Universiẗat, Diss., 1999--Saarbrücken. / Erscheinungsjahr auf der Haupttitels.: 1999.

This study deals with difficulties of different nature encountered in the process of translating a tourist booklet on Dover Castle, published by English Heritage. The study focuses on grammatical aspects as the translation of the passive and attributive adjectives, and lexical aspects as the translation of terminology and cultural words and expressions. The analysis is based on Peter Newmark's translation theory and translation strategies. Passives and adjectives have been divided into groups based on the translation models and strategies used, and the focus is on changes in the grammatical structure. Most passives were rendered as passives in Swedish by the use of s-passives, which involves a change in the grammatical structure. Most adjectives in the source text were translated by the use of a corresponding adjective in the target language, with no change in the grammatical structure. However, some adjectives were translated using for example a prepositional phrase or a compound noun. The analysis on terminology and cultural aspects focuses on what translation strategies have been used and problems that may occur concerning such aspects. Most source language terms were translated by the use of corresponding target language terms found in dictionaries or parallel texts. Some of the terms were transferred into the target text in their source text form, with an added explanation. The cultural aspects consisted mainly of names, which were dealt with in different ways. Some were transferred using the source language form, and in some cases conventionalised Swedish forms were used. The main aim of the translation has been to render the text as natural as possible in Swedish, and the main aim of the analysis has been to discuss certain aspects of difficulty that occurred in the translation process.

Thèse numérisée par la Direction des bibliothèques de l'Université de Montréal.

Cette thèse explore la question ouverte des mécanismes selon lesquels les galaxies lointaines évoluent au cours du temps. Elle se concentre sur l’étude des galaxies irrégulières et sur la cause de l’évolution en taille des galaxies passives et compactes. Bien que des régions de formation stellaire très lumineuses (clumps) soient observées dans les galaxies irrégulières depuis longtemps, leur nature et évolution sont encore débattues. Les instabilités gravitationnelles des disques ont été proposées comme la cause principale pour la formation in-situ des clumps, même si certains d’entre eux pourraient avoir une origine ex-situ. De plus, il n’est pas encore clair s’ils peuvent vivre longtemps ou si les vents stellaires les détruisent rapidement. À partir de l’étude détaillée d’un clump très jeune que nous avons découvert dans le disque d’une galaxie à redshift z~2 et de l’analyse d’un échantillon statistique, j’ai conclu que les clumps peuvent se former in-situ et qu’ils vivent typiquement 500 Myr. Ce résultat conforte les simulations numériques qui indiquent que les clumps ont un rôle important pour la croissance de leur noyau. Cela pourrait stabiliser le disque et y avoir un lien avec la formation des galaxies compactes et passives qui ont été decouvertes à haut redshift. Elles ont des tailles significativement plus petites, à masse égale, que celles de leurs homologues locales. Cette découverte a déclenché un débat concernant les possibles mécanismes qui peuvent augmenter leur taille sans altérer leur masse. J’ai analysé un échantillon de 32 galaxies et j’ai conclu que des multiples fusions mineures pourraient être les responsables principaux de leur evolution temporelle / This thesis explores the still unanswered question of how distant galaxies evolve through cosmic time: on one side it focuses on star-forming clumpy galaxies, on the other it investigates the size evolution of passive compact ones. Despite star forming clumps have been observed in high-redshift irregular galaxies since a while, their nature and fate are still highly debated. Violent gravitational disk instability in gas-rich, turbulent galaxies has been proposed as the main cause for in-situ clumps formation, although a fraction of them might have an ex-situ origin. Furthermore, clumps contribution to galaxy evolution is highly debated: it is not clear yet if they are long-lived or if stellar feedback rapidly disrupts them. From both the in-depth study of an extremely young clump that we discovered in the disk of a galaxy at redshift z ~ 2, and the analysis of a full statistical sample, I concluded that at least some clumps form in-situ due to violent disk instability and that they typically live ~ 500 Myr. This supports numerical simulations indicating that clumps are longlived and could play an important role in bulge growth. This might stabilize the disk, quench star formation and have therefore a link with the formation of the compact and passive galaxies that have been observed at high redshift. They have significant smaller sizes, at fixed stellar mass, than local counterparts. This discovery has ignited an important debate concerning the possible mechanisms that could inflate the galaxy sizes without altering much their mass. I analyzed a sample of 32 galaxies and I concluded that multiple minor mergers could be the main drivers of their observed time evolution

The purpose of this experimental study is to examine the effects of recasts on narrowing overgeneralized grammar in the second language (L2). The study involved testing three major hypotheses of the mechanisms underlying recasts: the direct contrast hypothesis (Saxton, 1997, 2000), the additional input hypothesis (Gass, 1997; Gass & Mackey, 2007; Long, 1996, 2007), and the enhanced salience hypothesis (Leeman, 2003). Two structures (adjective ordering and indirect passives) were selected for this study, mainly because it was assumed that Japanese learners of English might produce overgeneralized rules that allow incorrect structures because of their first language (L1) influence. The participants were 97 Japanese university students learning English as a foreign language in Japan. They were randomly assigned to four treatment groups: recast (the REC Group), non-contingent positive evidence (the POS Group), recast plus additional input (the REC+ Group), and input with enhanced salience (the SAL Group). A pretest, posttest, and delayed-posttest design were employed. The measurement instruments were an oral production task, elicited imitation task, and untimed grammaticality judgment task, each of which was designed to elicit participants’ implicit and explicit knowledge about adjective ordering and indirect passives. Thus, the independent variable was the treatment conditions, and the dependent variable were the test scores regarding ungrammaticality of the overgeneralized rules of the target structures. After data screening, the data from 75 of the 97 participants were analyzed for adjective ordering, and the data from 90 participants were analyzed for indirect passives. The results showed that the POS Group did not improve on all the measures for adjective ordering and indirect passives; thus, it was suggested that positive evidence was not sufficient for the participants to narrow overgeneralized rules for the target structures. The findings indicated that for adjective ordering, medium effect sizes for the comparison of the POS and REC Groups were obtained on the grammaticality judgment tests for the pretest-posttest and pretest-delayed posttest comparisons. Thus, based on these effect sizes, the direct contrast hypothesis was partially supported (i.e., for one of the two structures and one measure of the three tests). Furthermore, on the basis of the results that the REC and REC+ Groups did not differ significantly on any measure and that on the adjective-ordering grammaticality judgment tests, the comparison between the REC+ Group and the POS Group obtained a medium effect size for the pretest-delayed posttest comparison, the provision of recasts in the REC and REC+ Groups was effective at least for the adjective-ordering grammaticality judgment tests; however, additional input alone did not have an impact on L2 learners’ retreat from the overgeneralized rules. Finally, the results showed that the POS and SAL Groups did not differ significantly and that there existed a difference in the performance on the adjective-ordering grammaticality judgment tests between the REC Group and the SAL Groups. Salience might not be effective in helping L2 learners retreat from overgeneralized rules, and the enhanced salience hypothesis can be interpreted as being limited to learning some linguistic structures. In conclusion, the present study provides empirical evidence that L2 learners can persist in using incorrect overgeneralized rules due to L1 rules and that ten tokens of the target structures are not effective for L2 learners to retreat from the overgeneralized rules. Furthermore, the findings lend support to the direct contrast hypothesis as a mechanism underlying recasts for one of the two target structures. / Teaching & Learning

Around 19% of global power consumption and around 3% of global oil demand is attributable to lighting. After the first incandescent lamp was invented in 1879, more and more energy efficient lighting devices, such as gas discharge lamps, and light-emitting diodes (LED), have been developed during the last century. It is estimated that over 38% of future global lighting energy demand could be avoided by the use of more efficient lamps and ballasts [1]. High intensity discharge (HID) lamps, one category of gas discharge lamp, have been widely used in both commercial and residential lighting applications due to their merits of high efficacy, long life, compact size and good color rendition [2-4]. However, HID lamps require a well-designed ballast to stabilize the negative VI characteristics. A so-called ignitor is also needed to provide high voltage to initiate the gas discharge. Stringent input harmonic current limits, such as the IEC 61000-3-2 Class C standard, are set for lighting applications. It is well-known that high-frequency electronic ballasts can greatly save energy, improve lamp performance, and reduce the ballast size and weight compared with the conventional magnetic ballast. However, a unique phenomenon called acoustic resonance could occur in HID lamps under high-frequency operation. A low-frequency square wave current driving scheme has proved to be the only effective method to avoid acoustic resonance in HID lamps. A typical electronic HID ballast consist of three stages: power factor correction (PFC), DC/DC power regulation and low-frequency DC/AC inverter. The ignitor is usually integrated in the inverter stage. The three-stage structure results in a large size and high cost, which unfortunately offsets the merit of the HID lamp, especially in low-wattage applications. In order to make HID lamps more attractive in low-wattage and indoor applications, it is critical to reduce the size, weight and cost of HID ballasts. This dissertation is aimed at developing a compact HID with an ultra-compact ballast installed inside the lamp fixture. It is a similar concept to the compact fluorescent lamp (CFL), but it is much more challenging than the CFL. Two steps are explored to achieve high power density of the HID ballast. The first step is to improve the system structure and circuit topology. Instead of a three-stage structure, a two-stage structure is proposed, which consists of a single-stage power factor correction (SSPFC) AC/DC front-end and an unregulated DC/AC inverter/ignitor stage. An SSPFC AC/DC converter is proposed as the front-end. A DCM non-isolated flyback PFC semi-stage and a DCM buck-boost DC/DC semi-stage share the semiconductor switch, driver and PWM controller, so that the component count and cost can be reduced. The proposed SSPFC AC/DC front-end converter can achieve a high power factor, low THD, low bulk capacitor voltage, and the desired power regulation with a simple control circuit. Because the number of high-frequency switches is reduced compared to that of state-of-the-art two-stage HID ballast topologies, the switching frequency can be increased without sacrificing high efficiency, so the passive component size can be reduced. The power density of the whole ballast is increased using this two-stage structure. It results in a 2.5 times power density (6 W/in3) improvement compared to the commercial product (2.4 W/in3). The power density of the converter in discrete fashion usually suffers as a result of poor three-dimensional (3D) volume utilization due to a large component count and the different form factor of different components. In the second step, integration and packaging technologies are explored to further increase the power density. A 3D passive integrated HID ballast is proposed in this dissertation. All power passive components are designed in planar shape with a uniform form factor to fully utilize the three-dimensional space. In addition, electromagnetic integration technologies are applied to achieve structural, functional and processing integration to reduce component volume and labor cost. System partitioning, integration and packaging strategies, and implementation of major power passive integration, including an integrated EMI filter, and an integrated ignitor, will be discussed in the dissertation. The proposed integrated ballast is projected to double the power density of the discrete implementation. By installing the HID ballast inside the lamp fixture, the ambient temperature for the ballast will be much higher than the conventional separately installed ballast, and combined with a reduced size, the thermal condition for the integrated ballast will be much more severe. A thermal simulation model of the integrated ballast is built in the IDEAS simulation tool, and appropriate thermal management methods are investigated using the IDEAS simulation model. Experimental verification of various thermal management methods is provided. Based on the thermal management study, a new integrated ballast with improved thermal design is proposed. / Ph. D.

Résumé: Les lacs occupent environ 30% du territoire dans le nord du Canada. Ils peuvent avoir d’importants impacts sur le climat qui ont un effet par la suite sur les propriétés thermiques des lacs. L’étude d’un lac doit généralement débuter avec un bilan calorifique de la surface. La différentiation précise que l’énergie disponible en flux de chaleur sensible ou latent est un élément crucial pour la compréhension des interactions entre les processus climatiques à une échelle régionale. Dans cette étude, le flux de chaleur sensible et le ratio de Bowen sont obtenus à partir des données de la télédétection. La méthodologie se base sur l’utilisation de données microondes (SMM/I) afin de calculer la température de la surface du sol avec la méthode de Fily et al. (2003) et Mialon et al. (2007). La caractérisation de la surface du sol est effectuée à partir de données satellitaires optiques (capteurs SPOT & VGT). Les paramètres météorologiques utilisés pour estimer les flux proviennent de la base de données du North American Regional Reanalysis (NARR). Les résultats présentent un portrait quotidien, mensuel et saisonnier de l’énergie sensible à l’interface sol-air par rapport à la fraction de la surface de plan d’eau (FWS). Nous comparons ensuite nos résultats avec les flux du modèle NARR et quelques mesures in situ. Quatre sites couvrants différents types de surfaces à travers le Canada ont été étudiés pendant les étés 1998 et 2000 (de juin à septembre) : le nord du Québec (toundra), les Territoires du Nord-Ouest (le Grand lac des Esclaves [ou, en anglais, Great Slave Lake. GSLJ et le bassin de la rivière Mackenzie), le Manitoba (zone humide) et le Labrador (taïga). Les résultats montrent que les flux d’énergie sensible satellitaires sont semblables aux flux estimés par le NARR lorsque la FWS est petit (sans lac) ou pour des zones avec de larges surfaces d’eau (Mackenzie Great Slave Lake), mais diffèrent lorsque la FWS augmente à l’intérieur d’un pixel. Ceci signifie que les modèles climatiques régionaux devraient considérer la proposition du territoire occupée par des étendues d’eau. Nous déduisons que les effets de la taille des lacs sont reliés aux conditions environnementales du milieu. Les résultats de la comparaison avec des mesures in situ pour le GSL et la zone humide sont encourageants. Le ratio de Bowen sur le site du bassin de la rivière Mackenzie montre qu’il y a une augmentation de la chaleur sensible durant la deuxième moitié de l’été en comparaison à la première moitié. Il y a ainsi un plus grand stress hydrique pendant la deuxième moitié de l’été. Cependant, il n’y a pas de patrons clairs en comparaison avec le site GSL. La comparaison entre les différents sites indique une variation énergétique annuelle minimale pour la zone humide. || Abstract: Lakes occupy roughly 30% of Canada's northern landscape. They can have important impacts on the climate, and climate modification will affect lake thermal properties. Study of a lake most generally begins with heat budget at its surface. Accurate partitioning of the available energy at the surface into sensible and latent heat flux is crucial to the understanding of interactions between climate processes on a regional scale. In this study, sensible heat flux and the Bowen ratio for the summer period are retrieved using reanalysis and remote sensing data. The methodology is based on the use of microwave data for retrieving land surface temperature (SSM/I) with the method of Fily et al. (2003) and Mialon et al. (2007). The land cover characterization is derived from remotely sensed optical data (SPOT VGT sensor). Some meteorological parameters used for retrieving flux are derived from the North American Regional Reanalysis (NARR) database. The results from this study present a picture of the daily, monthly and seasonally sensible energy over summer period at the land-air interface versus the Fraction of Water Surface (FWS). We compare our results with the NARR model's flux in addition to few in situ measurements. Four sites covering different land cover types across Canada were investigated during the 1998 and 2000 summers (June to September): Northern Quebec (tundra), Northwest Territories (Great Slave Lake, or GSL, and Mackenzie River Basin), Manitoba (wetlands) and Labrador (taiga). The results show that the satellite-derived sensible flux is close to the NARR flux estimate when the fraction of water surface is small (no lakes) and over large open water areas (Mackenzie Great Slave Lake), but differs when the FWS increases within the pixel. This means that regional climate models should take into account lake cover fraction. We infer that effects of the lake-size are closely related to surrounding environmental conditions. The results of the comparison with in situ measurements for Great Slave Lake (1998) and the Wetland site are encouraging. For the Mackenzie River Basin site, the results of the Bowen ratio show that there is an increase of sensible heat partition during the second half of the summer in comparison with the first half, meaning that there is more water stress over the second half of the summer than the first. But for the Wetland site, there is no clear pattern, and in comparison with GSL, temporal variation is less significant. Comparisons between different sites indicate minimum year to year energy variation for the Wetland area.

Les cyclones tropicaux, gigantesques machines thermiques redistribuent l'énergie entre les régions chaude et froide du globe. Cette énergie (chaleur latente) provient surtout des processus de condensation. La connaissance de la chaleur latente devrait améliorer la prévision de l'évolution de ses systèmes dévastateurs détectables par satellite. TRMM lancé en 1997, permet la restitution des contenus en hydrométéores grâce à des mesures couplées radar/radiomètre (PR/TMI). Cette thèse porte sur la restitution de la chaleur latente. On a donc développé un algorithme original de type bayésien utilisant des bases de données, construites à partir des mesures du PR. Cette méthode, appliquée au cyclone Bret (1999) a été validée par comparaison avec des résultats de mesures aéroportées et sol. On reproduit assez bien sur ce cas une structure énergétique propre à différents stade de son évolution. Une étude statistique pourrait ainsi permettre un diagnostic de l'évolution.

The present thesis consists in a real complex case study of halotectonic deformation in the central portion of the Santos Basin passive margin, investigating the complex interactions between deformation and sedimentation and evaluating its impacts on the petroleum systems.Seismic data set and exploration wells were input to the study, which cover an area where halokinetics affect the active petroleum systems. The thickest depocenter of the Santos basin is partially in the study area. The sedimentary infilling history is associated with the Cabo Frio Fault development, linked with a gap of tens of kilometers in the Albian sediments occurrence. Minibasins are under halotectonic compression towards deep waters. The challenges of this work consists in find methods to build and then to restore the structural model in an effective way.Fifteen interpreted horizons were used to build a surface model. Five geological cross-sections were restored and interpolated to get salt and overburden volumes. The restoration successively removes the uppermost layer and decompacts the remaining layers through sedimentary unloading. Restoration also considers the unloading isostatic compensation by flexure of an elastic weak layered lithosphere. The isostasy plays an important role in the tectono-sedimentary control. Its interaction with bathymetry, faults geometry and deposition surface profile constrains the local and regional subsidence. Biostratigraphic data and seismic pattern defined trough time the absolute paleobathymetry.The cross-section restoration has differentiated the major rift phase sequence deformation from three other drift phase sequences above the sal layer.Several 1D geo-histories were reconstructed for different points along the main direction of transport, to control the thickness variations in the 2D sections. 1D and 2D restorations clearly show that salt moves out of the study area and that there is strong coupling between the halotectonic dynamic and the change in the accommodation space.The extrapolation of the restoration results to 3D allows a better compartmentalization between the hemigrabens in the rift phase; highlights the role of structural high in the center of the investigated area and confirms the relation between structural evolution and sedimentation. Two major depocenters were amalgamated by salt deposition over the accentuated relief inherited from the larger vertical tectonic subsidence of the rift phase. The sedimentary infilling reveals the temporal relations of NE-SW and NNW-SSE oriented structures within the higher subsidence compartment in the western sector of the area. Two different ways to restore the sections allowed the uncertainty estimation of the bathymetry and isostatic relationships.The salt layer material conservation assumption, an operational tool, was consistent with the results. The salt volume balance over time showed a good correlation to lateral extension of the continental drift stage.The implications of the restoration scenario on the petroleum systems were only qualitatively interpreted and then compared to the classical backstripping approach. Three main sectors appear: the proximal shelf, the slope and the deepwater sector of the continental rise. The relatively early halotectonics of the proximal sector induces a more fast and short generation whereas, beyond the continental slope region, the petroleum systems is distally more heterogeneous, develops slower, dissipated and differentiated by the halokinesis, with longer refills fronts. The direct impact of salt restoration on the maturity evaluation is actually relatively small, because about 90% of ductile deformation occurred before 65 Ma and a significant part of the generation is ongoing after 65 Ma. On the other hand, the halotectonic restoration is fundamental regarding its impact on the petroleum systems.

Els robots caminadors presenten alguns avantatges respecte als tradicionals robots mòbils amb rodes, que es resumeixen en major mobilitat i capacitat de superació d'obstacles. Un dels principals problemes actuals en el desenvolupament de robots caminadors és l'elevat consum energètic i la poca autonomia, com a conseqüència de l'elevada massa dels accionaments i del seu rendiment limitat.Un primer objectiu d'aquesta tesi és desenvolupar i caracteritzar energèticament un accionament rotatiu per a les articulacions de la pota d'un robot caminador. Un cop desenvolupat aquest accionament s'analitza el comportament energètic global de la pota.La pota dissenyada és de tipus insecte i l'aplicació considerada és un robot hexàpode. L'accionament rotatiu consta de tres elements: un motor de corrent continu, una transmissió per corretja dentada i un reductor Harmonic Drive. El conjunt d'aquest reductor s'ha dissenyat específicament per a aquesta aplicació, i materialitza també l'enllaç de cada articulació de la pota.Per tal de determinar els requeriments mecànics a les articulacions, s'ha portat a terme una simulació dinàmica del moviment de la pota. Aquesta simulació s'ha realitzat considerant que les articulacions del mecanisme són ideals. Les resistències passives d'aquestes es tracten posteriorment al analitzar el comportament del reductor. Els resultats obtinguts són els parells requerits a les diferents articulacions al llarg d'un cicle.Amb l'objectiu de determinar el comportament real dels components de l'accionament, per a diferents condicions de funcionament, s'han plantejat i realitzat diverses sèries d'experiments. Per portar a terme els experiments ha estat necessària la construcció d'un prototipus de l'accionament. A partir de les dades experimentals s'han definit uns models matemàtics que representen les resistències passives a cada component de l'accionament per a diferents condicions de funcionament. S'ha realitzat un ajust per mínims quadràtics per tal determinar els paràmetres dels models que millor ajusten els resultats a les dades experimentals.Com que cada motor treballa com tal o com fre, segons la fase del moviment, s'ha analitzat en detall el funcionament d'una màquina elèctrica de corrent continu d'imants permanents, des del punt de vista de la conversió de potència elèctrica en mecànica i viceversa, en els 4 quadrants i tenint en compte els diferents modes de funcionament: com a motor, com a fre-generador i com a fre.Posteriorment s'han definit uns models globals del comportament mecànic dels elements de transmissió de l'accionament: la corretja dentada i el reductor HD, que integren la modelització de les resistències passives i també inclouen la inèrcia de les parts mòbils de cada element.Finalment s'ha realitzat la simulació dels tres accionaments d'una pota al llarg d'un cicle de l'anadura del robot, a partir dels requeriments de parell i velocitat angular a cada articulació i utilitzant els models dels elements definits anteriorment. L'objectiu és estimar l'energia intercanviada a cada element de l'accionament, avaluant per separat l'energia transmesa en sentit directe, des de l'alimentació del motor a la sortida del reductor, i en sentit invers. El resultat més important és l'estimació de l'energia elèctrica que cal subministrar a cada motor per cada cicle de moviment de la pota.Dels diversos casos analitzats se'n treu la conclusió que l'energia elèctrica que cal subministrar als motors és sempre molt superior, de l'ordre de 30 vegades, a l'energia mecànica que l'accionament transmet a l'articulació. La despesa energètica dels motors depèn poc de les condicions de càrrega a les que està sotmesa la pota, en canvi el moviment que descriuen les articulacions al llarg del cicle hi té una influència important. Reduint el recorregut angular del moviment de les articulacions s'aconsegueix un considerable estalvi energètic, doncs disminueix l'energia dissipada per les resistències passives. / Walking robots have some advantages respect to the traditional mobile wheeled robots. These can be summarised in superior mobility and the capacity to overcome obstacles. One of the main problems in the development of walking robots is the high energy consumption and the low autonomy of operation, due to the high mass of the actuators and their limited efficiency.A first objective of this thesis is to develop a rotary actuator, for the leg pin joints of a walking robot, and to characterize it energetically. Once the actuator is developed, leg global energetic performance is analysed.The leg design is according to an insect type leg and the considered application is a hexapod robot. The rotary actuator consists of three elements: a DC electric motor, a synchronous belt transmission and a Harmonic Drive gear reducer. Gear assembly has been particularly designed for this application, and integrates pin joint implementation.In order to determine the mechanical specifications at the joints, a dynamic simulation of the leg movement has been carried out. For this simulation pin joints in the mechanism have been considered as ideal joints. Resistance to motion at the joints are taken into account later, when gear operation is analysed. Results from the simulation are the required torques at the different joints along a walking cycle.Several series of experiments have been planned and carried out in order to work out the real behaviour of the actuator components under different working conditions. The construction of an actuator prototype has been necessary for the experimentation.Then derived from the experimental data, mathematical models that estimate resistance to motion at every actuator component as a function of the working conditions, have been defined. Parameters in the models to best fit the experimental data have been identified by least square approximations.Every DC motor works as a motor or as a brake depending on the movement stage. Therefore the operation of a permanent magnet DC electric machine has been analysed, from the point of view of the conversion of electrical power into mechanical and vice versa. This was performed taking into account the different operation modes: as a motor, as a brake-generator and as a brake.Global models that represent the mechanical behaviour of the transmission components in the actuator have been defined: the synchronous belt and the HD gear. These models integrate resistance to motion estimation and also the inertia of the mobile parts of every element. Finally, starting from the torque and angular velocity requirements at every joint and using the models of the elements described above, the simulation of the three actuators of a leg along a gait cycle of the robot has been carried out. The objective was to estimate the exchanged energy at every element of the actuator, separately calculating the transmitted energy in both directions: from the motor power supply to the output of the gear reducer and in reverse. The main result was the estimation of the electrical energy that must be supplied to every motor per leg movement cycle. From the several cases that have been analysed, it can be concluded that the electrical energy that has to be supplied to the motors is always much higher, around 30 times higher, than the mechanical energy that the actuator transmits to the joint. Motor energetic consumption has low dependence on leg load condition, but pin joint movement along a cycle has a great influence on it. Reducing the angular run of pin joint movements is a way to obtain substantial energy savings, since it decreases the resistance to motion energy dissipation.