Plasma nitriding of AA2011 alloy and surface characterization by NDT techniques

Al-Cu alloys are widely used in the aerospace and automotive industries due to their high specific strength in some tempered conditions. However, due to poor corrosion and wear resistance, they are often anodized and/or painted. Plasma nitriding has been proposed as an alternative, though the developments in this technique are still in a recent stage for Al alloys. Electrical characterization techniques are well implemented NDTs in the industry because of good accuracy associated with lower cost, compared to other methods. Some, like eddy currents and 4-point probe techniques, are often used in coating inspection. The objective of this study was to perform Al nitriding at low temperatures to minimize the tempering initial condition damage and to assess the feasibility of eddy currents technique as a method for evaluating surface properties. The work developed can be divided in two stages. The first one was the process tuning, done at the Shibaura Institute of Technology, in Tokyo; and the second was the electrical characterization done in Faculdade de Ciências e Tecnologia, UNL. Low temperature nitriding of AA2011 alloy specimens was successfully achieved. Electrical conductivity results show that lift-off measurements by eddy currents testing can be related to surface properties.

Hammaspyörän hampaan täytehitsaus

Tämän työn tarkoituksen on perehtyä hammaspyörän hampaan täytehitsaukseen. Ennen hitsausta on selvitettävä hammaspyörän materiaali sekä mikrorakenne, jonka perusteella tutkitaan hitsattavuutta. Työtä voidaan soveltaa teollisuudessa korjaushitsauksiin, jos laite on saatava nopeasti toimintakuntoon tai uuden hammaspyörän hankinta on mahdotonta tai kallista. Työssä perehdytään lähinnä hitsausprosessin onnistumiseen ja hammaspyörässä tapahtuviin mikrorakennemuutoksiin sekä kerrotaan erilaisia hitsausmenetelmiä lopputuloksen parantamiseksi. Työn alussa kerrotaan lyhyesti hammaspyörän seostuksen selvittämisestä, joka on oleellinen osa hitsausparametrien valinnassa ja hitsauksen onnistumisessa. Seostuksen selvittämiseen on useita keinoja, joista paras ja tarkin menetelmä on etsiä hammaspyörän valmistajan ainestodistus. Yleensä se ei kuitenkaan ole mahdollista, koska hammaspyörän alkuperää ei tunneta. Tässä työssä on käytetty esimerkkinä Valmet 361-traktorin perän hammaspyörää, josta on murtunut hammas kovassa rasituksessa.

Hitsauksen lämpökäsittelyn vaikutus CrMo - terästen mekaanisiin ominaisuuksiin

Diplomityössä selvitettiin lujuusomaisuuksien muuttumista toimitustilassa olevilla ja hitsatuilla The American Society of Mechanical Engineering (ASME) SA335 P22 ja P5 teräksillä, kun hitsauksen lämpökäsittelylämpötila oli suurempi kuin materiaalin valmistuksen aikainen päästölämpötila. Diplomityön perusteella ei pystytä yksiselitteisesti todentamaan tai kumoamaan standardeissa ASME SA335: 2008a, ASME B31.3: 2010 ja SFS-EN 13445: 2009 esitettyjä varoituksia materiaalien mekaanisten ominaisuuksien huononemisesta, jos hitsien jälkilämpökäsittely ylittää materiaalien päästölämpötilan. Veto-, kovuus- ja iskukoetestitulosten perusteella todennettiin kuitenkin, ettei lämpökäsittely lämpötila-alueella 710–760 oC vaikuta suuresti materiaalien mekaanisiin ominaisuuksiin. Tuloksien perusteella todettiin myös, että koemateriaalivalikoima ja tehdyt testisarjat olivat suppeahkoja tarkkojen hankinta- ja valmistuskriteerien määrittämiseksi.

Effect of heat input on the mechanical properties of welded joints in high strength steels

Experiments were carried out to determine the properties of the welded joints in 8mm thick high-strength steels produced by quenching and tempering and thermomechanical rolling with accelerated cooling (tensile strength 821–835 MPa). The dependence of the strength, elongation, hardness, impact energy and crack opening displacement on the heat input in the range 1.0–0.7 kJ mm21 was determined. The results show that the dependence of the strength of the welded joints decreases and that of the elongation increases. The heat input has only a slight effect on the impact energy and crack opening displacement in the heat-affected zone.

Chocolate rheology

Rheology is the science that studies the deformation and flow of solids and fluids under the influence of mechanical forces. The rheological measures of a product in the stage of manufacture can be useful in quality control. The microstructure of a product can also be correlated with its rheological behavior allowing for the development of new materials. Rheometry permits attainment of rheological equations applied in process engineering, particularly unit operations that involve heat and mass transfer. Consumer demands make it possible to obtain a product that complies with these requirements. Chocolate industries work with products in a liquid phase in conching, tempering, and also during pumping operations. A good design of each type of equipment is essential for optimum processing. In the design of every process, it is necessary to know the physical characteristics of the product. The rheological behavior of chocolate can help to know the characteristics of application of the product and its consumers. Foods are generally in a metastable state. Their texture depends on the structural changes that occur during processing. Molten chocolate is a suspension with properties that are strongly affected by particle characteristics including not only the dispersed particles but also the fat crystals formed during chocolate cooling and solidification. Chocolate rheology is extensively studied, and it is known that chocolate texture and stability is strongly affected by the presence of specific crystals

Training deep convolutional architectures for vision

Les tâches de vision artificielle telles que la reconnaissance d’objets demeurent irrésolues à ce jour. Les algorithmes d’apprentissage tels que les Réseaux de Neurones Artificiels (RNA), représentent une approche prometteuse permettant d’apprendre des caractéristiques utiles pour ces tâches. Ce processus d’optimisation est néanmoins difficile. Les réseaux profonds à base de Machine de Boltzmann Restreintes (RBM) ont récemment été proposés afin de guider l’extraction de représentations intermédiaires, grâce à un algorithme d’apprentissage non-supervisé. Ce mémoire présente, par l’entremise de trois articles, des contributions à ce domaine de recherche. Le premier article traite de la RBM convolutionelle. L’usage de champs réceptifs locaux ainsi que le regroupement d’unités cachées en couches partageant les même paramètres, réduit considérablement le nombre de paramètres à apprendre et engendre des détecteurs de caractéristiques locaux et équivariant aux translations. Ceci mène à des modèles ayant une meilleure vraisemblance, comparativement aux RBMs entraînées sur des segments d’images. Le deuxième article est motivé par des découvertes récentes en neurosciences. Il analyse l’impact d’unités quadratiques sur des tâches de classification visuelles, ainsi que celui d’une nouvelle fonction d’activation. Nous observons que les RNAs à base d’unités quadratiques utilisant la fonction softsign, donnent de meilleures performances de généralisation. Le dernière article quand à lui, offre une vision critique des algorithmes populaires d’entraînement de RBMs. Nous montrons que l’algorithme de Divergence Contrastive (CD) et la CD Persistente ne sont pas robustes : tous deux nécessitent une surface d’énergie relativement plate afin que leur chaîne négative puisse mixer. La PCD à "poids rapides" contourne ce problème en perturbant légèrement le modèle, cependant, ceci génère des échantillons bruités. L’usage de chaînes tempérées dans la phase négative est une façon robuste d’adresser ces problèmes et mène à de meilleurs modèles génératifs.

Improving sampling, optimization and feature extraction in Boltzmann machines

L’apprentissage supervisé de réseaux hiérarchiques à grande échelle connaît présentement un succès fulgurant. Malgré cette effervescence, l’apprentissage non-supervisé représente toujours, selon plusieurs chercheurs, un élément clé de l’Intelligence Artificielle, où les agents doivent apprendre à partir d’un nombre potentiellement limité de données. Cette thèse s’inscrit dans cette pensée et aborde divers sujets de recherche liés au problème d’estimation de densité par l’entremise des machines de Boltzmann (BM), modèles graphiques probabilistes au coeur de l’apprentissage profond. Nos contributions touchent les domaines de l’échantillonnage, l’estimation de fonctions de partition, l’optimisation ainsi que l’apprentissage de représentations invariantes. Cette thèse débute par l’exposition d’un nouvel algorithme d'échantillonnage adaptatif, qui ajuste (de fa ̧con automatique) la température des chaînes de Markov sous simulation, afin de maintenir une vitesse de convergence élevée tout au long de l’apprentissage. Lorsqu’utilisé dans le contexte de l’apprentissage par maximum de vraisemblance stochastique (SML), notre algorithme engendre une robustesse accrue face à la sélection du taux d’apprentissage, ainsi qu’une meilleure vitesse de convergence. Nos résultats sont présent ́es dans le domaine des BMs, mais la méthode est générale et applicable à l’apprentissage de tout modèle probabiliste exploitant l’échantillonnage par chaînes de Markov. Tandis que le gradient du maximum de vraisemblance peut-être approximé par échantillonnage, l’évaluation de la log-vraisemblance nécessite un estimé de la fonction de partition. Contrairement aux approches traditionnelles qui considèrent un modèle donné comme une boîte noire, nous proposons plutôt d’exploiter la dynamique de l’apprentissage en estimant les changements successifs de log-partition encourus à chaque mise à jour des paramètres. Le problème d’estimation est reformulé comme un problème d’inférence similaire au filtre de Kalman, mais sur un graphe bi-dimensionnel, où les dimensions correspondent aux axes du temps et au paramètre de température. Sur le thème de l’optimisation, nous présentons également un algorithme permettant d’appliquer, de manière efficace, le gradient naturel à des machines de Boltzmann comportant des milliers d’unités. Jusqu’à présent, son adoption était limitée par son haut coût computationel ainsi que sa demande en mémoire. Notre algorithme, Metric-Free Natural Gradient (MFNG), permet d’éviter le calcul explicite de la matrice d’information de Fisher (et son inverse) en exploitant un solveur linéaire combiné à un produit matrice-vecteur efficace. L’algorithme est prometteur: en terme du nombre d’évaluations de fonctions, MFNG converge plus rapidement que SML. Son implémentation demeure malheureusement inefficace en temps de calcul. Ces travaux explorent également les mécanismes sous-jacents à l’apprentissage de représentations invariantes. À cette fin, nous utilisons la famille de machines de Boltzmann restreintes “spike & slab” (ssRBM), que nous modifions afin de pouvoir modéliser des distributions binaires et parcimonieuses. Les variables latentes binaires de la ssRBM peuvent être rendues invariantes à un sous-espace vectoriel, en associant à chacune d’elles, un vecteur de variables latentes continues (dénommées “slabs”). Ceci se traduit par une invariance accrue au niveau de la représentation et un meilleur taux de classification lorsque peu de données étiquetées sont disponibles. Nous terminons cette thèse sur un sujet ambitieux: l’apprentissage de représentations pouvant séparer les facteurs de variations présents dans le signal d’entrée. Nous proposons une solution à base de ssRBM bilinéaire (avec deux groupes de facteurs latents) et formulons le problème comme l’un de “pooling” dans des sous-espaces vectoriels complémentaires.

Monte Carlo phase diagram for diblock copolymer melts

The phase diagram for diblock copolymer melts is evaluated from lattice-based Monte Carlo simulations using parallel tempering, improving upon earlier simulations that used sequential temperature scans. This new approach locates the order-disorder transition (ODT) far more accurately by the occurrence of a sharp spike in the heat capacity. The present study also performs a more thorough investigation of finite-size effects, which reveals that the gyroid (G) morphology spontaneously forms in place of the perforated-lamellar (PL) phase identified in the earlier study. Nevertheless, there still remains a small region where the PL phase appears to be stable. Interestingly, the lamellar (L) phase next to this region exhibits a small population of transient perforations, which may explain previous scattering experiments suggesting a modulated-lamellar (ML) phase.

Interaction of heat-moisture conditions and physical properties in oat processing: II. Flake quality

Product quality is an important determinant of consumer acceptance. Consistent oat flake properties are thus necessary in the mill as well as in the marketplace. The effects of kilning and tempering conditions (30, 60 or 90 min at 80, 95 or 110 degrees C) on flake peroxidase activity, size, thickness, strength and water absorption were therefore determined. After kilning, some peroxidase activity remained but steaming and tempering effectively destroyed the activity of these enzymes. Thus the supposed protective effect of kilning or groat durability was not confirmed. Kilning resulted in an increase in flake specific weight, but no other significant effect on flake quality was observed. Tempering time and temperature interacted significantly to produce complex effects on flake specific weight, thickness and water absorption. Flake thickness and specific weight were significantly correlated (r = 0.808, n = 54). Longer tempering times resulted in an increased fines' fraction, from 1.45% at 30 min to 1.75% at 90 min. It is concluded that whilst kilning has little effect on flake quality, the heat treatment immediately prior to flaking, can be used to adjust flake quality independently of flake thickness.

Interaction of heat-moisture conditions and physical properties in oat processing: I. Mechanical properties of steamed oat groats

Research interest in oats has focussed on their nutritional value, but there have been few studies of their food processing. Heat treatment is characteristic of oat processing, as it is needed to inactivate lipase and to facilitate flaking. A Texture Analyser was used to characterise the mechanical properties of unkilned and kilned oat groats after steaming and tempering in an oven for 30, 60 and 90 min at 80, 95 and 110 degrees C. Maximum force, number of peaks before maximum and final force after 5s hold were used to characterise the behaviour of the groats during compression. Kilned groats were larger and softer before steaming. After steaming and tempering, the moisture content of the kilned groats was higher than for unkilned groats. Hot, steamed oats were softer than cold, unsteamed groats, indicated by a decrease in maximum force from 59 to 55 N, and there was no significant difference between kilned and unkilned groats. However, higher temperatures during tempering increased maximum force. These results suggest that mild steam treatment yields softer oat groats, whereas cold or over-treated groats tend to be harder. (c) 2007 Elsevier Ltd. All rights reserved.

Power to the people: working-class demand for household power in 1930s Britain

The 1930s witnessed an intense struggle between gas and electricity suppliers for the working class market, where the incumbent utility—gas—was also a reasonably efficient (and cheaper) General Purpose Technology for most domestic uses. Local monopolies for each supplier boosted substitution effects between fuel types—as alternative fuels constituted the only local competition. Using newly-rediscovered returns from a major national household expenditure survey, we employ geographically-determined instrumental variables, more commonly used in the industrial organization literature, to show that gas provided a significant competitor, tempering electricity prices, while electricity demand was also responsive to marketing initiatives.

Polymorphism, crystallinity and hydrophilic-lipophilic balance of stearic acid and stearic acid-capric/caprylic triglyceride matrices for production of stable nanoparticles

There is an increasing interest in lipid nanoparticles because of their suitability for several administration routes. Thus, it becomes even more relevant the physicochemical characterization of lipid materials with respect to their polymorphism, lipid miscibility and stability, as well as the assessment of the effect of surfactant on the type and structure of these nanoparticles. This work focuses on the physicochemical characterization of lipid matrices composed of pure stearic acid or of mixtures of stearic acid-capric/caprylic triglycerides, for drug delivery. The lipids were analyzed by Differential Scanning Calorimetry (DSC), Wide Angle X-ray Diffraction (WAXD), Polarized Light Microscopy (PLM) and hydrophilic-lipophilic balance (HLB) in combination with selected surfactants to determine the best solid-to-liquid ratio. Based on the results obtained by DSC and WAXD, the selected qualitative and quantitative composition contributed for the production of stable nanoparticles, since the melting and the tempering processes provided important information on the thermodynamic stability of solid lipid matrices. The best HLB value obtained for stearic acid-capric/caprylic triglycerides was 13.8, achieved after combining these lipids with accepted surfactants (trioleate sorbitan and polysorbate 80 in the ratio of 10:90). The proposed combinations were shown useful to obtain a stable emulsion to be used as intermediate form for the production of lipid nanoparticles. (C) 2011 Elsevier B.V. All rights reserved.

Six-Day System Test and Component test and System Simulation for Combitanks with Internal heat Exchangers

An international standard, ISO/DP 9459-4 has been proposed to establish a uniform standard of quality for small, factory-made solar heating systerns. In this proposal, system components are tested separatelyand total system performance is calculated using system simulations based on component model parameter values validated using the results from the component tests. Another approach is to test the whole system in operation under representative conditions, where the results can be used as a measure of the general system performance. The advantage of system testing of this form is that it is not dependent on simulations and the possible inaccuracies of the models. Its disadvantage is that it is restricted to the boundary conditions for the test. Component testing and system simulation is flexible, but requires an accurate and reliable simulation model.The heat store is a key component conceming system performance. Thus, this work focuses on the storage system consisting store, electrical auxiliary heater, heat exchangers and tempering valve. Four different storage system configurations with a volume of 750 litre were tested in an indoor system test using a six -day test sequence. A store component test and system simulation was carried out on one of the four configurations, applying the proposed standard for stores, ISO/DP 9459-4A. Three newly developed test sequences for intemalload side heat exchangers, not in the proposed ISO standard, were also carried out. The MULTIPORT store model was used for this work. This paper discusses the results of the indoor system test, the store component test, the validation of the store model parameter values and the system simulations.

Estudos eletroquímicos de permeação de hidrogênio e determinação de concentrações críticas de hidrogênio em um aço de alta dureza

Estudou-se, através do método eletroquímico de permeação, a difusão de hidrogênio no ferro puro e no aço DIN 90MnV8. Observou-se nestes materiais, que o coeficiente aparente de difusão de hidrogênio diminui no decorrer dos transientes de permeação, tendo-se interpretado este comportamento como causado pela presença de armadilhas reversíveis. Deduziu-se, para transientes de permeação obtidos entre dois estados estacionários de difusão, que a variação da concentração de hidrogênio preso a armadilhas reversiveis é dada por: Delta Ca = Na. In [(1+alfa)1/2 dividido por (1+beta)1/2], alfa = K.Co/p beta=K.Cl/p sendo Na a densidade de armadilhas; k e p, os parâmetros de captura e de liberação de uma armadilha e Co e Cl, a concentrasão inicial e final de H. Mediu-se, no ferro recozido, um coeficiente médio de difusão de hidrogênio de 6,4.10-5 cm²/s. A deformação com 50% de trabalho a frio no ferro reduz este valor em aproximadamente uma ordem de grandeza. Verificou-se no aço DIN 90MnV8 que o coeficiente de difusão de hidrogênio aumenta com o aumento dn temperatura de revenimento após a têmpera, atingindo o valor máximo de 7,0.10-6 cm²/s no aço esferoidizado. A dependência térmica do coeficiente de difusão no aço DIN 90MnV8 esferoidizado pode ser expressa por: D = 1,95 (+0,49) exp (-30,43kJ/mol.R.T.) (cm²/s) e neste mesmo aço temperado e revenido com 55RC de dureza por : D=0,29 (+- 0,05 ) exp (-28,47kJ/mol.R.T.) ( cm²/s). Uma nova metodolagia fractomecânica foi introduzida, a qual permite a determinação da cnncentração crítica de hidrogênio necessária para provocar o avanço de uma trinca estacionária. Observou-se no aço DIN 9OMnV8, que a concentração crítica de hidrogênio diminui com o aumento da dureza do aço e é menor quando um valor maior de KI é aplicado. O modo de fratura apresentado Por este aço revelou ser independente da concentração de hidrogênio e do valor de KI, sendo função unicamente de sua dureza.