaciones plásticas también. En definitiva, se trata de un modo de verificar si la ductilidad de los diseños previstos es suficiente, y en qué medida, para verificar el estado de rotura previsto, para unas determinadas cargas impuestas. Dentro del desarrollo teórico del problema, se encuentran ciertas precisiones importantes. Entre ellas, la verificación de que el estado de rotura a que se llega de manera determinada mediante el balance energético elasto-plástico satisface las condiciones de la solución de rotura que los teoremas de carga última predicen, asegurando, por tanto, que la solución determinada -unicidad del problema elásticocoincide con el teorema de unicidad de la carga de rotura, acotando además cuál es el sistema de equilibrio y cuál es la deformada de colapso, aspectos que los teoremas de rotura no pueden asegurar, sino sólo el valor de la carga última a verificar. Otra precisión se basa en la particularidad de los casos en que el sistema presenta una superficie de rotura plana, haciendo infinitas las posibilidades de equilibrio para una misma deformada de colapso determinada, lo que está en la base de, aparentemente, poder plastificar a antojo en vigas y arcos. Desde el planteamiento anterior, se encuentra entonces que existe una condición inherente a cualquier sistema, definidas unas leyes constitutivas internas, que permite al mismo llegar al inicio del estado de rotura sin demandar deformación plástica alguna, produciéndose la plastificación simultánea de todas las regiones que hayan llegado a su solicitación de rotura. En cierto modo, se daría un colapso de apariencia frágil. En tal caso, el sistema conserva plenamente hasta el final su capacidad dúctil y tal estado actúa como representante canónico de cualquier otra solución de equilibrio que con idéntico criterio de diseño interno se prevea para tal estructura. En la medida que el diseño se acerque o aleje de la solución canónica, la demanda de ductilidad del sistema para verificar la carga última será menor o mayor. Las soluciones que se aparten en exceso de la solución canónica, no verificarán el estado de rotura previsto por falta de ductilidad: la demanda de deformación plástica de alguna región plastificada estará más allá de la capacidad de la misma, revelándose una carga de rotura por falta de ductilidad menor que la que se preveía por mero equilibrio. Para la determinación de las deformaciones plásticas de las rótulas, se ha tomado un modelo formulado mediante el Método de los Elementos de Contorno, que proporciona un campo continuo de desplazamientos -y, por ende, de deformaciones y de tensiones- incluso en presencia de fisuras en el contorno. Importante cuestión es que se formula la diferencia, nada desdeñable, de la capacidad de rotación plástica de las secciones de hormigón armado en presencia de cortante y en su ausencia. Para las rótulas de fábrica, la diferencia se establece para las condiciones de la excentricidad -asociadas al valor relativo de la compresión-, donde las diferencias entres las regiones plastificadas con esfuerzo normal relativo alto o bajo son reseñables. Por otro lado, si bien de manera un tanto secundaria, las condiciones de servicio también imponen un límite al diseño previo en carga última deseado. La plastificación lleva asociadas deformaciones considerables, sean locales como globales. Tal cosa impone que, en estado de servicio, si la plastificación de alguna región lleva asociadas fisuraciones excesivas para el ambiente del entorno, la solución sea inviable por ello. Asimismo, las deformaciones de las estructuras suponen un límite severo a las posibilidades de su diseño. Especialmente en edificación, las deformaciones activas son un factor crítico a la hora de decidirse por una u otra solución. Por tanto, al límite que se impone por razón de ductilidad, se debe añadir el que se imponga por razón de las condiciones de servicio. Del modo anterior, considerando las condiciones de ductilidad y de servicio en cada caso, se puede tasar cada decisión de diseño con la previsión de cuáles serán las consecuencias en su estado de carga última y de servicio. Es decir, conocidos los límites, podemos acotar cuáles son los diseños a priori que podrán satisfacer seguro las condiciones de ductilidad y de servicio previstas, y en qué medida. Y, en caso de no poderse satisfacer, qué correcciones debieran realizarse sobre el diseño previo para poderlas cumplir. Por último, de las consecuencias que se extraen de lo estudiado, se proponen ciertas líneas de estudio y de experimentación para poder llegar a completar o expandir de manera práctica los resultados obtenidos. ABSTRACT This work deals with a main issue for the ultimate load design in reinforced concrete and masonry structures: the actual possibility that needed yield strains to reach a ultimate state could be reached by yielded regions on the structure that should develop their ultimate capacity to fulfill such a state. Thus, some statically determined design decisions are posed as a start for prescribed ultimate loads to be counteracted, but finding out the determined value of the strains needed to reach the ultimate load state. Therefore, ultimate load theorems are not taken as they are, but a full elasto-plastic formulation point of view is used. As a result, the path the structure must develop in a monotonus increasing loading procedure is not neglected, leading to the fact that non yielded regions will restrict the supposed totally free yield strains under a pure ultimate load theory. In work and energy terms, in the overall account of external forces work and internal strain energy, those domains in the body not reaching their ultimate state are considered. Once thus established the energy balance of the system as its potential, by imposing on it the stationary condition, both displacements and yield strains appear as determined values. Consequently, what proposed is a means for verifying whether the ductility of prescribed designs is enough and the extent to which they are so, for known imposed loads. On the way for the theoretical development of the proposal, some important aspects have been found. Among these, the verification that the conditions for the ultimate state reached under the elastoplastic energy balance fulfills the conditions prescribed for the ultimate load state predicted through the ultimate load theorems, assuring, therefore, that the determinate solution -unicity of the elastic problemcoincides with the unicity ultimate load theorem, determining as well which equilibrium system and which collapse shape are linked to it, being these two last aspects unaffordable by the ultimate load theorems, that make sure only which is the value of the ultimate load leading to collapse. Another aspect is based on the particular case in which the yield surface of the system is flat -i.e. expressed under a linear expression-, turning out infinite the equilibrium possibilities for one determined collapse shape, which is the basis of, apparently, deciding at own free will the yield distribution in beams and arches. From the foresaid approach, is then found that there is an inherent condition in any system, once defined internal constitutive laws, which allows it arrive at the beginning of the ultimate state or collapse without any yield strain demand, reaching the collapse simultaneously for all regions that have come to their ultimate strength. In a certain way, it would appear to be a fragile collapse. In such a case case, the system fully keeps until the end its ductility, and such a state acts as a canonical representative of any other statically determined solution having the same internal design criteria that could be posed for the that same structure. The extent to which a design is closer to or farther from the canonical solution, the ductility demand of the system to verify the ultimate load will be higher or lower. The solutions being far in excess from the canonical solution, will not verify the ultimate state due to lack of ductility: the demand for yield strains of any yielded region will be beyond its capacity, and a shortcoming ultimate load by lack of ductility will appear, lower than the expected by mere equilibrium. For determining the yield strains of plastic hinges, a Boundary Element Method based model has been used, leading to a continuous displacement field -therefore, for strains and stresses as well- even if cracks on the boundary are present. An important aspect is that a remarkable difference is found in the rotation capacity between plastic hinges in reinforced concrete with or without shear. For masonry hinges, such difference appears when dealing with the eccentricity of axial forces -related to their relative value of compression- on the section, where differences between yield regions under high or low relative compressions are remarkable. On the other hand, although in a certain secondary manner, serviceability conditions impose limits to the previous ultimate load stated wanted too. Yield means always big strains and deformations, locally and globally. Such a thing imposes, for serviceability states, that if a yielded region is associated with too large cracking for the environmental conditions, the predicted design will be unsuitable due to this. Furthermore, displacements must be restricted under certain severe limits that restrain the possibilities for a free design. Especially in building structures, active displacements are a critical factor when chosing one or another solution. Then, to the limits due to ductility reasons, other limits dealing with serviceability conditions shoud be added. In the foresaid way, both considering ductility and serviceability conditions in every case, the results for ultimate load and serviceability to which every design decision will lead can be bounded. This means that, once the limits are known, it is possible to bound which a priori designs will fulfill for sure the prescribed ductility and serviceability conditions, and the extent to wich they will be fulfilled, And, in case they were not, which corrections must be performed in the previous design so that it will. Finally, from the consequences derived through what studied, several study and experimental fields are proposed, in order to achieve a completeness and practical expansion of the obtained results.

Novel free-form optical surface design with spiral symmetry

Manufacturing technologies as injection molding or embossing specify their production limits for minimum radii of the vertices or draft angle for demolding, for instance. These restrictions may limit the system optical efficiency or affect the generation of undesired artifacts on the illumination pattern when dealing with optical design. A novel manufacturing concept is presented here, in which the optical surfaces are not obtained from the usual revolution symmetry with respect to a central axis (z axis), but they are calculated as free-form surfaces describing a spiral trajectory around z axis. The main advantage of this new concept lies in the manufacturing process: a molded piece can be easily separated from its mold just by applying a combination of rotational movement around axis z and linear movement along axis z, even for negative draft angles. The general designing procedure will be described in detail

Free-form Fresnel RXI-RR Köhler design for high-concentration photovoltaics with spectrum-splitting

Development of a novel HCPV nonimaging concentrator with high concentration (>500x) and built-in spectrum splitting concept is presented. It uses the combination of a commercial concentration GaInP/GaInAs/Ge 3J cell and a concentration Back-Point-Contact (BPC) silicon cell for efficient spectral utilization, and external confinement techniques for recovering the 3J cell's reflection. The primary optical element (POE) is a flat Fresnel lens and the secondary optical element (SOE) is a free-form RXI-type concentrator with a band-pass filter embedded in it - Both the POE and SOE performing Köhler integration to produce light homogenization on the receiver. The band-pass filter transmits the IR photons in the 900-1200 nm band to the silicon cell. A design target of an "equivalent" cell efficiency ~46% is predicted using commercial 39% 3J and 26% Si cells. A projected CPV module efficiency of greater than 38% is achievable at a concentration level larger than 500X with a wide acceptance angle of ±1°. A first proof-of concept receiver prototype has been manufactured using a simpler optical architecture (with a lower concentration, ~100x and lower simulated added efficiency), and experimental measurements have shown up to 39.8% 4J receiver efficiency using a 3J cell with a peak efficiency of 36.9%.

Analytic design method for optimal imaging: coupling three ray sets using two free-form lens profiles

In this work, a new two-dimensional optics design method is proposed that enables the coupling of three ray sets with two lens surfaces. The method is especially important for optical systems designed for wide field of view and with clearly separated optical surfaces. Fermat’s principle is used to deduce a set of functional differential equations fully describing the entire optical system. The presented general analytic solution makes it possible to calculate the lens profiles. Ray tracing results for calculated 15th order Taylor polynomials describing the lens profiles demonstrate excellent imaging performance and the versatility of this new analytic design method.

Analytic free-form lens design in 3D: coupling three ray sets using two lens surfaces

The two-dimensional analytic optics design method presented in a previous paper [Opt. Express 20, 5576–5585 (2012)] is extended in this work to the three-dimensional case, enabling the coupling of three ray sets with two free-form lens surfaces. Fermat’s principle is used to deduce additional sets of functional differential equations which make it possible to calculate the lens surfaces. Ray tracing simulations demonstrate the excellent imaging performance of the resulting free-form lenses described by more than 100 coefficients.

Efficient design and optimization of bio-photonic sensing cells (BICELLs) for label free biosensing

In previous works we demonstrated the benefits of using micro–nano patterning materials to be used as bio-photonic sensing cells (BICELLs), referred as micro–nano photonic structures having immobilized bioreceptors on its surface with the capability of recognizing the molecular binding by optical transduction. Gestrinone/anti-gestrinone and BSA/anti-BSA pairs were proven under different optical configurations to experimentally validate the biosensing capability of these bio-sensitive photonic architectures. Moreover, Three-Dimensional Finite Difference Time Domain (FDTD) models were employed for simulating the optical response of these structures. For this article, we have developed an effective analytical simulation methodology capable of simulating complex biophotonic sensing architectures. This simulation method has been tested and compared with previous experimental results and FDTD models. Moreover, this effective simulation methodology can be used for efficiently design and optimize any structure as BICELL. In particular for this article, six different BICELL's types have been optimized. To carry out this optimization we have considered three figures of merit: optical sensitivity, Q-factor and signal amplitude. The final objective of this paper is not only validating a suitable and efficient optical simulation methodology but also demonstrating the capability of this method for analyzing the performance of a given number of BICELLs for label-free biosensing.

Analytic free-form lens design for imaging applications with high aspect ratio

A new three-dimensional analytic optics design method is presented that enables the coupling of three ray sets with only two free-form lens surfaces. Closely related to the Simultaneous Multiple Surface method in three dimensions (SMS3D), it is derived directly from Fermat?s principle, leading to multiple sets of functional differential equations. The general solution of these equations makes it possible to calculate more than 80 coefficients for each implicit surface function. Ray tracing simulations of these free-form lenses demonstrate superior imaging performance for applications with high aspect ratio, compared to conventional rotational symmetric systems.

Analytic free-form lens design for tracking integration in concentrating photovoltaics

In this work the concept of tracking integration in concentrating photovoltaics (CPV) is revisited and developed further. With respect to conventional CPV, tracking integration eliminates the clear separation between stationary units of optics and solar cells, and external solar trackers. This approach is capable of further increasing the concentration ratio and makes high concentrating photovoltaics (> 500x) available for single-axis tracker installations. The reduced external solar tracking effort enables possibly cheaper and more compact installations. Our proposed optical system uses two laterally moving plano-convex lenses to achieve high concentration over a wide angular range of ±24°. The lateral movement allows to combine both steering and concentration of the incident direct sun light. Given the specific symmetry conditions of the underlying optical design problem, rotational symmetric lenses are not ideal for this application. For this type of design problems, a new free-form optics design method presented in previous papers perfectly matches the symmetry. It is derived directly from Fermat's principle, leading to sets of functional differential equations allowing the successive calculation of the Taylor series coeficients of each implicit surface function up to very high orders. For optical systems designed for wide field of view and with clearly separated optical surfaces, this new analytic design method has potential application in both fields of nonimaging and imaging optics.

Optimization of a label-free biosensor vertically characterized based on a periodic lattice of high aspect ratio SU-8 nano-pillars with a simplified 2D theoretical model

We have recently demonstrated a biosensor based on a lattice of SU8 pillars on a 1 μm SiO2/Si wafer by measuring vertically reflectivity as a function of wavelength. The biodetection has been proven with the combination of Bovine Serum Albumin (BSA) protein and its antibody (antiBSA). A BSA layer is attached to the pillars; the biorecognition of antiBSA involves a shift in the reflectivity curve, related with the concentration of antiBSA. A detection limit in the order of 2 ng/ml is achieved for a rhombic lattice of pillars with a lattice parameter (a) of 800 nm, a height (h) of 420 nm and a diameter(d) of 200 nm. These results correlate with calculations using 3D-finite difference time domain method. A 2D simplified model is proposed, consisting of a multilayer model where the pillars are turned into a 420 nm layer with an effective refractive index obtained by using Beam Propagation Method (BPM) algorithm. Results provided by this model are in good correlation with experimental data, reaching a reduction in time from one day to 15 minutes, giving a fast but accurate tool to optimize the design and maximizing sensitivity, and allows analyzing the influence of different variables (diameter, height and lattice parameter). Sensitivity is obtained for a variety of configurations, reaching a limit of detection under 1 ng/ml. Optimum design is not only chosen because of its sensitivity but also its feasibility, both from fabrication (limited by aspect ratio and proximity of the pillars) and fluidic point of view. (© 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

Freeform Fresnel RXI-RR Köhler design with spectrum-splitting for photovoltaics

The development of a novel optical design for the high concentration photovoltaics (HPCV) nonimaging concentrator (>500x) that utilizes a built-in spectrum splitting concept is presented. The primary optical element (POE) is a flat Fresnel lens and the secondary optical element (SOE) is a free-form RXI-type concentrator with a band-pass filter embedded in it. The POE and SOE perform Köhler integration to produce light homogenization on the receiver. The system uses a combination of a commercial concentration GaInP/GaInAs/Ge 3J cell and a concentration Back-PointContact (BPC) silicon cell for efficient spectral utilization, and an external confinement technique for recovering the 3J cell’s reflection. A design target of an “equivalent” cell efficiency ~46% is predicted using commercial 39% 3J and 26% Si cells. A projected CPV module efficiency of greater than 38% is achievable at a concentration level greater than 500X with a wide acceptance angle of ±1º. A first proof-of concept receiver prototype has been manufactured using a simpler optical architecture (with a lower concentration, ~100x and lower simulated added efficiency), and experimental measurements have shown up to 39.8% 4J receiver efficiency using a 3J cell with a peak efficiency of 36.9%

Free-Form, High Energy Performance, Transparent Envelope

This article examines, from the energy viewpoint, a new lightweight, slim, high energy efficient, light-transmitting envelope system, providing for seamless, free-form designs for use in architectural projects. The research was based on envelope components already existing on the market, especially components implemented with granular silica gel insulation, as this is the most effective translucent thermal insulation there is today. The tests run on these materials revealed that there is not one that has all the features required of the new envelope model, although some do have properties that could be exploited to generate this envelope, namely, the vacuum chamber of vacuum insulated panels (VIP), the monolithic aerogel used as insulation in some prototypes, reinforced polyester barriers. By combining these three design components — the high-performance thermal insulation of the vacuum chamber combined with monolithic silica gel insulation, the free-form design potential provided by materials like reinforced polyester and epoxy resins—, we have been able to define and test a new, variable geometry, energy-saving envelope system.

Advances in the Simultaneous Multiple Surface optical design method for imaging and non-imaging applications

Classical imaging optics has been developed over centuries in many areas, such as its paraxial imaging theory and practical design methods like multi-parametric optimization techniques. Although these imaging optical design methods can provide elegant solutions to many traditional optical problems, there are more and more new design problems, like solar concentrator, illumination system, ultra-compact camera, etc., that require maximum energy transfer efficiency, or ultra-compact optical structure. These problems do not have simple solutions from classical imaging design methods, because not only paraxial rays, but also non-paraxial rays should be well considered in the design process. Non-imaging optics is a newly developed optical discipline, which does not aim to form images, but to maximize energy transfer efficiency. One important concept developed from non-imaging optics is the “edge-ray principle”, which states that the energy flow contained in a bundle of rays will be transferred to the target, if all its edge rays are transferred to the target. Based on that concept, many CPC solar concentrators have been developed with efficiency close to the thermodynamic limit. When more than one bundle of edge-rays needs to be considered in the design, one way to obtain solutions is to use SMS method. SMS stands for Simultaneous Multiple Surface, which means several optical surfaces are constructed simultaneously. The SMS method was developed as a design method in Non-imaging optics during the 90s. The method can be considered as an extension to the Cartesian Oval calculation. In the traditional Cartesian Oval calculation, one optical surface is built to transform an input wave-front to an out-put wave-front. The SMS method however, is dedicated to solve more than 1 wave-fronts transformation problem. In the beginning, only 2 input wave-fronts and 2 output wave-fronts transformation problem was considered in the SMS design process for rotational optical systems or free-form optical systems. Usually “SMS 2D” method stands for the SMS procedure developed for rotational optical system, and “SMS 3D” method for the procedure for free-form optical system. Although the SMS method was originally employed in non-imaging optical system designs, it has been found during this thesis that with the improved capability to design more surfaces and control more input and output wave-fronts, the SMS method can also be applied to imaging system designs and possesses great advantage over traditional design methods. In this thesis, one of the main goals to achieve is to further develop the existing SMS-2D method to design with more surfaces and improve the stability of the SMS-2D and SMS-3D algorithms, so that further optimization process can be combined with SMS algorithms. The benefits of SMS plus optimization strategy over traditional optimization strategy will be explained in details for both rotational and free-form imaging optical system designs. Another main goal is to develop novel design concepts and methods suitable for challenging non-imaging applications, e.g. solar concentrator and solar tracker. This thesis comprises 9 chapters and can be grouped into two parts: the first part (chapter 2-5) contains research works in the imaging field, and the second part (chapter 6-8) contains works in the non-imaging field. In the first chapter, an introduction to basic imaging and non-imaging design concepts and theories is given. Chapter 2 presents a basic SMS-2D imaging design procedure using meridian rays. In this chapter, we will set the imaging design problem from the SMS point of view, and try to solve the problem numerically. The stability of this SMS-2D design procedure will also be discussed. The design concepts and procedures developed in this chapter lay the path for further improvement. Chapter 3 presents two improved SMS 3 surfaces’ design procedures using meridian rays (SMS-3M) and skew rays (SMS-1M2S) respectively. The major improvement has been made to the central segments selections, so that the whole SMS procedures become more stable compared to procedures described in Chapter 2. Since these two algorithms represent two types of phase space sampling, their image forming capabilities are compared in a simple objective design. Chapter 4 deals with an ultra-compact SWIR camera design with the SMS-3M method. The difficulties in this wide band camera design is how to maintain high image quality meanwhile reduce the overall system length. This interesting camera design provides a playground for the classical design method and SMS design methods. We will show designs and optical performance from both classical design method and the SMS design method. Tolerance study is also given as the end of the chapter. Chapter 5 develops a two-stage SMS-3D based optimization strategy for a 2 freeform mirrors imaging system. In the first optimization phase, the SMS-3D method is integrated into the optimization process to construct the two mirrors in an accurate way, drastically reducing the unknown parameters to only few system configuration parameters. In the second optimization phase, previous optimized mirrors are parameterized into Qbfs type polynomials and set up in code V. Code V optimization results demonstrates the effectiveness of this design strategy in this 2-mirror system design. Chapter 6 shows an etendue-squeezing condenser optics, which were prepared for the 2010 IODC illumination contest. This interesting design employs many non-imaging techniques such as the SMS method, etendue-squeezing tessellation, and groove surface design. This device has theoretical efficiency limit as high as 91.9%. Chapter 7 presents a freeform mirror-type solar concentrator with uniform irradiance on the solar cell. Traditional parabolic mirror concentrator has many drawbacks like hot-pot irradiance on the center of the cell, insufficient use of active cell area due to its rotational irradiance pattern and small acceptance angle. In order to conquer these limitations, a novel irradiance homogenization concept is developed, which lead to a free-form mirror design. Simulation results show that the free-form mirror reflector has rectangular irradiance pattern, uniform irradiance distribution and large acceptance angle, which confirm the viability of the design concept. Chapter 8 presents a novel beam-steering array optics design strategy. The goal of the design is to track large angle parallel rays by only moving optical arrays laterally, and convert it to small angle parallel output rays. The design concept is developed as an extended SMS method. Potential applications of this beam-steering device are: skylights to provide steerable natural illumination, building integrated CPV systems, and steerable LED illumination. Conclusion and future lines of work are given in Chapter 9. Resumen La óptica de formación de imagen clásica se ha ido desarrollando durante siglos, dando lugar tanto a la teoría de óptica paraxial y los métodos de diseño prácticos como a técnicas de optimización multiparamétricas. Aunque estos métodos de diseño óptico para formación de imagen puede aportar soluciones elegantes a muchos problemas convencionales, siguen apareciendo nuevos problemas de diseño óptico, concentradores solares, sistemas de iluminación, cámaras ultracompactas, etc. que requieren máxima transferencia de energía o dimensiones ultracompactas. Este tipo de problemas no se pueden resolver fácilmente con métodos clásicos de diseño porque durante el proceso de diseño no solamente se deben considerar los rayos paraxiales sino también los rayos no paraxiales. La óptica anidólica o no formadora de imagen es una disciplina que ha evolucionado en gran medida recientemente. Su objetivo no es formar imagen, es maximazar la eficiencia de transferencia de energía. Un concepto importante de la óptica anidólica son los “rayos marginales”, que se pueden utilizar para el diseño de sistemas ya que si todos los rayos marginales llegan a nuestra área del receptor, todos los rayos interiores también llegarán al receptor. Haciendo uso de este principio, se han diseñado muchos concentradores solares que funcionan cerca del límite teórico que marca la termodinámica. Cuando consideramos más de un haz de rayos marginales en nuestro diseño, una posible solución es usar el método SMS (Simultaneous Multiple Surface), el cuál diseña simultáneamente varias superficies ópticas. El SMS nació como un método de diseño para óptica anidólica durante los años 90. El método puede ser considerado como una extensión del cálculo del óvalo cartesiano. En el método del óvalo cartesiano convencional, se calcula una superficie para transformar un frente de onda entrante a otro frente de onda saliente. El método SMS permite transformar varios frentes de onda de entrada en frentes de onda de salida. Inicialmente, sólo era posible transformar dos frentes de onda con dos superficies con simetría de rotación y sin simetría de rotación, pero esta limitación ha sido superada recientemente. Nos referimos a “SMS 2D” como el método orientado a construir superficies con simetría de rotación y llamamos “SMS 3D” al método para construir superficies sin simetría de rotación o free-form. Aunque el método originalmente fue aplicado en el diseño de sistemas anidólicos, se ha observado que gracias a su capacidad para diseñar más superficies y controlar más frentes de onda de entrada y de salida, el SMS también es posible aplicarlo a sistemas de formación de imagen proporcionando una gran ventaja sobre los métodos de diseño tradicionales. Uno de los principales objetivos de la presente tesis es extender el método SMS-2D para permitir el diseño de sistemas con mayor número de superficies y mejorar la estabilidad de los algoritmos del SMS-2D y SMS-3D, haciendo posible combinar la optimización con los algoritmos. Los beneficios de combinar SMS y optimización comparado con el proceso de optimización tradicional se explican en detalle para sistemas con simetría de rotación y sin simetría de rotación. Otro objetivo importante de la tesis es el desarrollo de nuevos conceptos de diseño y nuevos métodos en el área de la concentración solar fotovoltaica. La tesis está estructurada en 9 capítulos que están agrupados en dos partes: la primera de ellas (capítulos 2-5) se centra en la óptica formadora de imagen mientras que en la segunda parte (capítulos 6-8) se presenta el trabajo del área de la óptica anidólica. El primer capítulo consta de una breve introducción de los conceptos básicos de la óptica anidólica y la óptica en formación de imagen. El capítulo 2 describe un proceso de diseño SMS-2D sencillo basado en los rayos meridianos. En este capítulo se presenta el problema de diseñar un sistema formador de imagen desde el punto de vista del SMS y se intenta obtener una solución de manera numérica. La estabilidad de este proceso se analiza con detalle. Los conceptos de diseño y los algoritmos desarrollados en este capítulo sientan la base sobre la cual se realizarán mejoras. El capítulo 3 presenta dos procedimientos para el diseño de un sistema con 3 superficies SMS, el primero basado en rayos meridianos (SMS-3M) y el segundo basado en rayos oblicuos (SMS-1M2S). La mejora más destacable recae en la selección de los segmentos centrales, que hacen más estable todo el proceso de diseño comparado con el presentado en el capítulo 2. Estos dos algoritmos representan dos tipos de muestreo del espacio de fases, su capacidad para formar imagen se compara diseñando un objetivo simple con cada uno de ellos. En el capítulo 4 se presenta un diseño ultra-compacto de una cámara SWIR diseñada usando el método SMS-3M. La dificultad del diseño de esta cámara de espectro ancho radica en mantener una alta calidad de imagen y al mismo tiempo reducir drásticamente sus dimensiones. Esta cámara es muy interesante para comparar el método de diseño clásico y el método de SMS. En este capítulo se presentan ambos diseños y se analizan sus características ópticas. En el capítulo 5 se describe la estrategia de optimización basada en el método SMS-3D. El método SMS-3D calcula las superficies ópticas de manera precisa, dejando sólo unos pocos parámetros libres para decidir la configuración del sistema. Modificando el valor de estos parámetros se genera cada vez mediante SMS-3D un sistema completo diferente. La optimización se lleva a cabo variando los mencionados parámetros y analizando el sistema generado. Los resultados muestran que esta estrategia de diseño es muy eficaz y eficiente para un sistema formado por dos espejos. En el capítulo 6 se describe un sistema de compresión de la Etendue, que fue presentado en el concurso de iluminación del IODC en 2010. Este interesante diseño hace uso de técnicas propias de la óptica anidólica, como el método SMS, el teselado de las lentes y el diseño mediante grooves. Este dispositivo tiene un límite teórica en la eficiencia del 91.9%. El capítulo 7 presenta un concentrador solar basado en un espejo free-form con irradiancia uniforme sobre la célula. Los concentradores parabólicos tienen numerosas desventajas como los puntos calientes en la zona central de la célula, uso no eficiente del área de la célula al ser ésta cuadrada y además tienen ángulos de aceptancia de reducido. Para poder superar estas limitaciones se propone un novedoso concepto de homogeneización de la irrandancia que se materializa en un diseño con espejo free-form. El análisis mediante simulación demuestra que la irradiancia es homogénea en una región rectangular y con mayor ángulo de aceptancia, lo que confirma la viabilidad del concepto de diseño. En el capítulo 8 se presenta un novedoso concepto para el diseño de sistemas afocales dinámicos. El objetivo del diseño es realizar un sistema cuyo haz de rayos de entrada pueda llegar con ángulos entre ±45º mientras que el haz de rayos a la salida sea siempre perpendicular al sistema, variando únicamente la posición de los elementos ópticos lateralmente. Las aplicaciones potenciales de este dispositivo son varias: tragaluces que proporcionan iluminación natural, sistemas de concentración fotovoltaica integrados en los edificios o iluminación direccionable con LEDs. Finalmente, el último capítulo contiene las conclusiones y las líneas de investigación futura.

Conceptual design of the liquid metal laboratory of the TECHNOFUSION facility

The application of liquid metal technology in fusion devices requires R&D related to many phenomena: interaction between liquid metals and structural material as corrosion, erosion and passivation techniques; magneto-hydrodynamics; free surface fluid-dynamics and any other physical aspect that will be needed for their safe reliable operation. In particular, there is a significant shortage of experimental facilities dedicated to the development of the lithium technology. In the framework of the TECHNOFUSION project, an experimental laboratory devoted to the lithium technology development is proposed, in order to shed some light in the path to IFMIF and the design of chamber's first wall and divertors. The conceptual design foresee a development in two stages, the first one consisting on a material testing loop. The second stage proposes the construction of a mock-up of the IFMIF target that will allow to assess the behaviour of a free-surface lithium target under vacuum conditions. In this paper, such conceptual design is addressed.

User-centric need-driven affect modeling for spoken conversational agents: design and evaluation

It is easy to get frustrated at spoken conversational agents (SCAs), perhaps because they seem to be callous. By and large, the quality of human-computer interaction is affected due to the inability of the SCAs to recognise and adapt to user emotional state. Now with the mass appeal of artificially-mediated communication, there has been an increasing need for SCAs to be socially and emotionally intelligent, that is, to infer and adapt to their human interlocutors’ emotions on the fly, in order to ascertain an affective, empathetic and naturalistic interaction. An enhanced quality of interaction would reduce users’ frustrations and consequently increase their satisfactions. These reasons have motivated the development of SCAs towards including socio-emotional elements, turning them into affective and socially-sensitive interfaces. One barrier to the creation of such interfaces has been the lack of methods for modelling emotions in a task-independent environment. Most emotion models for spoken dialog systems are task-dependent and thus cannot be used “as-is” in different applications. This Thesis focuses on improving this, in which it concerns computational modeling of emotion, personality and their interrelationship for task-independent autonomous SCAs. The generation of emotion is driven by needs, inspired by human’s motivational systems. The work in this Thesis is organised in three stages, each one with its own contribution. The first stage involved defining, integrating and quantifying the psychological-based motivational and emotional models sourced from. Later these were transformed into a computational model by implementing them into software entities. The computational model was then incorporated and put to test with an existing SCA host, a HiFi-control agent. The second stage concerned automatic prediction of affect, which has been the main challenge towards the greater aim of infusing social intelligence into the HiFi agent. In recent years, studies on affect detection from voice have moved on to using realistic, non-acted data, which is subtler. However, it is more challenging to perceive subtler emotions and this is demonstrated in tasks such as labelling and machine prediction. In this stage, we attempted to address part of this challenge by considering the roles of user satisfaction ratings and conversational/dialog features as the respective target and predictors in discriminating contentment and frustration, two types of emotions that are known to be prevalent within spoken human-computer interaction. The final stage concerned the evaluation of the emotional model through the HiFi agent. A series of user studies with 70 subjects were conducted in a real-time environment, each in a different phase and with its own conditions. All the studies involved the comparisons between the baseline non-modified and the modified agent. The findings have gone some way towards enhancing our understanding of the utility of emotion in spoken dialog systems in several ways; first, an SCA should not express its emotions blindly, albeit positive. Rather, it should adapt its emotions to user states. Second, low performance in an SCA may be compensated by the exploitation of emotion. Third, the expression of emotion through the exploitation of prosody could better improve users’ perceptions of an SCA compared to exploiting emotions through just lexical contents. Taken together, these findings not only support the success of the emotional model, but also provide substantial evidences with respect to the benefits of adding emotion in an SCA, especially in mitigating users’ frustrations and ultimately improving their satisfactions. Resumen Es relativamente fácil experimentar cierta frustración al interaccionar con agentes conversacionales (Spoken Conversational Agents, SCA), a menudo porque parecen ser un poco insensibles. En general, la calidad de la interacción persona-agente se ve en cierto modo afectada por la incapacidad de los SCAs para identificar y adaptarse al estado emocional de sus usuarios. Actualmente, y debido al creciente atractivo e interés de dichos agentes, surge la necesidad de hacer de los SCAs unos seres cada vez más sociales y emocionalmente inteligentes, es decir, con capacidad para inferir y adaptarse a las emociones de sus interlocutores humanos sobre la marcha, de modo que la interacción resulte más afectiva, empática y, en definitiva, natural. Una interacción mejorada en este sentido permitiría reducir la posible frustración de los usuarios y, en consecuencia, mejorar el nivel de satisfacción alcanzado por los mismos. Estos argumentos justifican y motivan el desarrollo de nuevos SCAs con capacidades socio-emocionales, dotados de interfaces afectivas y socialmente sensibles. Una de las barreras para la creación de tales interfaces ha sido la falta de métodos de modelado de emociones en entornos independientes de tarea. La mayoría de los modelos emocionales empleados por los sistemas de diálogo hablado actuales son dependientes de tarea y, por tanto, no pueden utilizarse "tal cual" en diferentes dominios o aplicaciones. Esta tesis se centra precisamente en la mejora de este aspecto, la definición de modelos computacionales de las emociones, la personalidad y su interrelación para SCAs autónomos e independientes de tarea. Inspirada en los sistemas motivacionales humanos en el ámbito de la psicología, la tesis propone un modelo de generación/producción de la emoción basado en necesidades. El trabajo realizado en la presente tesis está organizado en tres etapas diferenciadas, cada una con su propia contribución. La primera etapa incluyó la definición, integración y cuantificación de los modelos motivacionales de partida y de los modelos emocionales derivados a partir de éstos. Posteriormente, dichos modelos emocionales fueron plasmados en un modelo computacional mediante su implementación software. Este modelo computacional fue incorporado y probado en un SCA anfitrión ya existente, un agente con capacidad para controlar un equipo HiFi, de alta fidelidad. La segunda etapa se orientó hacia el reconocimiento automático de la emoción, aspecto que ha constituido el principal desafío en relación al objetivo mayor de infundir inteligencia social en el agente HiFi. En los últimos años, los estudios sobre reconocimiento de emociones a partir de la voz han pasado de emplear datos actuados a usar datos reales en los que la presencia u observación de emociones se produce de una manera mucho más sutil. El reconocimiento de emociones bajo estas condiciones resulta mucho más complicado y esta dificultad se pone de manifiesto en tareas tales como el etiquetado y el aprendizaje automático. En esta etapa, se abordó el problema del reconocimiento de las emociones del usuario a partir de características o métricas derivadas del propio diálogo usuario-agente. Gracias a dichas métricas, empleadas como predictores o indicadores del grado o nivel de satisfacción alcanzado por el usuario, fue posible discriminar entre satisfacción y frustración, las dos emociones prevalentes durante la interacción usuario-agente. La etapa final corresponde fundamentalmente a la evaluación del modelo emocional por medio del agente Hifi. Con ese propósito se llevó a cabo una serie de estudios con usuarios reales, 70 sujetos, interaccionando con diferentes versiones del agente Hifi en tiempo real, cada uno en una fase diferente y con sus propias características o capacidades emocionales. En particular, todos los estudios realizados han profundizado en la comparación entre una versión de referencia del agente no dotada de ningún comportamiento o característica emocional, y una versión del agente modificada convenientemente con el modelo emocional propuesto. Los resultados obtenidos nos han permitido comprender y valorar mejor la utilidad de las emociones en los sistemas de diálogo hablado. Dicha utilidad depende de varios aspectos. En primer lugar, un SCA no debe expresar sus emociones a ciegas o arbitrariamente, incluso aunque éstas sean positivas. Más bien, debe adaptar sus emociones a los diferentes estados de los usuarios. En segundo lugar, un funcionamiento relativamente pobre por parte de un SCA podría compensarse, en cierto modo, dotando al SCA de comportamiento y capacidades emocionales. En tercer lugar, aprovechar la prosodia como vehículo para expresar las emociones, de manera complementaria al empleo de mensajes con un contenido emocional específico tanto desde el punto de vista léxico como semántico, ayuda a mejorar la percepción por parte de los usuarios de un SCA. Tomados en conjunto, los resultados alcanzados no sólo confirman el éxito del modelo emocional, sino xv que constituyen además una evidencia decisiva con respecto a los beneficios de incorporar emociones en un SCA, especialmente en cuanto a reducir el nivel de frustración de los usuarios y, en última instancia, mejorar su satisfacción.