Dynamic power management: from portable devices to high performance computing

Electronic applications are nowadays converging under the umbrella of the cloud computing vision. The future ecosystem of information and communication technology is going to integrate clouds of portable clients and embedded devices exchanging information, through the internet layer, with processing clusters of servers, data-centers and high performance computing systems. Even thus the whole society is waiting to embrace this revolution, there is a backside of the story. Portable devices require battery to work far from the power plugs and their storage capacity does not scale as the increasing power requirement does. At the other end processing clusters, such as data-centers and server farms, are build upon the integration of thousands multiprocessors. For each of them during the last decade the technology scaling has produced a dramatic increase in power density with significant spatial and temporal variability. This leads to power and temperature hot-spots, which may cause non-uniform ageing and accelerated chip failure. Nonetheless all the heat removed from the silicon translates in high cooling costs. Moreover trend in ICT carbon footprint shows that run-time power consumption of the all spectrum of devices accounts for a significant slice of entire world carbon emissions. This thesis work embrace the full ICT ecosystem and dynamic power consumption concerns by describing a set of new and promising system levels resource management techniques to reduce the power consumption and related issues for two corner cases: Mobile Devices and High Performance Computing.

Ab-initio determination of x-ray absorption near edge structure (xanes) spectra in an ultrasoft and norm conserving pseudopotentials scheme

X-ray absorption spectroscopy (XAS) is a powerful means of investigation of structural and electronic properties in condensed -matter physics. Analysis of the near edge part of the XAS spectrum, the so – called X-ray Absorption Near Edge Structure (XANES), can typically provide the following information on the photoexcited atom: - Oxidation state and coordination environment. - Speciation of transition metal compounds. - Conduction band DOS projected on the excited atomic species (PDOS). Analysis of XANES spectra is greatly aided by simulations; in the most common scheme the multiple scattering framework is used with the muffin tin approximation for the scattering potential and the spectral simulation is based on a hypothetical, reference structure. This approach has the advantage of requiring relatively little computing power but in many cases the assumed structure is quite different from the actual system measured and the muffin tin approximation is not adequate for low symmetry structures or highly directional bonds. It is therefore very interesting and justified to develop alternative methods. In one approach, the spectral simulation is based on atomic coordinates obtained from a DFT (Density Functional Theory) optimized structure. In another approach, which is the object of this thesis, the XANES spectrum is calculated directly based on an ab – initio DFT calculation of the atomic and electronic structure. This method takes full advantage of the real many-electron final wavefunction that can be computed with DFT algorithms that include a core-hole in the absorbing atom to compute the final cross section. To calculate the many-electron final wavefunction the Projector Augmented Wave method (PAW) is used. In this scheme, the absorption cross section is written in function of several contributions as the many-electrons function of the finale state; it is calculated starting from pseudo-wavefunction and performing a reconstruction of the real-wavefunction by using a transform operator which contains some parameters, called partial waves and projector waves. The aim of my thesis is to apply and test the PAW methodology to the calculation of the XANES cross section. I have focused on iron and silicon structures and on some biological molecules target (myoglobin and cytochrome c). Finally other inorganic and biological systems could be taken into account for future applications of this methodology, which could become an important improvement with respect to the multiscattering approach.

Nonlinear Characterization and Modelling of GaN HEMTs for Microwave Power Amplifier Applications

Semiconductors technologies are rapidly evolving driven by the need for higher performance demanded by applications. Thanks to the numerous advantages that it offers, gallium nitride (GaN) is quickly becoming the technology of reference in the field of power amplification at high frequency. The RF power density of AlGaN/GaN HEMTs (High Electron Mobility Transistor) is an order of magnitude higher than the one of gallium arsenide (GaAs) transistors. The first demonstration of GaN devices dates back only to 1993. Although over the past few years some commercial products have started to be available, the development of a new technology is a long process. The technology of AlGaN/GaN HEMT is not yet fully mature, some issues related to dispersive phenomena and also to reliability are still present. Dispersive phenomena, also referred as long-term memory effects, have a detrimental impact on RF performances and are due both to the presence of traps in the device structure and to self-heating effects. A better understanding of these problems is needed to further improve the obtainable performances. Moreover, new models of devices that take into consideration these effects are necessary for accurate circuit designs. New characterization techniques are thus needed both to gain insight into these problems and improve the technology and to develop more accurate device models. This thesis presents the research conducted on the development of new charac- terization and modelling methodologies for GaN-based devices and on the use of this technology for high frequency power amplifier applications.

Developmental changes of BOLD signal correlations with global human EEG power and synchronization during working memory

In humans, theta band (5-7 Hz) power typically increases when performing cognitively demanding working memory (WM) tasks, and simultaneous EEG-fMRI recordings have revealed an inverse relationship between theta power and the BOLD (blood oxygen level dependent) signal in the default mode network during WM. However, synchronization also plays a fundamental role in cognitive processing, and the level of theta and higher frequency band synchronization is modulated during WM. Yet, little is known about the link between BOLD, EEG power, and EEG synchronization during WM, and how these measures develop with human brain maturation or relate to behavioral changes. We examined EEG-BOLD signal correlations from 18 young adults and 15 school-aged children for age-dependent effects during a load-modulated Sternberg WM task. Frontal load (in-)dependent EEG theta power was significantly enhanced in children compared to adults, while adults showed stronger fMRI load effects. Children demonstrated a stronger negative correlation between global theta power and the BOLD signal in the default mode network relative to adults. Therefore, we conclude that theta power mediates the suppression of a task-irrelevant network. We further conclude that children suppress this network even more than adults, probably from an increased level of task-preparedness to compensate for not fully mature cognitive functions, reflected in lower response accuracy and increased reaction time. In contrast to power, correlations between instantaneous theta global field synchronization and the BOLD signal were exclusively positive in both age groups but only significant in adults in the frontal-parietal and posterior cingulate cortices. Furthermore, theta synchronization was weaker in children and was--in contrast to EEG power--positively correlated with response accuracy in both age groups. In summary we conclude that theta EEG-BOLD signal correlations differ between spectral power and synchronization and that these opposite correlations with different distributions undergo similar and significant neuronal developments with brain maturation.

Analysis of nonlinear gain-induced effects on short-pulse amplification in doped fibers by use of an extended power equation

Optical pulse amplification in doped fibers is studied using an extended power transport equation for the coupled pulse spectral components. This equation includes the effects of gain saturation, gain dispersion, fiber dispersion, fiber nonlinearity, and amplified spontaneous emission. The new model is employed to study nonlinear gain-induced effects on the spectrotemporal characteristics of amplified subpicosecond pulses, in both the anomalous and the normal dispersion regimes.

Wind plant interaction with series-compensated power systems

Wind power based generation has been rapidly growing world-wide during the recent past. In order to transmit large amounts of wind power over long distances, system planners may often add series compensation to existing transmission lines owing to several benefits such as improved steady-state power transfer limit, improved transient stability, and efficient utilization of transmission infrastructure. Application of series capacitors has posed resonant interaction concerns such as through subsynchronous resonance (SSR) with conventional turbine-generators. Wind turbine-generators may also be susceptible to such resonant interactions. However, not much information is available in literature and even engineering standards are yet to address these issues. The motivation problem for this research is based on an actual system switching event that resulted in undamped oscillations in a 345-kV series-compensated, typical ring-bus power system configuration. Based on time-domain ATP (Alternative Transients Program) modeling, simulations and analysis of system event records, the occurrence of subsynchronous interactions within the existing 345-kV series-compensated power system has been investigated. Effects of various small-signal and large-signal power system disturbances with both identical and non-identical wind turbine parameters (such as with a statistical-spread) has been evaluated. Effect of parameter variations on subsynchronous oscillations has been quantified using 3D-DFT plots and the oscillations have been identified as due to electrical self-excitation effects, rather than torsional interaction. Further, the generator no-load reactance and the rotor-side converter inner-loop controller gains have been identified as bearing maximum sensitivity to either damping or exacerbating the self-excited oscillations. A higher-order spectral analysis method based on modified Prony estimation has been successfully applied to the field records identifying dominant 9.79 Hz subsynchronous oscillations. Recommendations have been made for exploring countermeasures.

Stable oxygen isotope record and lithogenic composition of ODP Hole 117-722B

The upper 38 m of Hole 722B sediments (Owen Ridge, northwest Arabian Sea) was sampled at 20 cm intervals and used to develop records of lithogenic percent, mass accumulation rate, and grain size spanning the past 1 m.y. Over this interval, the lithogenic component of Owen Ridge sediments can be used to infer variability in the strength of Arabian Sea summer monsoon winds (median grain size) and the aridity of surrounding dust source-areas (mass accumulation rate; MAR in g/cm**2/k.y). The lithogenic MAR has strong 100, 41, and 23 k.y. cyclicities and is forced primarily by changes in source-area aridity associated with glacial-interglacial cycles. The lithogenic grain size, on the other hand, exhibits higher frequency variability (23 k.y.) and is forced by the strength of summer monsoon winds which, in turn, are forced by the effective sensible heating of the Indian-Asian landmass and by the availability of latent heat from the Southern Hemisphere Indian Ocean. These forcing mechanisms combine to produce a wind-strength record which has no strong relationship to glacial-interglacial cycles. Discussion of the mechanisms responsible for production of primary Milankovitch cyclicities in lithogenic records from the Owen Ridge is presented elsewhere (Clemens and Prell, 1990, doi:10.1029/PA005i002p00109). Here we examine the 1 m.y. record from Hole 722B focusing on different aspects of the lithogenic components including an abrupt change in the monsoon wind-strength record at 500 k.y., core-to-core reproducibility, comparison with magnetic susceptibility, coherency with a wind-strength record from the Pacific Ocean, and combination frequencies in the wind-strength record. The Hole 722B lithogenic grain-size record shows an abrupt change at 500 k.y. possibly indicating decreased monsoon wind-strength over the interval from 500 k.y. to present. The grain-size decrease appears to be coincident with a loss of spectral power near the 41 k.y. periodicity. However, the grain-size decrease is not paralleled in the Globigerina bulloides upwelling record, an independent record of summer monsoon wind-strength (Prell, this volume). These observations leave us with competing hypotheses possibly involving: (1) a decrease in the sensitivity of monsoon windstrength to obliquity forcing, (2) decoupling of the grain size and G. bulloides records via a decoupling of the nutrient supply from wind-driven upwelling, and/or (3) a change in dust source-area or the patterns of dust transporting winds. Comparison of the lithogenic grain size and weight percent records from Hole 722B with those from a nearby core shows that the major and most minor events are well replicated. These close matches establish our confidence in the lithogenic extraction techniques and measurements. Further, reproducibility on a core-to-core scale indicates that the eolian depositional signal is regionally strong, coherent, and well preserved. The lithogenic weight percent and magnetic susceptibility are extremely well correlated in both the time and frequency domains. From this we infer that the magnetically susceptible component of Owen Ridge sediments is of terrestrial origin and transported to the Owen Ridge via summer monsoon winds. Because of the high correlation with the lithogenic percent record, the magnetic susceptibility record can be cast in terms of lithogenic MAR and used as a high resolution proxy for continental aridity. In addition to primary Milankovitch periodicities, the Hole 722B grain-size record exhibits periodicity at 52 k.y. and at 29 k.y. Both periodicities are also found in the grain-size record from piston core RC11-210 in the equatorial Pacific Ocean. Comparison of the two grain-size records shows significant coherence and zero phase relationships over both the 52 and 29 k.y. periodicities suggesting that the strengths of the Indian Ocean monsoon and the Pacific southeasterly trade winds share common forcing mechanisms. Two possible origins for the 52 and 29 k.y. periodicities in the Hole 722B wind-strength record are (1) direct Milankovitch forcing (54 and 29 k.y. components of obliquity) and (2) combination periodicities resulting from nonlinear interactions within the climate system. We find that the 52 and 29 k.y. periodicities show stronger coherency with crossproducts of eccentricity and obliquity (29 k.y.) and precession and obliquity (52 k.y.) than with direct obliquity forcing. Our working hypothesis attributes these periodicities to nonlinear interaction between external insolation forcing and internal climatic feedback mechanisms involving an interdependence of continental snow/ice-mass (albedo) and the hydrological cycle (latent heat availability).

Sedimentation rate, bulk density, accumulation rate of terrigenous material and grain sizes of ODP Leg 167 sites

The terrigenous mineral fraction of sediments recovered by drilling during Ocean Drilling Program Leg 167 at Sites 1018 and 1020 is used to evaluate changes in the source and transport of fine-grained terrigenous sediment and its relation to regional climates and the paleoceanographic evolution of the California Current system during the late Pleistocene. Preliminary time scales developed by correlation of oxygen isotope stratigraphies with the global SPECMAP record show average linear sedimentation rates in excess of 100 m/m.y., which provide an opportunity for high-resolution studies of terrigenous flux, grain size, and mineralogy. The mass flux of terrigenous minerals at Site 1018 varies from 5 to 30 g/(cm**2 x k.y.) and displays a general trend toward increased flux during glacials. The terrigenous record at Site 1020 shows a similar pattern of increased glacial input, but overall accumulation rates are significantly lower. Spectral analysis demonstrates that most of this variability is concentrated in frequency bands related to orbital cycles of eccentricity, tilt, and precession. Detailed grain-size analysis performed on the isolated terrigenous mineral fraction shows that sediments from Site 1018 are associated with higher energy transport and depositional regimes than those found at Site 1020. Grain-size data are remarkably uniform throughout the last 500 k.y., with no discernible difference observed between glacial and interglacial size distributions within each site. X-ray diffraction analysis of the <2-µm clay component suggests that the deposition of minerals found at Site 1020 is consistent with transport from a southern source during intervals of increased terrigenous input.

Physical properties and sedimentology on core GeoB1510-1

Detailed information about the sediment properties and microstructure can be provided through the analysis of digital ultrasonic P wave seismograms recorded automatically during full waveform core logging. The physical parameter which predominantly affects the elastic wave propagation in water-saturated sediments is the P wave attenuation coefficient. The related sedimentological parameter is the grain size distribution. A set of high-resolution ultrasonic transmission seismograms (ca. 50-500 kHz), which indicate downcore variations in the grain size by their signal shape and frequency content, are presented. Layers of coarse-grained foraminiferal ooze can be identified by highly attenuated P waves, whereas almost unattenuated waves are recorded in fine-grained areas of nannofossil ooze. Color-encoded pixel graphics of the seismograms and instantaneous frequencies present full waveform images of the lithology and attenuation. A modified spectral difference method is introduced to determine the attenuation coefficient and its power law a = kfn. Applied to synthetic seismograms derived using a "constant Q" model, even low attenuation coefficients can be quantified. A downcore analysis gives an attenuation log which ranges from ca. 700 dB/m at 400 kHz and a power of n = 1-2 in coarse-grained sands to few decibels per meter and n ? 0.5 in fine-grained clays. A least squares fit of a second degree polynomial describes the mutual relationship between the mean grain size and the attenuation coefficient. When it is used to predict the mean grain size, an almost perfect coincidence with the values derived from sedimentological measurements is achieved.

Upper Cretaceous sediments of ODP Hole 122-762C

Well-developed Campanian to Maestrichtian pelagic cyclic sediments were recovered from Hole 762C on the Exmouth Plateau, off northwest Australia, during Ocean Drilling Program Leg 122. The cycles consist of nannofossil chalk (light beds) and clayey nannofossil chalk (dark beds). Both light and dark beds are strongly to moderately bioturbated, alternate on a decimeter scale, and exhibit gradual boundaries. Bioturbation introduces materials from a bed of one color into an underlying bed of another color, indicating that diagenesis is not responsible for the cyclicity. Differences in composition between the light and dark beds, revealed by calcium carbonate measurement and X-ray diffraction analysis, together with trace fossil evidence, indicate that the cycles in the sediments are a depositional feature. Diagenetic processes may have intensified the appearance of the cycles. Spectral analysis was applied to the upper Campanian to lower Maestrichtian cyclic sediments to examine the regularity of the cycles. Power spectra were calculated from time series using Walsh spectral analysis. The most predominant wavelengths of the color cycles are 34-41 cm and 71-84 cm. With an average sedimentation rate of 1.82 cm/k.y. in this interval, we found the time durations of the cycles to be around 41 k.y. and 21 k.y., respectively, comparable to the obliquity and precession periods of the Earth's rotation, which strongly suggests an orbital origin for the cycles. On the basis of sedimentological evidence and plate tectonic reconstruction, we propose the following mechanism for the formation of the cyclic sediments from Hole 762C. During the Late Cretaceous, when there was no large-scale continental glaciation, the cyclic variations in insolation, in response to cyclic orbital changes, controlled the alternation of two prevailing climates in the area. During the wetter, equable, and warmer climatic phases under high insolation, more clay minerals and other terrestrial materials were produced on land and supplied by higher runoff to a low bioproductivity ocean, and the dark clayey beds were deposited. During the drier and colder climatic phases under low insolation, fewer clay minerals were produced and put into the ocean, where bioproductivity was increased and the light beds were deposited.

Measured Stark widths of several spectral lines of Pb III

The Stark full widths at half of the maximal line intensity (FWHM, ω) have been measured for 25 spectrallines of PbIII (15 measured for the first time) arising from the 5d106s8s, 5d106s7p, 5d106s5f and 5d106s5g electronic configurations, in a lead plasma produced by ablation with a Nd:YAG laser. The optical emission spectroscopy from a laser-induced plasma generated by a 10 640 Å radiation, with an irradiance of 2 × 1010 W cm− 2 on a lead target (99.99% purity) in an atmosphere of argon was analysed in the wavelength interval between 2000 and 7000 Å. The broadening parameters were obtained with the target placed in argon atmosphere at 6 Torr and 400 ns after each laser light pulse, which provides appropriate measurement conditions. A Boltzmann plot was used to obtain the plasma temperature (21,400 K) and published values of the Starkwidths in Pb I, Pb II and PbIII to obtain the electron number density (7 × 1016 cm− 3); with these values, the plasma composition was determined by means of the Saha equation. Local Thermodynamic Equilibrium (LTE) conditions and plasma homogeneity has been checked. Special attention was dedicated to the possible self-absorption of the different transitions. Comparison of the new results with recent available data is also presented.

Stark Broadening of Sn III Spectral Lines of Astrophysical Interest: Predictions and Regularities

The determination of the Stark broadening parameters of Sn ions is useful for astrophysicists interested in the determination of the density of electrons in stellar atmospheres. In this paper, we report on the calculated values of the Stark broadening parameters for 171 lines of Sn iii arising from 4d105sns (n= 6–9), 4d105snp (n= 5, 6), 4d105p2, 4d105snd (n= 5–7), 4d105s4f and 4d105s5g. Stark linewidths and line shifts are presented for an electron density of 1023 m−3 and temperatures T= 11 000–75 000 K. These have been calculated using a semi-empirical approach, with a set of wavefunctions obtained from Hartree–Fock relativistic calculations, including core polarization effects. The results obtained have been compared with available experimental data. These can be used to consider the influence of Stark broadening effects in A-type stellar atmospheres

Flujo de potencia entre estructuras complejas sometidas a excitaciones de alta frecuencia y aplicaciones a la detección de daño estructural

El tema central de investigación en esta Tesis es el estudio del comportamientodinámico de una estructura mediante modelos que describen la distribución deenergía entre los componentes de la misma y la aplicación de estos modelos parala detección de daños incipientes.Los ensayos dinámicos son un modo de extraer información sobre las propiedadesde una estructura. Si tenemos un modelo de la estructura se podría ajustar éstepara que, con determinado grado de precisión, tenga la misma respuesta que elsistema real ensayado. Después de que se produjese un daño en la estructura,la respuesta al mismo ensayo variará en cierta medida; actualizando el modelo alas nuevas condiciones podemos detectar cambios en la configuración del modeloestructural que nos condujeran a la conclusión de que en la estructura se haproducido un daño.De este modo, la detección de un daño incipiente es posible si somos capacesde distinguir una pequeña variación en los parámetros que definen el modelo. Unrégimen muy apropiado para realizar este tipo de detección es a altas frecuencias,ya que la respuesta es muy dependiente de los pequeños detalles geométricos,dado que el tamaño característico en la estructura asociado a la respuesta esdirectamente proporcional a la velocidad de propagación de las ondas acústicas enel sólido, que para una estructura dada es inalterable, e inversamente proporcionala la frecuencia de la excitación. Al mismo tiempo, esta característica de la respuestaa altas frecuencias hace que un modelo de Elementos Finitos no sea aplicable enla práctica, debido al alto coste computacional.Un modelo ampliamente utilizado en el cálculo de la respuesta de estructurasa altas frecuencias en ingeniería es el SEA (Statistical Energy Analysis). El SEAaplica el balance energético a cada componente estructural, relacionando la energíade vibración de estos con la potencia disipada por cada uno de ellos y la potenciatransmitida entre ellos, cuya suma debe ser igual a la potencia inyectada a cadacomponente estructural. Esta relación es lineal y viene caracterizada por los factoresde pérdidas. Las magnitudes que intervienen en la respuesta se consideranpromediadas en la geometría, la frecuencia y el tiempo.Actualizar el modelo SEA a datos de ensayo es, por lo tanto, calcular losfactores de pérdidas que reproduzcan la respuesta obtenida en éste. Esta actualización,si se hace de manera directa, supone la resolución de un problema inversoque tiene la característica de estar mal condicionado. En la Tesis se propone actualizarel modelo SEA, no en término de los factores de pérdidas, sino en términos deparámetros estructurales que tienen sentido físico cuando se trata de la respuestaa altas frecuencias, como son los factores de disipación de cada componente, susdensidades modales y las rigideces características de los elementos de acoplamiento.Los factores de pérdidas se calculan como función de estos parámetros. Estaformulación es desarrollada de manera original en esta Tesis y principalmente sefunda en la hipótesis de alta densidad modal, es decir, que en la respuesta participanun gran número de modos de cada componente estructural.La teoría general del método SEA, establece que el modelo es válido bajounas hipótesis sobre la naturaleza de las excitaciones externas muy restrictivas,como que éstas deben ser de tipo ruido blanco local. Este tipo de carga es difícil dereproducir en condiciones de ensayo. En la Tesis mostramos con casos prácticos queesta restricción se puede relajar y, en particular, los resultados son suficientementebuenos cuando la estructura se somete a una carga armónica en escalón.Bajo estas aproximaciones se desarrolla un algoritmo de optimización por pasosque permite actualizar un modelo SEA a un ensayo transitorio cuando la carga esde tipo armónica en escalón. Este algoritmo actualiza el modelo no solamente parauna banda de frecuencia en particular sino para diversas bandas de frecuencia demanera simultánea, con el objetivo de plantear un problema mejor condicionado.Por último, se define un índice de daño que mide el cambio en la matriz depérdidas cuando se produce un daño estructural en una localización concreta deun componente. Se simula numéricamente la respuesta de una estructura formadapor vigas donde producimos un daño en la sección de una de ellas; como se tratade un cálculo a altas frecuencias, la simulación se hace mediante el Método delos Elementos Espectrales para lo que ha sido necesario desarrollar dentro de laTesis un elemento espectral de tipo viga dañada en una sección determinada. Losresultados obtenidos permiten localizar el componente estructural en que se haproducido el daño y la sección en que éste se encuentra con determinado grado deconfianza.AbstractThe main subject under research in this Thesis is the study of the dynamic behaviourof a structure using models that describe the energy distribution betweenthe components of the structure and the applicability of these models to incipientdamage detection.Dynamic tests are a way to extract information about the properties of astructure. If we have a model of the structure, it can be updated in order toreproduce the same response as in experimental tests, within a certain degree ofaccuracy. After damage occurs, the response will change to some extent; modelupdating to the new test conditions can help to detect changes in the structuralmodel leading to the conclusión that damage has occurred.In this way incipient damage detection is possible if we are able to detect srnallvariations in the model parameters. It turns out that the high frequency regimeis highly relevant for incipient damage detection, because the response is verysensitive to small structural geometric details. The characteristic length associatedwith the response is proportional to the propagation speed of acoustic waves insidethe solid, but inversely proportional to the excitation frequency. At the same time,this fact makes the application of a Finite Element Method impractical due to thehigh computational cost.A widely used model in engineering when dealing with the high frequencyresponse is SEA (Statistical Energy Analysis). SEA applies the energy balance toeach structural component, relating their vibrational energy with the dissipatedpower and the transmitted power between the different components; their summust be equal to the input power to each of them. This relationship is linear andcharacterized by loss factors. The magnitudes considered in the response shouldbe averaged in geometry, frequency and time.SEA model updating to test data is equivalent to calculating the loss factorsthat provide a better fit to the experimental response. This is formulated as an illconditionedinverse problem. In this Thesis a new updating algorithm is proposedfor the study of the high frequency response regime in terms of parameters withphysical meaning such as the internal dissipation factors, modal densities andcharacteristic coupling stiffness. The loss factors are then calculated from theseparameters. The approach is developed entirely in this Thesis and is mainlybased on a high modal density asumption, that is to say, a large number of modescontributes to the response.General SEA theory establishes the validity of the model under the asumptionof very restrictive external excitations. These should behave as a local white noise.This kind of excitation is difficult to reproduce in an experimental environment.In this Thesis we show that in practical cases this assumption can be relaxed, inparticular, results are good enough when the structure is excited with a harmonicstep function.Under these assumptions an optimization algorithm is developed for SEAmodel updating to a transient test when external loads are harmonic step functions.This algorithm considers the response not only in a single frequency band,but also for several of them simultaneously.A damage index is defined that measures the change in the loss factor matrixwhen a damage has occurred at a certain location in the structure. The structuresconsidered in this study are built with damaged beam elements; as we are dealingwith the high frequency response, the numerical simulation is implemented witha Spectral Element Method. It has therefore been necessary to develop a spectralbeam damaged element as well. The reported results show that damage detectionis possible with this algorithm, moreover, damage location is also possible withina certain degree of accuracy.

Impact of N2 plasma power and duration on AlGaN/GaN HEMT

urface treatments have been recently shown to play an active role in electrical characteristics in AlGaN/GaN HEMTs, in particular during the passivation processing [1-4]. However, the responsible mechanisms are partially unknown and further studies are demanding. The effects of power and time N2 plasma pre-treatment prior to SiN deposition using PE-CVD (plasma enhanced chemical vapour deposition) on GaN and AlGaN/GaN HEMT have been investigated. The low power (60 W) plasma pre-treatment was found to improve the electronic characteristics in GaN based HEMT devices, independently of the time duration up to 20 min. In contrast, high power (150 and 210 W) plasma pretreatment showed detrimental effects in the electronic properties (Fig. 1), increasing the sheet resistance of the 2DEG, decreasing the 2DEG charge density in AlGaN/GaN HEMTs, transconductance reduction and decreasing the fT and fmax values up to 40% respect to the case using 60 W N2 plasma power. Although AFM (atomic force microscopy) results showed AlGaN and GaN surface roughness is not strongly affected by the N2-plasma, KFM (Kelvin force microscopy) surface analysis shows significant changes in the surface potential, trending to increase its values as the plasma power is higher. The whole results point at energetic ions inducing polarization-charge changes that affect dramatically to the 2-DEG charge density and the final characteristics of the HEMT devices. Therefore, we conclude that AlGaN surface is strongly sensitive to N2 plasma power conditions, which turn to be a key factor to achieve a good surface preparation prior to SiN passivation.

Theoretical Stark broadening parameters of Pb V spectral lines

In this paper, we report calculated values of the Stark widths and shifts for 72 spectral lines of Pb v. They were calculated using the Griem semi-empirical approach. A set of wavefunctions obtained from Hartree–Fock relativistic calculations including core polarization effects was used. Stark widths and shifts corresponding to lines arising from 5d9ns (n = 7, 8), 5d9 6p, 5d9 6d and 5d9 5f configurations of Pb v. Stark widths and shifts are presented for an electron density of 1017 cm−3 and temperatures T = 1.6–5.0 (104 K). The 2142.5, 2167.9 and 2278.6 Å lines of Pb v recently measured are included in our calculations. In this case, we have included for comparison calculations without core polarization effects. There is good agreement between our calculations and the above-cited experimental values.